A Feedback-corrected Collaborative Filtering for Personalized Real-world Service Recommendation

The emergence of Internet of Things (IoT) integrates the cyberspace with the physical space. With the rapid development of IoT, large amounts of IoT services are provided by various IoT middleware solutions. So, discovery and selecting the adequate services becomes a time-consuming and challenging task. This paper proposes a novel similarity-measurement for computing the similarity between services and introduces a new personalized recommendation approach for real-world service based on collaborative filtering. In order to evaluate the performance of proposed recommendation approach, large-scale of experiments are conducted, which involves the QoS-records of 339 users and 5825 real web-services. The experiments results indicate that the proposed approach outperforms other compared approaches in terms of accuracy and stability.