Optimization and intervention in point processes networks with application to social campaigning and fake news mitigation

Steering the activities on social networks to a desirable outcome is of many practical, economic, and societal interest. For example, can one model and exploit social network dynamic behaviors to steer the online community to a desired activity level? Specifically, can one drive the overall exposure to a campaign to a certain level (e.g., at least twice per day per user) by incentivizing a small number of users to take more initiatives? What about maximizing the overall service usage for a target group of users? How about making sure that a job opportunity news is shared uniformly and without discrimination in the social network? Moreover, is there any efficient way to mitigate a fake news propagating on the network?
We model social events using multivariate Hawkes processes, which can capture both endogenous and exogenous event intensities and formulate the problem as a Markov decision problem. We propose a convex dynamic programming framework to find the optimal policy that balances the high present reward and large penalty on low future outcome in the presence of extensive uncertainties. This talk will establish the fundamentals of intervention and control in networks by combining the rich area of temporal point processes and the well-developed framework of Markov decision processes.