An individual WEB search framework based on user profile and clustering analysis

The search engine has played an increasingly important role, but nowadays the search engine based on string matching has many intrinsic shortcomings such as a low accuracy and inadequate individual support and so on. The paper introduces an improved framework which is based on the traditional search engine, adds a two-level profile architecture to trace and analyze users’ previous searching history, recommends information according to some other people’s search history in the same group, evaluates search result on real time to update the user query model, does Clustering analysis on search result, submits search result to users in different categories. At last people can get information that they are really interested in. The improved framework can greatly improve individual search function and users’ search satisfaction.