Influence Maximization in an Uncertain World

Google map

Abstract: The potential of Artificial Intelligence to tackle challenging problems that afflict society is enormous, particularly in the areas of healthcare, conservation and public safety and security. Many problems in these domains involve harnessing social networks of under-served communities to enable positive change, e.g., using social networks of homeless youth to raise awareness about HIV (and other STDs). Unfortunately, most of these real-world problems are characterized by uncertainties about social network structure and influence models, and previous research in AI fails to sufficiently address these uncertainties, as they make several unrealistic simplifying assumptions for these domains. In this talk, I will describe my research on algorithmic interventions in social networks. In the first part of my talk, I will describe my work on developing new influence maximization algorithms which can handle various uncertainties in social network structure, influence models, etc., that commonly exist in real-world social networks. I will discuss how my algorithms utilize techniques from sequential planning problems and computational game theory to develop new kinds of algorithms in the sub-fields of multi-agent systems and reasoning under uncertainty. In the second part of my talk, I will discuss the real-world deployment of my algorithms to spread awareness about HIV among homeless youth in Los Angeles. This represents one of the first-ever deployments of computer science based influence maximization algorithms in this domain. I will discuss the challenges that I faced, and the lessons that can be gleaned for future deployment of AI systems.