An Experimental Comparison of Multiwinner Voting Rules on Approval Elections

Authors: Piotr Faliszewski, Martin Lackner, Krzysztof Sornat, Stanisław Szufa

IJCAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this paper, we experimentally compare major approval-based multiwinner voting rules.
Researcher Affiliation Academia 1AGH University, Poland 2TU Wien, Austria 3IDSIA, USI-SUPSI, Switzerland
Pseudocode No The paper describes algorithms conceptually but does not include structured pseudocode or algorithm blocks.
Open Source Code Yes The code for the experiments is available at https://github.com/ Project-PRAGMA/Map-of-Rules-IJCAI-2023.
Open Datasets Yes We also use real-life participatory budgeting (PB) data from Pabulib [Stolicki et al., 2020]
Dataset Splits No The paper describes generating instances from statistical cultures and selecting a subset from a real-life dataset, but does not specify explicit training/validation/test splits of these instances.
Hardware Specification No No specific hardware details (such as GPU/CPU models, processor types, or memory amounts) used for running experiments are mentioned in the paper.
Software Dependencies Yes in our experiments we use the implementations of the rules provided in the abcvoting library [Lackner et al., 2023]. The reference indicates 'Journal of Open Source Software, 8(81):4880, 2023'.
Experiment Setup Yes We generated 6000 instances with 100 candidates and 100 voters from the six following statistical cultures (1000 elections per culture): 1D-Euclidean with r = 0.05, 2DEuclidean with r = 0.2, resampling with p = 0.1 and ϕ {0, 1 999, 2 999, . . . , 998 999, 1}, disjoint with p = 0.1, ϕ {0, 1 999, 2 999, . . . , 998 999, 1}, and g = 10, party-list with g = 10, and the Pabulib model. For all instances, we use a committee size of k = 10.