Comment Data Mining to Estimate Student Performance Considering Consecutive Lessons

The purpose of this study is to examine different formats of comment data to predict student performance. Having students write comment data after every lesson can reflect students’ learning attitudes, tendencies and learning activities involved with the lesson. In this research, Latent Dirichlet Allocation (LDA) and Probabilistic Latent Semantic Analysis (pLSA) are employed to predict student grades in each lesson. In order to obtain further improvement of prediction results, a majority vote method is applied to the predicted results obtained in consecutive lessons. The research findings show that our proposed method continuously tracked student learning situations and improved prediction performance of final student grades.