Peng, Xian, et al. "Topic tracking model for analyzing student-generated posts in SPOC discussion forums." International Journal of Educational Technology in Higher Education 17 (2020): 1-22.

This study addresses the challenge of navigating and tracking valuable information in small private online courses (SPOCs) by analyzing student-generated forum posts with an advanced dynamic topic model, the Time Information-Emotion Behavior Model (TI-EBTM). This model incorporates time, emotion, and behavior into the topic modeling process, enabling effective tracking of topic evolution and understanding of emotional and behavioral changes over time. Applied to data from 30 SPOC courses, TI-EBTM demonstrated superior performance in extracting significant topics and provided insights into how students' focus and emotional responses shift, particularly highlighting a tendency for increased negative emotions towards certain topics towards the end of the semester. The findings can help instructors optimize forum interactions and enhance students' learning experiences.