Easier contagion and weaker ties make anger spread faster than joy in social media

Similar to face-to-face communication in daily life, more and more evidence suggests that human emotion also spread in online social media through virtual interactions. However, the mechanism underlying the emotion contagion, like whether different feelings diffuse unlikely or how the spread is coupled with the social network, is rarely investigated. Indeed, because of the costly expense and spatio-temporal limitations, it is challenging for conventional questionnaires or controlled experiments. While given the instinct of collecting natural affective responses of massive connected individuals, online social media offer us an ideal proxy to tackle this issue from the perspective of computational social science. In this paper, based on the analysis of millions of tweets in Weibo, a Twitter-like service in China, we surprisingly find that anger is more contagious than joy, indicating that it can sparkle more angry follow-up tweets and anger prefers weaker ties than joy for the dissemination in social network, indicating that it can penetrate different communities and break local traps by more sharings between strangers. We conjecture that easier contagion and weaker ties could together make anger spread faster than joy, which is further justified by the diffusion of bursty events with different dominant emotions. To our best knowledge, for the first time we quantificationally disclose the difference between joy and anger in dissemination mechanism and our findings would shed lights on personal anger management in human communication and collective outrage control in cyber space.