Cold-Start Web Service Recommendation Using Implicit Feedback

Service recommendation becomes an increasingly important issue when more and more Web services are published on the Internet. Existing Web service recommendation approaches based on collaborative filtering (CF) seldom consider recommending services based on users’ ratings on services since such kind of explicit feedback is difficult to collect. In addition, the new user cold-start problem is also an important issue due to the lack of accuracy in service recommendations since the new users have not yet cast a significant numbers of votes. In this paper, a dataset consisting of much user-service interaction data is reported. The interaction data created according to users’ behaviors can highly represent their preferences. Therefore, we construct pseudo ratings based on this kind of implicit feedback. We develop a novel service recommendation approach which can partially deal with cold-start problem using an online learning model. Experiments show the proposed approach can achieve satisfied results in prediction accuracy and time cost. Keywords-cold-start Web service recommendation; implicit feedback; probability matrix factorization;