During Knowledge Processing, the Knowledge Processing database is updated with layout data points, which consist of two types of information:
Information about what a page looks like, for example if it says "Invoice" at a specific position.
Information that has been added to the page, either automatically or by a user. Examples:
A field called "Total to pay" was interpreted with a value of "500.00".
A page is associated with a document type.
A field position was learned in Verification (ReadSoft Capture Framework).
Layout data points improve performance during Classification and Interpretation, because the information can be reused the next time the same type of document is processed.
For additional information, see How and when are layout data points added?