Identification and Prediction of Opinion Leaders in Large Scale Enterprise Social Networks (ESNs)

Abstract

Increased application of Enterprise Social Network (ESN) platforms at the workplace extended the sources of influence on corporate social communities from management to unofficial opinion leaders that can have a direct influence on individuals' decision-making using ESN platforms. This study has attempted to examine the relationship between network structural positions and behaviours of ESN users and their ability to influence other users via ESN platforms. By analysing the pattern of user interactions via the ESN platform (Yammer), this study revealed that over (𝐍≅𝟑𝟑%) of users’ who voluntary participated at online group discussion forums were triggered by less than (𝐧≅𝟑%) of opinion-makers in ESN platforms. Correspondingly, the results of this study showed that membership in various online workgroups, structural position of actors in social networks, and the degree of actors and groups’ online activities largely impact a user’s ability to become an opinion leader or an influencer within corporate social environments. Looking further into the results it is revealed that memberships in various workgroups significantly promote the users’ ability of influence, however, the number of direct ties do not promote the degree of influence with the same rate. This means users who are members of highly dynamic online groups are more likely to influence others compared with those who are members of limited communities regardless of the number of social communities’ members. Considering these facts, this study has proposed a model to compute an influence score which represents the ability of an ESN’s user to influence others in decision making. Subsequently, a set of predictor features were extracted to successfully predict this score.