Web User Prediction by Integrating Markov Model with Different Features

Project Discription

Abstract Prediction with the help of the different features is an important branch of data mining. This work utilizes all the features of the web mining such as web content, web log, and web structure. All these are utilize to understand the the user behavior and with the Ant colony technique. This technique for learning make the task easier to predict the next page of the user base of the pattern follow by an ant. Experiment shows efficiency that about 85 % for single and 82 % for two element are correct from this technique. It has observed that learnining time has also been improved, means ant required less iteration for generating correct sessions. Conclusions World Wide Web has necessitated the users to make use of automated tools to locate desired information resources and to follow and asses their usage pattern. Web user prediction introduce to reduce the user access latency problem of the internet; its success mainly relies on the accuracy of web page prediction. Ant colony is the most commonly used pattern finding technique. It include initial phenomenon values are set by the frequency of the links use user. Our results shows that by the utilization of all the features of the web ant behavior is utilize which also generate the same sequence of element as per the real user session with an accuracy of 85 %. In future one can consider average time duration on the web page a next page selection factor for increasing the accuracy of the work.

Project Sample Image