Invited Speakers
Niclas Boehmer (Hasso Plattner Institute, Potsdam, Germany)
"To a Man with a Hammer, Everything Looks Like a Nail": Seeing Multiwinner Elections in Multiobjective Optimization and Collective Response Systems
Abstract: In the spirit of Mark Twain's observation, I will present two problems that can be modeled as multiwinner elections: (1) selecting a representative set of Pareto-efficient solutions in a multiobjective optimization problem, and (2) selecting a representative set of statements reflecting the full spectrum of opinions in a user discussion. In both cases, I will begin with context from the application domain before turning to the axiomatic, algorithmic, and experimental perspectives that multiwinner voting brings to these problems. In turn, I will discuss new questions these applications pose for social choice theorists.
The talk's content is based on joint papers with Sara Fish, Ariel Procaccia, and Maximilian Wittmann.
Bio: Niclas Boehmer is an Assistant Professor for Algorithmic Decision Making and Society at the Hasso Plattner Institute (HPI), University of Potsdam. Previously, Niclas completed a postdoc at Harvard University, focusing on AI for Social Good, and a PhD in theoretical computer science at TU Berlin. In his work, Niclas designs and analyzes algorithms that make "fair" decisions for groups of individuals with competing interests.
Nick Mattei (Tulane University, USA)
Leveraging Data (PrefLib) and Computational Social Choice for Human Centered Artificial Intelligence
Abstract: In recent years there has been an explosion in interest in topics that sit at the intersection of applications of computing technology and societal issues. There has been significant work in the academic, industrial, and policy spaces to clarify and formalize best practices regarding the deployment of computational decision making (e.g., artificial intelligence and machine learning) at scale. Part of this work has been a new found interest in many age old conversations about the roles and limits of technology and society. In this talk I’ll give an overview of recent work covering several projects that use tools from computational social choice, data science, and artificial intelligence more generally to build systems and algorithms that are context aware, flexible, and human-centered. I'll do this through the lense of PrefLib, and efforts to increase the usage of data and experiment in Social Choice, and how we can continue to use these ideas across areas like recommender systems and community engaged AI projects.
Bio: Nicholas Mattei is an Associate Professor of Computer Science at Tulane University and Co-Director of the Tulane Center of Excellence for Community Engaged Artificial Intelligence. His research lies at the intersection of artificial intelligence and machine learning with topics in data science, economics, decision making, and psychology. Most of my projects leverage theory, data, and experiment to create novel algorithms, mechanisms, and systems that enable and support individual and group decision making. He has published over 100 articles in top conferences and journals focusing on the technical areas of computational social choice, preference learning, multi-agent systems and others. He has received funding from the NSF, IBM, Google, Dept. of Education, and the Naval Research Laboratory; including a 2024 NSF CAREER award. Before joining Tulane, he was a Research Staff Member at IBM Research, TJ Watson Research Laboratory in New York, USA; a Senior Researcher at Data61/CSIRO (Commonwealth Science Research Organization, formerly NICTA) in Sydney, Australia; and an aerospace technology engineer at NASA Ames Research Center in Mountain View, CA, USA. I completed my PhD in 2012 at the University of Kentucky.
Ulrike Schmidt-Kraepelin (Eindhoven University of Technology, The Netherlands)
Rolling the Dice, Carefully
Abstract: In recent years, the idea of best-of-both-worlds randomization has gained significant traction in the social choice literature, particularly in areas such as fair division, committee voting, and apportionment. The premise is simple yet powerful: design randomized mechanisms that achieve the strongest possible ex-post (realized) guarantees of deterministic algorithms, while simultaneously offering the strongest possible ex-ante (expected) guarantees of randomized algorithms.
The first part of this talk revisits randomized apportionment, a topic long thought to be settled. We show that careless randomization can introduce non-monotonicities in correlations, which can lead to strategic voting when parties care about joint events, such as jointly reaching a majority in parliament. We propose natural monotonicity axioms to rule out these paradoxes and analyze which randomized apportionment rules satisfy them.
While randomization is often seen as a purely theoretical tool, the second part of the talk turns to a practical challenge: designing sampling methods for inviting citizens to participate in citizens' assemblies in countries without a central register, such as Germany. By adapting ideas from randomized apportionment, we develop algorithms that are fair in expectation while also producing explainable outcomes in every realization.
The talk is based on joint work with José Correa, Paul Gölz, Jan Maly, Jamie Tucker-Foltz, Markus Utke, Victor Verdugo, and Philipp Verpoort.
Bio: Ulrike Schmidt-Kraepelin is an assistant professor at TU Eindhoven, working on topics such as multi-winner elections, apportionment, portioning, and fair division. Before moving to the Netherlands, she was a postdoctoral researcher at Universidad de Chile in Santiago, hosted by José Correa, and at SLMath (formerly MSRI) in Berkeley. She received her PhD from TU Berlin in November 2022, where she was advised by Markus Brill.
