Recommender systems on web service selection problems using a new hybrid approach

Recommender systems can reduce information overload and recommend better items for customers in different fields such as, electronic commerce, web service selection and so on. Recommender systems use different technologies that are classified into two broad groups: content-based filtering technologies and collaborative filtering technologies. In this paper, a new hybrid recommender system has been proposed to alleviate new user cold-start problem. The new user cold-start problem is when a number of users have a few numbers of ratings on items. Also, an unsupervised learning algorithm means dynamic K-means algorithm has been used to address scalability problem. In this paper, recommender system approaches have been used for web service selection. The changed weighted majority voting (CWMV) technique has been used to combine the output of six strategies. The experiment results show the proposed approach has recommendations with better accuracy than other approaches.