A New Method for Detecting Users Behavior from Web Access Logs

The development of the web has created a big challenge for directing the client to the website pages in their area of interest. One option is to capture the requirement of the client and provide them a list of recommendation. Most specifically, an online navigation behavior develops with each passing day, consequently extracting information intelligently from it is a difficult issue. Webmaster of an organization should utilize methods of web mining to fetch users requirement and Web usage mining (WUM) is one of the method. WUM is designed to operate on web server logs, logs contain client's navigation history which is very useful for the web recommendation. Recommendation is an application of web usage mining. Consequently, recommendation system can be utilized to forecast the navigation pattern of client and recommend those to client in a form of recommendation list. In this paper, we tend to propose a recommendation approach that recommends a listing of pages based mostly upon client's historic pattern (recorded within the web log). This approach brings the accuracy of the pages displayed to the client or users.