A big challenge for the web master of any e-commerce website is to match user needs and keep them continuously to browse their site. It is easy for any unsatisfied user to reach out the counterpart site in a single click. Many Web usage mining methods were adopted to work on web server log and capture the future browsing pattern of user. However, the accuracy of the methods can’t satisfy the user especially in huge site. This paper presents the novel Forager agent based architecture that uses its collective intelligence for identifying the future browsing pattern of user. Our practical implementation shows that accuracy and coverage measures are very much improved than existing methods.