Measuring Social Influence with Networked Synthetic Control

TL;DR

This paper introduces a novel measure called social value that combines machine learning and network science to quantify social influence, addressing the challenge of counterfactuals in social influence measurement.

Contribution

It presents a new synthetic control-based measure for social influence that accounts for external regressors and network structure, with theoretical properties and computational improvements.

Findings

Social value diverges from traditional centrality measures.

The generalized friendship paradox is validated through simulations.

Computational reduction techniques are proposed for ensemble models.

Abstract

Measuring social influence is difficult due to the lack of counter-factuals and comparisons. By combining machine learning-based modeling and network science, we present general properties of social value, a recent measure for social influence using synthetic control applicable to political behavior. Social value diverges from centrality measures on in that it relies on an external regressor to predict an output variable of interest, generates a synthetic measure of influence, then distributes individual contribution based on a social network. Through theoretical derivations, we show the properties of SV under linear regression with and without interaction, across lattice networks, power-law networks, and random graphs. A reduction in computation can be achieved for any ensemble model. Through simulation, we find that the generalized friendship paradox holds -- that in certain situations, your friends have on average more influence than you do.