A Collaborative Filtering Recommendation Algorithm Based on Probabilistic Latent Semantic Analysis

The information overload problem affects everyday experience in the search for valuable knowledge. To overcome this problem, people often rely on suggestions from others who have more experience on a topic. Many researchers and practitioners pay more attention on building a proper method which can help users obtain resources and services which wanted. Collaborative filtering systems can deal with large numbers of people and with many different items. However there is a problem of scalability result in low quality recommendations. In this paper, a collaborative filtering recommendation algorithm based on probabilistic latent semantic analysis is presented. This method uses the probabilistic latent semantic analysis model to predictin at first. And then, the presented method utilized the item based collaborative filtering algorithm to produce the recommendations. The personalized collaborative filtering approach based on probabilistic latent semantic analysis can alleviate the scalability problem in the collaborative recommendations.