Overview

This repository contains code for the paper:

Alan Tsang*, Bryan Wilder*, Eric Rice, Milind Tambe, Yair Zick. Group-Fairness in Influence Maximization. IJCAI 2019. [arXiv].

@inproceedings{tsang2019group,
  title={Group-Fairness in Influence Maximization},
  author={Tsang, Alan and Wilder, Bryan and Rice, Eric and Tambe, Milind and Zick, Yair},
  booktitle={International Joint Conference on Artificial Intelligence},
  year={2019}
}

Run check_fairness.py in order to compare the performance of the three algorithms in the paper on the synthetic Antelope Valley networks (included in the networks folder). check_fairness_intersectional.py runs the experiment where nodes have multiple group memberships.

Dependencies

• Optionally, you can use Gurobi to solve the inner maxmin LP instead of the mirror descent algorithm discussed in the paper. This may be faster and require less tuning for small problems.