Abstract— Online discussions serve as a bridge between face-to-face and virtual lessons in the distance education. Previous researches have investigated the impact of student participation in online discussions on learning performance. However, student opinions expressed in the posts are also good indicators to student learning. The recent achievements of the natural language processing technologies, especially sentiment analysis in Chinese makes it possible to automatically detect the polarity of each course post. In this study, we collected online posts from 62 students in a distance education course program and used machine learning algorithms to build a sentiment classifier. The features were extracted from a recent developed Chinese affective lexical resource. Finally, we examined the impact of the post opinion on the student final grade. As a result, it is found that the decision tree classifier got the highest Marco-average F1-score (0.693) and the correlation between the student post sentiments and final grade is moderate.

Index Terms— Sentiment analysis, student forum posts, language technologies.

Ming Liu, Weiwei Xu, Huiran Liu, Jiaqin Zheng is with Southwest University, Beibei District, Chongqing, China, 400715. Tel.: +86 15178719397; email: mingliu@swu.edu.cn

Cite: Ming Liu, Weiwei Xu, Huiran Liu, Jiaqin Zheng, " Sentiment Analysis in Chinese Web Discussion Forums," International Journal of e-Education, e-Business, e-Management and e-Learning vol. 5, no. 1, pp. 47-54, 2015.