Here the reader will learn about the purpose of Entities in Classification Models and how to associate Entities with Classification Models by providing the User a step-by-step guide.
Users would use Entities in Classification Models to store information used during or after the KAPTO processed the input documents.
As a categorization model does not extract entity information, entities defined in a Classification Model are not filled by AI by definition. On the contrary, they are implicitly marked as "not filled by AI" and thus can be only managed through extension points and API.
As a usage example that clarifies the functional meaning of entities in Classification Models, we can think of KAPTO receiving bulk pdf documents from a source - an external application - that "knows" to which policy those documents are referring but to which the specific content is unknown. Here the goal is to split out the document and send each segment, with the indication of the policy, to another service that stores them in a document management system with the proper metadata (the policy number). In this case, the external system can load the document through KAPTO API, valorizing the entity policy with the correct value. When KAPTO completed the segmentation task through a post-final extension, it could call the API of the document management system indicating in the metadata the policy's value stored in the entity.
How to Associate Entities with Classification Model
to
Note: It’s always possible to manage Entities in the Classification model by clicking the Pencil icon next to the Entity you want to modify or the Delete icon if you want to delete the entity.