Adaptive multi-resolution Modularity for detecting communities in networks
Abstract
Community structure is a common topological property of complex networks, which attracted much attention from various fields. Optimizing quality functions for community structures is a kind of popular strategy for community detection, such as Modularity optimization. Here, we introduce a general definition of Modularity, by which several classical (multi-resolution) Modularity can be derived, and then propose a kind of adaptive (multi-resolution) Modularity that can combine the advantages of different Modularity. By applying the Modularity to various synthetic and real-world networks, we study the behaviors of the methods, showing the validity and advantages of the multi-resolution Modularity in community detection. The adaptive Modularity, as a kind of multi-resolution method, can naturally solve the first-type limit of Modularity and detect communities at different scales; it can quicken the disconnecting of communities and delay the breakup of communities in heterogeneous networks; and thus it is expected to generate the stable community structures in networks more effectively and have stronger tolerance against the second-type limit of Modularity.
- Publication:
-
Physica A Statistical Mechanics and its Applications
- Pub Date:
- February 2018
- DOI:
- 10.1016/j.physa.2017.09.023
- Bibcode:
- 2018PhyA..491..591C
- Keywords:
-
- Complex networks;
- Community detection;
- Modularity;
- Multi-resolution