An advanced automated approach for community mining in signed social networks

摘要

A number of recent studies have focused on the properties of social networks. In this article, we highlight the property of community structure, in which the nodes of the network are joined together in tightly knit groups and there are looser connections between the groups. We propose an approach AACMA ( Advanced Automatic community Mining Approach ) for detecting such communities, build around the idea of using centrality measure to find community boundaries. We test AACMA on real-world graphs whose community is already detected and it detects this known structure with high sensitivity and reliability. We also compare the method with a different method named ABCD (Attractiveness-Based Community Detection) by using the same dataset. We find that AACMA provides more accurate results than the compared approach.