Custom Entities

As with preexisting entities, you can also train models to extract custom entities. These entities can belong to any domain and can be of any type. For example, you can use medical text to extract a list of diagnosis or medicines. The process of extracting custom entities includes:

1. Preparing data: Preparing the input file and test file
2. Configuring the options: Creating training options file that contains information about the model and the options to be applied while training the model
3. Training the model
4. Extracting entities

When you successfully perform all these steps, the new entity type gets added to the list in the Entity Extractor stage, and you can use it to extract details from an unstructured file.