English

Diverse Data Selection under Fairness Constraints

Data Structures and Algorithms 2020-10-20 v1

Abstract

Diversity is an important principle in data selection and summarization, facility location, and recommendation systems. Our work focuses on maximizing diversity in data selection, while offering fairness guarantees. In particular, we offer the first study that augments the Max-Min diversification objective with fairness constraints. More specifically, given a universe UU of nn elements that can be partitioned into mm disjoint groups, we aim to retrieve a kk-sized subset that maximizes the pairwise minimum distance within the set (diversity) and contains a pre-specified kik_i number of elements from each group ii (fairness). We show that this problem is NP-complete even in metric spaces, and we propose three novel algorithms, linear in nn, that provide strong theoretical approximation guarantees for different values of mm and kk. Finally, we extend our algorithms and analysis to the case where groups can be overlapping.

Keywords

Cite

@article{arxiv.2010.09141,
  title  = {Diverse Data Selection under Fairness Constraints},
  author = {Zafeiria Moumoulidou and Andrew McGregor and Alexandra Meliou},
  journal= {arXiv preprint arXiv:2010.09141},
  year   = {2020}
}
R2 v1 2026-06-23T19:26:11.483Z