PERSONALIZED USER FEATURE SIMILARITY SUPPORTED COLLABORATIVE FILTERING AND WEB USAGE MINING FOR PAGE RECOMMENDATION

Abstract

The rapid growth of the web has run to a considerable increase in information dissemination. Freshly, Research on hybrid personalized recommendation systems mainly addresses two problems: prediction and completion of sparse data, and the user s personalized recommendation. To address the issue of high data sparsity and low recommendation accuracy in the traditional service recommendation models, this study presents a hybrid personalized collaborative filtering model for page recommendation based on collaborative filtering by introducing user preferences. The example verified that the hybrid personalized recommendation system based model can effectively reduce the data sparsity and increase the accuracy of the prediction.In this work, the efficacy of collaborative filtering and web usage for recommender system is investigated and proposed. The experimental results demonstrate that the proposed WUM improves the performance of the recommendations.