Intelligent Tagging Semantic Metadata Tags

This document provides a list of all supported semantic metadata tags and their attributes, along with descriptions and examples.

For a conceptual explanation of these tags, see How Does Intelligent Tagging Work?

Use the x-calais-selectiveTags header (mandatory for new users, soon-to-be mandatory for all users) to specify which of the metadata types may appear in the API response.

Note to Internal Intelligent Tagging users: See the Supplementary Guide for Internal Intelligent Tagging (for internal customers who do not connect through API Gateway) for details about a few differences in the tagging output.

Additional resource: API User Guide.

Read about which metadata tags are actively enhanced and supported.

Instance Tag

Entity Markup Tags: Entities, Relations

Slugline Tag (premium feature)

Industry Tag (premium feature)

ForEndUserDisplay Attribute

PermID Attribute

Note: Intelligent Tagging also supports several non-English languages: Chinese, French, German, Japanese, Spanish. Each language supports a specific subset of metadata types. For details, see Intelligent Tagging for Non-English Languages.

Entity MarkupTags

Each Intelligent Tagging entity markup tag identifies an Entity or a Relation. Entities are straightforward things like companies, people, cities, telephone numbers; Relations are more complex mentions that indicate relationships between things. For example, deals, IPOs, analyst recommendations, company reorganizations.

Entities

Click a link for a description of the entity type and its attributes.

It is important to note that most attributes are optional; a tag can be extracted with some but not all of its attributes.

EntertainmentAwardEvent

*These are the actively enhanced and supported metadata types.

Relations

Click a link for a description of the relation type and its attributes.

It is important to note that most attributes are optional; a tag can be extracted with some but not all of its attributes.

Acquisition

Alliance

AnalystEarningsEstimate

AnalystRecommendation

CompanyAccountingChange

CompanyAffiliates

CompanyCompetitor

CompanyCustomer

CompanyEarningsAnnouncement

CompanyEarningsGuidance

CompanyEmployeesNumber

CompanyReorganization

PoliticalRelationship

TIP: Improve the search results of your application by mapping common search strings to entity types. A list of synonyms for financial Entity Relation types is available.

Note to Internal Intelligent Tagging users: Please see the Supplementary Guide for Internal Intelligent Tagging (for internal customers who do not connect through API Gateway) for a list of supported entities and relations, and for details about the differences in the tagging output.

Which Metadata Tags are Actively Enhanced and Supported?

Intelligent Tagging supports an extensive set of (over 100) metadata types. We actively focus development efforts on enhancing the focused list of concepts that are the most important to our customers.

Priority 1: The following metadata concepts are actively enhanced and tuned:

- Entities: Company, Country, Currency, CurrencyPair, MarketIndex, Person

- Classification: DocCat (topic tags), Slugline, SocialTag

Priority 2: The following metadata concepts may be tuned on demand, but not too often:

- Entities: City, PharmaceuticalDrug

- Relations: Acquisition, Alliance, Bankruptcy, Buybacks, CompanyAffiliates, CompanyEarningsAnnouncement, CompanyLayoffs, CreditRating, Deal, Dividend, IPO, JointVenture, Merger, NaturalDisaster, PersonCareer

Priority 3: The rest of the metadata concepts are experimental and are not maintained (P/R issues are not being actively improved).

Complete list of Entity and Relation types.

InstanceInfo Tag

For a conceptual explanation of this tag, see How Does Intelligent Tagging Work?

Name

InstanceInfo

Definition

Describes a mention of an Intelligent Tagging type found in the text.

Attributes

Detection: The text string in which the mention was identified.

docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format.

exact: The mention.

length: The length (in characters) of the mention.

offset: Offset of the mention (in characters) from the beginning of the document.

prefix: The portion of the text string that precedes the mention.

subject: A hash tag generated by Intelligent Tagging. This ID points to the associated extracted entity or relation.

suffix: The portion of the text string that follows the mention.

The tag also includes a comment that indicates the type of entity or relation extracted from this instance, and its main attribute values.

Note: In the JSON output format, because related tags are nested, the docid and subject attributes are not displayed in the Instance tag.

Example 1: A mention of a Company (Company type)

Input Text

Technology giant Apple (APPL.O) is looking beyond mobile devices to learn how to make a self-driving electric car.

Output

InstanceInfo tag in RDF output format:

<rdf:Description rdf:about="http://d.opencalais.com/dochash-1/7586b818-40af-3d55-ac16-bf3520cddfa6/Instance/60">

<rdf:type rdf:resource="http://s.opencalais.com/1/type/sys/InstanceInfo"/>

<c:docId rdf:resource="http://d.opencalais.com/dochash-1/7586b818-40af-3d55-ac16-bf3520cddfa6"/>

<c:subject rdf:resource="http://d.opencalais.com/comphash-1/705cd5cf-93e1-323c-8d4e-1ea3200d37e4"/>



<c:detection>[<Body> (Reuters) - Technology giant ]Apple (AAPL.O)[ is looking beyond mobile devices to learn how to]</c:detection>

<c:prefix> <Body> (Reuters) - Technology giant </c:prefix>

<c:exact>Apple (AAPL.O)</c:exact>

<c:suffix> is looking beyond mobile devices to learn how to</c:suffix>

<c:offset>151</c:offset>

<c:length>14</c:length>

</rdf:Description>

JSON output format:

{

"detection": "[\n\t<Body> \n(Reuters) - Technology giant ]Apple (AAPL.O)[ is looking beyond mobile devices to learn how to]",

"prefix": "\n\t<Body> \n(Reuters) - Technology giant ",

"exact": "Apple (AAPL.O)",

"suffix": " is looking beyond mobile devices to learn how to",

"offset": 151,

"length": 14

},

Example 2: A mention of a company location (CompanyLocation type)

Input Text

Technology giant Apple (AAPL.O) is looking beyond mobile devices to learn how to make a self-driving electric car, and is talking to experts at carmakers and automotive suppliers, a senior auto industry source familiar with the discussions said on Saturday.

The Cupertino, California-based maker of phones, computers and, soon, watches is exploring how to make an entire vehicle, not just designing automotive software or individual components, the auto industry source said.

Output

InstanceInfo tag in RDF output format:

<rdf:Description rdf:about="http://d.opencalais.com/dochash-1/7586b818-40af-3d55-ac16-bf3520cddfa6/Instance/33">

<rdf:type rdf:resource="http://s.opencalais.com/1/type/sys/InstanceInfo"/>

<c:docId rdf:resource="http://d.opencalais.com/dochash-1/7586b818-40af-3d55-ac16-bf3520cddfa6"/>

<c:subject rdf:resource="http://d.opencalais.com/genericHasher-1/f64cef62-b4a7-39f1-b92a-ef8834053ff1"/>



<c:detection>[with the discussions said on Saturday.]The Cupertino, California-based maker[ of phones, computers and, soon, watches is]</c:detection>

<c:prefix>with the discussions said on Saturday. </c:prefix>

<c:exact>The Cupertino, California-based maker</c:exact>

<c:suffix> of phones, computers and, soon, watches is</c:suffix>

<c:offset>395</c:offset>

<c:length>37</c:length>

</rdf:Description>

JSON output format:

"instances": [{

"detection": "[with the discussions said on Saturday. \n \n]The Cupertino, California-based maker[ of phones, computers and, soon, watches is]",

"prefix": "with the discussions said on Saturday. \n \n",

"exact": "The Cupertino, California-based maker",

"suffix": " of phones, computers and, soon, watches is",

"offset": 395,

"length": 37

}]

},

Entity Markup Tags - Entities

List of Supported Entities

List of Supported Relations

Anniversary

em/e/Anniversary

Definition

This tag is generated when Intelligent Tagging identifies a mention or indication of an anniversary.

Attributes

forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false).

name: The text string from the input document that represents the anniversary. When there are multiple mentions, the most complete mention is used.

Example 1

Input Text	Spain marked the first anniversary of the train bombings in Madrid that took 200 lives.
Extracted Entity	first anniversary of the train bombings in Madrid

Example 2

Input Text	The office of Serbian President Boris Tadic today said he will attend the 10^th anniversary of the massacre of Bosnian Muslims by Serbian troops.
Extracted Entity	10^th anniversary of the massacre of Bosnian Muslims by Serbian troops.

City

em/e/City
Definition	The name of a city or other municipality. Districts or neighborhoods within a city (e.g. Soho, London) are not extracted.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.
Related Tag	er/Geo/City (Disambiguation Tag)

Example1

Input Text	Finally, negotiations are expected to go through Friday night and into early Saturday morning in the Hague over a treaty to reduce global warming.
Extracted Entity	the Hague

Example2

Input Text	Three Russian soldiers died and six were wounded when a Russian military truck blew up a rebel mine near the southern village of Alkhan-Yurt.
Extracted Entity	Alkhan-Yurt

Company

This is one of the actively enhanced and supported metadata types.

em/e/Company
Definition	A full or partial company name. The term "Company" refers to any business organization, including newspapers, media companies, law firms, etc. Note: A company name in an index name is not extracted. For example, JPMorgan is not extracted from the phrase, “JPMorgan Alerian MLP Index ETN.” Bloomberg is not extracted from the phrase, “Bloomberg Consumer Comfort Index.” Barclays is not extracted from the phrase, “Barclays Global Aggregate Bond Index UCITSETF.” Etc.
Attributes	confidencelevel: A confidence score on a scale of 0 to 1. The value represents the probability that the extracted entity is indeed of the assigned type. For example: The confidence level score for an em/e/person tag indicates the probability that the extracted person is indeed a person. The confidence level score for an em/e/company tag indicates the probability that the extracted company is indeed a company. The confidence level score for an em/e/pharmaceuticalDrug tag indicates the probability that the extracted pharmaceutical drug is indeed a pharmaceutical drug. The higher the value, the higher the probability. The consuming application can use this score to achieve higher accuracy results by ignoring instances with confidence scores below a specified level. Note that boosting Precision in this manner is at the expense of Recall. Every em/e tag that displays this attribute also has a corresponding Confidence tag. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). The confidence score determines if the forenduserdisplay value is true or false. inlineric: The company Reuters Instrument Code (RIC), if it is mentioned in the text. name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used. nationality: The nationality of the extracted Person, Company, or Organization, if it is mentioned in the text. recognizedas: Indicates if the extraction was based on a company RIC, company ticker, or company name found in the text. Possible values are: ric, ticker, name. Ticker extraction is a premium feature, enabled with the x-calais-EnableTickerExtraction header. Please note that Intelligent Tagging identifies and extracts primary tickers in the text. For example, “Falabella,” the primary ticker of the company SACI Falabella, is identified as a company mention, whereas “FALABE-OSA,” a non-primary ticker, is not extracted. To extend ticker extraction to non-primary tickers, ALSO define the x-calais-EnableTickerRecallOriented header. Defining both headers triggers recall-oriented ticker extraction which is optimized for company tagging of research content such as analyst emails. (But not optimized for documents like long research reports.) All rics are identified and extracted.
Related Tags	Confidence, Relevance, er/Company (Disambiguation Tag)

Example1

Input Text	Zenith National Insurance Corp. (NYSE:ZNT) reported net income of $20.8 million for the fourth quarter of 2003.
Extracted Entity	Zenith National Insurance Corp.

Example2

Input Text	We increased our ownership in Advent Capital (Holdings) PLC ("Advent Capital") in the third quarter ... was reduced by approximately $0.9 million after tax for our share of adverse development recorded by Advent Capital in the third quarter of 2003.
Extracted Entity	Advent Capital (Holdings) LTD

Continent

em/e/Continent
Definition	A continent.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The entity name, based on a mention in the text and mapped to a value from an internal reference list.
Related Tag	er/Continent (Disambiguation Tag)

Example

Input Text	European Union countries now invest more in South America than does the United States.
Extracted Entity	South America

Country

This is one of the actively enhanced and supported metadata types.

em/e/Country
Definition	A country.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The entity name, based on a mention in the text and mapped to a value from an internal reference list.
Related Tag	er/Geo/Country (Disambiguation Tag)
Equivalent Topic Tags	You can also use Intelligent Tagging topic tags to identify documents about countries.

Example1

Input Text	He also dealt with refugees in Cyprus after the 1974 Turkish invasion.
Extracted Entity	Cyprus

Example2

Input Text

He gained widespread praise for overseeing East Timor's three-year transition to independence after Indonesia withdrew in 1999.

Extracted Entity

East Timor

Indonesia

Currency

This is one of the actively enhanced and supported metadata types.

This is a premium metadata type.

em/e/Currency
Definition	A reference to a traditional currency. Some crypto currencies are also covered.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The entity name, based on a mention in the text and mapped to a value from an internal reference list. permid: The Refinitiv unique ID (PermID) for this currency. ric: The Reuters Instrument Code (RIC) for the currency. rcscode: The code for this currency as defined in the RCS (Refinitiv Classification Services) taxonomy. confidencelevel: Indicates the probability with values between 0 and 1 that the extracted currency is indeed a currency.
Related Tag	Relevance

Example1

Input Text	The Singapore dollar surged its highest in three years.
Extracted Entity	Singapore Dollar

Example2

Input Text	The Chinese currency had been strengthening this year, fueling speculation on whether Beijing can intervene in markets to limit its rise.
Extracted Entity	Chinese Yuan Renminbi

CurrencyPair

This is one of the actively enhanced and supported metadata types.

This is a premium metadata type.

em/e/CurrencyPair
Definition	A mention of a currency pair. Please note that when a currency pair is extracted, additional em/e/currency tags are not output for the individual currencies.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The currency pair name, based on a mention in the text and mapped to a value from an internal reference list. permid: The Refinitiv unique ID (PermID) for this currency pair. ric: The Reuters Instrument Code (RIC) for the currency pair. confidencelevel: Not in use. The value will always be output as 99%.
Related Tag	Relevance

Example1

Input Text	Australian dollar rose to a two-month high against the U.S. dollar.
Extracted Entity	AUD/USD

Example2

Input Text	The yuan has been pegged against the U.S. dollar.
Extracted Entity	USD/CNY

Editor

em/e/Editor

Definition

The editor/s of the article, as mentioned in the Editor section (usually the last section of the document).

Note: Intelligent Tagging will not generate an em/e/person tag for a person mentioned only in the Editor section.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text

(Editing by Keith Weir and Sara Ledwith)

Extracted Entity

Keith Weir

Sara Ledwith

EmailAddress

em/e/EmailAddress

Definition

An email address.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text	For information contact: info@softwareshelf.co.uk.
Extracted Entity	info@softwareshelf.co.uk

EntertainmentAwardEvent

em/e/EntertainmentAwardEvent

Definition

An entertainment-related event or award, such as a film or music festival, or an award related to the entertainment industry.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example 1

Input Text	This week at the Venice Film Festival, a woman pushed past guards and tried to hug Pitt before being pulled away.
Extracted Entity	Venice Film Festival

Example 2

Input Text	Ledger, 28, and Williams, 27 received Oscar nominations for their performances in “Brokeback Mountain” directed by Ang Lee.
Extracted Entity	Oscar

Facility

em/e/Facility

Definition

The name of a man-made physical entity.

Things like courts, embassies, consulates, radio stations, and TV stations are assigned a Facility tag only if the text string contains one or more explicit location or building indicator, such as "the Parliament building," "The crowd outside the Stanislaus County Courthouse," etc.
Places like parks, borders, settlements, farms, parking lots, etc., are not assigned a Facility tag.
Theme parks and amusement parks are assigned a Facility tag.
Universities and hospitals are assigned a Facility tag.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example 1

Input Text

A walking group for seniors, sponsored by Suburban Hospital, meets at the Montgomery Mall in Bethesda.

Extracted Entity

Montgomery Mall

Suburban Hospital

Example 2

Input Text	Today's meeting took place at the Pratica di Mare military base, Europe's second largest, located 30 kilometers south of Rome.
Extracted Entity	Pratica di Mare military base

Example 3

Input Text	The launch took place at the Welsh High School hall and was attended by children whom Sister Aidan had delivered.
Extracted Entity	Welsh High School hall

Example 4

Input Text	Singapore said some of the suspects had received training at al-Qaida terrorist camps in Afghanistan.
Extracted Entity	al-Qaida terrorist camps

FaxNumber

em/e/FaxNumber

Definition

A fax number, including the prefix and extension if they are present in the input text.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example 1

Input Text	Fax the completed form to +49 (0) 89 613 676 33
Extracted Entity	+49 (0) 89 613 676 33

Example 2

Input Text	Complete the survey and send by fax to 334-473-6503.
Extracted Entity	334-473-6503

Holiday

em/e/Holiday

Definition

The name of a holiday, including the country or nationality if mentioned.

Attributes

name: The entity name, based on a mention in the text and mapped to a value from an internal reference list.

Example 1

Input Text	Israel had closed off the West Bank and Gaza Strip on Monday at the beginning of the four-day Muslim Feast of the Sacrifice holiday.
Extracted Entity	Eid Al-Adha (The phrase “Feast of the Sacrifice” was mapped to the official name of the holiday, Eid Al-Adha, as defined in the internal dataset.)

Example 2

Input Text	Parades will take place during the Latvian Foundation Day.
Extracted Entity	Latvia’s Foundation Day

IndustryTerm

em/e/IndustryTerm

Definition

A text string related to a specific industry, segment, product family, or business.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example 1

Input Text	Avionics has expanded its capability to fulfill its customers’ product needs, and now provides a complete line of avionics products.
Extracted Entity	avionics products

Example 2

Input Text

IBM Life Sciences brings together IBM resources, including research, services and e-business expertise, and high-performance computing to offer new solutions for the life sciences market, including e-health and pharmaceutical industries.

Extracted Entity

computing

e-business

e-health and pharmaceutical industries

Journalist

em/e/Journalist

Definition

The author of the article, as mentioned in the Journalist section (usually the first section of the document, indicated by the word “by”).

Note: Intelligent Tagging will not generate an em/e/person tag for a person mentioned only in the Journalist section.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text

By Nellie Andreeva and Kimberly Nordyke

Extracted Entity

Nellie Andreeva

Kimberly Nordyke

MarketIndex

This is one of the actively enhanced and supported metadata types.

em/e/MarketIndex
Definition	A stock market index.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The entity name, based on a mention in the text and mapped to a value from an internal reference list. permid:The Refinitiv unique ID (PermID) for this metadata tag type. Can be used when building a knowledge graph. ric: The Market Index Reuters Instrument Code (RIC), if it is mentioned in the text. For example: .AEX confidencelevel: Indicates the probability with values between 0 and 1 that the extracted market index is indeed a market index. The higher the value, the higher the probability. The consuming application can use this score to achieve higher accuracy results by ignoring instances with confidence scores below a specified level. Note that boosting Precision in this manner is at the expense of Recall.
Related Tag	RelevanceInfo Tag

Example 1

Input Text	The Nasdaq composite advanced 3.3 percent to 7,220.54 with Apple and Amazon both rising.
Extracted Entity	Nasdaq composite index, .IXIC

Example 2

Input Text	The Dow also had its biggest one-day percentage gain since August 2015
Extracted Entity	Dow Jones industrial average, .DJI

MedicalCondition

em/e/MedicalCondition

Definition

A medical condition such as a disease, disorder, or syndrome.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example 1

Input Text	After an operation for pancreatic cancer last year, Pavarotti had hoped to finish a world tour but he died on Thursday at the age of 71.
Extracted Entity	pancreatic cancer

Example2

Input Text

First they tried nudging. Now companies are penalizing workers who have high health risks such as obesity and high blood pressure as insurance costs climb.

Extracted Entity

obesity

high blood pressure

MedicalTreatment

em/e/MedicalTreatment

Definition

A treatment, procedure, or therapy.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text	Merck plans to conduct a Phase 3 trial of oral deforolimus in patients with metastatic soft-tissue and bone sarcomas following a favorable response to chemotherapy.
Extracted Entity	chemotherapy

Movie

em/e/Movie
Definition	A movie or film.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.
Equivalent Topic Tag	M:H1 Film You can also use this Intelligent Tagging classification topic tag to identify documents about Movies.

Example

Input Text	Tyler portrays Detective Wallis in the movie “Death Sentence,” which also stars Kevin Bacon as Nick Hume, a vigilante out for revenge.
Extracted Entity	Death Sentence

MusicAlbum

em/e/MusicAlbum
Definition	A recorded music release, such as a music album, CD, etc.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.
Equivalent Topic Tag	M:H0 Music You can also use this Intelligent Tagging classification topic tag to identify documents about Music Albums.

Example

Input Text	U.S. rapper 50 Cent released his new album, “Curtis,” on MySpace Friday – days before the hotly anticipated disc is set to drop in stores.
Extracted Entity	Curtis

MusicGroup

em/e/MusicGroup

Definition

A music group or band.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text	On November 24, 1991, the world lost one of the greatest musical voices of all times with the passing of Freddie Mercury, the lead singer for the group Queen.
Extracted Entity	Queen

NaturalFeature

em/e/NaturalFeature

Definition

A natural feature such as a river, sea, lake, or mountain, or the name of a geographical region.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text	The Congolese refugees in Tanzania said they were forced to flee southwards, along the shores of Lake Tanganyika.
Extracted Entity	Lake Tanganyika

OperatingSystem

em/e/OperatingSystem

Definition

An operating system.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text	Jay reported yesterday that the Google Maps for mobile for Palm OS update would be available today.
Extracted Entity	Palm OS

Organization

em/e/Organization
Definition	The name of an organization (governmental, military, or other organization). Note: A business organization that generates an em/e/Company tag (that is extracted as a company) will not generate an em/e/Organization tag.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The organization name. nationality: The nationality of the extracted Person, Company, or Organization, if it is mentioned in the text. organizationtype: An organization classification. Possible values: sports, governmental military, governmental civilian, political party, N/A. permid: The Refinitiv unique ID (PermID) for this metadata tag type. Can be used when building a knowledge graph.
Related Tag	er/Organization (Disambiguation Tag)

Example 1

Input Text	Bond players will be focused on the BOJ's September Tankan corporate sentiment survey due out next Thursday.
Extracted Entity	Bank of Japan

Example 2

Input Text	Chief of the General Staff, Lt. Gen. Gabi Ashkenazi, met earlier this evening with the Commander of the Kenyan Military, General Jeremiah Mutinda.
Extracted Entity	Kenyan Military

Person

This is one of the actively enhanced and supported metadata types.

em/e/Person
Definition	The name of a person. If the person is mentioned more than once in the document, the most complete mention is extracted.
Attributes	commonname: The commonly used person name. confidencelevel: A confidence score on a scale of 0 to 1. The value represents the probability that the extracted entity is indeed of the assigned type. For example: The confidence level score for an em/e/person tag indicates the probability that the extracted person is indeed a person. The confidence level score for an em/e/company tag indicates the probability that the extracted company is indeed a company. The confidence level score for an em/e/pharmaceuticalDrug tag indicates the probability that the extracted pharmaceutical drug is indeed a pharmaceutical drug. The higher the value, the higher the probability. The consuming application can use this score to achieve higher accuracy results by ignoring instances with confidence scores below a specified level. Note that boosting Precision in this manner is at the expense of Recall. Every em/e tag that displays this attribute also has a corresponding Confidence tag. firstname: Most complete mention of the first name. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). lastname: Most complete mention of the last name. middlename: Most complete mention of the middle name. name: The name of the person as it appears in the text. If the person is mentioned more than once in the text, then the most complete mention is used. nationality: Not currently in use. The attribute value is always “N/A.” permid: The Refinitiv unique ID (PermID) for this metadata tag type. Can be used when building a knowledge graph. persontype: Not currently in use. The attribute value is always “N/A.”
Related Tags	Confidence, Relevance, er/Person (Disambiguation Tag)

Example 1

Input Text

Prof. George Brown Tindall is survived by his wife Blossom Tindall, son, Bruce Tindall of San Diego; daughter, Blair Tindall of Santa Monica, Calif., and one grandson.

Extracted Entity

Blair Tindall

Blossom Tindall

Bruce Tindall

George Brown Tindall

Example 2

Input Text	The Palestinian Prime Minister Mahmoud Abbas, also known as Abu Mazen, visited Gaza yesterday. Abu Mazen told the reporters…
Extracted Entity	Mahmoud Abbas (Abu Mazen)

PharmaceuticalDrug

This is a premium metadata type.

em/e/PharmaceuticalDrug
Definition	A pharmaceutical drug.
Attributes	confidencelevel: A confidence score on a scale of 0 to 1. The value represents the probability that the extracted entity is indeed of the assigned type. For example: The confidence level score for an em/e/person tag indicates the probability that the extracted person is indeed a person. The confidence level score for an em/e/company tag indicates the probability that the extracted company is indeed a company. The confidence level score for an em/e/pharmaceuticalDrug tag indicates the probability that the extracted pharmaceutical drug is indeed a pharmaceutical drug. The higher the value, the higher the probability. The consuming application can use this score to achieve higher accuracy results by ignoring instances with confidence scores below a specified level. Note that boosting Precision in this manner is at the expense of Recall. Every em/e tag that displays this attribute also has a corresponding Confidence tag. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.
Related Tags	Confidence, Relevance

Example 1

Input Text	Novartis AG's acquisition of a majority stake in Speedel Holding AG is a positive strategic move that will bring control over the blood pressure drug Tekturna and enhance its development pipeline, according to analysts.
Extracted Entity	Tekturna

PhoneNumber

em/e/PhoneNumber

Definition

A phone number, complete with the prefix and extension if they are present in the input text.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text	For further information call (203) 661-1926, ext. 6668.
Extracted Entity	(203) 661-1926, ext. 6668

PoliticalEvent

em/e/PoliticalEvent

Definition

A political event such as an election, primary, vote, rally, demonstration, debate, speech, etc.

Attributes

date: An absolute date (in the YYYY-MM-DD format). This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date.

datestring: A mention of a date or a reference to a date (e.g. yesterday, next month), taken directly from the text.

location: Location of the event if it is mentioned in the text.

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

politicaleventtype: A political event classification. For example, elections, primaries.

Example

Input Text	Conservative billionaire Silvio Berlusconi won Italy's general elections.
Extracted Entity	general elections

Position

em/e/Position

Definition

A position that a person held, holds, or will hold.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example 1

Input Text	Dave Rivello, special agent with Drug Enforcement Agency in Salisbury, said the drugs first were distributed mostly in the Salisbury area.
Extracted Entity	special agent

Example 2

Input Text	Former Paterson Mayor Martin G. Barnes pleaded guilty in July.
Extracted Entity	Mayor

Example 3

Input Text	A judge investigating suspicions that Total paid bribes to access markets has opened legal proceedings against one of the oil giant’s Middle East directors.
Extracted Entity	judge

Product

em/e/Product
Definition	A product.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used. producttype: A product classification. Possible values: aircraft, car, drug, electronics, weapon, other.
Related Tag	er/Product (disambiguation tag)
Equivalent Topic Tag	E:40 Products You can also use this Intelligent Tagging classification topic tag to identify documents about Products.

Example

Input Text	Dealers cannot keep fuel-efficient cars like the Ford Focus in stock, and automakers cannot roll them off the assembly line fast enough to keep up with demand.
Extracted Entity	Ford Focus

ProgrammingLanguage

em/e/ProgrammingLanguage

Definition

A programming language.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text	Resolver One uses the Python programming language, which was first released in 1991.
Extracted Entity	Python

ProvinceOrState

em/e/ProvinceOrState
Definition	A province, state, county, or other jurisdiction that is part of a country.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.
Related Tag	er/Geo/ProvinceOrState (Disambiguation Tag)

Example

Input Text

The private placement will be units of flow-through common shares and common share purchase warrants to investors in Alberta and British Columbia.

Extracted Entity

Alberta, Canada

British Columbia, Canada

PublishedMedium

em/e/PublishedMedium

Definition

A publication such as a newspaper, journal, or magazine.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example 1

Input Text	The company also declined to deny or confirm a report in The Boston Globe last month, which quoted a letter to State Street clients alerting them to a 42 percent decline this year in the State Street Limited Duration Bond Fund for institutional investors.
Extracted Entity	The Boston Globe

Example 2

Input Text	The Oakland Tribune reported Monday night the deal will pay Russell at least $31 million guaranteed.
Extracted Entity	The Oakland Tribune

Example 3

Input Text	The 400 richest Americans, including U.S. citizens now living abroad, as compiled by Forbes magazine.
Extracted Entity	Forbes

RadioProgram

em/e/RadioProgram

Definition

A radio program.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text	When Americans want to be a part of the national conversation, they turn to Talk of the Nation, NPR’s mid-day news-talk show.
Extracted Entity	Talk of the Nation

RadioStation

em/e/RadioStation

Definition

A radio station. This entity type primarily identifies U.S. radio stations and may identify non-U.S. radio stations as well.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text	WBUR is a public radio station broadcasting from the campus of Boston University in Boston, Massachusetts.
Extracted Entity	WBUR

Region

em/e/Region
Definition		A non-politically defined and not natural geographical region of the world (excluding continents). For example, the Middle East.
Attributes		forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.
Related Tag		er/Region (disambiguation tag)

Example

Input Text	The Company facilitates the movement of freight to and from anywhere in the world, with particular focus on business between the Far East and the United States.
Extracted Entity	Far East

SportsEvent

em/e/SportsEvent

Definition

A sports event.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example 1

Input Text	The San Antonio Silver Stars may have fallen short of the ultimate goal of playing for the WNBA championship.
Extracted Entity	WNBA championship

Example 2

Input Text	The World Anti-Doping Agency yesterday welcomed the holding of an international summit on doping in cycling next month after this year's Tour de France was rocked by scandals.
Extracted Entity	Tour de Frances

SportsGame

em/e/SportsGame

Definition

A reference to a sport such as Sumo, Swimming, Yachting, Football, etc.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text	The Packers' Tracy White came out of the pile with the football for a touchdown.
Extracted Entity	Football

SportsLeague

em/e/SportsLeague

Definition

The name of a sports league. Some sports leagues may also generate an em/e/organization tag.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text	The team is being considered as one of six cities to host an expansion International Basketball League team for play in 2006.
Extracted Entity	International Basketball League

Technology

em/e/Technology

Definition

Name or description of a type of technology.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example 1

Input Text	Vairex was founded in 1987 to develop advanced proprietary oil-free compressor technology.
Extracted Entity	oil-free compressor technology

Example 2

Input Text

The software's ability to stream over multiple formats has also been improved to now include MPEG-1, MPEG-2, WindowsMedia, RealVideo and QuickTime with MPEG-4 to be added soon.

Extracted Entity

MPEG-1

MPEG-2

MPEG-4

TVShow

em/e/TVShow

Definition

A TV show.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text	The actress, who stars in the new ABC series Pushing Daisies, fractured a few ribs, according to E!Online.com.
Extracted Entity	Pushing Daisies

TVStation

em/e/TVStation

Definition

A TV station. This entity type primarily identifies U.S. TV stations and may identify non-U.S. TV stations as well.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example

Input Text

The Minnesota News Council narrowly upheld two complaints against KBJR-TV of Duluth and upheld one complaint against KSTP-TV of the Twin Cities on Thursday.

Extracted Entity

KBJR-TV

KSTP-TV

URL

em/e/URL

Definition

A URL or FTP address.

Attributes

name: The text string from the input document that represents the entity. When there are multiple mentions that refer to the same entity, the most complete mention is used.

Example 1

Input Text	http://www.macromedia.com/go/buymx2004
Extracted Entity	http://www.macromedia.com/go/buymx2004

Example 2

Input Text	http://reports.huginonline.com/915745/122470.pdf
Extracted Entity	http://reports.huginonline.com/915745/122470.pdf

Example 3

nput Text	ftp://ftp.demon.co.uk/pub/mirrors/garbo
Extracted Entity	ftp://ftp.demon.co.uk/pub/mirrors/garbo

Entity Markup Tags - Relations

List of Supported Relations

List of Supported Entities

Acquisition

This is a premium metadata type.

em/r/Acquisition
Definition	A reference to an acquisition (past, present, or future) of shares in a company or of assets of another company.
Attributes	company_acquirer: Company that is acquiring the shares or assets. company_beingacquired: The company whose shares or assets are being acquired. confidencelevel: A confidence score on a scale of 0 to 1. The value represents the probability that the extracted relation is indeed of the assigned type. For example: The confidence level score for an em/r/deal tag indicates the probability that the deal is indeed a deal between the companies specified by the Acquirer and Target attributes. The confidence level score for an em/r/bankruptcy tag indicates the probability that text actually refers to a bankruptcy involving the company specified by the Company attribute. The confidence level score for an em/r/IPO tag indicates the probability that the text actually refers to an IPO involving the company specified by the Company attribute. The higher the value, the higher the probability. The consuming application can use this score to achieve higher accuracy results by ignoring instances with confidence scores below a specified level. Note that boosting Precision in this manner is at the expense of Recall. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text.
Equivalent Topic Tag	E:J Mergers/Acquisitions/Takeovers You can also use this Intelligent Tagging classification topic tag to identify documents about Acquisitions.

Example

Input Text

Ericsson Business Innovation, the innovation development company recently set up by Ericsson (NASDAQ:ERICY), is buying a 29 percent stake in Mediatude, a company specializing in targeted mobile marketing services over voice, SMS and WAP.

Extracted Relation

company_acquirer = Ericsson Business Innovation

company_beingacquired = Mediatude Ltd

status = announced

Alliance

em/r/Alliance
Definition	A reference to an alliance agreement signed recently or in the past between two or more companies.
Attributes	company: Company that is a member of the alliance. Note that the company attribute appears two or more times in the em/r/alliance tag. Each company attribute indicates a different company that is a member of the alliance. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text.
Equivalent Topic Tag	E:4G Strategic Combinations You can also use this Intelligent Tagging classification topic tag to identify documents about Alliances.

Example

Input Text

Digital Marketing Services, Inc. (DMS), the leading provider of online marketing research and a division of America Online Inc. (AOL), today announced an alliance with Netcentives Inc. (Nasdaq: NCNT)

Extracted Relation

company = Digital Marketing Services, Inc.

company = Netcentives Inc.

status = announced

datestring = today

Analyst EarningsEstimate

em/r/AnalystEarningsEstimate
Definition	An estimation by external analysts regarding the future (i.e., not-yet-reported) financial results of a publicly-traded company.
Attributes	company_rated: Company that is the subject of the estimate, recommendation, or rating. company_source: The financial or other body publishing the estimate, rating, or recommendation. financialmetric: The reported metric, based on a mention in the text and mapped to one of the following values: Revenues Financial Results EARNINGS EBIT EBITDA FFO Loss All but Revenues and Financial Results may be suffixed by: (BASIC), (BASIC_AND_DILUTED), or (DILUTED). forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). person_source: The person who issued the estimate, rating, or recommendation. quarter: Time period to which the extracted relation refers. year: The relevant year.
Equivalent Topic Tags	M:1XC Estimate Changes; M:1XD Estimate Increases; M:1XE Estimate Decreases You can also use these Intelligent Tagging classification topic tags to identify documents about Analyst Earnings Estimates.

Example 1

Input Text

Shares of Oplink Communications Inc., a provider of optical manufacturing, design and integration services, rocketed in afternoon trading on Friday after the company … Analysts polled by Thomson Financial expect earnings of 60 cents per share on $75.7 million in sales for 2007.

Extracted Relation

company_rated = Oplink Communications, LLC

financialmetric = earnings per share (diluted)

quarter = FY

year = 2007

Extracted Relation

company_rated = Oplink Communications, LLC

financialmetric = Revenues

quarter = FY

year = 2007

Example 2

Input Text

After casual dining chain California Pizza Kitchen Inc. boosted its second-quarter profit outlook on better-than-expected sales, ThinkEquity analyst Nicole M. Miller said the company is positioned to continue growing long-term with new menu items. Miller increased her second-quarter estimate a penny to 30 cents per share and reiterated her "Buy" rating on the stock.

Extracted Relation

company_source = ThinkEquity Capital Advisors, LLC

person_source = Nicole M. Miller

company_rated = California Pizza Kitchen, Inc.

financialmetric = Earnings Per Share (Diluted)

quarter = Q2

AnalystRecommendation

em/r/AnalystRecommendation
Definition	A recommendation by an analyst (of an investment bank or similar organization) regarding shares of a publicly traded company. This tag type extracts recommendations regarding shares (equity) only. A recommendation regarding bonds or other credit instruments is not extracted by this tag type.
Attributes	company_rated: Company that is the subject of the estimate, recommendation, or rating. company_source: The financial or other body publishing the estimate, rating, or recommendation. financialtrend: The trend of the current recommendation with respect to the previous recommendation. Possible values: Downgraded Initiated (new recommendation as far as we know.) Reiterated (same recommendation given again) Upgraded forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). person_source: The person who issued the estimate, rating, or recommendation. price_new: Target price quoted by the analyst. price_old: Previous target price, quoted by the same analyst. rank_new: New or current rank given to the company stock by the analyst. (For example, “Strong Buy,” “Hold,” etc.) rank_old: Previous company rank (as published by the same analyst).
Equivalent Topic Tags	M:1X6 Rating Changes; M:1X7 Rating Upgrades; M:1X8 Rating Downgrades You can also use these Intelligent Tagging classification topic tags to identify documents about Analyst Recommendations.

Example 1

Input Text

RBC analyst Jordan Rohan upgraded aQuantive to "Outperform" from "Sector Perform" and raised the price target to $27 per share, from $25.

Extracted Relation

company_source = Banque Royale Du Canada

person_source = Jordan Rohan

company_rated = aQuantive, Inc.

financialtrend = upgraded

rank_new = Outperform

rank_old = Sector Perform

price_new = $27

price_old = $25

Example2

Input Text

Prudential analyst John McPeake initiated coverage of three software companies Tuesday, …
McPeake also initiated coverage of Salesforce.com, which provides customer relationship management software and services via Web browsers.

Extracted Relation

company_source = Prudential Public Limited Company

person_source = John McPeake

company_rated = Salesforce.com, Inc.

financialtrend = initiated

ArmedAttack

em/r/ArmedAttack
Definition	An armed attack by a country or an organization, or an armed attack on a leader or political figure.
Attributes	attacker: The name of the attacker. (Could be the name of a group, a party, a person, etc.). attacktype: Type of attack. (For example: war, incursion, skirmish, attach, explosion, armed patrol, coup, etc.) casualties: Number and degree of casualties in the attack. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). locationstring: Location of the armed attack, as mentioned in the text. partyinvolved1: Country, person, or group involved in the attack. partyinvolved2: Country, person, or group involved in the attack. If it is not clear which side is the attacker and which side is the attacked, they will be extracted as partyinvolved1 and partyinvolved2 in random order. Otherwise, partyinvolved1 is the attacker, partyinvolved2 is the attacked, and the attacker attribute will have the same value as the partyinvolved1 attribute. status: The status, based on an indication in the text. weaponused: Weapons used to attack.
Equivalent Topic Tags	M:EL Conflicts/War/Peace; M:F7 Bombing (Non-Military); M:9 Civil Unrest; M:NX Insurgencies; M:LK International/National Security; M:2 Military Coflicts. You can also use these Intelligent Tagging classification topic tags to identify documents about Armed Attacks.

Example

Input Text

About an hour later, the family heard Russian warplanes overhead and the missiles struck. Raghad, her grandfather and another relative were killed.

Extracted Relation

attacktype= attack

weaponused=the missiles

status=confirmed

ArmsPurchaseSale

em/r/ArmsPurchaseSale
Definition	A reference to an arms deal.
Attributes	armsdescription: Types, amounts, and models of weaponry and vehicles included in the deal. armsseller: The party selling the arms. countryarmspurchaser: Country which is purchasing the arms. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). moneyamount: The value of the deal. organizationarmspurchaser: The organization purchasing arms. personarmspurchaser: The person purchasing the arms. status: The status, based on an indication in the text.
Equivalent Topic Tag	M:N9 Military Procurement You can also use this Intelligent Tagging classification topic tag to identify documents about Arms Purchase and Sales.

Example 1

Input Text

A book by an FBI consultant on international terrorism says that Osama bin Laden's al-Qaeda terrorist network purchased 20-suitcase nuclear weapons for $30 million in 1998.

Extracted Relation

armsdescription = 20-suitcase nuclear weapons

organizationarmspurchaser = al-Qaeda

status = completed

date = 1998-00-00

Example2

Input Text

The Saudi arms deal stems from the American proposal to offer Gulf Arab states some $20 billion in weapons, including Joint Direct Attack Munition (JDAM) bomb kits for the Saudis.

Extracted Relation

armsdescription = weapons, including Joint Direct Attack Munition (JDAM) bomb kits

countryarmspurchaser = Saudi Arabia

status = proposed

Arrest

em/r/Arrest
Definition	A reference to the arrest of one or more people.
Attributes	charge: Phrase describing an offense/charge. In case of multiple charge attributes, the first charge attribute is the primary charge. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). othercharges: Additional charges. person: The person who is the subject of this extracted relation.
Equivalent Topic Tag	M:EV Judicial Process/Court Cases/Court Decisions You can also use this Intelligent Tagging classification topic tag to identify documents about Arrests.

Example 1

Input Text

Issac's brother Ramzi Issac, who was arrested along with the London suspect, was also charged with possessing false documents.

Extracted Relation

person = Ramzi Issac

charge = possessing false documents

Example2

Input Text

Also arrested Monday were his 33-year-old son Justin Zivojinovich and his wife, Michelle, 30.

Extracted Relation

person = Justin Zivojinovich

datestring = Monday

Extracted Relation

person = Michelle

datestring = Monday

Bankruptcy

em/r/Bankruptcy
Definition	Extracts references to bankruptcy issues that involve companies: Corporate insolvencies and bankruptcies, where a company has filed for bankruptcy or been declared insolvent. Voluntary and involuntary insolvency proceedings including creditor protection actions such as Chapter 11 federal bankruptcy code proceedings in the U.S. and administration orders elsewhere as well as receiverships. Company ceasing trading, bankruptcy court rulings and a company's emergence from a bankruptcy protection or administration process.
Attributes	company: The company that is the subject of the relation type. confidencelevel: A confidence score on a scale of 0 to 1. The value represents the probability that the extracted relation is indeed of the assigned type. For example: The confidence level score for an em/r/deal tag indicates the probability that the deal is indeed a deal between the companies specified by the Acquirer and Target attributes. The confidence level score for an em/r/bankruptcy tag indicates the probability that text actually refers to a bankruptcy involving the company specified by the Company attribute. The confidence level score for an em/r/IPO tag indicates the probability that the text actually refers to an IPO involving the company specified by the Company attribute. The higher the value, the higher the probability. The consuming application can use this score to achieve higher accuracy results by ignoring instances with confidence scores below a specified level. Note that boosting Precision in this manner is at the expense of Recall. Note that this em/r/tag also has a corresponding Confidence tag. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text.
Related Tags	Confidence Tag
Equivalent Topic Tag	E:H Corporate Bankruptcy/Insolvency You can also use this Intelligent Tagging classification topic tag to identify documents about Bankruptcy.

Example 1

Input Text

German holding company WCM Inc. said on Wednesday it had applied for insolvency.

Extracted Relation

company = WCM Beteiligungs- und Grundbesitz-Aktiengesellschaft

status = applied

datestring = Wednesday

Example 2

Input Text

Toys R Us may file for bankruptcy before the holidays..

Extracted Relation

company = Toys “R” Us, Inc.

status = possible apply

Example 3

Input Text

Winn-Dixie ready to seek OK to exit bankruptcy.

Extracted Relation

company = Winn-Dixie Stores, Inc.

status = expected to emerge

BonusSharesIssuance

em/r/BonusSharesIssuance
Definition		A reference to bonus shares being issued (in the past, present, or future).
Attributes		bonussharesratio: The ratio of the number of shares held by the shareholders to the number of shares they will receive. company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text.
Equivalent Topic Tag	E:5B Bonus Share Issues You can also use this Intelligent Tagging classification topic tag to identify documents about Bonus Shares Issuance.

Example

Input Text

The Capital Market Authority said Saudi British Bank, an affiliate of HSBC Holdings, could issue one bonus share for every two shares held, increasing the number of the bank’s outstanding shares to 1.5 billion from 1 billion.

Extracted Relation

company = HSBC Holdings PLC

bonussharesratio = 1:2

status = rumored

BusinessRelation

em/r/BusinessRelation

Definition

A reference to a business agreement signed recently or in the past, between two or more companies.

Note that agreements phrased as "alliances" or "partnerships" are extracted as an Alliance. A "weaker" contact is extracted as a BusinessRelation.

Attributes

businessrelationtype: Defines the business relationship, as mentioned in the text. For example: licensing agreement.

company: Company that is involved in the business relation.

Note that the company attribute appears multiple times in the em/r/BusinessRelation tag. Each company attribute indicates a different company that is involved in the business activity.

date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date.

datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text.

status: The status, based on an indication in the text.

Example 1

Input Text

On October 6, 2016, Natera, Inc. ("the Company") entered into an amendment (the "Amendment") to its lease agreement with BMR-201 Industrial Road LP (the "Landlord") dated October 26, 2015 (the "Lease").

Extracted Relation

company = Natera, Inc.

company = BMR Group PLC

businessrelationtype = lease agreement

status = known

Example 2

Input Text

Johnson & Johnson today announced that Ethicon, a medical device company in the Johnson & Johnson family of companies, has executed a definitive agreement to enter into a strategic collaboration with Google, Inc., working with the Life Sciences team on advancing surgical robotics to benefit surgeons, patients and health care systems.

Extracted Relation

company = Johnson & Johnson

company = Google Inc.

businessrelationtype = strategic collaboration

status = known

Example 3

Input Text

Depomed Inc <DEPO.O> on Monday reached a deal with activist investor Starboard Value LP to avoid a shareholder vote due next month to oust the drugmaker's entire board.

Extracted Relation

company = Depomed, Inc.

company = Starboard Value LP

businessrelationtype = deal

datestring = Monday

status = announced

Buybacks

em/r/Buybacks
Definition	Extracts references to a company's plan to repurchase its own securities in the open market (buyback events).
Attributes	company: The company that is the subject of the relation type. date: Date related to the buyback, or the announcement date (in the YYYY-MM-DD format). This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text.
Equivalent Topic Tag	E:4P Share Buybacks/Repurchases You can also use this Intelligent Tagging classification topic tag to identify documents about Buyback.

Example 1

Input Text

EarthLink, Inc. (Nasdaq: ELNK), one of the nation’s leading internet service providers (ISPs), today announced that its Board of Directors has expanded its share repurchase program, authorizing the purchase of an additional $100 million of its outstanding shares of common stock.

Extracted Relation

company = EarthLink, Inc.

datestring = today

status = announced

Example 2

Input Text

The program will take effect upon completion of the Commercial Capital Bancorp’s stock repurchase program which was authorized in May 2004.

Extracted Relation

company = Commercial Capital Bancorp, Inc.

datestring = May 2004

date = 2004-05-00

status = planned

CandidatePosition

em/r/CandidatePosition

Definition

Extracts information about political election candidates, including mayoral candidates, for current or past elections and the office (position) in question. The position can be one that the candidate aspires to, or one currently or previously held.

Attributes

candidate: Name of the candidate.

positionaspirational: The position to which the person aspires.

positioncurrent: Candidate’s current position.

positionpast: A candidate’s past position.

Example

Input Text

Center-left Chilean presidential candidate Ricardo Lagos has dropped out of the nation's 2017 presidential campaign, he said in televised remarks on Monday, after the nation's powerful Socialist Party endorsed a different candidate on Sunday.

Extracted Relation

candidate = Ricardo Lagos

positionaspirational = President

CompanyAccountingChange

em/r/CompanyAccountingChange
Definition	Extracts references to a company's change in its accounting method or policy. For example, a changed fiscal year, a change in the depreciation policy, or new accounting standards.
Attributes	accountingchangetype: The changed accounting method or policy. Possible values: FY (fiscal year) Depreciation policy Accounting standard. company: The company that is the subject of the relation type. date: Date of the announcement of the change, (in the YYYY-MM-DD format). This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text.
Equivalent Topic Tag	E:4E Accounting Issues You can also use this Intelligent Tagging classification topic tag to identify documents about Company Accounting Changes.

Example 1

Input Text

On June 2, 2004, Elizabeth Arden, Inc.'s Board of Directors approved a fiscal year-end change from January 31 to June 30.

Extracted Relation

company = Elizabeth Arden, Inc.

accountingchangetype = FY

date = 2004-06-02

datestring = June 2, 2004

status = announced

Example 2

Input Text

Lockheed Martin Corporation (NYSE: LMT) today announced that, effective January 1, 1999, it will adopt a new accounting standard pertaining to costs related to start-up activities.

Extracted Relation

company = Lockheed Martin Corporation

accountingchangetype = accounting standard

date = 1999-01-01

datestring = today

status = announced

CompanyAffiliates

em/r/CompanyAffiliates
Definition	Extracts references to a parent company or subsidiary relationship between two companies.
Attributes	affiliaterelationtype: Nature of the relationship between CompanyAffiliate and Company_Parent. For example: subsidiary, division, spinoff, child, parent. company_affiliate: Company that is a subsidiary or division of another company. company_parent: Company that is a parent of another company. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false).
Equivalent Topic Tag	E:4G Strategic Combinations You can also use this Intelligent Tagging classification topic tag to identify documents about Company Affiliates.

Example 1

Input Text

Bharti Airtel's wholly-owned arm, Bharti Airtel International (Netherlands) BV, has priced the Euro 750 million 4% guaranteed senior notes due in 2018, which will be fully and unconditionally guaranteed by the parent company.

Extracted Relation

company_parent = Bharti Airtel Limited

company_affiliate = Bharti Airtel International (Netherlands) B.V.

affiliaterelationtype = wholly-owned arm

Example 2

Input Text

SK Global Chemical International Trading Co., an affiliate of SK Innovation Co., the group's refining unit, was slapped with the fine by China's foreign currency authorities, according to the sources.

Extracted Relation

company_parent = SK Innovation Co., Ltd.

company_affiliate = SK Global Chemical International Trading Co.

affiliaterelationtype = affiliate

Example 3

Input Text

Altria announced that it would spin off its international division, Philip Morris International.

Extracted Relation

company_parent = Altria Group, Inc.

company_affiliate = Philip Morris International

affiliaterelationtype = spinoff

Example 4

Input Text

Kalamazoo, Michigan - December 31, 2013 - Stryker Corporation (NYSE:SYK) announced today a definitive agreement to acquire Patient Safety Technologies, Inc. (PSTX) for $2.22 per share, with an aggregate purchase price of $120 million. Patient Safety Technologies conducts its business through its wholly owned subsidiary, SurgiCount Medical, Inc.

Extracted Relation

company_parent = Patient Safety Technologies, Inc.

company_affiliate = Surgicount Medical, Inc.

affiliaterelationtype = subsidiary

CompanyCompetitor

em/r/CompanyCompetitor
Definition	A competitor of the company.
Attributes	company: The company that is the subject of the relation type. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false).
Equivalent Topic Tag	M:GL Competition You can also use this Intelligent Tagging classification topic tag to identify documents about Company Competitors.

Example

Input Text	Shares of EDS's closest competitors - Infosys Technologies (nasdaq:INFY - news - people), Computer Sciences Corporation (nyse: CSC - news - people) and Accenture (nyse: ACN - news - people) - were all down Tuesday as investors did the math.
Extracted Relation	company = EDS Informatica, S.A. company = Infosys Limited
Extracted Relation	company = EDS Informatica, S.A. company = Computer Sciences Corporation
Extracted Relation	company = EDS Informatica, S.A. company = Accenture Public Limited Company

CompanyCustomer

em/r/CompanyCustomer

Definition

A customer of the company.

Attributes

company_customer: Company that purchases/purchased products or services from the company_provider.

company_provider: Company that sells the product or service.

organization_customer: Organization that purchases/purchased products or services from the company_provider.

Example

Input Text	Curtis's customers include mass merchandisers such as Bi-Way Stores, Wal-Mart and Family Dollar.
Extracted Relation	company_provider = Curtis company_customer = Bi-Way Stores LTD (Dylex Ltd)
Extracted Relation	company_provider = Curtis company_customer = Walmart Stores, Inc.
Extracted Relation	company_provider = Curtis company_customer = Family Dollar Stores, Inc.

CompanyEarningsAnnouncement

em/r/CompanyEarningsAnnouncement
Definition	Extracts references to an official announcement, made by the publicly traded company itself, regarding its financial earnings results.
Attributes	company: The company that is the subject of the relation type. financialmetric: The reported metric, based on a mention in the text and mapped to one of the following values: Revenues Financial Results EARNINGS EBIT EBITDA FFO Loss All but Revenues and Financial Results may be suffixed by: (BASIC), (BASIC_AND_DILUTED), or (DILUTED). forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). permid:The Refinitiv unique ID (PermID) for this metadata tag type. Can be used when building a knowledge graph. quarter: Time period to which the extracted relation refers. year: The relevant year.
Equivalent Topic Tag	E:2P Earnings Releases You can also use this Intelligent Tagging classification topic tag to identify documents about Company Earnings Announcements.

Example 1

Input Text

Announcing the publication itself of the financial results: Stein Mart, Inc. (Nasdaq: SMRT - News) today announced financial results for its first quarter.

Extracted Relation

company = Stein Mart, Inc.

quarter = Q1

financialmetric = Financial Results

Example 2

Input Text	The Hershey Company (NYSE: HSY) today announced sales and earnings for the fourth quarter and year ended December 31, 2008. Consolidated net sales were $1,377,380,000 compared with $1,342,222,000 for the fourth quarter of 2007.
Extracted Relation	company = The Hershey Company quarter = FY year = 2008 financialmetric = Earnings
Extracted Relation	company = The Hershey Company quarter = Q4 year = 2008 financialmetric = Earnings
Extracted Relation	company = The Hershey Company quarter = Q4 year = 2008 financialmetric = Revenues

CompanyEarningsGuidance

em/r/CompanyEarningsGuidance
Definition	Extracts references to projected earnings, made by the publicly traded company itself.
Attributes	company: The company that is the subject of the relation type. financialmetric: The reported metric, based on a mention in the text and mapped to one of the following values: Revenues Financial Results EARNINGS EBIT EBITDA FFO Loss All but Revenues and Financial Results may be suffixed by: (BASIC), (BASIC_AND_DILUTED), or (DILUTED). financialtrend: Indicates whether projected earnings are higher or lower than the previous earnings announcement. Possible values: Higher Lower forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). quarter: Time period to which the extracted relation refers. year: The relevant year.
Equivalent Topic Tag	E:Q Results Forecasts/Warnings You can also use this Intelligent Tagging classification topic tag to identify documents about Company Earnings Guidance.

Example

Input Text

The Hershey Company (NYSE: HSY) today announced sales and earnings for the fourth quarter and year ended December 31, 2008. “Hershey's strong fourth quarter results represent a solid end to the year" said David J. West, President and Chief Executive Officer. “For 2009, we expect net sales growth of 2-3 percent."

Extracted Relation

company = The Hershey Company

quarter = FY

year = 2009

financialmetric = Revenues

financialtrend = higher

CompanyEmployeesNumber

em/r/CompanyEmployeesNumber

Definition

The number of workers in a company.

Attributes

company: The company that is the subject of the relation type.

employeesnumber: The number of people employed by the company.

location: Location related to the extracted relation (address or other physical description).

unit: The name of the department or unit.

Example 1

Input Text

Wachovia has 120,000 employees nationwide.

Extracted Relation

company = Wachovia Commodities, LLC.

employeesnumber = 120,000

location = nationwide

Example 2

Input Text

The Coca-Cola Co. says it is cutting information technology jobs in a bid to become more efficient. Company spokesman Dana Bolden would not say Thursday how many jobs will be eliminated, but said there will not be a major reduction in staff. The IT department has 1,100 employees.

Extracted Relation

company = The Coca-Cola Company

employeesnumber = 1,100

unit = IT

Example 3

Input Text

Nokia is focused on wireless and wired telecommunications, with 128,445 employees in 120 countries.

Extracted Relation

company = Nokia Oyj

employees = 128,445

location = in 120 countries

CompanyExpansion

em/r/CompanyExpansion
Definition	A company entering a new market or creating a new unit.
Attributes	company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. expansiontype: “New Market” or “New Unit.” forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). location: Location related to the extracted relation (address or other physical description). status: The status, based on an indication in the text.
Equivalent Topic Tag	E:5D Expansions/New Markets/New Units You can also use this Intelligent Tagging classification topic tag to identify documents about Company Expansion.

Example 1

Input Text

Lenovo Company enters netbook market.

Extracted Relation

company = Lenovo Group Limited

expansiontype = New Market

status = known

Example 2

Input Text

FC Holdings (M) Bhd expects sales to increase following the launch of its new product, 'Olé Olé Rice Wrap'. KFC Marketing Sdn Bhd Senior General Manager, Angelina Villanueva, said the new product, priced from RM5.90, offered a unique combination of Mexican rice, salsa sauce and the signature Zinger strips wrapped in a toasted tortilla. "The new product, which is inspired by the Mexican flavours, is targeted at busy young executives as well as millennials. "It is convenient and perfect for busy people who are always on the go," she told reporters at the launch of 'Olé Olé Rice Wrap' here today. Villanueva said the latest wrap will be available from today until September or while stocks last. Meanwhile, General Manager of Operation, Chandrasagran Munusamy, said the company will add another 15 stores in the second half of this year, after opening 15 new stores in the first half. To-date, KFC has 645 stores nationwide with the addition of 15 new stores since January this year.

Extracted Relation

company = KFC Marketing Sdn Bhd

expansiontype = New Unit

status = known

Example 3

Input Text

DGSE Companies, Inc. Announces Creation of Superior Precious Metals, Inc.

Extracted Relation

company = DGSE Companies, Inc.

expansiontype = New Unit

status = announced

CompanyForceMajeure

em/r/CompanyForceMajeure

Definition

Unexpected external event that interrupts the regular activity of a company.

Attributes

company: The company that is the subject of the relation type.

datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text.

forcemajeure: The cause of the disruption of normal activity.

location: Location related to the extracted relation (address or other physical description).

Example

Input Text

A fire broke out at Finnish oil refiner Neste Oil's 50,000 barrels per day Naantali refinery in western Finland on Saturday, delaying restart of the plant by several days, the plant spokesman said.

Extracted Relation

company = Neste Oyg

location = Finland

forcemajeure = fire

date = 2008-10-11

datestring = on Saturday

CompanyFounded

em/r/CompanyFounded

Definition

The year in which the company was founded.

Attributes

company: The company that is the subject of the relation type.

year: Year in which the extracted relation took place.

Example 1

Input Text

Mr. Tarasofsky serves as the chief executive officer of CSI Diagnostics, a business management consulting firm, which he co-founded in 2010.

Extracted Relation

company = CSI Diagnostics

year = 2010

Example 2

Input Text

Hindustan Aeronautics Limited (HAL) came into existence on 1st October 1964.

Extracted Relation

company = Hindustan Aeronautics Limited

year = 1964

CompanyInvestigation

em/r/CompanyInvestigation
Definition	Reference to a company being investigated (for accounting or other issues).
Attributes	company_investigated: Company under investigation. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). investigationtype: The type of investigation being conducted. Possible values: Accounting Others organization_regulator: The regulator involved in the investigation. person_investigated: The person who is under investigation.
Equivalent Topic Tags	E:D Corporate/Market Regulation; E:45 Antitrust Regulation; E:64 Banking Capital and liquidity Requirements; M:1QX MiFID II; M:1M3 Medical Regulatory Issues You can also use these Intelligent Tagging classification topic tags to identify documents about Company Investigations.

Example

Input Text

EchoStar Communications Corp (DISH.O) *on Wednesday said the U.S. Securities and Exchange Commission has inquired into its recent record-keeping and internal controls.

Extracted Relation

company_investigated = Dish Network Corporation

investigationtype = Others

organization_regulator = U.S. Securities and Exchange Commission

CompanyInvestment

em/r/CompanyInvestment

Definition

An investment in a company.

Attributes

company: Company in which the investment was made.

company_investor: Company making the investment.

datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text.

status: The status, based on an indication in the text.

Example

Input Text

Health Web site Drkoop.com Inc. said on Tuesday it received $20 million in equity financing, a new management team and a reconfigured board.

Extracted Relation

company = Drkoop.com, Inc.

status = known

datestring = Tuesday

CompanyLaborIssues

em/r/CompanyLaborIssues
Definition	Production stoppages brought about by labor disagreements, negotiations, etc.
Attributes	company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). location: Location related to the extracted relation (address or other physical description). status: The status, based on an indication in the text.
Equivalent Topic Tags	E:C Labor/Personnel; M:FW Hiring; E:3Z Job Cuts; E:3Y Labor Disputes;M:FX Workers Pay You can also use these Intelligent Tagging classification topic tags to identify documents about Company Labor Issues.

Example 1

Input Text

Workers at Greek telecoms group OTE staged a 24-hour strike on Tuesday to protest against planned changes they say will affect job security, as well as the group's further privatization.

Extracted Relation

company = Hellenic Telecommunications Organization S.A.

datestring = on Tuesday

status = known

Example 2

Input Text

A strike at Chile's Escondida, the world's biggest copper mine, entered a third week on Monday with the union rejecting a new wage offer and the company inviting workers to negotiate individual contracts.

Extracted Relation

company = Escondida

location = Chile

datestring = on Monday

status = continued

CompanyLayoffs

em/r/CompanyLayoffs
Definition	An occurrence of a company reducing its workforce. Intelligent Tagging extracts this metadata type based on mentions in the text, of the company cutting the number of employees, jobs, positions, etc.
Attributes	company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. employeesnumber: The number of employees being laid off by the company. employeespercentage: The percentage of employees (out of the total company workforce) being laid off by the company. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text.
Equivalent Topic Tag	E:3Z Job Cuts You can also use this Intelligent Tagging classification topic tag to identify documents about Company Layoffs.

Example 1

Input Text

Tanium is the latest highflying tech startup to be discovered pushing the envelope to speed growth, in a Silicon Valley environment where aggressive tactics often are seen as essential to success. Health-benefits broker Zenefits last year said that its co-founder had written software to complete an insurance-license qualification course in less time than was permitted. He later resigned under pressure, and Zenefits has since cut its staff by two-thirds and its valuation by half. Andreessen is also among the biggest investors in Zenefits.

Extracted Relation

Company=Zenefits Insurance Services

Status=known

Employeespercentage=67%

Example 2

Input Text

During 2008, Hewlett-Packard laid off 2 percent of its staff, or about 1,770 workers.

Extracted Relation

company = HP Inc.

status = known

date = 2008-00-00

datestring = During 2008

employeespercentage = 2%

Extracted Relation

company = HP Inc.

status = known

date = 2008-00-00

datestring = During 2008

employeesnumber = 1,770

Example 3

Input Text

Door-to-door newspaper sales will be dropped, and more than 200 telemarketing jobs will be eliminated, The Los Angeles Times said.

Extracted Relation

company = Los Angeles Times

status = planned

employeesnumber = over 200

CompanyLegalIssues

em/r/CompanyLegalIssues
Definition	A reference to Corporate Litigation or a Class Action involving a company.
Attributes	company_plaintiff: Company that is a plaintiff in a legal issue. company_sued: Company that is being sued. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). lawsuitclass: Classification of a legal issue. Possible values: Class Action Lawsuit person_plaintiff: The person who is a plaintiff in a legal issue. status: The status, based on an indication in the text. sueddescription: Description of the entity being sued.
Equivalent Topic Tags	E:57 Corporate Litigation; E:7 Class Actions You can also use these Intelligent Tagging classification topic tags to identify documents about Company Legal Issues.

Example 1

Input Text

A Dutch court ruled against navigation systems company TomTom on Thursday in a patent infringement lawsuit against rival navigation device maker IBM from the United States.

Extracted Relation

company_sued = International Business Machines Corporation

company_plaintiff = TomTom N.V.

lawsuitclass = lawsuit

datestring = Thursday

Example 2

Input Text

Eastman Kodak Co., the world's top maker of photographic film, said a federal court wins summary judgment in favor of the company in a class action lawsuit brought on behalf of shareholders.

Extracted Relation

company_sued = Eastman Kodak Co.

lawsuitclass = class action

status = wins summary judgment

CompanyListingChange

em/r/CompanyListingChange
Definition	A reference to a company entering or leaving any known stock-exchange.
Attributes	changetype: Classification of the change. Possible values: Enters Leaves company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). stockexchange: The name of the relevant stock market. ticker: The company’s ticker symbol if mentioned in the text.
Equivalent Topic Tag	E:5G Security Listings/Delistings You can also use this Intelligent Tagging classification topic tag to identify documents about Company Listing Changes.

Example 1

Input Text

Camtek Ltd. chose to delist its shares from the Tel Aviv Stock Exchange in 2003.

Extracted Relation

company = Camtek Ltd

stockexchange = TASE

changetype = leaves

date = 2003-00-00

datestring = in 2003

Example 2

Input Text

Marconi's American Depository Receipts Ltd. will trade on Nasdaq under the ticker symbol MRCIY.

Extracted Relation

company = Marconi's American Depository Receipts Ltd.

stockexchange = NASDAQ

changetype = enters

ticker = MRCIY

CompanyLocation

em/r/CompanyLocation

Definition

The location of a company’s headquarters (city, state, and country, if mentioned in the text).

Attributes

city: City related to the extracted relation.

company: The company that is the subject of the relation type.

companylocationtype: Functional description of the company location. For example: center of operations, headquarters, branch office.

country: Country related to the extracted relation.

provinceorstate: The ProvinceOrState, related to the extracted relation.

Example 1

Input Text

RC Networks, headquartered in San Diego, Calif., USA, is a privately held, venture-backed company.

Extracted Relation

company = RC Networks Inc.

city = San Diego, California, United States

provinceorstate = California, United States

companylocationtype = headquarters

Example 2

Input Text

Ore.-based PixelWorks Inc.

Extracted Relation

company = PixelsWorks, Inc.

provinceorstate = Oregon, United States

companylocationtype = N/A

CompanyMeeting

em/r/CompanyMeeting
Definition	A meeting held by a company for its shareholders.
Attributes	city: City related to the extracted relation. company: The company that is the subject of the relation type. companymeetingtype: The nature of the company meeting, based on topics to be discussed. For example: Annual General Meeting (AGM), Extraordinary General Meeting (EGM), Shareholders Meeting, etc. country: Country related to the extracted relation. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. facility: The facility related to the extracted relation. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). provinceorstate: The ProvinceOrState, related to the extracted relation. status: The status, based on an indication in the text.
Equivalent Topic Tag	E:4D Shareholder Meetings You can also use this Intelligent Tagging classification topic tag to identify documents about Company Meetings.

Example 1

Input Text

Separately, ABN Amro said it would hold a shareholder meeting Sept. 20 to debate the merits of the two bids.

Extracted Relation

company = ABN Amro Bank N.V.

companymeetingtype = Shareholders’ Meeting

status = announced

datestring = Sept. 20

Example 2

Input Text

Amedia Networks, Inc. (OTC Bulletin Board: AANI - News), a provider of next generation media gateways to distribute and manage ultra-broadband triple-play services in the home, today announced that its 2007 Annual meeting of Stockholders will be held on Tuesday, August 14, 2007, at 8:30 AM at the Sheraton Eatontown Hotel.

Extracted Relation

company = Amedia Networks Inc.

companymeetingtype = Shareholders’ Meeting

status = announced

date = 2007-08-14

datestring = Tuesday, August 14, 2007

facility = Sheraton Eatontown Hotel

CompanyNameChange

em/r/CompanyNameChange
Definition	A reference to a company name change.
Attributes	company_formername: The name of the company before the name change. company_newname: The new name of the company. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text. stockexchange: The name of the relevant stock market. ticker: The company’s new or updated stock symbol if mentioned in the text.
Equivalent Topic Tag	E:4C Name Changes You can also use this Intelligent Tagging classification topic tag to identify documents about Company Name Changes

Example:

Input Text

Fitch does not rate the credit or financial strength of Cenlar FSB or its parent Cenlar Capital Corporation (CCC). Cenlar, an operating subsidiary of Cenlar Capital Corporation, is a privately held (employee-owned), federally chartered savings bank. However, Fitch's financial institutions group reviewed Cenlar's financial statements to provide an internal assessment, as a company's financial condition is a component of Fitch's servicer rating analysis.

Cenlar operates from four locations in Ewing, NJ and has been in existence since 1958 operating under the name of Larson Mortgage Company (Larson) as a traditional mortgage banking company. In 1984, Larson acquired Centennial Savings and Loan Association and changed its name to Cenlar. In early 1997, the servicer exited the retail banking and loan origination business and began performing interim servicing, third-party subservicing and primary servicing for various financial entities.

Extracted Relation

company_formername = Cenlar FSB

company_newname = Cenlar Capital Corp.

status = known

CompanyProduct

em/r/CompanyProduct
Definition	Extracts a reference to a company and its product/s.
Attributes	company: The company that is the subject of the relation type. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). product: The product that is the subject of the extracted relation. producttype: Product classification. Possible values: aircraft car drug electronics weapon other
Equivalent Topic Tag	E:40 Products You can also use this Intelligent Tagging classification topic tag to identify documents about Company Products.

Example

Input Text

Boeing is introducing the 787 Dreamliner which includes new technologies to create better environmental performance for commercial jetliners.

Extracted Relation

company = The Boeing Company

product = 787 Dreamliner

producttype = aircraft

CompanyReorganization

em/r/CompanyReorganization
Definition	A reference to a corporate reorganization or restructuring.
Attributes	company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text.
Equivalent Topic Tag	E:56 Organizational Restructuring You can also use this Intelligent Tagging classification topic tag to identify documents about Company Reorganization.

Example 1

Input Text

Badger Income Fund said it will delay a planned corporate reorganization.

Extracted Relation

company = Badger Income Fund

status = postponed

Example 2

Input Text

Avis Budget Group, Inc. announced Thursday the restructuring of its Budget Truck Rental subsidiary.

Extracted Relation

company = Avis Budget Group, Inc.

status = announced

datestring = Thursday

CompanyRestatement

em/r/CompanyRestatement

Definition

A reference to a company restatement of its financial statement for a specific time period.

Attributes

company: The company that is the subject of the relation type.

datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text.

quarter: Time period to which the extracted relation refers.

status: The status, based on an indication in the text.

year: The relevant year.

Example

Input Text	AMIS Holdings, Inc. announced that the audit committee has completed its review, and as a result, the Company has decided to restate its consolidated financial statements for the first and second quarters of fiscal year ended December 31, 2007.
Extracted Relation	company = AMIS Holdings, Inc. quarter = Q1 year = 2007 status = announced
Extracted Relation	company = AMIS Holdings, Inc. quarter = Q1 year = 2007 status = announced
Extracted Relation	company = AMIS Holdings, Inc. quarter = Q2 year = 2007 status = announced

CompanyTechnology

em/r/CompanyTechnology

Definition

A technology associated with a company.

Attributes

company: Company associated with the technology.

technology: The technology related to the company.

Example 1

Input Text

Their expertise means that the company is ideally suited to distributing RC Networks' SDSL solutions, which complement the existing offerings.

Extracted Relation

company = RC Networks, Inc.

technology = SDSL

Example 2

Input Text

Ampliphi Biosciences Corporation develops gene therapy products for the treatment of acquired and inherited diseases.

Extracted Relation

company = Ampliphi Biosciences Corporation

technology = gene therapy

CompanyTicker

em/r/CompanyTicker

Definition

Extracted instance of a company name followed by its stock or ticker symbol/s in parentheses.

Attributes

company: The company that is the subject of the relation type.

stockexchange: The name of the relevant stock market.

ticker: The company’s ticker symbol if mentioned in the text.

Example

Input Text

The Board of Globalworth Real Estate Investments Limited (AIM: GWI) announces that the registered office address of the Company has changed to PO Box 156, Frances House, Sir William Place, St Peter Port, Guernsey, GY1 4EU with effect from 18 December 2013.

Extracted Relation

company = Globalworth Real Estate Investments Limited

ticker = GWI

stockexchange = AIM

CompanyUsingProduct

em/r/CompanyUsingProduct

Definition

A relationship that includes a company or an organization and a product that they are using.

Attributes

company_customer: Company that purchases/purchased products or services from the company_provider.

company_provider: Company that sells the product or service.

datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text.

organization_customer: Organization that purchases/purchased products or services from the company_provider.

product: The product that is the subject of the extracted relation.

producttype: Product classification.

Possible values:

aircraft
car
drug
electronics
weapon
other

status: The status, based on an indication in the text.

Example 1

Input Text

BakBone Software(R) (TSX: BKB, OTC Pink Sheets: BKBO), a global provider of data protection software, is using the AppExchange on-demand platform.

Extracted Relation

company_customer = BakBone Software Incorporated

product = AppExchange

producttype = Other

status = announced

Example 2

Input Text

After a year of negotiations, the West Australian government has formally chosen the Leighton Properties/Mirvac consortium as the preferred developer for the $3 billion-plus worth of work above the Perth City Link.

Extracted Relation

organization_customer = West Australian government

product = Leighton Properties/Mirvac

producttype = Other

status = announced

ConferenceCall

em/r/ConferenceCall
Definition	A conference call held by a company (in the past, present, or future).
Attributes	company: The company that is the subject of the relation type. conferencecalltype: The nature of the conference call, based on the topics to be discussed. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). quarter: Time period to which the extracted relation refers. status: The status, based on an indication in the text.
Equivalent Topic Tag	E:58 Earnings Calls/Guidance Calls You can also use this Intelligent Tagging classification topic tag to identify documents about Conference Calls.

Example 1

Input Text

“The July-September quarter was our best by production on like-to-like basis,”

Tata Steel Europe managing director and chief executive, Karl-Ulrich Kohler,

said in a post-earnings conference call. Rise in quarterly volumes represent a

big step forward, he added. This has lifted Tata Steel’s stock 16 per cent in

the last one month and the price has more than doubled from its lows in August

this year.

In its September 2013 earnings conference call, Tata Steel said its production

in Europe had increased due to stabilisation of the new blast furnace at Port

Talbot in the UK. Its efforts to raise efficiency have also lifted operating

profit for every tonne.

Extracted Relation

company = Tata Steel Europe Limited

conferencecalltype = earnings

status = announced

Extracted Relation

company = Tata Steel Europe Limited

date = 2013-09-00

datestring = September 2013

conferencecalltype = earnings

status = announced

Example 2

Input Text

Raiffeisen Group and Bank Vontobel to hold joint conference call.

Extracted Relation

company = Raiffeisen Grp – Global Custody

status = planned

Extracted Relation

company = Bank Vontobel Cayman

status = planned

ContactDetails

em/r/ContactDetails

Definition

The contact details of a company or person.

Note: Entities of type Person, Company, City, Country and ProvinceOrState, which appear only within a ContactDetails instance, will not generate em/e/Person, Company, City, Country, or Provinceorstate tags.

Attributes

address: The contact address.

companycontactentity: The name of the company whose contact details are provided.

emailaddress1: The contact e-mail address.

emailaddress2: A second contact e-mail address.

fax: The fax number of the contact.

messenger: The contact’s Instant Messaging name.

personcontactentity1: The first person whose contact details are provided.

personcontactentity2: The second person whose contact details are provided.

telephone1: The contact phone number.

telephone2: A second contact phone number.

url: Website of the contact.

Example

Input Text

For more information, see www.soldiersangels.org or call 615-676-0239.

Extracted Relation

telephone = 615-676-0239

url = www.soldiersangels.org

Conviction

em/r/Conviction
Definition	Reference to the conviction of one or more people.
Attributes	charge: Phrase describing an offense/charge. In case of multiple charge attributes, the first charge attribute is the primary charge. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). othercharges: Additional charges. person: The person who is the subject of this extracted relation.
Equivalent Topic Tag	M:EV Judicial Process/Court Cases/Court Decisions You can also use this Intelligent Tagging classification topic tag to identify documents about Convictions.

Example 1

Input Text

Mr. al-Sirri himself was sentenced to death in absentia in an Egyptian court in 1994 for his role in an assassination attempt on the Egyptian prime minister.

Extracted Relation

person = al-Sirri

charge = assassination attempt on the Egyptian prime minister

datestring = in1994

date = 1994-00-00

Example 2

Input Text

When José María del Nido took over as president of Sevilla Football Club in May 2002, he promised to "clean up the accounts and win trophies." That he certainly did, through astute transfer business that balanced the books and reaped six

titles. On Monday, Del Nido bade farewell to the club at a press conference after the Supreme Court upheld his conviction for embezzlement and perverting the course of justice in the so-called Minutas case, in which Del Nido and jailed former Marbella mayor Julián Muñoz systematically overbilled the town authorities for services.

Extracted Relation

person = José María del Nido

charge = embezzlement

charge = perverting the course of justice

CreditRating

em/r/CreditRating
Definition	A recent rating by a rating agency of a company, financial institution, or country.
Attributes	company_rated: Company that is the subject of the estimate, recommendation, or rating. company_source: The financial or other body publishing the estimate, rating, or recommendation. country_rated: Country which is being rated. financialtrend:The trend of the current rating with respect to the previous rating. Possible values: Affirms Assigns Changes Cuts Expects to change Puts Raises Rates Removes Says Withdraws forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). organization_rated: The organization that is being rated. rank_new: New or current ranking of the company. rank_old: Previous ranking of the company.
Equivalent Topic Tag	E:N Credit Rating Issued You can also use this Intelligent Tagging classification topic tag to identify documents about Credit Ratings.

Example 1

Input Text

HONG KONG (Standard & Poor's) Sept. 27, 2005--Standard & Poor's Ratings Services said today it had assigned its 'BB' long-term corporate credit rating to Fosun International Ltd. (Fosun).

Extracted Relation

company_source = Standard and Poor’s Financial Services LLC

company_rated = Fosun International Limited.

financialtrend = Assigns

rank_new = BB

Example 2

Input Text	This includes Fitch's 'A-' long-term and senior debt ratings and 'BBB+' preferred stock rating on RenRe Ltd., as well as its 'A+' insurer financial strength rating on lead operating subsidiary.
Extracted Relation	company_source = Fitch Ratings Ltd. company_rated = RenRe North America Insurance Holdings, Inc. rank_new = A-
Extracted Relation	company_source = Fitch Ratings Ltd. company_rated = RenRe North America Insurance Holdings, Inc. rank_new = BBB+
Extracted Relation	company_source = Fitch Ratings Ltd. company_rated = RenRe North America Insurance Holdings, Inc. rank_new = A+

Example 3

Input Text

Fitch Ratings-New York-September 23, 2005: Fitch has upgraded UNOVA, Inc.'s (UNOVA) issuer default rating to 'BB-' from 'B-'

Extracted Relation

company_source = Fitch Ratings, Inc.

company_rated = Intermec, Inc.

financialtrend = Raises

rank_new = BB-

rank_old = B-

Deal

This is a premium metadata type.

em/r/Deal
Definition	A deal between two companies. The tags that describe a particular deal (one or more instance tags, em/r/deal, er/deal, confidence tag) are excluded from the output if Intelligent Tagging is not able to find the deal in the Refinitiv Deals Authority dataset. Note: The same text may also be identified by Intelligent Tagging as a Merger or an Acquisition. Intelligent Tagging generates em/r/merger and em/r/acquisition tags for identified Mergers and Acquisitions without attempting to locate them in the Refinitiv dataset.
Attributes	acquirer: The purchasing company. confidencelevel: A confidence score on a scale of 0 to 1. The value represents the probability that the extracted relation is indeed of the assigned type. For example: The confidence level score for an em/r/deal tag indicates the probability that the deal is indeed a deal between the companies specified by the Acquirer and Target attributes. The confidence level score for an em/r/bankruptcy tag indicates the probability that text actually refers to a bankruptcy involving the company specified by the Company attribute. The confidence level score for an em/r/IPO tag indicates the probability that the text actually refers to an IPO involving the company specified by the Company attribute. The higher the value, the higher the probability. The consuming application can use this score to achieve higher accuracy results by ignoring instances with confidence scores below a specified level. Note that boosting Precision in this manner is at the expense of Recall. Note that this em/r/tag also has a corresponding Confidence tag. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). permid: The Refinitiv unique ID (PermID) for this metadata tag type. Can be used when building a knowledge graph. target: Company being acquired.
Related Tags	er/Deal (Disambiguation Tag), Confidence Tag, em/e/company
Equivalent Topic Tag	E:J Mergers/Acquisitions/Takeovers You can also use this Intelligent Tagging classification topic tag to identify documents about Deals.

Example 1

Input Text

Amazon.com Inc (AMZN.O) on Wednesday cleared two of the biggest hurdles it needed to close its $13.7 billion acquisition of Whole Foods Market Inc WFM.O, with approvals from a U.S. regulator and the grocery chain’s shareholders.

Extracted Relation

acquirer = Amazon.com, Inc.

target = Whole Foods Market, Inc.

Example 2

Input Text

The Anthem (ANTM)-Cigna (CI) merger trial has fallen behind schedule, which potentially could delay a ruling by at least a couple of days. Judge Amy Berman Jackson had previously divided the Justice Department's challenge to the health-insurance merger into two phases.

Extracted Relation

acquirer = Anthem, Inc.

target = Cigna Corporation

DebtFinancing

em/r/DebtFinancing
Definition	A mention of companies seeking to finance their debt.
Attributes	company: Company raising money through debt financing. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. debtaction: The action taken. Extracted based on the context. debttype: The type of debt expressed as a financing instrument. For example, shares, bonds, notes, etc. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text.
Equivalent Topic Tag	A:T Corporate Debt You can also use this Intelligent Tagging classification topic tag to identify documents about Debt Financing.

Example 1

Input Text

Verizon entered into a $7.55 billion 364-day credit facility.

Extracted Relation

company = Verizon Communications Inc.

status = announced

debttype = credit facility

debtaction = entered

Example 2

Input Text

Bharti Airtel said its Euro 750-million ($1 billion) corporate bond issue elicited enthusiastic response and was oversubscribed more than five times with participation from 370 high-quality accounts.

"The notes have been priced 300 basis points over the five-year Euro Mid Swap with a fixed coupon of 4% per annum to yield 4.055%. Bharti will apply the net proceeds to refinance existing debt, " the company said in a media statement on Wednesday. The notes were distributed across a mix of high-quality fixed income accounts, including fund managers (77%), SSA/insurance (9%), private banks (9%) and banks (5%).

Extracted Relation

company = Bharti Airtel Limited

status = known

debttype = senior notes

debtaction = has priced

date = 2018-00-00

datestring = in 2018

DelayedFiling

em/r/DelayedFiling

Definition

A mention of a delay in the filing of financial reports by a company.

Attributes

company: The company that is the subject of the relation type.

filingtype: Name/s of the report/s filed, based on a mention in the text and mapped to one of the following values:

Form 10-K (for Form 10-K or Form 10-KSB)

Form 20-F

Form 10-Q (for Form 10-Q, 10-QSB, or other quarterly report)

Annual report

Interim report

If the mention in the text cannot be mapped to one of the above values, the mention itself is used as the attribute value.

quarter: Time period to which the extracted relation refers.

status: The status, based on an indication in the text.

year: The relevant year.

Example

Input Text

Hollinger International Inc. filed its delayed quarter financial report with U.S. securities regulators on Friday without the required review of either its auditor or its top executives and said it expected the statement to be declared "deficient" as a result. The newspaper company - at the centre of widespread controversy over $32-million (U.S.) in unauthorized payments received by a number of executives, including former chief executive officer Conrad Black - had foreshadowed the move earlier in the week, following Lord Black's early departure from the chief executive job.

Extracted Relation

company = Chicago Newspaper Liquidation Corp.

filingtype = financial report

status = known

DiplomaticRelations

em/r/DiplomaticRelations
Definition	Information about diplomatic relations between diplomatic entities (e.g. countries, governments, minority groups, regions, international organizations, a spokesperson for a group, etc.).
Attributes	date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. diplomaticaction: Diplomatic action taken. diplomaticentity1: An organization or person relevant to the diplomatic action. diplomaticentity2: An organization or person relevant to the diplomatic action. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). location: Location related to the extracted relation (address or other physical description).
Equivalent Topic Tag	M:C Diplomacy/Foreign Policy; M:1SJ Belt and Road Initiative You can also use this Intelligent Tagging classification topic tags to identify documents about Diplomatic Relations.

Example

Input Text

Poland and the Baltic states came together on Saturday to issue a harsh condemnation of Russia for what they described as aggression in Georgia and urged NATO and the European Union to oppose it.

Extracted Relation

diplomaticentity1 = Poland

diplomaticaction = condemnation

diplomaticentity2 = Russia

Extracted Relation

diplomaticentity1 = Baltic States

diplomaticaction = condemnation

diplomaticentity2 = Russia

Dividend

Definition

Information about dividends paid to shareholders by a company.

Attributes

company: The company that is the subject of the relation type.

dividendtype: The type of dividend.

Possible values:

cash
shares

quarter: Time period to which the extracted relation refers.

status: The status, based on an indication in the text.

year: The relevant year.

Equivalent Topic Tag

E:F Dividends

You can also use this Intelligent Tagging classification topic tag to identify documents about Dividends.

Example 1

Input Text

South African short-term insurer Santam reported annual results that beat expectations on Wednesday and said it will pay a special cash dividend, sending its stock nearly seven percent higher.

Extracted Relation

company = Santam Limited

status = announced

dividendtype = cash

Example 2

Input Text

"We have a good level of yearly free cash flow, about $800 million for the entire Philippine Long Distance Telephone Co. group. Most likely we are enhancing our dividend payout to 70 percent of core earnings in 2007," he added.

Extracted Relation

company = PLDT Inc.

status = planned

quarter = FY

year = 2007

EmploymentChange

em/r/EmploymentChange
Definition	An announcement about a recent appointment, retirement, departure (including dismissal), or promotion of one or more officers in a company or organization.
Attributes	changetype: Classification of the change. Possible values: Enters Leaves Retired company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. employmentchangestatus: Status of the change in employment. Possible values: announced planned forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). organization: The organization that is a participant in the extracted relation. person: The person who is the subject of this extracted relation. position: Position that the person held, holds, or will hold.
Equivalent Topic Tags	E:6T People Moves; E:6V Change of Board Member; E:68 Change of CEO; E:69 Change of CFO; E:6A Change of Chairperson; E:6B Change of President. You can also use these Intelligent Tagging classification topic tags to identify documents about Employment Changes.

Example 1

Input Text

Electronic component developer Panja Inc. said it has appointed Berry Cash as chairman of the company.

Extracted Relation

company = Panja Inc.

person = Berry Cash

position = chairman

changetype = enters

employmentchangestatus = announced

Example 2

Input Text

Bank One Corp. said that president Verne Istock will retire Sept. 30.

Extracted Relation

company = Bank One Corporation

person = Verne G. Istock

position = president

changetype = leaves

datestring = Sept. 30

employmentchangestatus = planned

EmploymentRelation

em/r/EmploymentRelation

Definition

Extracts a reference to a person’s employment by another person.

Attributes

datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text.

person_employee: The person who is the employee.

person_employer: The person who is the employer.

position: Position that the person held, holds, or will hold.

status: The status, based on an indication in the text.

Example

Input Text

Jan Scheck ended up being hired by Mr. E.J. Ridings at Trump Mortgage as national sales manager, but left the firm this year after his one-year contract was up.

Extracted Relation

person_employer = E.J. Ridings

person_employee = Jan Scheck

position = national sales manager

EnvironmentalIssue

em/r/EnvironmentalIssue
Definition	An issue that has an effect on the environment.
Attributes	date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. effect: The impact of the environmental issue. environmentalissue: The environmental issue mentioned in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). location: Location related to the extracted relation (address or other physical description).
Equivalent Topic Tags	M:H Environment; M:NI Nature/Wildlife; M:JZ Pollution; U:F1 Water. You can also use these Intelligent Tagging classification topic tags to identify documents about Environmental Issues.

Example

Input Text

Habitat destruction and poaching have decimated populations of the Sumatran rhino, a species once widespread in Asia. Today, fewer than 100 Sumatran rhinos are believed to still exist in the wild. All of them are in Indonesia, in either Kalimantan or the island of Sumatra. For years, it was assumed that the Sumatran rhino had gone extinct in Kalimantan. But in 2013, camera trap evidence suggested that at least a few of the species had returned. It wasn't until Najaq's capture in March, however, that conservationists could confirm their existence in the area.

Extracted Relation

environmentalissue=Habitat destruction

EquityFinancing

em/r/EquityFinancing
Definition	A secondary issuance of securities or a new issuance of other stocks.
Attributes	company: The company that is the subject of the relation type. currency: Currency in which the issuance is offered. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text.
Equivalent Topic Tags	E:M Initial Public Offerings; E:5C Secondary Share Offerings/Issues You can also use these Intelligent Tagging classification topic tags to identify documents about Equity Financing.

Example 1

Input Text

Last week SandRidge Energy, Inc. Announced Pricing of Public Offering of 15.2 Million Shares of Common Stock, at the price of AUD0.4344 per share, totaling to a purchase price of AUD198,000.

Extracted Relation

company = SandRidge Energy, Inc.

status = announced

datestring = Last week

Example 2

Input Text

Inversiones Copper Holding Chile Limitada will issue 1,993,750 common shares at a deemed price of $0.075 per share in satisfaction of the purchase price of $159,500.

Extracted Relation

company = Inversiones Copper Holding Chile Ltda – Manto Medio Claim Assets

status = planned

Extinction

em/r/Extinction

Definition

Extracted mention/s of an extinct or endangered animal or plant.

Attributes

datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text.

location: Location related to the extracted relation (address or other physical description).

species: The affected plant or animal species.

Example 1

Input Text

The Chinese paddle fish and the dog-eating catfish in Southeast Asia are on the brink of extinction because of pollution, overfishing and dam building.

Extracted Relation

species = The Chinese paddle fish

location = Southeast Asia

Extracted Relation

species = the dog-eating catfish

location = Southeast Asia

Example 2

Input Text	A tall, heavy beast with long, forward-curving horns faced down a smaller bull. Its head was held high as if in challenge. The smaller animal seemed to recoil in submission. Even the cave lion, the largest of predators, looked unlikely to challenge the dominant bull. Behind them a herd of giant deer ran from unseen danger. This scene was depicted in a painting on a cavern wall at Lascaux, France, 17,000 years ago. Megaloceros, the giant deer in the background, are extinct. The cave lion is extinct. The bulls - a species called aurochs (pronounced aur-ox) - have disappeared. Their genes, however, are still present in modern cattle, and scientists have been trying to bring them back to life.
Extracted Relation	species = The cave lion
Extracted Relation	species = The giant deer

FamilyRelation

em/r/FamilyRelation

Definition

A family relationship between two people.

Attributes

familyrelationtype: Describes the relationship of the person_relative to the person.

person: The person who is the subject of this extracted relation.

person_relative: The person who is the family relative.

Example 1

Input Text

The wife of an Australian Greenpeace activist released from a Russian jail says she is relieved, but aware the family's ordeal is not yet over. Colin Russell was released from a St Petersburg jail on Friday after two months in custody over a protest against Russia's Prirazlomnaya oil rig in the Arctic. The Tasmanian was the last of the so-called "Arctic 30" protesters to be granted bail. His wife Christine was in St Petersburg for his release, having vowed not to leave Russia without her husband.

Extracted Relation

person = Colin Russell

familyrelationtype = wife

person_relative = Christine

Example 2

Input Text

Mahathir Mohamad was to be accompanied by his wife Siti Hasmah Mohamad Ali.

Extracted Relation

person = Mahathir Mohamad

person_relative = Siti Hasmah Mohamad Ali

familyrelationtype = wife

FDAPhase

em/r/FDAPhase

Definition

The stage of a drug in the US Food and Drug Administration’s approval process.

Attributes

company: Name of the drug manufacturer.

fdastage: The stage of the drug in the FDA approval process.

product: The product that is the drug in question.

producttype: Product classification.

Possible values:

aircraft
car
drug
electronics
weapon
other

status: The status, based on an indication in the text.

Example

Input Text

Provectus Pharmaceuticals, Inc. (OTC Bulletin Board: PVCT) announced today it has filed an Investigational New Drug (IND) application for Provecta(TM), an advanced drug therapy designed to treat breast, liver, prostate and other potentially deadly cancers.”

Extracted Relation

company = Provectus Pharmaceuticals, Inc.

product = Provecta(TM)<

fdastage = INDA

status = NA

producttype = drug

IndicesChanges

em/r/IndicesChanges
Definition	References to changes in financial indices, e.g. companies added to or removed from one of the indices. IndicesChanges are extracted only if a company is removed from an index for lack of presentation, and not because of a merger or acquisition.
Attributes	changetype: Classification of the change. Possible values: Enters Leaves company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). marketindex: The index in which a change was made.
Equivalent Topic Tag	E:10 Indices Changes You can also use this Intelligent Tagging classification topic tag to identify documents about Indices Changes.

Example 1

Input Text	Software giant Synopsys (SNPS) will be added to the S&P 500 Index before the market opens on Thursday, S&P Dow Jones Indices said today. It replaces Harman International (HAR), a company bought by Samsung on Friday that makes car audio systems and connected-car technology. The swap-out comes after S&P Dow Jones Indices on Friday said that Advanced Micro Devices (AMD), whose chips and graphics processors lifted its stock last year, investment firm Raymond James Financial (RJF) and real estate investment trust Alexandria Real Estate Equities (ARE) would be added to the S&P 500 on March 20.
Extracted Relation	Marketindex = S&P 500 company = Alexandria Real Estate Equities, Inc. changetype = enters datestring = March 20
Extracted Relation	Marketindex = S&P 500 company = Raymond James Financial, Inc. changetype = enters datestring = March 20
Extracted Relation	Marketindex = S&P 500 company = Synopsys, Inc. changetype = enters

Example 2

Input Text

Just witness the effect on troubled telecom WorldCom on the day after the research firm said it would boot the troubled telecom company off the S&P 500 index.

Extracted Relation

marketindex = S&P 500

company = WorldCom

changetype = leaves

Indictment

em/r/Indictment
Definition	An indictment of a person or group of people (charged with criminal offenses).
Attributes	charge: Phrase describing an offense/charge. In case of multiple charge attributes, the first charge attribute is the primary charge. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). othercharges: Additional charges. person: The person who is the subject of this extracted relation.
Equivalent Topic Tag	M:EV Judicial Process/Court Cases/Court Decisions You can also use this Intelligent Tagging classification topic tag to identify documents about Indictments.

Example 1

Input Text

BISMARCK, N.D. — A special commissioner has recommended that suspended McKenzie County Sheriff Gary Schwartzenberger be allowed to return to his post. Karen Klein says in her recommendation to Gov. Doug Burgum that the state hasn't provided enough evidence for Schwartzenberger's removal. Schwartzenberger is accused of bullying, retaliation and misusing a county credit card. He's scheduled for trial in September on a misdemeanor charge.

Extracted Relation

person = Gary Schwartzenberger

charge = bullying

charge = retaliation

charge = misusing a county credit card

Extracted Relation

person = Gary Schwartzenberger

charge = a misdemeanour charge

datestring = in September

Example 2

Input Text

Jermaine Wallace, 25, of Salisbury, and Natasha Adams, 25, of Bowie, were charged with conspiracy to distribute cocaine and money laundering.

Extracted Relation

person = Jermaine Wallace

charge = conspiracy to distribute cocaine

charge = money laundering

Extracted Relation

person = Natasha Adams

charge = conspiracy to distribute cocaine

charge = money laundering

IPO

em/r/IPO
Definition	An initial public offering (IPO) of stock by a company.
Attributes	company: The company that is the subject of the relation type. confidencelevel: A confidence score on a scale of 0 to 1. The value represents the probability that the extracted relation is indeed of the assigned type. For example: The confidence level score for an em/r/deal tag indicates the probability that the deal is indeed a deal between the companies specified by the Acquirer and Target attributes. The confidence level score for an em/r/bankruptcy tag indicates the probability that text actually refers to a bankruptcy involving the company specified by the Company attribute. The confidence level score for an em/r/IPO tag indicates the probability that the text actually refers to an IPO involving the company specified by the Company attribute. The higher the value, the higher the probability. The consuming application can use this score to achieve higher accuracy results by ignoring instances with confidence scores below a specified level. Note that boosting Precision in this manner is at the expense of Recall. Note that this em/r/tag also has a corresponding Confidence tag. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). permid:The Refinitiv unique ID (PermID) for this metadata tag type. Can be used when building a knowledge graph. status: The status, based on an indication in the text.
Related Tag	Confidence Tag
Equivalent Topic Tag	E:M Initial Public Offering You can also use this Intelligent Tagging classification topic tag to identify documents about Initial Public Offerings.

Example 1

Input Text

SM Investments Corp. has set the price of the Philippines' largest initial public offering at 250 pesos ($4.6) per share to raise a total 28.75 billion pesos ($530 million), traders said on Friday.

Extracted Relation

company = SM Investments Corporation

status = announced

datestring = Friday

Example 2

Input Text

Science Applications International Corp. on Monday set its planned initial public offering at 75 million shares at an estimated price of $13 to $15 each.

Extracted Relation

company = Leidos Holdings, Inc.

status = planned

datestring = Monday

JointVenture

Definition

A joint-venture agreement, signed recently or in the past.

Attributes

company: The company that is the subject of the relation type.

company_newname: The new name of the company.

datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text.

status: The status, based on an indication in the text.

Equivalent Topic Tag

E:4G Strategic Combinations

You can also use this Intelligent Tagging classification topic tag to identify documents about Joint Ventures.

Example

Input Text

L-3 Communications (NYSE:LLL) and Thales Avionics, a wholly-owned subsidiary of Thales (formerly Thomson-CSF) specializing in avionics and aircraft cabin electronics, announced that they have signed an agreement to form a joint venture company to operate the assets of L-3's Aviation Communications & Surveillance Systems (ACSS).

Extracted Relation

company = L-3 Technologies, Inc.

company = Thales Avionics SAS

status = announced

ManMadeDisaster

em/r/ManMadeDisaster
Definition	A disaster or incident that is not due to an act of nature. For example fire, nuclear disaster, oil spill, explosion.
Attributes	date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. facility: The facility in which the incident occurred. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). location: Location related to the extracted relation (address or other physical description). manmadedisaster: Phrase describing a man-made disaster.
Equivalent Topic Tags	M:1LY Disasters/Accidents; M:NA Air Accidents; M:KF Fires; M:NC Ground Accidents/Collisions; M:ND Sea Accidents; M:KM Wildfires/Forest Fires. You can also use these Intelligent Tagging classification topic tags to identify documents about Man-made Disasters.

Example

Input Text

Rehabilitated Penguins are Released Following Oil Spill in Uruguay.

Extracted Relation

manmadedisaster = oil spill

location = Uruguay

Merger

This is a premium metadata type.

em/r/Merger
Definition	A reference to a merger transaction (past, present, or future).
Attributes	company: Company involved in the merger. Note: The company attribute appears two or more times in the em/r/merger tag. Each company attribute indicates a different company that is involved in the merger. confidencelevel: A confidence score on a scale of 0 to 1. The value represents the probability that the extracted relation is indeed of the assigned type. For example: The confidence level score for an em/r/deal tag indicates the probability that the deal is indeed a deal between the companies specified by the Acquirer and Target attributes. The confidence level score for an em/r/bankruptcy tag indicates the probability that text actually refers to a bankruptcy involving the company specified by the Company attribute. The confidence level score for an em/r/IPO tag indicates the probability that the text actually refers to an IPO involving the company specified by the Company attribute. The higher the value, the higher the probability. The consuming application can use this score to achieve higher accuracy results by ignoring instances with confidence scores below a specified level. Note that boosting Precision in this manner is at the expense of Recall. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text.
Equivalent Topic Tag	E:J Mergers/Acquisitions/Takeovers You can also use this Intelligent Tagging classification topic tag to identify documents about Mergers.

Example

Input Text

Extracted Relation

company = Anthem, Inc.

company = Cigna Corporation

status = known

MilitaryAction

em/r/MilitaryAction
Definition	A non-combative action taken by the military.
Attributes	action: Non-combative action taken by the military. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). locationstring: Location of the military action as mentioned in the text. military: A military group, or the country or person a military group serves. When the name of a capital city is used to refer to a government, the attribute value will be the country name. In the following example, “Moscow deployed 10,000 extra troops,” Russia is extracted as the Military attribute, and not Moscow. purpose: Purpose of the action.
Equivalent Topic Tag	M:2 Military Conflicts You can also use this Intelligent Tagging classification topic tag to identify documents about Military Action.

Example 1

Input Text

Ministry head Shota Utiashvili said Moscow was deploying 10,000 extra troops in Georgia.

Extracted Relation

military = Russia

action = was deploying 10,000 extra troops

locationstring = Georgia

Example 2

Input Text

Thousands of Turkish troops have crossed into northern Iraq to hunt Kurdish rebels.

Extracted Relation

military = Turkish troops

action = have crossed into northern Iraq

purpose = to hunt Kurdish rebels

locationstring = Iraq

MovieRelease

em/r/MovieRelease
Definition	A movie release (past, present, or future).
Attributes	date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). movie: The movie. status: The status of the movie release based on an indication in the text. Possible values: · New (for a movie that was released during the current month or for a future release) · Known (all other instances, where there is no clear indication that it is a new or future release)
Equivalent Topic Tag	M:H1 Film You can also use this Intelligent Tagging classification topic tag to identify documents about Movie Releases.

Example

Input Text

The latest Will Smith movie, "Hancock", will hit theaters Wednesday, July 2.

Extracted Relation

movie = Hancock

datestring = Wednesday, July 2

status = New

MusicAlbumRelease

em/r/MusicAlbumRelease
Definition	A music album release (past, present, or future)
Attributes	date: An absolute date (in the YYYY-MM-DD format). This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date or a reference to a date (e.g. yesterday, next month), taken directly from the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). musicalbum: The album. musicgroup_performer: The musician or group releasing the album. person_performer: The person who is releasing the music album. status: The status of the filing based on an indication in the text. Possible values: New (for an album that was released during the current month or for a future release) Known (for all other instances, where there is no clear indication that it is a new or future release)
Equivalent Topic Tag	M:H0 Music You can also use this Intelligent Tagging classification topic tag to identify documents about Music Album Releases.

Example

Input Text

Björk‘s new album is expected to be released in November, according to reports. The Icelandic musician recently announced a new single called ‘The Gate’, which will be available digitally from September 18 and on vinyl from September 22. Björk’s new album, her follow-up to 2015’s ‘Vulnicura’, has long been expected to be released later this year. Now The New York Times reports that it will be released in November.

Extracted Relation

status = new

person_performer = Björk

NaturalDisaster

em/r/NaturalDisaster
Definition	A natural disaster. For example, a hurricane, an earthquake, a flood.
Attributes	date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). location: Location related to the extracted relation (address or other physical description). naturaldisaster: Phrase describing a natural disaster.
Equivalent Topic Tags	M:1LY Disasters/Accidents; M:KN Droughts; M:KL Earthquakes; M:KF Fires; M:KJ Floods; M:NK Tsunami; M:KM Wildfilres/Forest Fires; U:D2 Wind/Hurricanes/Tyhpoons/Tornadoes You can also use these Intelligent Tagging classification topic tags to identify documents about Natural Disasters.

Example

Input Text

More than 250 people died when torrential rains and flooding swept China last month, according to the China Daily.

Extracted Relation

naturaldisaster = flood

location = China

PatentFiling

em/r/PatentFiling
Definition	A company or organization filing for patent registration, and information about the patent.
Attributes	company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). organization: The organization that is a participant in the extracted relation. patentdescription: Phrase describing the patent. status: The status of the filing based on an indication in the text. Possible values: New (filed within the past 12 months) Known (filed more than a year ago)
Equivalent Topic Tag	E:6C Intellectual Property You can also use this Intelligent Tagging classification topic tag to identify documents about Patent Filing.

Example

Input Text

Lixte Biotechnology Holdings announces filing of a New patent Application for Cancer Therapy Based on Targeting a Biomarker by Novel Investigational Agents Tuesday August 19, 8:15 am ET.

Extracted Relation

company = Lixte Biotechnology Holdings, Inc.

patentdescription = Cancer Therapy Based on Targeting a Biomarker by Novel Investigational Agents

datestring = Tuesday August 19

status = New

PatentIssuance

em/r/PatentIssuance
Definition	Reference to a patent being issued/granted to a company or organization, and information about the patent.
Attributes	company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). organization: The organization that is a participant in the extracted relation. patentdescription: Phrase describing the patent. patentnumber: The patent number. status: The status of the patent issuance based on an indication in the text. Possible values: New (issued within the past 12 months) Known (issued more than a year ago
Equivalent Topic Tag	E:6C Intellectual Property You can also use this Intelligent Tagging classification topic tag to identify documents about Patent Issuance.

Example

Input Text

Method patent protects the entire Proterro sugar-making process that integrates a patented, highly productive microorganism with a novel, robust photo bioreactor made from off-the-shelf materials EWING, N.J.--(Business Wire)—

The United States Patent and Trademark Office has issued U.S. Patent No. 8,597,914 to Proterro, Inc., protecting Proterro`s unique, biosynthetic sugar-making process.

Extracted Relation

company = Proterro Inc

patentnumber = 8,597,914

status = new

PersonAttributes

em/r/PersonAttributes

Definition

Extracts references to a person’s age, birth date, birth place, and/or gender.

Attributes

age: The person’s age.

birthdate: The person’s date of birth.

birthplace: The person’s place of birth.

gender: The person’s gender. Possible values: “F” (female), “M” (male).

person: The person who is the subject of this extracted relation.

Example 1

Input Text

Leonard Pennario was born in Buffalo on July 9, 1924. When he was 10 he and his family moved to Los Angeles

Extracted Relation

person = Leonard Pennario

birthdate = 1924-07-09

birthplace = Buffalo

gender = M

Example 2

Input Text

Mercedes-Benz Korea announced on Wednesday that Harald Behrend, 46, has been named as the new head of the company.

Extracted Relation

person = Harald Behrend

age = 46

gender = M

PersonCareer

em/r/PersonCareer

Definition

A reference to a person and the position they hold or held in the past. The position may be that of a judge or other civil servant, a political office, or a professional position in a company or organization.

Attributes

careertype: Indicates whether the career is political or professional.

city: City related to the extracted relation.

company: Company affiliated with the position.

country: Country related to the extracted relation.

organization: The organization that is affiliated with the position.

person: The person who is the subject of this extracted relation.

position: Position that the person held, holds, or will hold.

positionnormalized: A role in the legal profession.

Possible values:

Attorney

Judge

provinceorstate: The ProvinceOrState, related to the extracted relation.

status: The status, based on an indication in the text.

Example

Input Text	(Virginia Beach, VA) - Ronald Liu has been named Chief Financial Officer (CFO) of AMSEC LLC, a Virginia Beach-based company that provides a broad range of engineering and technical services to the Department of Defense, various governmental agencies and the commercial maritime industry. “Ron Liu has a terrific background in a broad spectrum of significant corporate positions. Bringing Ron aboard adds great leadership to our team. As CFO, he will play a major part in continuing the significant growth AMSEC LLC has enjoyed over the past several years,” said Gary Lisota, AMSEC LLC president. Prior to joining AMSEC LLC, Liu was director of Operations and Financial Controls at Johnson Controls Government Systems LLC where he was responsible for all financial aspects of a business unit that experienced revenue growth of more than 400 percent in a five-year period.
Extracted Relation	person = Ronald Liu position = Chief Financial Officer (CFO) company = Amsec LLC careertype = professional status = current
Extracted Relation	person = Ronald Liu position = Director of Operations and Financial Controls company = Johnson Controls Govt Systems careertype = professional status = past
Extracted Relation	person = Gary Lisota position = President company = Amsec LLC careertype = professional status = current

PersonCommunication

em/r/PersonCommunication

Definition

Communication (meeting, phone call, direct verbal communication, etc.) between people or groups of people. The meetings can be confirmed to have taken place, alleged to have taken place, or planned for the future. The meeting participants may be identified in this tag by any of the following attributes: person, persondescription, organizationorcompany, facility.

Attributes

datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text.

facility: The facility related to the extracted relation.

organizationorcompany: An organization or company participating in the communication.

person: The person who is a participant in this extracted relation.

persondescription: Description of participant/s in the communication.

status: The status, based on an indication in the text.

Example

Input Text

Sears Holdings (SHLD) wants sister chain Sears Hometown and Outlet Stores to participate in the previously announced process of considering strategic alternatives for SHLD's Kenmore, Craftsman and DieHard brands. SHLD CEO Eddie Lampert communicated as much to representatives of Sears Hometown Thursday, according to a securities filing by ESL Partners, Lampert's hedge fund. Lampert told the representatives he and other reporting persons would consider buying additional stock in Sears Hometown in "connection with any such transaction," the filing says.

Extracted Relation

person=Edward S. Lampert

persondescription=representatives

status=announced

PersonEducation

em/r/PersonEducation

Definition

A person’s academic degree/s or certification/s.

Attributes

certification: A mention of professional certification.

degree: Academic degree awarded, not including the specific department or field of study.

person: The person who is the subject of this extracted relation.

schoolororganization: Academic institution (typically a university or similar institution) that the person attended or attends, or that awarded the degree.

Example

Input Text	(Virginia Beach, VA) - Ronald Liu has been named Chief Financial Officer (CFO) of AMSEC LLC, a Virginia Beach-based company that provides a broad range of engineering and technical services to the Department of Defense, various governmental agencies and the commercial maritime industry. Mr. Liu was awarded a juris doctor degree by Northrop University School of Law and is licensed to practice law in the state of California. He is a certified public accountant and a graduate of The Anderson School of Management at the University of California at Los Angeles. Mr. Liu graduated summa cum laude from UCLA with a B.A. in Economics in 1974.
Extracted Relation	person = Ronald Liu certification = Certified Public Accountant
Extracted Relation	person = Ronald Liu degree = Juris Doctor (J.D.) schoolororganization = Northrup University School of Law
Extracted Relation	person = Ronald Liu degree = Bachelor of Arts schoolororganization = UCLA

PersonEmailAddress

em/r/PersonEmailAddress

Definition

A person’s e-mail address.

Attributes

emailaddress: The e-mail address.

person: The person who is the subject of this extracted relation.

Example

Input Text

Wells Fargo & Company Shareholder inquiries, 1-800-730-6001 Financial advisor inquiries, 1-888-877-9275 or Media contact: John Roehm, 415-222-5338 john.o.roehm@wellsfargo.com

Extracted Relation

person = John Roehm

emailaddress = john.o.roehm@wellsfargo.com

PersonLocation

em/r/PersonLocation

Definition

A person’s location. The person may be visiting the location, or residing there permanently or temporarily.

Attributes

locationstring: Location as mentioned in the text.

person: The person who is the subject of this extracted relation.

persongroup: A phrase that describes the participants in the extracted relation. For example, “twenty five Americans,” “thousands of soldiers,” “officials from the Ministry of Transportation,” “the ambassador to Pakistan.”

Example

Input Text

President Hamid Karzai met with reporters in Kabul.

Extracted Relation

persongroup = Hamid Karzai

location = Kabul

Extracted Relation

person = reporters

location = Kabul

PersonParty

em/r/PersonParty

Definition

The political party affiliation of a person.

Attributes

party: The political party with which the person is associated. May be a full party name or a generic name such as Liberals, or Democrats.

partydescription: A phrase from the text that describes the political party.

person: The person belonging to or associated with a political party.

Example 1

Input Text

Republican candidates, including Rick Santorum and Newt Gingrich, attend a 2012 Primary Gala event, just days before votes are cast in the influential New Hampshire primary.

Extracted Relation

person = Rick Santorum

party = Republican

Extracted Relation

person = Newt Gingrich

party = Republican

Example 2

Input Text

French President Nicolas Sarkozy clashes with Socialist former prime minister Laurent Fabius in a first face-to-face debate.

Extracted Relation

person = Laurent Fabius

party = Socialist

Example 3

Input Text

Communist Party of Nepal (Maoist) -- founded in 1994 by Pushpa Kamal Dahal, widely known as Prachanda.

Extracted Relation

person = Pushpa Kamal Dahal

party = Communist Party of Nepal

partydescription = Maoist

PersonRelation

em/r/PersonRelation

Definition

A business, friend, academic, military-service, political, or romantic relationship between people.

Attributes

datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text.

person: The person who is a participant in this extracted relation.

personalrelationtype: Classifies the relationship between two people.

Possible values:

Business
Friends
Academic
Military_Service
Politics
Romantic

Example

Input Text

Mark McDonald and Ray Odierno served together in combat in 2008, and McDonald said Odierno was an "absolute joy" to work with.

Extracted Relation

person = Mark McDonald

person = Ray Odierno

datestring = in 2008

date = 2008-00-00

personrelationtype = Military_service

PersonTravel

em/r/PersonTravel

Definition

A mention of a person travelling, past, present, or future.

Attributes

datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text.

locationdestination: The travel destination.

locationorigin: The place of travel origin.

person: The person who is the subject of this extracted relation.

status: The status, based on an indication in the text.

Example

Input Text

In another sign of easing tensions, Indian officials said that Prime Minister Atal Bihari Vajpayee might visit Pakistan early next year for a regional summit.

Extracted Relation

person = Atal Bihari Vajpayee

locationdestination = Pakistan

datestring = early next year

status = possible_future

PoliticalEndorsement

em/r/PoliticalEndorsement
Definition	An endorsement of a political entity (person, party, lobby) by another political entity.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). groupendorsee: The political entity (i.e. person, party, lobby) that received the endorsement. groupendorser: The political entity (i.e. person, party, lobby) that made the endorsement.
Equivalent Topic Tag	M:N Government/Politics You can also use this Intelligent Tagging classification topic tag to identify documents about Political Endorsement.

Example

Input Text

Obama wins endorsement of government employees union.

Extracted Relation

groupendorsee = Barack Obama

groupendorser = government employees union

PoliticalRelationship

em/r/PoliticalRelationship
Definition	A relationship (alliance or rivalry) between two political entities (people, political parties, lobbies). Note: Endorsements are not extracted as PoliticalRelationships.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). politicalentity1: First political entity of the relationship. politicalentity2: Second political entity in the relationship. politicalrelationshiptype: Classifies the relationship. Possible values: Alliance Rivalry
Equivalent Topic Tag	M:N Government/Politics You can also use this Intelligent Tagging classification topic tag to identify documents about Political Relationships.

Example

Input Text

The Democrats' expected coalition partners include the Liberal Democratic Party, whose leader Cedomir Jovanovic continued to his second term.

Extracted Relation

politicalentity = Democrats

politicalentity = Liberal Democratic

politicalrelationshiptype = Alliance

PollsResult

em/r/PollsResult
Definition	Indication of voting results as determined by a poll.
Attributes	date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). location: Location related to the extracted relation (address or other physical description). opponent: The opposing organization or party. politicalevent: The political event. pollconductor: Name of the company/organization conducting the poll. winningcandidate: The name of the winning person or political party.
Equivalent Topic Tags	M:G Elections/Voting; M:RA Presidential Elections; M:RB Upper House Elections; M:RC Lower House Elections; M:RD Gubernatorial Elections You can also use these Intelligent Tagging classification topic tags to identify documents about Poll Results.

Example

Input Text

(AP) Exit polls showed conservative challenger Angela Merkel's party leading in German parliamentary elections Sunday.

Extracted Relation

politicalevent = parliamentary elections

winningcandidate = Angela Merkel

pollconductor = AP

datestring = Sunday

location = German

ProductIssues

em/r/ProductIssues
Definition	An issue or problem with a product. This tag may be generated if product issues are mentioned in association with a company, even if no specific product is mentioned.
Attributes	company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. product: The product that is the subject of the extracted relation. producttype: Product classification. Possible values: aircraft car drug electronics weapon other status: The status, based on an indication in the text.
Equivalent Topic Tag	E:40 Products You can also use this Intelligent Tagging classification topic tag to identify documents about Product Issues.

Example 1

Input Text

Mazda has recalled more than 60,000 cars throughout the U.S. and Canada due to a wiring problem that can impede power-assisted steering, as well as the passenger airbag.

Extracted Relation

product = cars

producttype = Other

company = Mazda Motor Corporation

status = known

Example 2

Input Text

Toyota is recalling about 533,000 of its Sequoia sports cars.

Extracted Relation

product = Sequoia

producttype = car

company = Toyota Motor Corporation

status = announced

ProductRecall

em/r/ProductRecall
Definition	A product recall.
Attributes	company_distributor: Company that distributes the product, or the name of the manufacturer’s agent. company_recalling: Company that is recalling a product. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). product: The product that is the subject of the extracted relation. producttype: Product classification. Possible values: aircraft car drug electronics weapon other recalledquantity: The recalled quantity. status: The status, based on an indication in the text.
Equivalent Topic Tag	E:66 Product Recalls You can also use this Intelligent Tagging classification topic tag to identify documents about Product Recalls.

Example 1

Input Text

Toyota is recalling about 533,000 of its Sequoia sports cars.

Extracted Relation

company_recalling = Toyota Motor Corporation

product = Sequoia

producttype = Car

recalledquantity = about 533,000

status = announced

Example 2

Input Text

Samsung will use the lithium-ion batteries made by Sony in its smart phones to avoid a repeat of the unprecedented recall of its devices in 2016 over spontaneous combustions of its Note 7 smart phones.

Extracted Relation

product = devices

producttype = Other

date = 2016-00-00

datestring = in 2016

status = known

ProductRelease

em/r/ProductRelease
Definition	A reference to a product release.
Attributes	company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). product: The product that is the subject of the extracted relation. producttype: Product classification. Possible values: aircraft car drug electronics weapon other status: The status, based on an indication in the text.
Equivalent Topic Tag	E:6Q Product Launch You can also use this Intelligent Tagging classification topic tag to identify documents about Product Releases.

Example

Input Text

Samsung's first opportunity to recover from the Galaxy Note 7 debacle will come with the launch of its new flagship phone, the Galaxy S8, in the spring of 2017. While we don't know too much about the device, Samsung has announced that it will include a new digital assistant powered by Viv, the artificial intelligence startup Samsung bought earlier this year. Oh, and Viv was founded by the same people who built Siri.

Extracted Relation

company = Samsung Electronics Co., Ltd

product = Galaxy S8

producttype = Other

status = known

Quotation

em/r/Quotation

Definition

A direct or indirect quotation.

Attributes

datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text.

quotation: Direct or indirect quotation, taken directly from the text.

quotationtype: Classifies the quotation as direct or indirect.

Possible values:

Primary (the mention appears within quotation marks)
Paraphrase (the mention does not appear within quotation marks)

speaker: The person, company, organization, or country being quoted.

Example

Input Text

"It's going to have a real impact on the level of services," acting Treasurer David Rousseau said at a briefing for reporters.

Extracted Relation

speaker = David Rousseau

quotation = It's going to have a real impact on the level of services

quotationtype = Primary

SecondaryIssuance

em/r/SecondaryIssuance
Definition	Stock issuance by a company that is not an initial issuance. (Initial issuance of stock generates the em/r/IPO tag.)
Attributes	company: The company that is the subject of the relation type. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). status: The status, based on an indication in the text.
Equivalent Topic Tag	E:5C Secondary Share Offerings/Issues; E:1J Rights Issue You can also use these Intelligent Tagging classification topic tags to identify documents about Secondary Issuance.

Example

Input Text

Lumera Corporation plans to raise $17 mln from private offer.

Extracted Relation

company = Lumera Corporation

status = planned

StockSplit

em/r/StockSplit
Definition	A stock split declared by a company.
Attributes	company: The company that is the subject of the relation type. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). stocksplitratio: The ratio of the number of shares before the split to the number of shares after the split, if indicated in the text.
Equivalent Topic Tag	E:4N Share Splits; E:5B Bonus Shares Issues You can also use these Intelligent Tagging classification topic tags to identify documents about Stock Splits.

Example 1

Input Text

The Board of Directors of Curtiss-Wright Corporation (NYSE: CW; CW.B) today declared a two-for-one split of the company's Common Stock.

Extracted Relation

company = Curtiss-Wright Corporation

stocksplitratio = 2:1

Example 2

Input Text	Poniard Pharmaceuticals to Seek Shareholder Approval for Reverse Stock Split
Extracted Relation	company = Poniard Pharmaceuticals

Trial

em/r/Trial
Definition	A trial of one or more people.
Attributes	charge: Phrase describing an offense/charge. In case of multiple charge attributes, the first charge attribute is the primary charge. date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). othercharges: Additional charges. person: The person who is the subject of this extracted relation.
Equivalent Topic Tag	M:EV Judicial Process/Court Cases/Court Decisions You can also use this Intelligent Tagging classification topic tag to identify documents about Trials.

Example 1

Input Text

Judge Brinkema announced the schedule for the penalty phase of Zacarias Moussaoui's trial Friday.

Extracted Relation

person = Zacarias Moussaoui

datestring = Friday

Example 2

Input Text

Jon Horvath, a former SAC Capital Advisors LP analyst, named in court Tuesday two people who he said were sources of illegal stock tips at technology companies. Testifying in the insider-trading trial of SAC portfolio manager Michael Steinberg, Mr. Horvath said that he had a source at Sun Microsystems Inc. and another source at Ingram Micro Inc., and that both provided important confidential information about the performance of their companies.

Extracted Relation

person = Jon Horvath

VotingResult

em/r/VotingResult
Definition	Information about the results of a vote.
Attributes	date: An absolute date (in the YYYY-MM-DD format) that is related to the extracted relation. This date is either mentioned in the text, or calculated based on the datestring attribute value and relative to the text publication date. datestring: A mention of a date, or a reference to a date (e.g. yesterday, next month) relating to the extracted relation, as it appears in the text. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). location: Location related to the extracted relation (address or other physical description). opponent: The opposing organization or party. politicalevent: The political event. winningcandidate: The name of the winning person or political party.
Equivalent Topic Tag	M:G Elections/Voting; M:RA Presidential Elections; M:RB Upper House Elections; M:RC Lower House Elections; M:RD Gubernatorial Elections You can also use these Intelligent Tagging classification topic tags to identify documents about Voting Results.

Example

Input Text

Boris Johnson defeated Ken Livingstone in the London mayoral election.

Extracted Relation

politicalevent = mayoral election

winningcandidate = Boris Johnson

opponent = Ken Livingstone

location = London, Greater London, United Kingdom

RelevanceInfo Tag

For a conceptual explanation of this tag, see How Does Intelligent Tagging Work?

RelevanceInfo
Definition	Indicates how centric the associated entity is to the document.
Attributes	docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. subject: A hash tag generated by Intelligent Tagging. This ID points to the associated extracted entity. In the JSON output format, because related tags are nested, the docid and subject attributes are not displayed in the RelevanceInfo tag. relevance: The relevance score. Currently, the following values are supported: 1.0 – This value is reserved for the company identified as the reporting company in a document with a predefined format such as an SEC report. 0.8 – Entities defined as having high relevance receive this score. For example: a company that is mentioned in the title; a company mentioned prominently or frequently in the document. 0.5, 0.2 –These scores are assigned to entities that are not highly relevant to the story. For example: a company that is mentioned only once towards the end of a long document; a company mentioned in passing; a company mentioned as a representing law firm, financial advisor, underwriter, etc. Note: If the tagging output contains multiple companies with relevance 0.5, or 0.2, it may be an indication that the input document is made up of multiple paragraphs or sections, where each one discusses a single company in detail (for example, Top News, or Breaking News type stories), or that the input document includes charts or tables that display multiple companies. In this case, your use case determines if the company tag is of interest. 0.0 – This value is reserved for entities identified as irrelevant to the story. For example, mentions of companies as rating agencies, reporting agencies, stock exchanges, and social applications receive a score of 0. relevancecont: A more granular relevance score. This attribute is relevant only to Company entities. This attribute is not relevant to JSON output. A score that is 0.8 and above indicates a high relevance. A score of 0 indicates zero relevance. All other scores indicate not high relevance. For bothrelevance and relevancecont scores, the three following buckets are the most indicative ones: A score equal to or greater than 0.8 indicates high relevance. A score greater than zero and less than 0.8 indicates that the entity is not highly relevant to the story. A score of 0 indicates zero relevance. Recommendations for Using Relevance Scores Related to Company Tags For the highest accuracy, we recommend ranking companies based on the relevance score and not on therelevancecont score. A score of 0.2 should be considered the same as a score of 0.5. If your use case requires high precision, you can filter for High Relevance company tags only. If you are monitoring a specific company and do not want to miss any hints/indications, you should not filter company tags by relevance score.
Example	RelevanceInfo tag in RDF output format: <rdf:Description rdf:about="http://d.opencalais.com/dochash-1/7586b818-40af-3d55-ac16-bf3520cddfa6/Relevance/47"> <rdf:type rdf:resource="http://s.opencalais.com/1/type/sys/RelevanceInfo"/> <c:docId rdf:resource="http://d.opencalais.com/dochash-1/7586b818-40af-3d55-ac16-bf3520cddfa6"/> <c:subject rdf:resource="http://d.opencalais.com/comphash-1/705cd5cf-93e1-323c-8d4e-1ea3200d37e4"/> <c:relevance>0.8</c:relevance> <c:relevancecont>0.72</c:relevance> </rdf:Description> JSON output format: "relevance": 0.8 },

Confidence Tag

For a conceptual explanation of this tag, see the section about Confidence Scoring in How Does Intelligent Tagging Work?

Currently, Intelligent Tagging generates Confidence tags for the following entity and relation types: Company, Person, Pharmaceutical Drug, Bankruptcy, Deal, IPO.

The consuming application can achieve higher accuracy results by ignoring instances with confidence scores below a specified level. Note, however, that raising the specified level boosts Precision at the expense of Recall.

Note: Both the Confidence tag and the entity markup tag itself contain the confidence score. So you can retrieve the (same) confidence score from either tag.

Confidence
Definition	Indicates the probability that the associated e.g. extracted person or company is indeed a person or company.
Attributes	aggregate: The confidence score. The confidencelevel attribute of the related entity markup tag has the same value. So you can retrieve the confidence score from either tag. docid: The unique ID of the containing document, generated by Intelligent Tagging, and formatted as a URI. subject: A hash tag generated by Intelligent Tagging. This ID points to the associated extracted entity or relation. The following attributes are for internal use by Intelligent Tagging: dblookup: A value used internally by Intelligent Tagging to calculate the confidence score. resolution: A value used internally by Intelligent Tagging to calculate the confidence score. statisticalfeature: A value used internally by Intelligent Tagging to calculate the confidence score. *Note*: In the JSON output format, because related tags are nested, the docid and subject attributes are not displayed in the Confidence tag.
Example	RDF output format Note: The "aggregate" attribute exposes the confidence score. <rdf:Description rdf:about="http://d.opencalais.com/conf/comphash-1/705cd5cf-93e1-323c-8d4e-1ea3200d37e4"> <rdf:type rdf:resource="http://s.opencalais.com/1/type/tag/Confidence"/> <c:docId rdf:resource="http://d.opencalais.com/dochash-1/7586b818-40af-3d55-ac16-bf3520cddfa6"/> <!--Apple--> <c:subject rdf:resource="http://d.opencalais.com/comphash-1/705cd5cf-93e1-323c-8d4e-1ea3200d37e4"/> <c:statisticalfeature>0.997</c:statisticalfeature> <c:dblookup>0.0</c:dblookup> <c:resolution>0.9928677</c:resolution> <c:aggregate>0.996</c:aggregate> </rdf:Description> JSON output format: "confidence": { "statisticalfeature": "0.997", "dblookup": "0.0", "resolution": "0.9928677", "aggregate": "0.996" },

Disambiguation Tags

Currently, Intelligent Tagging supports Disambiguation for the following entity and relation types:

TopmostPublicParentCompany

Click a link for a description of the tag and its attributes.

Important: Most attributes are optional; a tag can be extracted with some but not all of its attributes.

For a conceptual explanation of this tag, see How Does Intelligent Tagging Work?

Note to Internal Intelligent Tagging users: Please see the Supplementary Guide for Internal Intelligent Tagging (for internal customers who do not connect through API Gateway) for a list of supported disambiguation tags, and for details about the differences in the tagging output.

City

er/Geo/City
Definition	This tag is generated when an extracted entity of the type City is successfully mapped to a city in the RCS (Refinitiv Classification Services) taxonomy.
Attributes	containedbycountry: The name of the country in which the geographical entity is located. containedbystate: The name of the state in which the geographical entity is located. docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. latitude: The latitude of the geographical entity, as defined in Freebase. longitude: The longitude of the geographical entity, as defined in Freebase. name: The official name as it appears in the Refinitiv dataset. permid: The Refinitiv unique ID (PermID) for this city. rcscode: The code of this geographical entity, as defined in the RCS (Refinitiv Classification Services) taxonomy. shortname: A name commonly used to refer to the geographical entity, as defined in Freebase. subject: The hash tag generated by Intelligent Tagging. This ID points to the corresponding extracted entity and related tags within this document. Note that this attribute is not relevant to the JSON output format.
Related Tag	em/e/City

Company

Note to Internal Intelligent Tagging users: Internal Intelligent Tagging output exposes different attributes for the er/Company tag. See the Supplementary Guide for Internal Intelligent Tagging (for internal customers who do not connect through API Gateway).

er/Company
Definition	This tag is generated when an extracted entity of the type Company is mapped to a company in the Refinitiv dataset. The dataset is dynamic, and includes companies that are involved in financial transactions as well as many other companies that are important to investors or to tax and legal professionals. Note that the openpermid attribute gives you direct access to high quality, curated Refinitiv company data. (In JSON output, the attribute name is id.)
Attributes	commonname: The commonly used name, as it appears in the Refinitiv dataset. For example, "BMW" is the common name for the company Bayerische Motoren Werke Aktiensgesellschaft (Bavarian Motor Works Corporation). docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. id: See the openpermid attribute description, below. ispublic: Indicates whether the company is public (true) or private (false). legacyid: For Intelligent Tagging users, this attribute is for internal use only. For Open Calais users: This is the link to the company as it would have been displayed in the legacy Open Calais er/company tag. This attribute is relevant to users who upgraded from the legacy Intelligent Tagging to the new Open Calais. The purpose of this attribute is to enable you to map the old links, which are no longer supported, to the corresponding new links. This attribute is not relevant to the JSON output format. An example of an old company link in the legacy Intelligent Tagging er/company tag (RDF output): <rdf:Description rdf:about="http://d.opencalais.com/er/company/ralg-tr1r/39d9a86f-6241-3688-b05b-4595eeaba2d1"> The company link and the legacyid attribute as they appear in the new Intelligent Tagging er/company tag (RDF output): <rdf:Description rdf:about="http://d.opencalais.com/er/company/ralg-oa/5000608903"> <c:legacyid rdf:resource="http://d.opencalais.com/er/company/ralg-tr1r/39d9a86f-6241-3688-b05b-4595eeaba2d1"/> The company link and the legacyid attribute as they appear in the new Intelligent Tagging er/company tag (N3 output): <http://d.opencalais.com/er/company/ralg-oa/5000608903> c:legacyid < http://d.opencalais.com/er/company/ralg-tr1r/39d9a86f-6241-3688-b05b-4595eeaba2d1> ; name: The official name as it appears in the Refinitiv dataset. openpermid:A direct link to the relevant company page on the Open PermID website (https://permid.org). This link gives you direct access to high quality, curated Refinitiv company data. (In the JSON output, the attribute name isid.) permid: The Refinitiv unique ID (PermID) for this company. primaryric: The company’s primary Reuters Instrument code (RIC). score: A score representing the probability that the extracted company was accurately mapped to the corresponding company in the Refinitiv dataset. Scores range from 0 - 1; the higher the score, the higher the probability. The consuming application can use this score to achieve higher accuracy results. subject: The hash tag generated by Intelligent Tagging. This ID points to the corresponding extracted entity and related tags within this document. Note that this attribute is not relevant to the JSON output format. The tag may contain multiple subject attributes if the same company was extracted multiple times, and recognized as a single company on resolution. ticker: The company stock symbol.
Related Tags	em/e/Company er/TopmostPublicParentCompany

Continent

er/Continent
Definition	This tag is generated when an extracted entity of the type Continent is successfully mapped to a continent in the RCS (Refinitiv Classification Services) taxonomy.
Attributes	docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. name: The official name as it appears in the Refinitiv dataset. permid: The Refinitiv unique ID (PermID) for this geographical entity. rcscode: The code of this geographical entity, as defined in the RCS (Refinitiv Classification Services) taxonomy. subject: The hash tag generated by Intelligent Tagging. This ID points to the corresponding extracted entity and related tags within this document. Note that this attribute is not relevant to the JSON output format.
Related Tag	em/e/Continent

Country

er/Geo/Country
Definition	This tag is generated when an extracted entity of the type Country is successfully mapped to a country in the RCS (Refinitiv Classification Services) taxonomy. .
Attributes	docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. latitude: The latitude of the geographical entity, as defined in Freebase. longitude: The longitude of the geographical entity, as defined in Freebase. name: The official name as it appears in the Refinitiv dataset. permid: The Refinitiv unique ID (PermID) for this country. rcscode: The code of this geographical entity, as defined in the RCS (Refinitiv Classification Services) taxonomy. shortname: A name commonly used to refer to the geographical entity, as defined in Freebase. subject: The hash tag generated by Intelligent Tagging. This ID points to the corresponding extracted entity and related tags within this document. Note that this attribute is not relevant to the JSON output format.
Related Tag	em/e/Country

Deal

This is a premium metadata type.

Note to Internal Intelligent Tagging users: Internal Intelligent Tagging output exposes additional attributes for the er/Deal tag. See the Supplementary Guide for Internal Intelligent Tagging (for internal customers who do not connect through API Gateway).

er/Deal
Definition	This tag is generated when an extracted relation of the type Deal is successfully mapped to a deal in the Refinitiv Deals Authority dataset.
Attributes	dealsdcnum: The deal ID as defined in the Refinitiv Deals Authority dataset. docid: The unique ID of the containing document, generated byIntelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. permid: The Refinitiv unique ID (PermID) for this deal. subject: The hash tag generated byIntelligent Tagging. This ID points to the corresponding extracted entity and related tags within this document. Note that this attribute is not relevant to the JSON output format.
Related Tag	em/r/Deal

Organization

er/Organization
Definition	This tag is generated when an extracted entity of the type Organization is successfully mapped to an organization in the Refinitiv Organization Authority dataset.
Attributes	docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. name: The official name as it appears in the Refinitiv dataset. permid: The Refinitiv unique ID (PermID) for this organization. To link directly to related, high-quality, curated Refinitiv data on the Open PermID website, add this prefix to the permid value: “https://permid.org/1-“ For example: https://permid.org/1-8589934571 rcscode: The organization code, as defined in the RCS (Refinitiv Classification Services) taxonomy. subject: The hash tag generated by Intelligent Tagging. This ID points to the corresponding extracted entity and related tags within this document. Note that this attribute is not relevant to the JSON output format.
Related Tag	em/e/Organization

Person

er/Person
Definition	This tag is generated when an extracted entity of the type Person is successfully mapped to a person in the Refinitiv People Authority dataset. Currently this tag is only generated for executives. Note to Internal Intelligent Tagging users: Internal Intelligent Tagging output exposes slightly different attributes for the er/Person tag. See the Supplementary Guide for Internal Intelligent Tagging (for internal customers who do not connect through API Gateway).
Attributes	commonname: The commonly used name, as it appears in the Refinitiv dataset. docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. name: The official name as it appears in the Refinitiv dataset. officerid: A unique ID (as defined in the Refinitiv Officers and Directors dataset) that specifies a position held (currently or in the past) by the person, in a particular company. The er/person tag may include multiple officerid attributes. openpermid: A direct link to the relevant person page on the Open PermID website (https://permid.org). This link gives you direct access to high quality, curated Refinitiv people data. (In the JSON output, the attribute name is id.) paid: The Refinitiv unique ID (from the People Authority) for this person. personid: The Person ID as defined in the Refinitiv Officers and Directors dataset. score: A score representing the probability that the extracted person was accurately mapped to the corresponding person in the Refinitiv dataset. Scores range from 0 - 1; the higher the score, the higher the probability. The consuming application can use this score to achieve higher accuracy results. subject: The hash tag generated by Intelligent Tagging. This ID points to the corresponding extracted entity and related tags within this document. Note that this attribute is not relevant to the JSON output format.
Related Tag	em/e/Person

Product

er/Product
Definition	This tag is generated when an extracted entity of the type Product is successfully mapped to a product in the Refinitiv dataset.
Attributes	docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. name: The official name as it appears in the Refinitiv dataset. ownername: The name of the organization that owns the product. The organization name is as it appears in the Refinitiv dataset. ownerpermid: The permid of the organization that owns the product. subject: The hash tag generated by Intelligent Tagging. This ID points to the corresponding extracted entity and related tags within this document. Note that this attribute is not relevant to the JSON output format.
Related Tag	em/e/Product

ProvinceOrState

er/Geo/ProvinceOrState
Definition	This tag is generated when an extracted entity of the type ProvinceOrState is successfully mapped to a province or state in the RCS (Refinitiv Classification Services) taxonomy.
Attributes	containedbycountry: The name of the country in which the geographical entity is located. containedbystate: The name of the state in which the geographical entity is located. docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. latitude: The latitude of the geographical entity, as defined in Freebase. longitude: The longitude of the geographical entity, as defined in Freebase. name: The official name as it appears in the Refinitiv dataset. permid: The Refinitiv unique ID (PermID) for this province or state. rcscode: The code of this geographical entity, as defined in the RCS (Refinitiv Classification Services) taxonomy. shortname: A name commonly used to refer to the geographical entity, as defined in Freebase. subject: The hash tag generated by Intelligent Tagging. This ID points to the corresponding extracted entity and related tags within this document. Note that this attribute is not relevant to the JSON output format.
Related Tag	em/e/ProvinceOrState

Region

er/Region
Definition	This tag is generated when an extracted entity of the type Region is successfully mapped to a region in the RCS (Refinitiv Classification Services) taxonomy.
Attributes	docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. name: The official name as it appears in the Refinitiv dataset. permid: The Refinitiv unique ID (PermID) of the extracted region. rcscode: The code of this region, as defined in the RCS (Refinitiv Classification Services) taxonomy. subject: The hash tag generated by Intelligent Tagging. This ID points to the corresponding extracted entity and related tags within this document. Note that this attribute is not relevant to the JSON output format.
Related Tag	em/e/Region

TopmostPublicParentCompany

This is a premium metadata type.

Note to Internal Intelligent Tagging users: Internal Intelligent Tagging output exposes different attributes for the er/TopmostPublicParentCompany tag. For details, see the Supplementary Guide for Internal Intelligent Tagging (for internal customers who do not connect through API Gateway)

er/TopmostPublicParentCompany
Definition	This tag is output when a private company that is extracted and resolved to the Refinitiv dataset has a publicly traded parent company. Note that the openpermid attribute gives you direct access to high quality, curated Refinitiv company data. (In JSON output, the attribute name is id.)
Attributes	commonname: The commonly used name, as it appears in the Refinitiv dataset. docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. id: See the openpermid attribute description, below. ispublic: Indicates whether the company is public (true) or private (false). legacyid: For Intelligent Tagging, this attribute is for internal use only. For Open Calais: The link to the company as it would have been displayed in the legacy Open Calais output. This attribute is relevant to users who upgraded from the legacy Open Calais to the new Open Calais. The purpose of this attribute is to enable you to map the old links, which are no longer supported, to the corresponding new links. This attribute is not relevant to the JSON output format. name: The official name as it appears in the Refinitiv dataset. openpermid: A direct link to the relevant company page on the Open PermID website (https://permid.org). This link gives you direct access to high quality, curated Refinitiv company data. (In the JSON output, the attribute name isid.) permid: The Refinitiv unique ID (PermID) for this company. primaryric: The company’s primary Reuters Instrument code (RIC). subject: The hash tag generated by Intelligent Tagging and assigned to the extracted child company. This ID points to the relevant em/e/company and related tags within the document. If the parent company has multiple extracted child companies within the document, additional instances of the subject attribute appear in the tag. Note that this entity is not relevant to the JSON output format. ticker: The company stock symbol.
Related Tag	em/e/Company

SocialTag

This is one of the actively enhanced and supported metadata types.

For a conceptual explanation of this tag, see How Does Intelligent Tagging Work?

SocialTag
Definition	Classifies the document based on Wikipedia folksonomy.
Attributes	docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). importance: Indicates how centric the topic identified by the social tag is to the document as a whole. Possible values: 1(very centric); 2 (somewhat centric); 3 (less centric). name: A word or phrase that describes the document as a whole. (The title of the associated Wikipedia article.) originalvalue: The original title of the associated Wikipedia article, if the title has changed. socialTag: The unique identifier for the socialTag across documents.
Example	SocialTag in RDF output format: <rdf:Description rdf:about="http://d.opencalais.com/dochash-1/7586b818-40af-3d55-ac16-bf3520cddfa6/SocialTag/4"> <rdf:type rdf:resource="http://s.opencalais.com/1/type/tag/SocialTag"/> <c:docId rdf:resource="http://d.opencalais.com/dochash-1/7586b818-40af-3d55-ac16-bf3520cddfa6"/> <c:socialtag rdf:resource="http://d.opencalais.com/genericHasher-1/083d56d1-2fed-3063-b59a-963d4fdaee36"/> <c:forenduserdisplay>true</c:forenduserdisplay> <c:importance>2</c:importance> <c:originalValue>Electric car</c:originalValue> </rdf:Description> JSON output format: "http://d.opencalais.com/dochash-1/268b4e49-098d-389d-9d76-816cb722a630/SocialTag/4": { "_typeGroup": "socialTag", "id": "http://d.opencalais.com/dochash-1/268b4e49-098d-389d-9d76-816cb722a630/SocialTag/4", "socialTag": "http://d.opencalais.com/genericHasher-1/083d56d1-2fed-3063-b59a-963d4fdaee36", "forenduserdisplay": "true", "name": "Electric car", "importance": "2", "originalValue": "Electric car" },

Topic Tag (DocCat)

This is one of the actively enhanced and supported metadata types.

Note to Internal Intelligent Tagging users: Internal Intelligent Tagging output exposes different attributes for the DocCat (topic) tag. For details, see the Supplementary Guide for Internal Intelligent Tagging (for internal customers who do not connect through API Gateway).

For a conceptual explanation of this tag, see How Does Intelligent Tagging Work?

DocCat
Definition	Identifies the topics discussed in the document. The reference list of topics is drawn from the RCS (Refinitiv Classification Services) taxonomy and the IPTC (International Press Telecommunications Council) news taxonomy. RCS topic classification is a premium feature.
Attributes	docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The topic name. This attribute is relevant to RCS topics only. permid: The Refinitiv unique ID (PermID) for this topic. rcscode: The topic code, as defined in the RCS (Refinitiv Classification Services) taxonomy. This attribute is relevant to RCS topics only. score: A confidence score on a scale of 0 to 1. The value indicates the probability that the topic is indeed discussed in the text and also how centric the topic is to the text. The higher the value, the higher the probability. The consuming application can use this score to achieve higher accuracy results by ignoring instances with scores below a specified level. shortName: Output for TRCS topics only. If the name attribute includes the string “(TRBC),” then the shortName attribute is the same, minus “(TRBC).” If the name attribute does not include the “(TRBC)” string, then the name and shortName attributes are identical. For example, if the topic name is “Technology Equipment (TRBC),” then the shortName is “Technology Equipment.”
Example	An IPTC taxonomy topic tag in RDF output format: <rdf:Description rdf:about="http://d.opencalais.com/dochash-1/7586b818-40af-3d55-ac16-bf3520cddfa6/cat/1"> <rdf:type rdf:resource="http://s.opencalais.com/1/type/cat/DocCat"/> <c:docId rdf:resource="http://d.opencalais.com/dochash-1/7586b818-40af-3d55-ac16-bf3520cddfa6"/> <c:forenduserdisplay>false</c:forenduserdisplay> <c:score>0.988</c:score> <c:name>Technology_Internet</c:name> </rdf:Description> IPTC taxonomy topic tag in JSON output format: "http://d.opencalais.com/dochash-1/268b4e49-098d-389d-9d76-816cb722a630/cat/1": { "_typeGroup": "topics", "forenduserdisplay": "false", "score": 0.988, "name": "Technology_Internet" }, A TRCS taxonomy topic tag in RDF output format: <rdf:Description rdf:about="http://d.opencalais.com/dochash-1/3c07707e-c5b7-367d-8cb6-48b06b1adcae/cat/4"> <rdf:type rdf:resource="http://s.opencalais.com/1/type/cat/DocCat"/> <c:docId rdf:resource="http://d.opencalais.com/dochash-1/3c07707e-c5b7-367d-8cb6-48b06b1adcae"/> <c:forenduserdisplay>true</c:forenduserdisplay> <c:rcscode>M:1YL</c:rcscode> <c:name>Industry Risk</c:name> <c:permid>1003952436</c:permid> <c:score>0.964</c:score> </rdf:Description> TRCS taxonomy topic tag in JSON output format: "http://d.opencalais.com/dochash-1/cb004505-4b12-339a-a722-70b71089a68c/cat/6": { "_typeGroup": "topics", "forenduserdisplay": "true", "rcscode": "M:1YN", "name": "Macroeconomics", "permid": "1003952440", "score": 0.945

Slugline Tag

This is one of the actively enhanced and supported metadata types.

This is a premium metadata type.

For a conceptual explanation of this tag, see How Does Intelligent Tagging Work?

Slugline
Definition	A Reuters slug line that describes what the document is about as a whole. To trigger slugline tagging, use the x-calais-selectiveTags:slugline header value. When slugline tagging is enabled, each Intelligent Tagging output document may contain up to 8 slugline tags.
Attributes	forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false).All slugline tags are assigned the value, true slugline: The slug line assigned to the document by Intelligent Tagging. isactive: Currently not supported. creationdate: Currently not supported. slugid: Currently not supported. confidencelevel: A confidence score on a scale of 0 to 1. The value represents the probability that the document is indeed about the event described by the assigned slug line. The higher the value, the higher the probability. The consuming application can use this score to achieve higher accuracy results by ignoring documents with slugline tags with confidence scores below a specified level. Note that boosting Precision in this manner is at the expense of Recall.
Example	Slugline tag in JSON output format: "http://d.opencalais.com/dochash-1/0091c2a4-0d9f-3545-b6e0-1b0a8a7139d9/Slugline/3": { "_typeGroup": "sluglines", "id": "http://d.opencalais.com/dochash-1/0091c2a4-0d9f-3545-b6e0-1b0a8a7139d9/Slugline/3", "forenduserdisplay": "true", "slugline": "UNITED RENTALS-RESULTS/PREVIEW", "isactive": "true", "creationdate": "2017-01-01T00:00:00.000Z", "slugid": "sluglines/UNITEDnbspRENTALS-RESULTS/PREVIEW", "confidencelevel": "0.5042" Slugline tag in RDF output format: <rdf:Description rdf:about="http://d.opencalais.com/dochash-1/2771afb0-3124-3a21-ac80-ba4a94fbc8d3/Slugline/1"> <rdf:type rdf:resource="http://s.opencalais.com/1/type/tag/Slugline"/> <c:forenduserdisplay>true</c:forenduserdisplay> <c:slugline>BRITAIN-EU/BANKS</c:slugline> <c:isactive>true</c:isactive> <c:creationdate>2017-01-01T00:00:00.000Z</c:creationdate> <c:slugid>sluglines/BRITAIN-EU/BANKS</c:slugid> <c:confidencelevel>0.2617</c:confidencelevel> </rdf:Description>

Industry Tag

For a conceptual explanation of this tag, see How Does Intelligent Tagging Work?

This is a premium metadata type.

Industry
Definition	Identifies the industries related to the companies mentioned in the document.
Attributes	docid: The unique ID of the containing document, generated by Intelligent Tagging and formatted as a URI. Note that this attribute is not relevant to the JSON output format. forenduserdisplay: A recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false). name: The name of the industry. permid: The Refinitiv unique ID (PermID) for this Industry. To link directly to related, high-quality, curated Refinitiv data on the Open PermID website, add this prefix to the permid value: “https://permid.org/1-“ For example: https://permid.org/1-4294951707 rcscode: The industry code, as defined in the RCS (Refinitiv Classification Services) taxonomy. relevance: Indicates how relevant the industry is to the story. Values range from 0 to 1. The higher the score, the higher the relevance. trbccode: The industry code, as defined in the Refinitiv Business Classification (TRBC) taxonomy.
Example	Industry tag in RDF output format: <rdf:Description rdf:about="http://d.opencalais.com/dochash-1/8f9c80e8-66fc-39a5-847c-517148a862ac/Industry/5"> <rdf:type rdf:resource="http://s.opencalais.com/1/type/tag/Industry"/> <c:docId rdf:resource="http://d.opencalais.com/dochash-1/8f9c80e8-66fc-39a5-847c-517148a862ac"/> <c:forenduserdisplay>false</c:forenduserdisplay> <c:name>Phones & Smart Phones</c:name> <c:rcscode>B:1769</c:rcscode> <c:trbccode>5710602011</c:trbccode> <c:permid>4294951232</c:permid> <c:relevance>0.800</c:relevance> </rdf:Description> Industry tag in JSON output format: "http://d.opencalais.com/dochash-1/8f9c80e8-66fc-39a5-847c-517148a862ac/Industry/5": { "_typeGroup": "industry", "forenduserdisplay": "false", "name": "Phones & Smart Phones", “trbccode”: "5710602011” ; "rcscode": "B:1769", "permid": "4294951707", "relevance": 0.8 },

ForEndUserDisplay Attribute

The forEndUserDisplay attribute is present in Topic (DocCat), Entity (em/e), Relation (em/r), Slugline tags, and SocialTags, and Industry tags. The attribute value (true or false) is a recommendation of whether the tag is suitable as a search item for a specific document (true) or whether the metadata is primarily of use for aggregation and analytics on large quantities of documents (false).

Some metadata types have a set forenduserdisplay value, while for other metadata types, the forenduserdisplay value is dynamic:

DocCat (Topic) tags– RCS topics optimized with (DIY) Self Service Classification - the forenduserdisplay value is always true.

SocialTags – the forenduserdisplay value is always true.

Slugline tags – the forenduserdisplay value is always true.

Entities and Relations - Some entity and relation metadata types have set forenduserdisplay values, while other entity and relation metadata types have dynamic forenduserdisplay values, determined per instance.

For example:

Every instance of the em/r/Dividend tag (the Dividend relation tag) defines forenduserdisplay as true.

Each instance of the em/e/Company tag is assigned a forenduserdisplay value based on a confidence score calculated for that specific company tag during extraction.

See ForEndUserDisplay attribute in Entity and Relation Tags for further information.

Generally speaking, the forenduserdisplay=true status is assigned to metadata types that consistently provide high precision (>80%) results based on our tests.

Continue reading this section for more detailed information and best practice recommendations for each metadata type.

Best Practice Recommendations

If your use case requires high precision, you can make use of the ForEndUserDisplay attribute to filter the tagging output according to your needs.

The ForEndUserDisplay value is not relevant (there is no need to filter out tags with ForEndUserDisplay =false value) to the following types of use cases:

Aggregation/graph use cases, or use cases where you can gain confidence by processing multiple examples.

For example, looking for trends over time.

Use cases where you may not want to lose hints/indications and can afford a higher rate of precision errors.

For example, monitoring a specific company, industry or country where you don’t want to miss anything.

ForEndUserDisplay Attribute in DocCat (Topic) Tags

Topics optimized with (DIY) Self Service Classification are assigned the forenduserdisplay=true value.

The DocCat tag also assigns a confidence score on a scale of 0 to 1. The value indicates the probability that the topic is indeed discussed in the text and is centric to the text. The higher the value, the higher the probability.

For DocCat (topic) tags, we suggest using the forenduserdisplay value together with the score value to filter the tagging output to suit your use case.

For example, if your use case requires a high level of precision, we suggest the following:

If the forenduserdisplay=true and the score is greater than 0.5, use the tag.

If the forenduserdisplay=false and the score is greater than 0.9, use the tag.

For a semi-automated editorial workflow, if the forenduserdisplay=false and the score is less than 0.9, we suggest marking the tags as maybe being about the topic.

For a fully automated workflow, we suggest ignoring tags with the forenduserdisplay=false and a score of less than 0.9.

Read about Topic Tags

ForEndUserDisplay Attribute in Social Tags

The forenduserdisplay value is always true for social tags.

ForEndUserDisplay Attribute in Entity and Relation Tags

For entities and relations, the forenduserdisplay value is determined as follows:

em/e/Company, em/e/Person, em/e/PharmaceuticalDrug, em/r/Bankruptcy, em/r/Deal, em/r/IPO: forenduserdisplay value is defined per tag instance, based on the confidencelevel value. For example, Intelligent Tagging determines and assigns the appropriate forenduserdisplay value to each instance of the em/e/Company tag in the tagging output.
All other entities and relations: Each metadata tag type has a set value (listed below).

For example, every em/r/Dividend tag is assigned the forenduserdisplay=true value.

As previously mentioned, the forenduserdisplay=true status is assigned to metadata types that consistently provide high precision (>80%) results.

forenduserdisplay Values: Entities

Entity Type	Forenduserdisplay Value
Anniversary	false
City	true
Company	Defined per tag instance, based on the confidencelevel value.
Continent	false
Country	true
Currency	true
CurrencyPair	true
Editor	true
EmailAddress	false
EntertainmentAwardEvent	false
Facility	false
FaxNumber	false
Holiday	false
IndustryTerm	false
Journalist	true
MarketIndex	true
MedicalCondition	false
MedicalTreatment	false
Movie	false
MusicAlbum	false
MusicGroup	false
NaturalFeature	false
OperatingSystem	false
Organization	false
Person	Defined per tag instance, based on the confidencelevel value.
PharmaceuticalDrug	Defined per tag instance, based on the confidencelevel value.
PhoneNumber	false
PoliticalEvent	false
Position	false
Product	false
ProgrammingLanguage	false
ProvinceOrState	false
PublishedMedium	false
RadioProgram	false
RadioStation	false
Region	false
SportsEvent	false
SportsGame	false
SportsLeague	false
Technology	false
TVShow	false
TVStation	false
URL	False

forenduserdisplay Values: Relations

Relation Type	Forenduserdisplay Value
Acquisition	true
Alliance	true
AnalystEarningsEstimate	false
AnalystRecommendation	false
ArmedAttack	true
ArmsPurchaseSale	false
Arrest	true
Bankruptcy	Defined per tag instance, based on the confidencelevel value.
BonusSharesIssuance	true
BusinessRelation	true
Buybacks	true
CandidatePosition	true
CompanyAccountingChange	true
CompanyAffiliates	true
CompanyCompetitor	true
CompanyCustomer	true
CompanyEarningsAnnouncement	true
CompanyEarningsGuidance	true
CompanyEmployeesNumber	true
CompanyExpansion	true
CompanyForceMajeure	true
CompanyFounded	true
CompanyInvestigation	true
CompanyInvestment	true
CompanyLaborIssues	true
CompanyLayoffs	true
CompanyLegalIssues	true
CompanyListingChange	true
CompanyLocation	true
CompanyMeeting	true
CompanyNameChange	true
CompanyProduct	true
CompanyReorganization	true
CompanyRestatement	true
CompanyTechnology	true
CompanyTicker	true
CompanyUsingProduct	true
ConferenceCall	true
ContactDetails	true
Conviction	true
CreditRating	true
Deal	Defined per tag instance, based on the confidencelevel value.
DebtFinancing	true
DelayedFiling	true
DiplomaticRelations	true
Dividend	true
EmploymentChange	true
EmploymentRelation	true
EnvironmentalIssue	true
EquityFinancing	true
Extinction	true
FamilyRelation	true
FDAPhase	true
IndicesChange	true
Indictment	true
IPO	Defined per tag instance, based on the confidencelevel value.
JointVenture	true
ManMadeDisaster	true
Merger	true
MilitaryAction	true
MovieRelease	true
MusicAlbumRelease	true
NaturalDisaster	true
PatentFiling	true
PatentIssuance	true
PersonAttributes	true
PersonCareer	true
PersonCommunication	true
PersonEducation	true
PersonEmailAddress	true
PersonLocation	true
PersonParty	true
PersonRelation	true
PersonTravel	true
PoliticalEndorsement	true
PoliticalRelationship	true
PollsResult	true
ProductIssues	true
ProductRecall	true
ProductRelease	true
Quotation	true
SecondaryIssuance	true
StockSplit	true
Trial	true
VotingResult	true

ForEndUserDisplay attribute in Slugline Tags

The forenduserdisplay value is always true for slugline tags.

(The Slugline tag is a premium metadata type.)

Recommendations

If your use case requires a high level of precision, we suggest using the forenduserdisplay value to filter the tagging output as follows:

forenduserdisplay=true: Use all the tags that define forenduserdisplay=true.

forenduserdisplay=false: For a semi-automated editorial workflow, we suggest marking these tags as maybe being about the entity or relation; for a fully automated workflow, we suggest ignoring these tags.

PermID Attribute

The PermID attribute is output by multiple Intelligent Tagging metadata types. A PermID is a Refinitiv unique ID which is applied consistently across all documents processed by Intelligent Tagging.

The PermID value identifies an entity type (for example, the Organization entity), or a particular entity (for example, Bank of Japan), depending on the metadata type.

PermID Attribute in Entity Markup Tags

PermID Attribute in Resolution and Industry Tags

PermID Attribute in DocCat (Topic) Tags

Linking to Open PermID

In Entity Markup Tags

With two exceptions (em/e/currency, and em/e/currencyPair), Entity Markup tags output a PermID attribute value that identifies the entity type.

The following Entity Markup tags output a PermID attribute value that identifies the entity type:

em/e/MarketIndex

em/e/Organization

em/e/Person

em/r/CompanyEarningsAnnouncement

em/r/Deal

em/r/IPO

For example, the PermID attribute value of the following em/e/Organization tag is the Refinitiv unique ID of the entity type (organization), and not of the organization itself (Bank of Japan).

JSON Output example:

    	
             "http://d.opencalais.com/genericHasher-1/c458a04c-6eb1-35df-b5ba-45b5ee4a97a4": {
        "_typeGroup": "entities",
        "_type": "Organization",
        "forenduserdisplay": "false",
        "name": "Bank of Japan",
        "organizationtype": "central bank",
        "nationality": "Japanese",
        "_typeReference": "http://s.opencalais.com/1/type/em/e/Organization",
        "permid": "https://permid.org/1-404010",

RDF Output Example:

    	
            <rdf:Description rdf:about="http://d.opencalais.com/genericHasher-1/c458a04c-6eb1-35df-b5ba-45b5ee4a97a4">
 
<rdf:type rdf:resource="http://s.opencalais.com/1/type/em/e/Organization"/>
 
<c:permid rdf:resource="https://permid.org/1-404010"/>
 
<c:forenduserdisplay>false</c:forenduserdisplay>
 
<c:name>Bank of Japan</c:name>
 
<c:organizationtype>central bank</c:organizationtype>
 
<c:nationality>Japanese</c:nationality>
 
</rdf:Description>

In fact, every em/e/Organization tag exposes the same PermID attribute value—the Refinitiv unique ID (PermID) for the Organization entity type. This enables linking all extractions of a particular metadata type, and can be used to build a knowledge graph.

The following Entity Markup tags output a PermID attribute value that identifies the extracted entity (for example Japanese Yen) and not the metadata type (currency):

em/e/Currency

em/e/CurrencyPair

For example, the PermID attribute value of the following em/e/Currency tag is the Refinitiv unique ID for the currency, Japanese Yen.

JSON Output Example:

    	
            "http://d.opencalais.com/genericHasher-1/dfc86e3b-662a-3cbb-8a9f-cb78040af075": {
"_typeGroup": "entities",
"_type": "Currency",
"forenduserdisplay": "true",
"name": "Japanese Yen",
"permid": "500186",
"rcscode": "C:5",
"ric": "JPY=",
"confidencelevel": "0.62978184",
"_typeReference": "http://s.opencalais.com/1/type/em/e/Currency",

RDF Output Example:

    	
            <rdf:Description rdf:about="http://d.opencalais.com/genericHasher-1/dfc86e3b-662a-3cbb-8a9f-cb78040af075"">
 
    <rdf:type rdf:resource="http://s.opencalais.com/1/type/em/e/Currency"/>
 
    <c:forenduserdisplay>true</c:forenduserdisplay>
 
    <c:name>Japanese Yen</c:name>
 
    <c:permid>500186</c:permid>
 
    <c:rcscode>C:5</c:rcscode>
 
    <c:ric>JPY=</c:ric>
 
    <c:confidencelevel>0.993723</c:confidencelevel>
 
</rdf:Description>

In Resolution and Industry Tags

In an entity resolution (disambiguation) tag or an Industry tag, the PermID attribute value identifies the specific entity or industry.

For example, the PermID attribute value in the following er/Organization tag is the Refinitiv unique ID for Bank of Japan.

JSON Output Example:

    	
             "resolutions": [
            {
                "rcscode": "M:6",
                "name": "Bank of Japan",
                "permid": "8589934442"
            }

RDF Output Example:

    	
            <rdf:Description rdf:about="http://d.opencalais.com/er/organization/comphash-1/221e4540-7b2b-3d89-87ff-fe9de1843d80">
 
<rdf:type rdf:resource="http://s.opencalais.com/1/type/er/Organization"/>
 
<c:docId rdf:resource="http://d.opencalais.com/dochash-1/93734939-2e1c-344b-8d0e-d8a3bb4568b9"/>
 
<c:rcscode>M:6</c:rcscode>
 
<c:name>Bank of Japan</c:name>
 
<c:permid>8589934442</c:permid>
 
<!--Bank of Japan-->
 
<c:subject rdf:resource="http://d.opencalais.com/genericHasher-1/c458a04c-6eb1-35df-b5ba-45b5ee4a97a4"/>
 
</rdf:Description>

The ID can be used to extract information about the entity or industry from the Refinitiv dataset. The ID also supports linkage across documents processed by Intelligent Tagging.

The following resolution (disambiguation) tags output the PermID attribute value that identifies the resolved entity (i.e. the unique ID of a specific company, city, deal, person, etc.

er/geo/City

er/Company

er/Continent

er/geo/Country

er/Deal

er/Organization

er/geo/ProvinceOrState

er/geo/Region

er/TopmostPublicParentCompany

In DocCat (Topic) Tags

A DocCat tag’s PermID attribute identifies the topic.

Linking to Open PermID

For the entity types supported by Open PermID—Person, Company, Currency, Industry, Organization, you can use the PermID exposed in the tagging output to access related, high-quality, curated data from the Refinitiv data sets.

Simply add the prefix “https://permid.org/1-“ to the permid value to create a direct link to the relevant page on the Open PermID website:

For example: https://permid.org/1-8589934442

The company and person resolution tags also include the openpermid attribute (called the "id" attribute in the JSON output), a direct link to Open PermID. (This attribute is not included in the internal Intelligent Tagging output.)

JSON Output Example:

    	
            "resolutions": [
            {
                "name": "BANK OF ENGLAND",
                "permid": "8589934365",
                "ispublic": "false",
                "commonname": "Bank of England",
                "score": 0.9442827,
                "id": "https://permid.org/1-8589934365"

RDF Output Example:

    	
            <rdf:Description rdf:about="http://d.opencalais.com/er/company/ralg-oa/4295905573">
 
<rdf:type rdf:resource="http://s.opencalais.com/1/type/er/Company"/>
 
<c:docId rdf:resource="http://d.opencalais.com/dochash-1/7586b818-40af-3d55-ac16-bf3520cddfa6"/>
 
<c:permid>4295905573</c:permid>
 
<c:score>0.9928677</c:score>
 
<!--Apple-->
 
<c:subject rdf:resource="http://d.opencalais.com/comphash-1/705cd5cf-93e1-323c-8d4e-1ea3200d37e4"/>
 
<c:legacyid rdf:resource="http://d.opencalais.com/er/company/ralg-tr1r/23d07771-c50b-315b-8050-3cdaf47ac0d0"/>
 
<c:name>Apple Inc</c:name>
 
<c:commonname>Apple</c:commonname>
 
<c:ticker>AAPL</c:ticker>
 
<c:primaryric>AAPL.OQ</c:primaryric>
 
<c:ispublic>true</c:ispublic>
 
<c:openpermid rdf:resource="https://permid.org/1-4295905573"/>
 
</rdf:Description>