Entity extraction stage can be used to extract entities like date, location, company name, etc from the documents. Entity extraction helps recognize specific entities within unstructured text and when an entity is identified, it enables you to take different actions based on the entity type.
Entity Extraction Configuration
Add a new stage of type Entity Extraction and configure the following properties for this stage.
- Stage Name – Name with which this stage would be referred to.
- Condition – The condition which defines the documents or records on which the entity extraction definition will be performed. The condition can be defined in two ways: basic and script. In the basic method, you can perform some basic checks on one or more index fields. If the conditions are satisfied, the entity extraction rules defined are applied on the documents. In case of script method, use a custom script to identify the records for entity extraction.
- Stage Definition – The definition of the source and target of the entities. You can add one or more rules to find the entities. The entity extraction will be performed in the order in which the rules are defined.
| Source Field | The field from which entities are to be extracted. |
| Add Entities | Entities to be identified from the source field. You can select one or more entities from the supported list of entities. |
| Target Field | Field to be used to store the extracted entities. |