Influence Maximization for Social Good: Use of Social Networks in Low Resource Communities
This addresses the problem of efficiently using social networks to aid low-resource communities like homeless shelters, representing an incremental advancement in applying influence maximization to social good.
The paper tackles the problem of influence maximization for social good by defining the Dynamic Influence Maximization Under Uncertainty (DIME) problem to model homeless shelter challenges, and proposes scalable POMDP algorithms (PSINET and HEALER) that were tested in a pilot study with homeless youth in Los Angeles, showing promise for larger-scale application.
This thesis proposal makes the following technical contributions: (i) we provide a definition of the Dynamic Influence Maximization Under Uncertainty (or DIME) problem, which models the problem faced by homeless shelters accurately; (ii) we propose a novel Partially Observable Markov Decision Process (POMDP) model for solving the DIME problem; (iii) we design two scalable POMDP algorithms (PSINET and HEALER) for solving the DIME problem, since conventional POMDP solvers fail to scale up to sizes of interest; and (iv) we test our algorithms effectiveness in the real world by conducting a pilot study with actual homeless youth in Los Angeles. The success of this pilot (as explained later) shows the promise of using influence maximization for social good on a larger scale.