RESEARCH AND APPLICATION ON SET PAIR ENTITY SIMILARITY MODEL OF SOCIAL NETWORK

Yanjun Zhao, Chunying Zhang

College of Science, Hebei Untied University, Tangshan, 063009, China

In allusion to the certain and uncertain value which exist in the node and relationship attributes of social network, From the attributes and relation angle of the entity to analyse the similarity degree of the affirmative, negative and uncertain between them, then build the set pair entity similarity model based on the set pair analytical method and apply it to the network association detection. First of all, applying the generalized pair close potential and the generalized set loose potential in social network based on set pair analysis method, and see it as the basis of association detection; secondly, giving the set similarity calculation method based on the entity attribute and relation, from the point of view of node attribute and relations attribute to calculate respectively, by setting the weight to consolidated calculate the set pair similarity of the entity; thirdly, utilizing entity set similarity to divide network association into clustering problem, then give the association partitioning algorithm; finally, integrating with the network instance to verify the effectiveness of the new network association partitioning algorithm.