Sociotechnical Trajectories: A Structural Framework for Analyzing Event Log Data

• In Event: Online Media II: Analyses of Communication Networks

Abstract

Sociotechnical systems like Wikipedia and GitHub encode a variety of relationships within the digital traces of their user contributions. While traditional social network analysis approaches capture explicitly declared relationships, they often ignore the implicit relations within sequential interactions such as commits. We propose a method using temporal adjacencies within event log data to produce structural representations of sequence data to identify the trajectories of both users and artifacts as they progress through the system over time. These sociotechnical trajectories reduce large quantities of digital trace data into more parsimonious network graphs. These trajectories reveal the structure of sequences such as regularities in action or anomalies in the behavior of users and artifacts that can motivate follow-on qualitative investigations of social roles, communities of practice, and institutional routines. We evaluate the potential of this approach with case studies of user and project activity on two large and complex sociotechnical systems, GitHub and Wikipedia.

Author

• Brian Christopher Keegan, Northeastern University