METHOD OF AND SYSTEM FOR AUTOMATIC DOCUMENT RETRIEVAL, CATEGORIZATION AND PROCESSING

摘要

A system and method are presented for performing word sense disambiguation using semantic networks with a mathematical formalism including probabilities and equivalent metrics. Also presented are techniques for automatically creating a knowledge base (network) to be used in the WSD process. Enhanced retrieval and categorization systems are developed. Information entropy theory is applied to determine importance o senses identified. Numerous applications of these basic techniques in improving accuracy of existing systems are described.