Social affinity and facial emotion based social correlation

Discovering the hidden correlation among the people help us to understand the relationship amongst them as well as the influence propagating between them. Further, this knowledge of correlation could be helpful in consensus building or locating the most influential person in the group. And could be used in different application; such as, negotiation, viral marketing, surveys, etc. In any network the correlation exist mainly because of three reasons; namely, homophily, cofounding and influence. Homophily is seen when any two node share some social ties among them,so their actions and reaction are correlated. Similarly cofounding exists when two nodes share hobbits, belief, interests, etc., so there is some correlation between them. Whereas, in order to discover the influence in the network, action-reaction pair between any nodes are need to be tracked; the action and reaction can be online activities, body pose, spoken words, emotions, etc. There are many similar works to discover the correlation in social networks; the analysis of social behaviour propagation in a social network is the integration of social theories with computational methods [1], [2]. The goal our work is to analyse the ways in which these group members induce their actions (emotions) to other members and based on that building the correlation amongst them. Whereas, there are three basic sources of human emotions: facial, speech or written text. As far as this work is concerned, we will discuss about building correlation based on the facial emotions.