Human-Centered Data Science
Crisis informatics is a study of how people converge, propagate information, and cooperate around the tasks they deem important on social media in crisis. The socio-behavioral focus of crisis informatics necessitates that research methodology accounts for the social context of users’ activity. On the other hand, the volume of the social media data requires the use of data science approaches, which can often decontextualize the social activity. I strive to work with datasets and computational methods that also account for highly situated and contextual nature of the social activity in disruption.
Network science is a powerful framework for understanding and modeling relational data. In the case of the social media data, network science allows us to both retain the context of individual activity (the links) and lift out more coarse-grained patterns in the network structure produced through social interaction.
Natural Language Processing
Because most of the social media data is textual, I also work with the Natural Language Processing (NLP) methods such as topic modeling and sentiment analysis. I am also interested in applying more complex NLP methods to the dynamic social media data, such as the Mixed Membership Markov Model.
Utilizing ideas from the theory of random walks, we can model aspects of social activity as as stochastic processes. This can be a fruitful approach for understanding and modeling social dynamics, especially if the temporal or spatial dimensions of the activity are of interest.