Disclaimer: Any similarity or resemblance of the project’s name to a certain cinematic universe is purely non-coincidental. That said, I personally have no idea about what goes on in MCU. I use this analogy to stress that research is as fun as watching an MCU movie (assuming most people love the latter).

The DiRe Research Universe (DiRe - Diversity and Representation) is a research project that uses insights from theoretical computer science to delineate the notions of diversity and representation. Currently, the project focuses on understanding the key differences between diversity and representation in the context of multiwinner elections. Specifically, on the surface, diversity and representation in multiwinner elections may seem mathematically equivalent just like the vertex cover problem on d-regular hypergraphs and on k-uniform hypergraphs, respectively. However, on zooming in, these are as different as the vertex cover problem on d-regular hypergraphs and on k-uniform hypergraphs, respectively. The vision is to extrapolate this understanding to a more general setting.

What does DRU currently encompass?

At a high level, DRU extends the DiRe committee paper [1] from two perspectives: fairness and theoretical. The former aims to showcase the systematic unfairness that may be caused by a diverse and representative outcome, and the latter aims to study the existence of DiRe outcomes and its corresponding complexity (read: is finding a diverse and representative outcome (of a certain size) PPAD( or PPA)-complete (under certain realistic assumptions)?).

Why care about DRU?

The need to reduce inequality is ubiquitous. At least three of the United Nations’ Sustainable Development Goals directly focus on this aim. However, the incorrect conflation of diversity and representation may be doing more harm than good. For instance, a paper published in Nature and a resolution presented in the United States Senate incorrectly conflate these two terms. Hence, in the pursuit to reduce inequality, DRU intends to highlight the importance of delineating diversity and representation. Overall, if you care about reducing inequality or fair AI or responsible AI, this universe will interest you. (If you are a technical geek, it may be an added reason!!)

Releases

[1] DiRe Committee: Diversity and Representation Constraints in Multiwinner Elections.
Relia, K.
Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, IJCAI-22. link

[2] Fairly Allocating Utility in Constrained Multiwinner Elections.
Relia, K.
Working paper, 2022. pre-print

[3] On the Complexity of Finding a Diverse and Representative Committee using a Monotone, Separable Positional Multiwinner Voting Rule.
Relia, K.
Working paper, 2022. pre-print (Some random musing - forthcoming)

[4] Next preprint coming soon in arXiv near you!
AY 2023-24

Acknowledgement

The universe was produced by (a.k.a. research was generously supported by) Julia Stoyanovich’s NSF grants.

(Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation or the PI.)