Approval-Based Multi-Winner Rules and Strategic Voting
Authors: Martin Lackner, Piotr Skowron
IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Both our axiomatic and experimental analysis show that approval-based multiwinner rules are generally very susceptible to strategic voting with one exception: Multiwinner Approval Voting. |
| Researcher Affiliation | Academia | 1 Technische Universität Wien, Austria 2 University of Warsaw, Poland |
| Pseudocode | No | The paper does not contain any pseudocode or clearly labeled algorithm blocks. |
| Open Source Code | No | The paper does not provide any statements or links regarding the availability of open-source code for the described methodology. |
| Open Datasets | No | The paper describes generating instances: "In randomly generated instances we investigated the possibility of voters to SD-manipulate, i.e., we computed the relative number of instances in which strategic voting allows voters to change the set of winning committees to one that stochastically dominates the original one." It specifies the parameters for generation (n=24, m=8, k=4) and how approval sets were sampled, but does not refer to a publicly available dataset with concrete access information. |
| Dataset Splits | No | The paper mentions generating instances and the parameters used (n=24, m=8, k=4) but does not specify any training, validation, or test dataset splits. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware used to run the experiments. |
| Software Dependencies | No | The paper does not list any specific software dependencies or their version numbers used for the experiments. |
| Experiment Setup | Yes | In Table 1 the results of two experiments are displayed. In the first experiment, it is assumed that voters approve of exactly two candidates, in the second that voters approve of exactly three candidates. For both experiments, approval sets were sampled uniformly at random to generate profiles with n = 24 voters, m = 8 candidates, and assuming a committee size of k = 4. Both experiments are based on 1,000 instances. ... for the (more uniform) class of p-geometric rules, p ≥ 1. These are Thiele methods defined by f(x, y) = Px i=1 1/pi. For this setting, we repeated our experiments (due to computational limitations with smaller profiles) and could clearly see that the number of SD-manipulable profiles increased with increasing p (cf. Figure 1). |