I develop interaction techniques for mining context in large-scale web data to create/support learning opportunities in the wild. I study computational representations to structure the collective contexts mined, based on which I build interfaces for expanding people's abilities to search, analyze and make decisions to improve individual and collective learning experiences.
I'm interested in the interaction bounds of interfaces, and the resulting strategic decisions users are forced to make. For example, I've looked at
I build alternative systems to these computationally bounded interfaces by using techniques from interaction design, human computation, and machine learning. At the same time, I aim to leverage these situations to learn about diverse explanations, reasonings, relationships between the choices users make with a vision that they will not only expand what people can learn but also expand what machines can learn.
I also spend a lot of my time thinking about ways we can improve learning at scale (i.e. MOOC) with peer learning and with interactions that leverage peer dynamics.