Title: Communities, modules, and large-scale structure in networks

Mark Newman

Short Abstract
Many networks, including social and technological networks, are found to divide into communities or modules, groups of network nodes with dense connections within groups and sparser connections between groups. The ability to detect and identify communities plays an important role in network visualization, data analysis, and other areas. In this talk I will discuss the theory and practice of community detection and demonstrate applications to networks from a range of fields. I will also discuss some recent results showing that in certain cases it is impossible to detect communities in networks, even though they are present, implying that there are fundamental limits to our ability to pry understanding from network data. Finally I will discuss other forms of large-scale structure in networks, such as hierarchical structure and overlapping communities, and methods for their detection. These studies are in their infancy but may lead us to new understanding of the connection between the structure and function of networked systems.