Research on Community Detection Algorithm Based on the UIR-Q
Abstract: Aiming
at the current problems of community detection algorithm in which user’s
property is not used; the community structure is not stable and the efficiency
of the algorithm is low, this paper proposes a community detection algorithm
based on the user influence and its parallelization method. In terms of the
concept of user influence in the subject communication and the PageRank
algorithm, this paper uses the properties of nodes of users in social networks
to form the user influence factors. Then, the user with the biggest influence
is set as the initial node of new community and and the local modularity is
introduced into detecting the community structure. in order to make the result of community
detection quick and efficient. Many experiments show that the improved
algorithm can efficiently detect the community structure with large scale users
and the results are stable. Therefore, this algorithm will have a wide applied
prospect.
Keywords: social networks;
community detection; user influence; PageRank algorithm; local modularity
Author: Zilong Jiang, Wei Dai,
Liangchen Chen, Xiufeng Cao, Yanling Shao
Journal Code: jptkomputergg160207
