Graph based Prediction Model to Improve Web Prefetching P.Venketesh R.Venkatesan

Web prefetching is an effective technique used to mitigate the user perceived latency by making predictions about the user’s future requests and prefetching them before the user actually demands them. In this paper, we present an algorithm that learns from user access patterns and builds a Precedence Graph (PG) that is used to generate the predictions. The difference in the relationship between objects of the same web page and the objects of different web pages are reflected in the graph implementation. It uses simple data structure to implement the graph, which is cost effective and consumes less computational resources. The proposed approach significantly improves the performance of web prefetching by utilizing limited amount of resources as compared to other existing algorithms used for prefetching