Political Leaning and Politicalness Classification of Texts
–arXiv.org Artificial Intelligence
This paper addresses the challenge of automatically classifying text according to political leaning and politicalness using transformer models. We compose a comprehensive overview of existing datasets and models for these tasks, finding that current approaches create siloed solutions that perform poorly on out-of-distribution texts. To address this limitation, we compile a diverse dataset by combining 12 datasets for political leaning classification and creating a new dataset for politicalness by extending 18 existing datasets with the appropriate label. Through extensive benchmarking with leave-one-in and leave-one-out methodologies, we evaluate the performance of existing models and train new ones with enhanced generalization capabilities.
arXiv.org Artificial Intelligence
Jul-21-2025
- Country:
- Asia
- Middle East
- Jordan (0.04)
- UAE > Abu Dhabi Emirate
- Abu Dhabi (0.04)
- Singapore (0.04)
- Middle East
- Europe
- Czechia > Pardubice Region
- Pardubice (0.04)
- France > Provence-Alpes-Côte d'Azur
- Bouches-du-Rhône > Marseille (0.04)
- Italy > Tuscany
- Florence (0.04)
- Netherlands (0.04)
- Portugal (0.04)
- Slovakia > Bratislava
- Bratislava (0.04)
- Spain (0.04)
- Czechia > Pardubice Region
- North America
- Canada (0.04)
- United States
- Maryland > Baltimore (0.14)
- Michigan > Washtenaw County
- Ann Arbor (0.04)
- Oregon > Multnomah County
- Portland (0.04)
- Washington > King County
- Seattle (0.04)
- Asia
- Genre:
- Research Report > New Finding (1.00)
- Industry:
- Government (1.00)
- Media > News (0.93)
- Technology: