A Cross Domain Recommendation Algorithm Based on Joint Distribution

• 1、Beijing Key Laboratory of Intelligent Telecommunications Software and Multimedia(Beijng University of Posts and Telecommunication), Beijing 100876

Abstract：With the rapid development of the mobile Internet, online services such as shopping platforms, social networks and homestay platforms have also attracted more people to use. When users use these services, they often leave too many data, and through analysis of the data, information such as behavior preferences hidden by users in various fields can be obtained. Then, how to use this information to recommend new items that match the user\'s preferences becomes a challenge. Now, many studies use cross-domain recommendation to make recommendations from knowledge of auxiliary domains, but many methods require heuristic network design methods. In this paper, a cross-domain recommendation algorithm based on federated distribution is proposed to learn the characteristics preferred by users in the two domains in a more intuitive way, and to model the joint distribution of cross-domain user behavior to improvin the effect of recommendation.

Keywords： Computer Application Technology Recommendation System Joint Distribution Cross Domain Recommendation