Centrality, the measure of influence

Centrality is an important concept in social networking and measures the influence of one node in its network. There are many approaches to measure centrality, such as degree centrality, betweenness centrality, eigenvector centrality and etc. Degree centrality may be the simplest and the most intuitive measure of centrality; however, why other measures also exist? This is because the imperfection in these centrality measures. I am not going to talk about the technical details about these centrality measures. It is very complex when we are trying to measure the influence. There are two questions that we need to answer. The first one is what influence is and the second one is what factors determine influence.

Influence is the power that determine the outcome of the entire network. When one is more influential, it means this person (or party) has greater power to determine the outcome. This is not a tricky question, and can be solved by a dictionary. However, the second question really makes the entire thing much more complicated. There are so many factors determining the influence of a node in the network. For example, the influence can be determined by how many other nodes are connected this node. The direction of information flow can be another determinant. There are other factors. Even if we are possible to list out all the determinants, there will be another issue, which is the most important determinant and which is the least important and how they are different from each other in terms of their power determining the influence. Different people can have different answers to these questions, which means there is no universally agreed measure of influence.