The social network friend recommendation method and system provided by the invention can accurately obtain friends in a social network. The interest similarity and social intimacy of users are taken as attributes to construct a weighted heterogeneous network, so that the friendship recommendation set has both the interest similarity and social intimacy with the target user, and the social intimacy is improved based on the interest similarity. Therefore, the obtained friend recommendation set is more accurate. In addition, the extracted friends also conform to the user’s habits, and are more in line with the user’s preferences. The social network friend recommendation system of the invention includes a first acquisition device, a second acquisition device, a first calculating device, a second calculating device and a third calculating device. The first acquisition device, the second acquisition device, the first calculating device and the second calculating device build a weighted heterogeneous network according to the interest similarity and social intimacy, and obtain the community set of the target user by mining the community set of the target user in the weighted heterogeneous network, and extract a set number of users from the weighted heterogeneous network according to the membership value of the target user to different communities, and then extract a set number of users from the weighted heterogeneous network according to the membership value. The extracted friend recommendation set is accurate and reliable, can meet the requirements of users on diversification.
SOCIAL NETWORK FRIEND RECOMMENDATION METHOD AND SYSTEM
