When you configure document separation, you should know how your classification settings work with your selected document separation type because they are dependent on each other.

If you choose the "Do not use layout classification" or "Do not use content classification" on the Classification tab, separation does not use the corresponding classifier.

When you select Trainable Document Separation, all pages of the training documents are trained so that TDS recognizes first, middle, and last pages. This training for separation occurs regardless of the layout and content classification behavior selected on the Classification tab. After separation, the classification settings are used based on the "How should the TDS classification result be used for document classification" option located on the Separation tab.

For example, if you choose to "Discard TDS classification result and classify document again", the classification setting you selected on the Classification tab is then used. Similarly, if you select Classify first page only only the first page is trained. If you select Classify each page, or Classify all pages at once, all pages are trained.

When Standard Document Separation is selected, the layout and content classification behavior is used to train the classifiers. This means that if "Classify first page only" is selected, then only the first page is trained. If "Classify each page" or "Classify all pages at once" are selected, all pages are trained.

Selecting No document separation does not affect the classifiers.

This behavior shows how document separation training and classification training are dependent on each other, and how changing one setting can effect the other.