Web Page Prediction Using Genetic Algorithm and Logistic Regression based on Weblog and Web Content Features

As website users are increasing day by day, user behaviour analysis for improving the website performance attracts many researchers. This paper introduces the web page prediction model by involving the Logistic Regression (LR), which takes web page content similarity as input in group of similar weblog rules. To get association rules, this work introduce Feed Forward Counter (FFC) model for identifying the association rule with single data iteration technique. Both regression and FFC increased the learning rate for enhancing the accuracy of page recommendation. For adopting the dynamic situation in the work, final prediction is done by Particle Swarm Optimization (PSO) algorithm, which uses learned regression value in fitness function value evaluation. Experiment is performed on instantaneous dataset attained from ProjectTunnel website, which is collection of weblog and keywords. Results shows that proposed Logistic Regression based Web Page Prediction Model (LWPPM) for next page prediction system has improved various evaluation parameters like precision, coverage, m-metric.