Social Network Construction and Analysis Based on Community Photo Collections with Face Recognition

Abstract

With the explosion development of multimedia information on the Internet and online sharing community, social network and Community multimedia data have received plenty of attention. In order to analyze and mining the structure and characteristics of online network communities with their large number of multimedia data automatically, we look into community photos as an basic data source for community social network construction and community organization structure analysis. By constructing the social network with community photo collection through face recognition algorithms, social network analysis method is used to analysis the community properties and individual characteristics. The effectiveness of different face recognition algorithms on community network construction is also calculated by three defined variables. Experimental analysis results of a real community album demonstrate the community social network characteristics and the influence brought by face recognition algorithms. The analysis method will provide new approach for using the online multimedia information to understanding the community organization structure and characteristics.