Architect for Software Development
In February 2013, Northwestern University partnered with Coursera to begin delivering Massive Open Online Courses (MOOCs) as part of our diversified online learning strategy. The “Massive” scale of these courses make tools showing learner activity in aggregate and by cohort extremely useful to our faculty. The Coursera platform provides some of these analytics tools, allowing instructors to track broad reach, engagement, and performance in their MOOCs. A valuable part of the completely online learning experience, however, is participation in discussion forums. The free text discourse that occurs in the forums is not easily tracked by quantitative, event-oriented, learning analytics. The Northwestern dashboard provides to faculty a view into MOOC-scale forum activity by combining forum metadata with text analysis and visualization techniques. The project performs entity extraction (people, places, organizations, and keywords) and sentiment analysis on forum posts, combined with geographic location information to enhance forum posts metadata for faceted searching and subsetting. Search results display facet values in aggregate using a variety of visualizations techniques, such as world map, heat map, word clouds, network graphs, stream graphs, and pie charts. Discussion forum activity can then be probed in aggregate by a variety of demographic, curricular, temporal, and topical characteristics.