A System for Extracting Discussion Circumstances on Social Media to Promote Inexperienced Students Understand Discussion Cases

Abstract

Opportunities are increasing for students to conduct discussions in problem-based learning (PBL) while observing discussions on social media. However, they often lack experience in examining actual discussion cases. We aim to develop methods for extracting discussion circumstances on social media to provide them to students so that they can use them as clues to understanding actual discussion cases. Specifically, we extract what kind of topics exist and how utterances are exchanged as discussion circumstances. In this paper, we describe methods for extracting the discussion circumstances and an overview of a support system that we developed. Moreover, we describe an experiment on extracting the circumstances from actual conversation data and discuss the effectiveness of our methods on the basis of the results.