Design of a structural search engine using a graph-based knowledge discovery system.

摘要

The World Wide Web (WWW) is an immense source of unstructured information. A structural search engine uses the hyperlink structure of the web site, as search criteria. The WWW can be represented in the form of a graph, giving structural form to the unstructured data. SUBDUE is a graph-based data mining tool, that discovers repetitive substructures in structural data. We hypothesize that SUBDUE can form the heart of a structural search engine. SUBDUE can discover structure formed by a user query within a graph representation of the WWW. WordNet is a tool that allows dictionaries to be searched conceptually. A second hypothesis of this research is that our structural search engine can be enhanced by using features supplied by WordNet. By integrating WordNet's capabilities, SUBDUE can increase its search capacity by searching for words similar to the query term, rather than just searching for the query term itself.