Automatic Recommendation of Web Pages for Online Users Using Web Usage Mining

Now a days, user rely on the web for information, but the currently available search engines often gives a long list of results, much of which are not always relevant to the user's requirement. Since a user has a specific goal when searching for information, personalized search may provide the results that accurately satisfy user's specific goal and intent for the search. Personalization of web search is to retrieve information according to user's interests which may be inferred from user's action, browsed documents or past query history etc. A real problem before web master of a website is to match the user needs and keep their attention focused to his website. In this paper, we propose a web recommendation approach which recommends user a list of pages based upon user's historic pattern and a list of web pages which have not been visited yet. This approach results in the improvement in the accuracy of the pages presented to the users.