SentBuk: Sentiment analysis for e-learning environments

This paper presents SentBuk, a Facebook application that extracts information about the user sentiment automatically, in a non-intrusive way. It performs sentiment analysis of user writings in Facebook walls, classifying each sentence as positive, neutral or negative. Finally, the overall user sentiment is calculated. On one hand, this information is useful to enrich user models for adaptive e-learning systems, so that these systems can adapt any of their aspects (tasks to be proposed to the student, contents, and so on) according to each student sentiment, among other criteria. On the other hand, the polarity of the emotions transmitted by the students enrolled in a course can constitute a useful feedback for the course teacher.