zentity relies on objects called entity models to control the execution of entity resolution jobs.
Entity models serve multiple purposes:
- They define which attributes map to which fields of which Elasticsearch indices.
- They define how to match an attribute for any index field it maps to.
- They define which combinations of matching attributes imply a resolution.
Read the entity model specification for specific details on the contents of entity models.
You must provide an entity model when making a resolution request. You can provide it in two ways:
- You can embed the entity model in the request under a field called
- You can index the entity model using the Models API and reference it by its
Option (2) gives you the ability to share, reuse, and build upon existing entity
models. You can manage entity models using the Models API.
zentity stores entity models in an Elasticsearch index called
Each document in this index represents the entity model for a distinct entity
type. The entity type is listed in the
field of the document. There can be only one entity model for a given entity type.
Once you have indexed an entity model, you can use it by setting the
parameter in your requests.
|Entity Model Specification