Automatically Constructing a Domain Ontology for Document Classification

For document classification, a domain ontology is first automatically constructed based on formal concept analysis. A document classification system, DCSO, is then proposed based the domain ontology. The features of DCSO are automatic construction of the system using the theorem of formal concept analysis, proposition of an XML knowledge-based schema for documents storage and quick search, and utilization of the hierarchy's property of ontology offering the accuracy of document classification. Experimental results for the DCSO are illustrated: the domain ontology is automatically constructed for document classification; the behavior of the accuracy for classification with DCSO is well; the searching time for DCSO steadily and slightly even though the queried documents increase rapidly.