Interscience Open Access Journals
The information extracted from Web pages can be used for effective query expansion. The aspect needed to improve accuracy of web search engines is the inclusion of metadata, not only to analyze Web content, but also to interpret. With the Web of today being unstructured and semantically heterogeneous, keyword-based queries are likely to miss important results. Using data mining methods, the authors' system derives dependency rules and applies them to concept-based queries. This paper presents a novel approach for query expansion that applies dependence rules mined from a large Web World, combining several existing techniques for data extraction and mining, to integrate the system into COMPACT, their prototype implementation of a concept-based search engine.