Sridhar, D., Bacco, C., & Blei, D. (2022). Estimating Social Influence from Observational Data. CLeaR.
Sridhar, Dhanya, C. Bacco, and D. Blei. “Estimating Social Influence from Observational Data.” CLeaR (2022).
Sridhar, Dhanya, et al. “Estimating Social Influence from Observational Data.” CLeaR, 2022.
We consider the problem of estimating social inﬂuence, the effect that a person’s behavior has on the future behavior of their peers. The key challenge is that shared behavior between friends could be equally explained by inﬂuence or by two other confounding factors: 1) latent traits that caused people to both become friends and engage in the behavior, and 2) latent preferences for the behavior. This paper addresses the challenges of estimating social inﬂuence with three contribu-tions. First, we formalize social inﬂuence as a causal effect, one which requires inferences about hypothetical interventions. Second, we develop Poisson Inﬂuence Factorization (PIF), a method for estimating social inﬂuence from observational data. PIF ﬁts probabilistic factor models to networks and behavior data to infer variables that serve as substitutes for the confounding latent traits. Third, we develop assumptions under which PIF recovers estimates of social inﬂuence. We empirically study PIF with semi-synthetic and real data from Last.fm, and conduct a sensitivity analysis. We ﬁnd that PIF estimates social inﬂuence most accurately compared to related methods and remains robust under some violations of its assumptions.