Identifying influential nodes in a social network

Abstract

Detecting influential nodes in a complex network is an on-going field of research with much practical interest. It has applications in a variety of fields, such as neuroscience, zoology, marketing, sociology, cybernetics, and more. These areas benefit from a method to study how information is diffused in complex networks. One could, for example, use such methods to model the spread of a pandemic and track the people most prone to the infection (Lu, et al., 2016). Nodes which diffuse a lot of information are called influential nodes. This means they are good receivers as well as senders of information, and therefore are the main conduits for information in their network. It is important to decide on a measure of influence to be able to identify and understand influential nodes. The problem is further complicated by the complexity of the network at hand. Large networks are usually formed from many smaller networks, which are termed communities within the network. This is analogous to Facebook (a complex network) being composed of several smaller, more localised groups of users (communities). Another common issue in real-life application is that the complete network structure may not be known (Tsugawa & Kimura, 2018).