A MSE model with learning mechanism and merging module based on FCA
Qinhua Dong, Yajun Du, Fugui Wang · 2008
Meta search engine improves the coverage of the search results, but itpsilas hard to ensure the accuracy of the search results. In order to improve search quality we propose a MSE model with learning mechanism and merging module based on FCA. Firstly, learning mechanism can adjust the expertness of member search engine in a certain domain by analyzing userpsilas behavior. Only when the expertness of member search engine reaches a certain value, can the member search engine be employed by meta search engine. Above all, we employ FCA to merge all the search results on the assumption that if a web page is retrieved by more member search engines it is more important. In the concept lattice, the intent of the concept includes member search engines and the extent of the concept includes the web pages retrieved by those member search engines. Because of its hierarchy structure, we can rank concepts by the number of its intents, and then rank the web pages included by the same concept according to their original places in member search engines and the expertness of member search engine which retrieved them.