Diversity, Agreement, and Polarization in Elections
Authors: Piotr Faliszewski, Andrzej Kaczmarczyk, Krzysztof Sornat, Stanisław Szufa, Tomasz Wąs
IJCAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | To evaluate our indices, we use the map of elections framework of Szufa et al. [2020], Boehmer et al. [2021], and Boehmer et al. [2022], applied to a dataset of randomly generated elections. In our experiments, we focused on elections with a relatively small number of candidates (8 candidates and 96 voters). |
| Researcher Affiliation | Academia | AGH University, Poland IDSIA, USI-SUPSI, Switzerland Pennsylvania State University, PA, USA |
| Pseudocode | No | The paper does not contain any pseudocode or clearly labeled algorithm blocks. |
| Open Source Code | Yes | the code of our experiments is available at https://github.com/ Project-PRAGMA/diversity-agreement-polarization-IJCAI23. |
| Open Datasets | Yes | We also include Sushi and Grenoble elections, similarly generated using different real-life data [Mattei and Walsh, 2013]. We also consider elections generated based on real-life data from a 2002 political election in Dublin [Mattei and Walsh, 2013]. |
| Dataset Splits | No | The paper describes evaluating its indices on a dataset but does not specify train, validation, or test splits. It states: |
| Hardware Specification | No | The paper does not specify any particular hardware (e.g., CPU, GPU models, memory) used for running the experiments. It only mentions that checking conclusions for all sizes of elections 'would require quite extensive computation'. |
| Software Dependencies | No | The paper mentions using 'an open-source Python library mapel', but it does not specify any version numbers for Python or mapel, nor for any other key software components. |
| Experiment Setup | Yes | In our experiments, we focused on elections with a relatively small number of candidates (8 candidates and 96 voters). The elections have 1000 voters instead of 96, so that the pictures look similar each time we draw an election from the model. First, we compared three ways of computing k-Kemeny distances: the greedy approach, the local search with swap size equal to 1, and a combined heuristic where we first calculate the greedy solution and then try to improve it using the local search. We ran all three algorithms for all k [96] and for every election in our dataset. Hence, in further computations, we used the former two algorithm and took the smaller of their outputs. |