Relationship Extractor

https://analytics.babelstreet.com/rest/v1/relationships

https://raw.githubusercontent.com/rosette-api/curl-examples/develop/examples/relationships.curl

Relationship Extractor builds on the results of the Entity Extractor and Linker to identify how entities are related to each other.

Relationship extractor performs targeted relationship extraction by finding anything within the text that looks like one of the supported relationship types. This allows you to filter and process the results according to your needs.

Using linguistic tools and Wikidata to bootstrap its knowledge, the exact action connecting the entities and other related information are extracted. It uses machine learning methods over parse trees and entity mentions to analyze the connection, and then return the components within the relationships.

Table 12. Relationship Types

Relation	Description
ORG-FOUNDED	Holds between an organization and a person or another organization that founded it
ORG-HEADQUARTERS	Holds between an organization and the location where its headquarters are located
ORG-SUBSIDIARY-OF	Holds between two organizations where the first is a subsidiary of the second
ORG-COLLABORATORS	Holds between two organizations the are working together toward a project or a goal
ORG-ACQUIRED-BY	Holds between two organizations where the first organization was acquired by the second one
ORG-PROVIDER-TO	Holds between two organization where the first organization is providing services or products to the second organization
ORG-TOP-EMPLOYEES	Holds between an organization and a person who is one of the organization’s top employees
PER-BIRTH-PLACE	Holds between a person and the location in which he was born
PER-EMPLOYEE-MEMBER-OF	Holds between a person and an organization with which the person is affiliated
PER-RESIDENCE	Holds between a person and its current or past location of residence
PER-PARENTS	Holds between two person entities where the second is the parent of the first
PER-SIBLINGS	Holds between two person entities which are siblings of one another
PER-SPOUSE	Holds between two person entities which are the life partners of one another
PER-OTHER-FAMILY	Holds between two person entities that belong to the same family but are not siblings, parents, children, or spouse of one another
CIT-OF	Holds between a person and his country of citizenship
EDU-AT	Holds between a person and an academic institution he is affiliated with

Tip

Try it in the interactive documentation

Request

Name	Type	Description
`content`	string	Text to process
`contentUri`	string	URI to accessible content
`language`	string	ISO 639 language code

Notice

content and contentUri are mutually exclusive; only one can be specified per call.

{
  "content": "string",
  "contentUri": "string,
  "language": "string"
}

Query Parameters

Name	Value	Description
output	rosette	Returns the response in ADM format.

Note

All input parameters, including the text being analyzed and any relevant options, are defined in the request body.

Response

The response is a list of extracted relationships. A relationship consists of the following elements:

Predicate: The main action or verb acting on the first argument, or the action connecting multiple arguments.
PredicateID: (optional) Label of the identified relationship type.
Argument(s): One or more subjects in the relationship.
Argument ID(s): (optional) The entity ID returned for arguments that can be linked to Wikidata.
Confidence: A score for each relationship result, ranging from 0 to 1. You can use this measurement as a threshold to filter out undesired results.

{
  "relationships": [
  {  
     "predicate": "string", 
     "predicateId": "string",
     "arg1": "string",
     "argId": "string",
     "arg2": "string",
     "arg2Id": "string",
     "confidence": number
  }
 ]
}

Supported languages

GET /relationships/supported-languages

Returns the list of supported languages and scripts for the endpoint, along with whether you have a license for the language.

Relationship Extractor currently only supports English language input.

Response

Field	Type	Description
`language`	string	ISO 639 language code
`script`	string	Four-letter ISO-15924 script code
`licensed`	boolean	Indicates if you are licensed for this language

{
  "supportedLanguages": [
    {
      "language": "string",
      "script": "string",
      "licensed": boolean
    }
  ]
}

Babel Street Analytics API