Predicting & Analyzing the Web User Navigational Behavior

Mainly focus on the problem of building models to represent past users behavior, that are able to predict the most likely links a user will request when viewing a page. WUM specifically designed to carry out applications by analyzing the usage data. The results of the Grey clustering algorithm used as the input for Grey Moving Probability Markov Model to predict users¿ next visit. This approach models navigation sessions and for predicting the next navigation step using transition probabilities with two estimation approaches. Prediction is a way of analyzing historical information to calculate the most possible probability of next request; the browsing pattern when the future users and customers are surfing the web site is matched. There are many web based advantages to implement the prediction, such as, web site personalization, structure appropriate web site, business intelligence etc. The transaction probabilities are more suitable for predicting the users¿ next request. In the prediction model, Variable Length Markov Chain is used to predict the category of users¿ next state with the transaction probability.