Effective information retrieval using co operative clients and transparent proxy cache with prediction and prefetching

Abstract

Searching in the web is an interesting application Information retrieval using web is an ever green field for research and is always the choice of researchers due to its importance in any real time application Every day the search engines processes billions of queries The increase in web users and web applications has led to increase in latency network congestion and server overloading Many research activities are going on to improve the search performance One well known traditional methodology is web caching When web caching is deployed it caches popular objects at the locations close to the clients newlineCache can be deployed at three levels client level proxy level and original server level Client level local cache is a part of hard disk in local browser allocated for web object caching But the current browser cache cannot share their contents with other clients Each browser cache is isolated and its resources are not utilized completely A co operative browser level web cache sharing system based on chord is suggested for performance improvement A proxy server is residing between client machines and the web server It can service request from a number of clients For each user request the system first checks the result in the local cache or in the authenticated cooperative caches In both the cases if a miss occurs the proxy checks whether it has a valid retrieved result newline newline