Lecture 12: Centrality Measures

Having built our own social graphs, we are all familiar with graphs and network visualizations now. But, it is important to understand that building colorful and good-looking graphs is not the main objective, we need to analyze and interpret the graphs in a proper way to make full use of it. So, I thought I should write about something that is significant and really helpful for all of us. This blog is about one important property called Centrality and how different centrality measures can be used to analyze a graph.

There are different measures of centrality available, to determine the relative importance of a node in a graph. They are:

    1. Degree centrality: This considers the node that has the most number of interactions in a graph as the most important node. So, when this measure is used, all the nodes that have relatively high number of interactions will occupy the center portion of the graph. The picture below illustrates this concept:

     2. Closeness centrality: The node which is close to most of the other nodes in a graph is considered as the most important node. It also implies that this node can communicate easily and quickly to the other nodes in the network. When this measure is used, the node that is most closely connected to other nodes will be at the center of the graph.

    3. Betweenness centrality: The most important node in this measure is the one that appears in most number of shortest paths in the network. This could be a very important measure as it indicates the node that increases the connectivity. Below is an interesting image to illustrate this concept:

[Find out which of these has a high betweenness centrality :) ]

These three measures are highly different from each other. You should use the one that is most appropriate to the nature of the network and kind of analysis involved.