An Improved User Browsing Behavior Prediction using Regression Analysis on Web Logs

Abstract

Web usage mining is widely used to discover the usage patterns from web log files. It deals with web log data which are taken from web servers, proxy server or client’s cache. By analyzing user’s browsing behavior, next web page prediction can be made. Various types of mining algorithms proposed over the years based on different techniques. But prediction of future request of the user mainly concern with its accuracy and efficiency. In this paper, we have proposed a new model for predicting the next web page. K-means clustering and Regression Analysis algorithms are used to predict the future request. These two algorithms in combination produce efficient and accurate results.