Normalized Cuts and Image Segmentation
IEEE Transactions on Pattern Analysis and Machine Intelligence
A Min-max Cut Algorithm for Graph Partitioning and Data Clustering
ICDM '01 Proceedings of the 2001 IEEE International Conference on Data Mining
SCAN: a structural clustering algorithm for networks
Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining
DENGRAPH: A Density-based Community Detection Algorithm
WI '07 Proceedings of the IEEE/WIC/ACM International Conference on Web Intelligence
An Algorithm to Find Overlapping Community Structure in Networks
PKDD 2007 Proceedings of the 11th European conference on Principles and Practice of Knowledge Discovery in Databases
Comparison of online social relations in volume vs interaction: a case study of cyworld
Proceedings of the 8th ACM SIGCOMM conference on Internet measurement
User interactions in social networks and their implications
Proceedings of the 4th ACM European conference on Computer systems
Microcosm: visual discovery, exploration and analysis of social communities
Proceedings of the companion publication of the 19th international conference on Intelligent User Interfaces
| Hi-index | 0.00 |
In this paper, we propose a density-based community detection method, CMiner, which exploits the interaction graph of online social networks to identify overlapping community structures. Based on the average reciprocated interactions of a node in the network, a new distance function is defined to find the similarity between a pair of nodes. The proposed method also provides a basic solution for automatic determination of the neighborhood threshold, which is a non-trivial problem for applying density-based clustering methods. Considering the local neighborhood of a node p, the distance function is used to determine the distance between the node p and its neighbors in the interaction graph to identify core nodes, which are then used to define overlapping communities. On comparing the experimental results with clique percolation and other related methods, we found that CMiner is comparable to the state-of-the-art methods and is also computationally faster.