Research And Application Of Metasearch Engine Based On Automatic Classification

Search engines are main tools for user to locate information on Internet. After inputting a keyword in any of these search engines, a long list of results will certainly be returned which may include such a vast amount of results for user to check one by one and locate which one is really needed. With respect to this embarrassed situation, this paper proposed a meta-search, which offers a unified way to access multiple search engines and increase the coverage and precision. While user searching with the Meta-search, it inspects the user's reaction to analyze users' interesting of information needs. This paper proposed an integrated results fusion strategy, using automatic classifying method and ranking approach to combine and integrate the results from different underlying engines into a unique collection. With the meta-search we can help users to locate quickly what they need on the Internet.This paper is part of project, the synthetically analyzing system of penalty information. The project is supported by the Ministry of Justice Republic of China(No. 03SFB3021), and it take penalty information as its main domain.