20 years of network community detection

A fundamental technical challenge in the analysis of network data is the automated discovery of communities -- groups of nodes that are strongly connected or that share similar features or roles. In this commentary we review progress in the field over the last 20 years. Community detection is a rich and challenging problem, partly because it is not very well posed: what exactly do we mean by a community? In most cases, communities are defined as non-overlapping groups of nodes such that there are more edges within groups than between them, but this definition still leaves open many possibilities, and there are correspondingly many computational approaches. The most common approaches are based on optimization.