The future of both the computational social sciences and consumer electronics lies in an understanding of the "data exhaust trails" of networked people. Our current understanding of human dynamics is still poor, because those who have access to the data and those who have insight about the data speak different languages.

I introduce an approach that identifies the generative multi-agent models of human dynamics from the social and management sciences as stochastic processes, and fits these multi-agent models to big data with variational Bayesian methods and Monte Carlo methods to make inferences. This bridges the gap between data collectors and modelers.

As an example of this approach, I describe how our sensor network data track people. Our computational tool combines these data sets with social sciences models, enabling a new research path.

Wen Dong focuses on modeling human interaction dynamics with stochastic process theory. He has published dozens of papers on the network influence model, group dynamics, and human problem-solving, and has earned recognition from several big businesses for constructing successful data products. Wen received his Ph.D. from the MIT Media Laboratory, with the thesis Modeling the Structure of Collective Intelligence.