A Multi-resolution Community Partition Method based on Node Degree. Issue 1 (May 2021)

• Record Type: Journal Article
• Authors: Xia, Quan; Wu, Fan; Chen, Kebin
• Is Part Of: Journal of physics. Volume 1924:Issue 1(2021)
• Publication Date: 2021-05
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1924/1/012011
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: Community detection is one of the significant methods to study the properties of complex networks. Community resolution, as an important index of detection algorithm, is often used to evaluate the performance of the algorithm, but many classical algorithms have shortcomings in this respect. Based on this, this paper proposes a multi-resolution community partition method based on node degree. By introducing similarity coefficient and propagation coefficient of degree, the resolution of network is effectively controlled. Experiments on a scale-free network with 100 nodes show that the proposed algorithm is stable and efficient, with low time complexity. The number of communities can vary from 3 to 27 with different combinations of two coefficients, which can well discover the rules and characteristics of the network.