Abstract:Social influence pervades our everyday lives and lays the foundation for complex social phenomena. In a crisis like the COVID-19 pandemic, social influence can determine whether life-saving information is adopted. Existing literature studying online social influence suffers from several drawbacks. First, a disconnect appears between psychology approaches, which are generally performed and tested in controlled lab experiments, and the quantitative methods, which are usually data-driven and rely on network and event analysis. The former are slow, expensive to deploy, and typically do not generalize well to topical issues (such as an ongoing pandemic); the latter often oversimplify the complexities of social influence and ignore psychosocial literature. This work bridges this gap and presents three contributions towards modeling and empirically quantifying online influence. The first contribution is a data-driven Generalized Influence Model that incorporates two novel psychosocial-inspired mechanisms: the conductance of the diffusion network and the social capital distribution. The second contribution is a framework to empirically rank users' social influence using a human-in-the-loop active learning method combined with crowdsourced pairwise influence comparisons. We build a human-labeled ground truth, calibrate our generalized influence model and perform a large-scale evaluation of influence. We find that our generalized model outperforms the current state-of-the-art approaches and corrects the inherent biases introduced by the widely used follower count. As the third contribution, we apply the influence model to discussions around COVID-19. We quantify users' influence, and we tabulate it against their professions. We find that the executives, media, and military are more influential than pandemic-related experts such as life scientists and healthcare professionals.

Estimating Social Influence Using Latent Space Adjusted Approach in R

Evaluating methods for estimating influence effects in social networks: a simulation study

Estimating Social Influence from Observational Data

Estimating Causal Peer Influence in Homophilous Social Networks by Inferring Latent Locations

How Social Networks Influence Human Behavior: An Integrated Latent Space Approach for Differential Social Influence

Estimation of Peer Influence Effect in Online Games Using Machine Learning Approaches.

Individual-level Social Influence Identification in Social Media: A Learning-Simulation Coordinated Method.

Latent Space Approaches to Social Network Analysis

A Popularity Scaled Latent Space Model for Large-Scale Directed Social Network

Bias reduction of peer influence effects with latent coordinates and community membership

Estimating Node Influenceability in Social Networks

Distinguishing Homophily from Peer Influence Through Network Representation Learning

Learning hidden influences in large-scale dynamical social networks: A data-driven sparsity-based approach

Conductance and Social Capital: Modeling and Empirically Measuring Online Social Influence

Influence Networks: Bayesian Modeling and Diffusion

A Data-Based Approach to Social Influence Maximization

Statistical Inference on Latent Space Models for Network Data

DeepInf: Modeling influence locality in large social networks

Link prediction via latent space logistic regression model

Identifying Peer Influence in Online Social Networks Using Transfer Entropy.

An Empirical Evaluation of Social Influence Metrics.