The Usefulness of Logical Structure in Flexible Document Categorization

This paper presents a new approach for automatic document categorization. Exploiting the logical structure of the document, our approach assigns a HTML document to one or more categories (thesis, paper, call for papers, email, ...). Using a set of training documents, our approach generates a set of rules used to categorize new documents. The approach flexibility is carried out with rule weight association representing your importance in the discrimination between possible categories. This weight is dynamically modified at each new document categorization. The experimentation of the proposed approach provides satisfactory results. Keywords-- categorization rule, document categorization, flexible categorization, logical structure.

• Type: Conference
• Year: 2004
• Where: IJIT
• Authors: Jebari Chaker, Habib Ounelli
• Keywords: Artificial Intelligence, Automatic Document Categorization, Document Categorization, IJIT 2004, Logical Structure