PROSAR-AIDA Free-form Capture...

Capturing The Data

Finding relevant index data without having to know the position or the layout of the data on the page.

In order to find the relevant data automatically, regardless of its position on the document, it must be possible to give a sufficiently precise definition for the desired data. PROSAR-AIDA provides various mechanisms for this purpose, including the search for data items with a known format (e. g. amounts), geometric and logical relationships of typical key words (e. g. key word "VAT" and "Total" in the same line), or the search for content from existing database tables (e.g. checking whether a client"s address exists on the document). All search methods can be combined with each other. This, for instance, makes it possible to filter out the total amount from amongst the many other amounts which may be on the document.

By using such definitions, PROSAR-AIDA checks all possible text areas with the help of the defined extraction rule and delivers all text areas found in the document which fulfill the checking criteria. In this way, data such as contract numbers, bank sort codes and account numbers, invoice dates, sums etc. can be filtered out and used for automatic indexing or for workflow routing purposes.

Should the document type detector recognize an actual form, all the features of a standard forms processing system are also available. Thus, processing all kinds of documents (freeform, formatted documents and forms) is possible in a single system without first having to sort the documents.