Like-minded detector to solve the cold start problem

Collaborative filtering systems effectively provide recommendations for users and items that the system has enough historical interactions. However, collaborative filter systems suffer from the cold start problem for both items cold start, when a new item added to the system with no or few historical preferences, and user cold start, when a new user visits the system with no historical preferences of the particular user. In this paper, we address the cold- start problem that is giving recommendations to novel users and items that no user of the community has rate it yet. Lots of studies tried to solve the cold start problem, but it solves either item-cold start or user-cold start, or the provided solutions suffer from privacy problem. Therefore, privacy protect model is suggested to solve the cold start problem (in both cases user and item cold start). We suggested two types of recommendation (node recommendation and batch recommendation), and we compared the suggested method with three other alternative methods (Triadic Aspect Method, Naive Filterbots Method, and MediaScout Stereotype Method), we used a dataset collected from online web news to generate recommendations based on our method and based on the other three alternative methods. We calculated the levels of novelty, coverage, and precision. We found that our method achieved higher levels of novelty in batch recommendation while achieved higher levels of coverage and precision in node recommendations compared with these three alternative methods.