No abstract

Web caching and pre fetching plays a significant role among the users and websites in reducing the response time of user requests and saving the network bandwidth. The websites which are most likely visited are stored under web cache and web pre-fetch to retrieve them later shortly. Thus web caching and web pre-fetch are one of the most thriving resolutions for enhancing the performance of the web-based systems. However, the complexity in formatting the ideal Web objects that will be re-visited in the prospect is still a trouble faced by an existing conventional cluster based caching techniques. In this work, Bayesian classifier is used for web object filtering process which enhances the performance of caching techniques. Apart from that, the validation process is also being taken place to improve the user navigation process under web based systems. The validation is made of user identity on web product navigation is done with User-profile relativity to national identification number. Web services enable the product vendor regarding the demand of their products by specific valid users. The validation is done with both the existing cluster based web object filters and the proposed classifier based web object filters and compares the results with it. An experimental evaluation is done with the several log files to estimate the performance of the proposed validation check for Bayesian classifier and cluster based web object filters for web user navigation in terms of efficiency of classification, validation, performance rate and compared with an existing cluster based web object filters and identified the better web caching technique.