Characterizing collective knowledge sharing behaviors in social network

Knowledge sharing behaviors play important roles in the spread of new concepts, ideas, and emerging technologies. However, how knowledge is disseminated in large-scale populations and what kinds of factors are associated with the collective knowledge sharing behaviors are still open issues. In this paper, we characterize the knowledge sharing behaviors via collective interaction in large-scale Question and Answer (Q&A) social networks. Specifically, we study the knowledge sharing behaviors in Zhihu and Quora, and utilize kendall coefficient and logistic regression model to quantify the influence of factors linked with sharing behaviors including friend's influence and the users' interest. We find that the knowledge sharing behavior is contagious among friends. The contagion is asymmetric. The individuals who are less active or less popular are more likely to be affected. Meanwhile the trusted users in the community are more likely to infect others.