An empirical study of behavior unconformity in web-based and non-web-based social networks

Social Networking Sites have recently been popular with researchers due to its wide coverage and heavy usage patterns. Much work has been done to investigate the interactions between web-based and non-web-based social networks, but not much attention is paid on the behavior unconformity between them. By applying nonnegative matrix factorization algorithm and visualization method to data collected from online and real-life social networks, we discover the link-patterns of web-based and non-web-based social networks among a certain group of students. By comparing the networking patterns, we prove the existence of behavior unconformity and show how behavior unconformity might strengthen the ties between the individuals.