User Portrait-based Hybrid Recommendation Method of Web Services

User Portrait-based Hybrid Recommendation Method of Web Services

As an effective means to solve information overload, service recommendation has become one of the main research directions of the scholars belonging to service computing domain. However, there are still some problems in traditional service recommendation methods, such as low accuracy of recommendation result and cold start for new users. For this reason, this paper proposes a hybrid recommendation method of web services based on user portrait. Firstly, this method establishes the user portrait model by considering the user’s natural attributes and interest attributes; Secondly, this method calculates the user portrait similarity on the basis of user portrait model and the user rating similarity on the basis of user rating matrix simultaneously, then combines social trust degree together by considering the trust relationship between users in user’s social networks; finally, the user portrait model, user similarity and social trust degree are integrated into the hybrid recommendation method of web services, so as to obtained the more accurate recommendation result for users. The experimental results show that the proposed method improves the accuracy of the recommendation result, and effectively alleviates the cold start problem.