Group Discussion Analysis and Digression Intervention

Abstract

It is in common knowledge that reading is one of the richest sources of knowledge in this world. Reading empowers you with the light that leads you through the dark. Therefore, we attempt to promote this valuable skill with this study. In this paper, a platform is developed that facilitates the exchange of thoughts and information among students. We have leveraged NLP to develop this application and categorize texts into various categories. Further, various text   classification methods are introduced to derive meaningful insights from written communication among students regarding books. We go on to apply the information drawn from text classification to a technology that engages readers through interactive games and discussions, IMapbook. The conversational text acquired through these discussions is further classified into various categories based on the context. Here, we aim to build a classifier that can predict these categories. Our study shows that the fine-tuned BERT, outperforms all the other methods used in this research.