The invention relates to a recommendation model integrating a social network and user evaluation. In order to reflect the influence of the user social relationship on the user decision and improve theaccuracy of the recommendation result, a recommendation model integrating the social network and the user score comment data is provided. The model extracts theme information from comments as characteristics of users and merchants through a text processing technology and a data mining algorithm, utilizes a community discovery algorithm to divide communities for the users, and finally utilizes a machine learning method to model the communities and recommend the communities. The model not only considers the preference of the individual users, but also contains the overall characteristics of thecommunity users, and can provide personalized recommendation for the community users.