Automatic extraction and classification of data items from library cataloging cards by a knowledge-based approach

An experimental approach to extracting and classifying data items automatically from library cataloging cards is presented. A basic strategy in this approach is to utilize various kinds of knowledge cooperatively. For example, for cataloging cards this would include the card structure, relationships among data items, and data format information. Unlike many traditional character recognition approaches, the authors´ approach is adaptable to cataloging cards composed of blurred and indistinct characters and/or described by various layout structures