Intelligent Preloading of Websites Resources Based on Clustering Web User Sessions

This paper presents a new approach to accelerate website loading. Today´s web users are impatient and therefore it is necessary to render each web page as soon as possible. However, the ever increasing average size of web pages and insufficient transmission speed of some networks do not reach the desired rendering time. In this work we proposed a solution for intelligent preloading of website resources based on a combination of the Markov model, a utility function, and a data mining technique which is based on clustering of users. Clustering is performed according to the behavior patterns of each user.