ABSTRACT:
In online knowledge communities (OKCs), humans and bots collaborate at massive scale, yet how their distinct resources shape collaboration patterns remains insufficiently understood. In particular, it is unclear whether humans consistently direct collaborative processes or whether bots can assume leading roles. Grounded in coordination theory, this study applies gSpan and local process model mining to 200 Wikipedia articles to examine how flow, fit, and sharing dependencies shape human-leadoff and bot-leadoff collaboration patterns. The results show that bots can function as initiating actors, guiding task flows in ways comparable to humans. Fit and sharing dependencies are key pathways linking human and bot resources to collaboration patterns. Task complexity exerts a moderating influence, revealing boundary conditions for coordination in human–bot collaboration. This study extends coordination theory to open, human–bot collaboration contexts and provides actionable guidance for managing task dependencies and fostering effective collaboration between humans and bots in OKCs.
Key words and phrases: Knowledge communities, human-bot teams, task dependencies, coordination theory, human-bot collaboration, bot functioning