Excalibur: a personalized meta search engine

General purpose Web search engines are becoming ineffective due to the rapid growth and changes in the contents of the World Wide Web. Meta-search engines help a bit by having a better coverage of the WWW. However, users are still overwhelmed by the large amount of irrelevant results returned by a search. A promising approach to tackle the problem is personalized search. Thus the problem of capturing users´ personal information need and re-organizing the results has attracted a lot of attention. In this paper, we present a meta-search engine that extracts users´ preference implicitly and provides immediate response by re-ranking the results. Re-ranking is done by using the Naive Bayesian classifier and the resemblance measure. Moreover, we show that the users´ preference can be succinctly represented by a few keywords.