Impact and degree of user sociability in social media

Users' posts receive feedback from other users in the form of comments, trackback, or recommendation in social media. These interactions form a graph in which the vertices represent a set of users, while the edges represent a set of feedback. Thus, the problem of users' rankings can be approached in terms of the link analysis of the social interactions between the users themselves within this graph. Link analysis algorithms, such as PageRank and HITS, have often been applied for users' rankings, especially for users' reputation, but no consideration has been given to how the user's sociability can affect the user's reputation. We propose two factors that affect the score of every user, the user's reputation, and the user's sociability, to address this problem. We present novel schemes that can effectively and separately estimate the reputation and sociability of the users. Furthermore, we present schemes to measure the degree of the user's sociability in a social network. Our experiments show that: (1) our schemes can effectively separate the user's pure reputation from the user's sociability (2) the pure reputation is capable of producing superior user ranking results than can previous work (3) the degree of user sociability for each social network varies and reveals significant characteristics of the corresponding network.