The research of personalized search engine based on users´ access interest

In this research, users´ access interests were introduced into the design of personalized search engine by using Web mining technology. Firstly, the users´ access interest transactions were gained by interest algorithm via mining the users´ logs. Secondly, it presents a method to compute session similarity of transactional unit and transaction and sets up an interest similarity matrix for clustering by setting the suitable threshold value. At last, the result of clustering was applied in improving the PageRank algorithm for more accuracy. The personalized search engine can recommend pages which have more access interest to users who have similar interest with previous users. So the search engine´s efficiency can be further improved and it can provide more accurate search service for users.