Multiple Partitioning of Multiplex Signed Networks: Application to European Parliament Votes

Presentation. For more than a decade, graphs have been used to model the voting behavior taking place in parliaments. However, the methods described in the literature suffer from several limitations. The two main ones are that 1) they rely on some temporal integration of the raw data, which causes some information loss; and/or 2) they identify groups of antagonistic voters, but not the context associated with their occurrence. In this article, we propose a novel method taking advantage of multiplex signed graphs to solve both these issues. It consists in first partitioning separately each layer, before grouping these partitions by similarity. We show the interest of our approach by applying it to a European Parliament dataset. Particularly, we study the voting behavior of French and Italian MEPs on "Agriculture and Rural Development" (AGRI) during the 2012-13 legislative year.

These are the data used in the following paper:

• N. Arınık, R. Figueiredo, and V. Labatut, “Multiple partitioning of multiplex signed networks: Application to European Parliament votes,” Social Networks, vol. 60, pp. 83–102, 2020. DOI: 10.1016/j.socnet.2019.02.001 ⟨hal-02082574⟩

Source code. The code source is accessible on GitHub: https://github.com/CompNet/MultiNetVotes

Citation. If you use these data our this source code, please cite the above paper.

@Article{Arinik2020,
  author    = {Arınık, Nejat and Figueiredo, Rosa and Labatut, Vincent},
  title     = {Multiple Partitioning of Multiplex Signed Networks: Application to {E}uropean {P}arliament Votes},
  journal   = {Social Networks},
  year      = {2020},
  volume    = {60},
  pages     = {83-102},
  doi       = {10.1016/j.socnet.2019.02.001},
}

----------------------------------------------
Details.

# RAW INPUT FILES
The 'itsyourparliament' folder contains all raw input files for further data processing. This is the same raw data that can be found in our previous Figshare repository: https://doi.org/10.6084/m9.figshare.5785833
The folder structure is as follows:
* itsyourparliament/
** domains: There are 28 domain files. Each file corresponds to a domain (such as Agriculture, Economy, etc.) and contains corresponding vote identifiers and their "itsyourparliament.eu" links.
** meps: There are 870 Members of Parliament (MEP) files. Each file contains the MEP information (such as name, country, address, etc.)
** votes: There are 7513 vote files. Each file contains the votes expressed by MEPs

# ROLLCALL NETWORKS
This folder contains two separate zip files regarding rollcall networks:
- rollcall-networks: This folder contains only the rollcall networks that are used in the article.
- all-rollcall-networks: For those who are interested in other countries or domains, we make available all rollcall networks that we can extract from raw data.
Note that these rollcall networks constitute the layers of the input signed multplex network, as illustrated in Figure 1 of the article. Note also that we consider three vote types in our network extraction process: FOR, AGAINST and ABSTAIN.

# ROLLCALL PARTITIONS
Note that MEPs who voted similarly are connected together by positive links, and are connected by negative links to MEPs that voted differently from them. MEPs who did not vote at all (ABSENT) are isolates (nodes without any
neighbor). We identify the factions of similarly voting MEPs in the graph by solving the Correlation Clustering problem (CC).
The rollcall partitions correspond to voting patterns, as illustrated in Figure 1 of the article.

# ROLLCALL CLUSTERING
This folder contains the results of Steps 3 and 4 of our workflow (see Figure 1 in the article). The structure of this folder is as follows:
|__ votetypes=FAA/: 'FAA' means we consider three vote types in our analysis: FOR, AGAINST and ABSTAIN.
|__ F.purity-k=2-sil=SILHOUETTE_SCORE
|__ clu=CLUSTER_NO/
|__ network: It corresponds to the network created through the similarity network-based approach, as explained in Section 4.4 of the article.
|__ partition: It corresponds to the characteristic voting pattern, as explained in Section 4.4 of the article.
----------------------------------------------

Funding: this research benefited from the support of the Agorantic FR 3621, as well as the FMJH Program PGMO and from the support to this program from EDF-THALES-ORANGE-CRITEO.