Developing an algorithm for defining key players in social networks by considering both individual characteristics and structure of relations between network’s members
Subject Areas :فرناز برزین پور 1 , سمیه علیزاده 2 , بنت الهدی علی احمدی 3
1 -
2 -
3 - University of Science and Technology
Keywords:
Abstract :
Social Network Analysis (SNA) is associated with the mapping and measuring of relationships and flows between people, groups, organizations, etc. Centrality measures in this context are defined to identify the most important actors in social network. One of the main centrality measures is closeness centrality; in this measure a node is more "central" if it is closer to many more nodes than any other node. In this paper, we develop a conceptual framework for defining centrality measures in complex networks. It is to be noted that one of the major limitations of determining centrality measures is concerning with the structure and effects of relations among people, groups or organizations in principle, and largely ignoring psychological attributes of the individuals. Therefore the proposed framework is based on the combination of two approaches: social network analysis and traditional social science approach by considering both structure of relations and individual characteristics. Detecting communities in complex networks is of considerable importance for understanding both the structure and function of the networks and it is necessary to interpret radial centrality measures. Therefore, we propose spectral clustering by determining the best number of communities as a prerequisite stage before finding closeness measures of centrality. Based on the proposed framework, an algorithm to compute the closeness centrality in complex networks is developed. We test the proposed algorithm on Zachary’s karate club network which is considerably used as a benchmark for community detection in a network. Experimental results indicate that the new algorithm is efficient at detecting both good inter-cluster closeness centrality and the appropriate number of clusters.
