Ties in Multiwinner Approval Voting

Authors: Łukasz Janeczko, Piotr Faliszewski

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

Reproducibility Variable Result LLM Response
Research Type Experimental 3. We generate a number of elections, both synthetic and based on real-life data, and evaluate the frequency of ties. It turns out to be surprisingly high. ... Our code is available at https://github.com/ Project-PRAGMA/Ties-IJCAI-2023.
Researcher Affiliation Academia Łukasz Janeczko , Piotr Faliszewski AGH University, Poland {ljaneczk,faliszew}@agh.edu.pl
Pseudocode No The paper describes algorithms in prose and mathematical notation but does not include structured pseudocode or algorithm blocks.
Open Source Code Yes Our code is available at https://github.com/ Project-PRAGMA/Ties-IJCAI-2023.
Open Datasets Yes Pabu Lib Data. Pabu Lib is a library of real-life participatory budgeting (PB) instances, mostly from Polish cities [Faliszewski et al., 2023].
Dataset Splits No The paper describes data generation models (Resampling Model, Interval Model, Pabu Lib Data) and explains how elections are generated, but it does not specify train/validation/test splits.
Hardware Specification No The paper describes experimental setups related to data generation and evaluation of tie frequencies but does not provide any specific hardware specifications.
Software Dependencies No The paper makes no mention of specific software dependencies or their version numbers.
Experiment Setup Yes In a basic experiment we fix the number of candidates m, the committee size k, and a statistical culture. Then, for each number n of voters between 20 and 100 (with a step of 1) we generate 1000 elections with m candidates and n voters, and for each of them compute whether our rules have a unique winning committee (we omit Greedy CCAV).