Community Discovery Algorithm for Social Networks Based on Parallel Recommendation

The distribution of user community in social network has the problem of similar feature distribution, which leads to the high viscosity between community and community, and it is difficult to mine community features. A social network community discovery algorithm is proposed based on parallel recommendation. The fuzzy decision model of social network community discovery is established by taking the user behavior value, user consumption value and loyalty of social network as independent variables. The relationship between the interaction degree and the recommendation effect in the social network is analyzed by extracting the characteristic quantity of the community association attribute in the social network, and the community group recommendation in the social network is carried out by using the parallel recommendation algorithm, and the number of visits between users is analyzed. The number of messages is used as the weighted weight coefficient, the difference factor is introduced to evaluate the influence of the community, the synchronous label of the social network community is established, and the community feature mining and parallel recommendation are realized according to the label location. The simulation results show that, the proposed method has good accuracy in community discovery, high accuracy in community attribute feature mining, and low cost of the algorithm. It has obvious advantages compared with similar algorithms.