Extraction d'Entites

Extracts named entities from text (people, organizations, places, dates, amounts). Uses an LLM for extraction (NER).

Parameters

Parameter	Type	Required	Variable	Description
input	dynamic value	Yes	Yes	Source text for named entity recognition (NER) extraction.
entityTypes	array	No	No	Entity types to extract (PERSON, ORGANIZATION, LOCATION, DATE, TIME, MONEY, EMAIL, PHONE, URL, PRODUCT, CUSTOM).
customEntities	array	No	No	Custom entity definitions with name, description, and examples.
minConfidence	number	No	No	Minimum confidence threshold (0 to 1) to include an entity in results. (Default: 0.5, min 0, max 1)
deduplicate	boolean	No	No	Remove duplicate entities from results. (Default: false)
normalize	boolean	No	No	Normalize extracted entities (e.g. standardize date formats). (Default: false)
linkToOntology	boolean	No	No	Link extracted entities to existing ontology nodes. (Default: false)
model.provider	text	No	No	Language model provider (e.g. openai, anthropic).
model.model	text	No	No	Model identifier to use (e.g. gpt-4o).
parameters.temperature	number	No	No	Model creativity. Low value recommended for extraction. (Default: 0.2, min 0, max 2)
parameters.maxTokens	number	No	No	Maximum number of tokens in the response. (Default: 4096, min 1, max 128000)
outputVariable	text	No	No	Output variable name containing the extracted entities.

Parameters marked Variable = Yes accept the {{blockName.field}} syntax.

Output

Output variable : entities

{
  "entities": [],
  "count": 0
}

Example

Extract entities from a business email.

Input :

{"text": "Jean Dupont de la societe Acme a signe le contrat le 15 mars a Paris."}

Output :

{"entities": [{"text": "Jean Dupont", "type": "PERSON"}, {"text": "Acme", "type": "ORGANIZATION"}, {"text": "Paris", "type": "LOCATION"}], "count": 3}

Tip

{{entities.entities}} is an array of objects with text and type. Standard types: PERSON, ORGANIZATION, LOCATION, DATE, MONEY.