Group Activity Recognition to Support Collaboration in Creative Digital Space

In order to orchestrate collaborative group work, such as a brainstorming workshop or active learning class, a facilitator must monitor all groups simultaneously. In this paper, we propose a method for recognizing group activities, based on machine learning, where the input features are derived from the analysis of raw system events generated by a collaboration system known as "creative digital space" that we developed before. To verify the effectiveness of the proposed method, we also evaluated the results obtained by using the test data collected from active learning classes.