Content Type - Rebuild Training

Removes all existing training and retrains the classification model from the current training set documents.

Remarks

Purpose

The Rebuild Training command first purges all existing training data and then retrains the classification model using the current set of training documents. This ensures that the model reflects the latest training samples and configuration.

The Training Set Batch

The training set batch is a special Batch object associated with the Content Model. It serves as the source of training samples for all Document Types in the model.

• Creation:
  The training set batch is created automatically the first time you perform training using commands such as Train As or Train From on a Batch Folder. It can also be created manually by right-clicking the Content Model and selecting "Create Training Set" (if available).

• Populating with Samples:
  Training samples are added to the training set batch by using the Train As or Train From commands on Batch Folder items in any batch. These commands copy the selected folder (and its pages) into the training set batch, associating it with the appropriate Document Type. Each sample appears as a folder under the training set batch, and is used during retraining.

• Location:
  The training set batch can be found in the Local Resources Folder of the Content Model, typically named "Training Set".

Usage

• Use after adding, removing, or updating training samples, or after making changes to the Content Model or Classify Method.
• Can be executed from the context menu or as part of a training workflow.

Related Concepts

• Training Set: The batch of documents used for retraining.
• Training Page: Stores individual training samples.
• Classify Method: The classification logic being retrained.
• Purge Training: Used automatically before retraining.

See Also

Content CategoryContent ModelDocument TypeForm TypePage Type