Discovery of Community Structures in a Heterogeneous Professional Online Network

Abstract

Socio-technical networks that are heterogeneous in composition of actors and the media through which they interact are becoming common, but opportunities to study the emergent community structure of such networks are rare. We report a study of an international online network of educators involved in many forms of professional development and peer support, including sponsored and volunteer-driven activities taking place in both synchronous and asynchronous media, with participants from diverse career stages and occupations in education. A modularity-partitioning algorithm was applied to a directed, weighted, multimodal graph that represents associations between actors and the artifacts (chats, discussions and files) through which they interact. This analysis simultaneously detects cohesive subgroups of actors and artifacts, providing rich information about how communities are technologically embedded. Researchers deeply familiar with the network validated the interpretability of the partitions as corresponding to known activities, while also identifying new findings. The paper describes this interpretative validation, summarizes findings concerning the distribution and nature of communities and groups found within the larger heterogeneous network, and discusses open research questions and implications for practitioners.