4.1.3 Model Training

This topic describes the systematic instructions to train the model on the annoted training corpus.

The annotated training corpus is a collection of annotated training files created using the annotator.

Model training is iterative and is carried out over increasing corpus size depending on the model parameters.

Each defined use case have its own training corpus available in the path set up in the use case definition.

Specify User ID and Password, and login to Home screen.

1. On Home screen, click Machine Learning. Under Machine Learning, click NLP Toolkit.
2. Under NLP Toolkit, click Model Training.
   The Model Training screen displays.

   Figure 4-14 Model Training

   
   

   
   Description of "Figure 4-14 Model Training"

   
   

For Training NER Models:

3. Select the Model Type as NLP(NER).
4. Select the type of Use Case Name from the drop-down list.
5. Click Train Model.

For Training Document Classifier Model:

6. Select the Model Type as Document Classification.
7. Select the type of Use Case Name from the drop-down list.
8. Click Train Model.
   If the model parameters acceptable, the user can save the model by clicking Save Model. The alternative is to add more annotated training files for the use case and repeat model training, till satisfactory parameters are achieved.