With the rise of smart cities, the workforce component also need to adapt to its new requirements. Current students and workers must transition to a mode of continuous learning, which goes beyond the confines of the usual University 4-year education. Our project examines how to employ educational technology at scale, and with personalization. These goals, while seemingly contradictory, are necessary to achieve better learning outcomes while minimizing the cost of faculty involvement.
Our research takes the approach on leveraging massive educational data that has been provisioned through large online learning platform such as massive online open courses, as well as freely available in instructional materials available on video and instructional websites.
We focus on two research tasks. The first is on building tools for personalized learning and instruction, including tools for instructor intervention, class management and courseware organization. Research on instructor intervention aims to better predict which conversations in a discussion forum merit intervention from instructional staff. Work on courseware organization aims at building knowledge structures to support learners, inclusive of course pathway prerequisite chains and concept knowledge graphs. The second research task aims to create a chat agent that can help students in online courseware system. It offers advice to guide students with behavioral modelling that is then ploughed into interventions such as course recommendation, video navigation, and topical question answering.