Relation extraction

Identifies and extracts connections between entities.

POST https://rex7.p.rapidapi.com/rex

Designed to uncover valuable insights by extracting structured information from unstructured text data.

Headers

NameTypeDescription

X-RapidAPI-Key*

String

Your API key

X-RapidAPI-Host*

String

rex7.p.rapidapi.com

content-type*

String

application/json

Request Body

NameTypeDescription

text*

String

Pass text for processing via this parameter. The maximum size in tokens is 3000, where 1 token is approximately 4 characters.

relations*

String list

Relations which will be used in the processing of the text.

scorer

Boolean

If set to true, the response will contain probability scores for each relation. The default value is set to false.

all

Boolean

This parameter controls the number of entities included in the response. In most cases, the best results are achieved by including only the target entities. However, if necessary, the parameter can be set to true and more entities will be included. The default value is set to false. Ignored if entities provided.

threshold

Float

It sets a relation threshold. If the score is below this threshold, the relation will not be included in the response. The default value is 0.0. Ignored if the scorer parameter is set to false.

entities

Objects list

Define what specific entities will be processed, and others will be skipped.

pairs filter

Objects list

Each tuple represents the source and target entity labels. If this parameter is provided and not empty, any relation that doesn't have a specified pair of source and target will be ignored. Parameter ignored if 'entities' parameter not provided.

relations filter

String

Map relations to entity pairs. Each relation can be mapped to multiple pairs. Each tuple represents the source and target entity labels. Any relation with a pair of sources and target which not included in to filter will be ignored.

The model has successfully analyzed the input text and labels.

Last updated