Online team-based game development discussions patterns summarised using probabilistic models

Communications between team members sent as messages about collaboration strategies during online game development were analysed to extract contributions to the discussion of their online activity and their reflections. The task was the collaborative development of online games by a team of university students. Their social media communications were classified into four functions: Proposal, Permission, Encouragement, and Acknowledgement based on the participant's frequency of contribution of reflections and the frequency of their communication activity. In the results, some significant relationships between participant's reflections and communication frequencies of the four categories were extracted in the later discussion cycles. Also, the appearance probabilities of communication activity in each of the four categories were calculated using Bayesian networks and the scores of participant's characteristics, such as personality and information literacy.