Help > Knowledge Processing > Learning classification offline

"Learning" classification offline

When set up to do so, the system automatically learns classification and interpretation during production. For this you need a Knowledge Processing activity step in your XBOUND process.

It is also possible to manually "learn" classification so that you already have a trained system when you start processing documents.

In order to do this, you add layouts (sample documents) using the Capture Components Administration module.

Use this procedure:

  1. Ensure that the ReadSoft Knowledge Processing Service is running.

  2. Using the Capture Components Administration module, open a document specification.

  3. Add a sample document that corresponds to the document specification – one for each typical appearance of this type of document.

  4. Select the sample documents and click Sample set.

  5. In the new dialog that is displayed, click Training.

  6. Click OK. The sample has now been added to the Knowledge Processing database.

Tip: These steps are described in more detail in Capture Components Administration Help.

Knowledge Processing: Overview