Enhancing Domain Knowledge with Semantic Models of Web Documents
Anna Rozeva
Full-Text PDF
XML 149 Views
DOI:10.17265/2159-5291/2013.07.001
Department of Applied Mathematics and Informatics, Technical University-Sofia, 1000 Sofia, Bulgaria
The paper considers the problem of semantic processing of web documents by designing an approach, which combines extracted semantic document model and domain- related knowledge base. The knowledge base is populated with learnt classification rules categorizing documents into topics. Classification provides for the reduction of the dimensionality of the document feature space. The semantic model of retrieved web documents is semantically labeled by querying domain ontology and processed with content-based classification method. The model obtained is mapped to the existing knowledge base by implementing inference algorithm. It enables models of the same semantic type to be recognized and integrated into the knowledge base. The approach provides for the domain knowledge integration and assists the extraction and modeling web documents semantics. Implementation results of the proposed approach are presented.
Semantic model, knowledge base, document classification, domain ontology, knowledge integration.