Teachers play a key role in CSCL (Computer-Supported Collaborative Learning) settings in developing collaboration skills among students. This support demands teachers to be aware of students' activities to identify students in need for intervention purposes. Students while participating in CSCL leave learning traces (e.g., digital logs, posts, chat) and Learning Analytics (LA) has enabled analysis of those traces to understand collaborative learning with the aim to support it. This approach, however, can only offer a partial picture of collaboration behaviour. For example, imagine a group working on a problem and drafting the solution in a collaborative text editor. There is a possibility that every member of the group participated in the discussion towards solving the given problem and then at the end of the activity only one member drafted the solution in the editor. Digital traces-based LA will tell us that only one member was active, however, this was not the case. To holistically understand the CSCL, there is a need to collect multimodal data to capture students’ interaction in physical spaces as well.
Pankaj Chejara is a PhD fellow at the University of Tallinn. His research focuses on building a system to support teachers in monitoring collaboration behaviour among students. This research involves first identifying features from various types of data (e.g., audio, video, digital logs) which can be indicative of the quality of collaboration among students, second using this knowledge to build an automated feedback system to support teachers in tracking and monitoring students during the activity using machine learning techniques. His visit is part of the student exhange between University of Tallinn and SLATE through the SEIS: Scaling up Educational Innovation in Schools project.