A Social Network Graphics Segmentation Algorithm based on community-detection

Today people use multiple online social networks. If we can identify all the accounts which belong to the same individual across different online social networks, we will get a richer understanding of social dynamics and apply it to a number of practical applications. Fortunately, some researchers have got some satisfactory achievements. Many related algorithms need to be calculated the “score” of a pair of nodes which will take much time, but some of those calculations may be redundant. In this paper, we come across a problem that reduces redundant calculations. The graphics segmentation methods give us some enlightenment to solve the problem. We can reduce redundant calculation by “cutting” social network graph to some small subgraphs. Then how to “cut” appears to be a challenge that how to “cut”. Considering the characteristics of the social network community, we will utilize the community segmentation algorithm to split large social networks and refactor subgraphs. Our experiments indicate that our algorithm takes 70% time of original algorithm.