A website recommender system based on an analysis of the user's access log

Abstract

Recommender Systems assists users in decision-making process when a potentially overwhelming set of choices of products or services is available. These systems automate the process of recommending products and services to consumers by analyzing data about items, users, and transactions and find associations between items and users. The results obtained are represented as recommendations. Collaborative filtering and content based filtering are widely used methods of providing recommendations. The website recommender system proposes the use of web mining techniques to recommend a personalized set of websites that might be of interest to a user. This approach exploits user access patterns as registered by proxy web server to underpin the relation between users and websites. These patterns are determined by clustering websites based on preferences given to them by different users. Hence, the websites visited by different users similarly and those visited by active user, forms the basis for the website recommender system to generate recommendations for the active user.