An Information Filtering Approach

Web mining is used to automatically discover and extract information from Web-related data source such as documents, services and user profiles. Although standard data mining methods are applied for mining on the Web, specific algorithms need to be developed and applied for Web based information processing in Web resources. In our paper, we develop a method to filter the relevant information to users based on user profile. Our methods reduce the probability of filtering out the relevant boundary documents and increase the probability of filtering out the irrelevant boundary documents. In this paper, for filtering the documents, first, the user profile is presented as a set of weighted concepts; then, mapping the words onto concepts is done so that all the words representing user profile´s concepts can be found. Later, the documents are weighted based on the presence of the user profile´s concepts according to their weights and with respect to the location of the concepts; finally, after the filtering process, user profile is updated. The results clearly show further improvements and the accomplishment of the present goal.