Discerning Affect in Student Discussions

Students' emotions and attitudes are discernible in messages posted to online question and answer boards. Understanding student sentiment may help instructors identify students with potential course issues, optimize help-seeking, and potentially improve student achievement, as well as identify both positive and negative actions by instructors and provide them with valuable feedback. Towards this end, we present a set of context-independent emotion acts that were used by students in a university-level computer science course to express certainty and uncertainty, frustration, and politeness in an online Q&A board and develop viable classification approaches. To explore the potential of sentiment-based profiling, we present a heuristic-driven analysis of thread resolution and detail future research.