Prediction of Consumer’s Future Demand in Web Page Personalization System

Abstract

Furtherance in Information Technology has progressed usage of Web services to extensive and exhaustive. From the gigantic volume of data, predicting optimal Web page is the cumbersome process. This paper presents an innovative methodology of minifying Web page recommendation systems by employing hybrid Levenberg–Marquardt firefly neural network algorithm along with improved fuzzy c-means clustering. Hybrid Levenberg–Marquardt firefly neural network algorithm is used to categorize the prospective and non-prospective consumer data of the Web log where improved fuzzy c-means clustering clusters the prospective data. Before a user begins the path of navigating through Internet for fetching relevant Web page, the cluster recommends the most relevant Web pages by analyzing the interests of similar users. A comparative analysis of proposed system with prevailing fuzzy k-means clustering technique exhibits that the performance of projected system is better.