An Empirical Evaluation Of Social Influence Metrics

TL;DR

This paper empirically compares various social influence metrics to evaluate their effectiveness in predicting individual influence adoption, considering classifier choice and data imbalance for practical application.

Contribution

It provides a comprehensive direct comparison of multiple social influence metrics and analyzes their predictive performance across different classifiers and data ratios.

Findings

Neighborhood-based measures perform well in influence prediction.

Structural diversity and temporal measures add predictive value.

Classifiers and data balance significantly impact prediction accuracy.

Abstract

Predicting when an individual will adopt a new behavior is an important problem in application domains such as marketing and public health. This paper examines the perfor- mance of a wide variety of social network based measurements proposed in the literature - which have not been previously compared directly. We study the probability of an individual becoming influenced based on measurements derived from neigh- borhood (i.e. number of influencers, personal network exposure), structural diversity, locality, temporal measures, cascade mea- sures, and metadata. We also examine the ability to predict influence based on choice of classifier and how the ratio of positive to negative samples in both training and testing affect prediction results - further enabling practical use of these concepts for social influence applications.