Automatic Detection of Cognitive Presence in Online Collaborative Learning Discussions Using Knowledge Information

Learners' cognition is crucial in online collaborative learning, and cognitive presence is an important indicator of cognition. Therefore, how to automatically detect cognitive presence in student discussions is a hot topic in current research. However, existing studies have not considered the knowledge information in collaborative learning discussions, which is relevant to students' cognition. Based on this, this study proposed an automatic detection approach that utilizes knowledge information and deep learning models to automatically detect cognitive presence in student discussions. We conducted our experiments on a dataset containing 8795 pieces of data. The experimental results show that the proposed model achieves an accuracy of 77.63%, which outperforms all baseline models. Our work provides an accurate and efficient tool for automatically analyzing cognitive presence and helps teachers implement interventions.