Online Discussion Transition Analysis for Group Learning Support

Group learning has been recognized as an effective learning method. Recently, group learning using online communication tools and face-to-face group learning have been increasing. This change in the learning environment has highlighted the importance of supporting teachers in monitoring the learning progress of each group. In this study, we propose an algorithm to analyze the discussion transitions. The algorithm first extracts the entities and Wikipedia categories from the discussion text. It then applies a queue model, which limits the analysis to the most recent topic, or a memory model, which emphasizes repeated topics. Subsequently, it triggers another algorithm that allocates a Wikipedia category as the main theme for each discussion text. To evaluate the effectiveness of the proposed algorithms, we tested them on the W2E dataset, which includes topics comprising multiple events and an actual conversational dataset. The test results confirmed that the algorithms analyzed the discussion transitions and main themes with high accuracy.