Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Political Leaning Inference through Plurinational Scenarios

Joseba Fernandez de Landa, Rodrigo Agerri

TL;DR

This paper tackles political leaning inference on Twitter data within three Spanish regions that exhibit plurinationality. It introduces Relational Embeddings learned from retweet interactions as unsupervised user representations and evaluates them in both binary left-right and seven-party multi-party settings. Across strongly supervised and weakly supervised scenarios, RE consistently outperforms baseline representation methods, with pronounced gains in the multi-party task and few-shot learning. The work also provides error analyses and visualizations that illuminate intra- and inter-party affinities, and suggests future extensions to text data and misinformation detection.

Abstract

Social media users express their political preferences via interaction with other users, by spontaneous declarations or by participation in communities within the network. This makes a social network such as Twitter a valuable data source to study computational science approaches to political learning inference. In this work we focus on three diverse regions in Spain (Basque Country, Catalonia and Galicia) to explore various methods for multi-party categorization, required to analyze evolving and complex political landscapes, and compare it with binary left-right approaches. We use a two-step method involving unsupervised user representations obtained from the retweets and their subsequent use for political leaning detection. Comprehensive experimentation on a newly collected and curated dataset comprising labeled users and their interactions demonstrate the effectiveness of using Relational Embeddings as representation method for political ideology detection in both binary and multi-party frameworks, even with limited training data. Finally, data visualization illustrates the ability of the Relational Embeddings to capture intricate intra-group and inter-group political affinities.

Political Leaning Inference through Plurinational Scenarios

TL;DR

This paper tackles political leaning inference on Twitter data within three Spanish regions that exhibit plurinationality. It introduces Relational Embeddings learned from retweet interactions as unsupervised user representations and evaluates them in both binary left-right and seven-party multi-party settings. Across strongly supervised and weakly supervised scenarios, RE consistently outperforms baseline representation methods, with pronounced gains in the multi-party task and few-shot learning. The work also provides error analyses and visualizations that illuminate intra- and inter-party affinities, and suggests future extensions to text data and misinformation detection.

Abstract

Social media users express their political preferences via interaction with other users, by spontaneous declarations or by participation in communities within the network. This makes a social network such as Twitter a valuable data source to study computational science approaches to political learning inference. In this work we focus on three diverse regions in Spain (Basque Country, Catalonia and Galicia) to explore various methods for multi-party categorization, required to analyze evolving and complex political landscapes, and compare it with binary left-right approaches. We use a two-step method involving unsupervised user representations obtained from the retweets and their subsequent use for political leaning detection. Comprehensive experimentation on a newly collected and curated dataset comprising labeled users and their interactions demonstrate the effectiveness of using Relational Embeddings as representation method for political ideology detection in both binary and multi-party frameworks, even with limited training data. Finally, data visualization illustrates the ability of the Relational Embeddings to capture intricate intra-group and inter-group political affinities.
Paper Structure (20 sections, 6 figures, 5 tables)

This paper contains 20 sections, 6 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related work
  3. Datasets
  4. Political Regions
  5. Data collection strategy
  6. Method
  7. User Representations
  8. Methodology
  9. Experimental setup
  10. Strongly supervised scenario
  11. Weakly supervised scenario
  12. Evaluation
  13. Strongly supervised scenario
  14. Binary framework.
  15. Multi-party framework.
  16. ...and 5 more sections

Figures (6)

  • Figure 1: Confusion matrices for Logistic Regresion classifier using RE and DW user representation models in the strongly supervised scenario on the EUS dataset.
  • Figure 2: Confusion matrices for Logistic Regresion classifier using RE and DW user representation models in the strongly supervised scenario on the GAL dataset.
  • Figure 3: Confusion matrices for Logistic Regresion classifier using RE and DW user representation models in the strongly supervised scenario on the CAT dataset.
  • Figure 4: Visualization of PCA, UMAP and t-SNE 2 dimension reduction for EUS Relational Embedding representation.
  • Figure 5: Visualization of PCA, UMAP and t-SNE 2 dimension reduction for GAL Relational Embedding representation.
  • ...and 1 more figures