Schedule

9:00 - 9:15 Gather, Settle, and Opening Remarks!

9:15 - 10:30 Technical Session (4 talks) Each talk: 10 minutes + QA

- The Algorithmic Landscape of Fair and Efficient Distribution of Delivery Orders in the Gig Economy
  Hadi Hosseini, Šimon Schierreich
- Maximum Welfare Allocations under Quantile Valuations
  Haris Aziz, Shivika Narang, Mashbat Suzuki
- Temporal Fair Division
  Benjamin Cookson, Soroush Ebadian, Nisarg Shah
- Centralized Group Equitability and Individual Envy-Freeness in the Allocation of Indivisible Items
  Ying Wang, Jiaqian Li, Tianze Wei, Hau Chan, Minming Li

Coffee break (with Posters)

11:00 -11:45 Keynote (Dr. Dominik Peters)

11:45 -12:30 Technical Session (4 talks) Each talk: 10 minutes + QA

Don’t Try This at Home: Examining How LLMs Perform Fair Division
Benjamin Cookson, Soroush Ebadian, Nisarg Shah
Achieving Fairness in Zoning Laws with Machine Learning
William Schimitsch, Farhad Mohsin
Fair Allocation of Service Tasks with Two-Dimensional Costs
Hadi Hosseini, Minming Li, Shivika Narang, Zhehan Yu
How to Resolve Envy by Adding Goods
Eva Deltl, Robert Bredereck, Leon Kellerhals, Pallavi Jain, Matthias Bentert

Lunch break

14:00 - 15:30 Technical Session (4 talks + 1 demo) Each talk: 10 minutes + QA

Fairly Stable Two-Sided Matching with Indifferences
Benjamin Cookson, Nisarg Shah
Metric Distortion in Peer Selection
Javier Cembrano, Golnoosh Shahkarami
Equitable Mechanism Design for Facility Location
Toby Walsh (video)
Mechanism Design for Facility Location using Predictions
Toby Walsh (video)
MatchU.ai and MatchXplain (demo)
James Ferris, Hadi Hosseini, Yubo Jing, Ronak Singh
Overview: The demonstration will showcase MatchU.ai and MatchXplain, two interactive websites developed by Prof. Hosseini's research group. These platforms integrate fair division and social choice algorithms with explainable AI (XAI) and visualizations to make a suite of tools for fair division, preference analysis, and two-sided matching accessible to the public. The demo will walk through the standard user journey on each site, from entering preferences to viewing potential solutions. Throughout, the demo will highlight the novel visualizations, XAI-driven insights, and educational resources that allow users to utilize the power of fair division and social choice algorithms, regardless of their prior expertise. Both platforms are currently available for public use.

Coffee Break (with Posters)

16:00-16:30 Poster Session

16:30 - 17:30 Panel Discussion (co-organized with SCaLa workshop)

Overview: Social Choice – including the study of agent interactions, fair division, allocation, voting, computational aspects, and all the other sub-areas, has long thought about what it means to acquire, manipulate, and decide from preferences. The growing use of data and GenAI systems across society has unveiled new ways to acquire data, make decisions, and apply our theoretical and algorithmic tools to new and important domains. In this panel, we welcome four researchers from across social choice and we’ll focus on discussing how social choice, broadly construed, interacts with our current environment, including increased data-fication, GenAI, and even more agents than we ever thought possible.

Panelists:

Edith Elkind: Ginni Rometty Professor of Computer Science, Northwestern University
Dominik Peters: Computer Science Researcher at LAMSADE
Andrzej Kaczmarczyk: Postdoctoral Fellow at The Strategic IntelliGence for Machine Agents (SIGMA) Lab at Chicago University
Alan Tsang (to confirm): Associate Professor of Computer Science, Carleton University

Moderated by: Nicholas Mattei

Keynote Address:

What Private Goods Can Teach Us About Public Goods: Computing Lindahl Equilibrium

Dr. Dominik Peters

Lindahl equilibrium is a solution concept for allocating a fixed budget across several divisible public goods. It always lies in the core, meaning that the equilibrium allocation satisfies desirable stability and proportional fairness properties. But little was known about how to compute Lindahl equilibrium. In the simplest case (sometimes known as the portioning or fair mixing setting), the fastest algorithm in practice is based on computing a proportional response dynamics, but this did not have a convergence rate guarantee. In more involved cases (which we call the "capped public goods setting" and which captures problems like fractional committee elections and the cake sharing model), no polynomial time algorithm was known, and the computational complexity of finding a Lindahl equilibrium was a long-standing open problem. In this talk, I will explain how we can use techniques developed to analyze Fisher markets to make progress on Lindahl equilibrium computation. We introduce a new convex program for this task which is structurally similar to Shmyrev's program for Fisher markets. We use it to prove that the dynamics converges at a 1/t rate, and show that it can be used to compute equilibria in polynomial time even in the capped setting. Based on joint work with Christian Kroer.

Page updated

Google Sites

Report abuse