2026-07-02
POLAR: A benchmark for multilingual, multicultural, and multi-event online polarization
Publication
Publication
Online polarization poses a growing challenge for democratic discourse, yet most computational social science research remains monolingual, culturally narrow, or event-specific. We introduce POLAR, a multilingual, multicultural, and multi-event dataset with over 110K instances in 22 languages drawn from diverse online platforms and real-world events. Polarization is annotated along three axes, namely detection, type, and manifestation, using a variety of annotation platforms adapted to each cultural context. We conduct two main experiments: (1) fine-tuning six pretrained small language models; and (2) evaluating a range of open and closed large language models in few-shot and zero-shot settings. The results show that, while most models perform well in binary polarization detection, they achieve substantially lower performance when predicting polarization types and manifestations. These findings highlight the complex, highly contextual nature of polarization and demonstrate the need for robust, adaptable approaches in NLP and computational social science. We release all resources to support research and effective mitigation of digital polarization globally.
| Additional Metadata | |
|---|---|
| doi.org/10.18653/v1/2026.findings-acl.1433 | |
| 64th Annual Meeting of the Association for Computational Linguistics | |
| creativecommons.org/licenses/by/4.0/ | |
| Organisation | Human-Centered Data Analytics |
|
Naseem, Usman, Geislinger, Robert, Ren, Juan, Kohail, Sarah, Veliz, Rudy Alexandro Garrido, Sahil, P Sam, Zhang, Yiran, Abdulmumin, Idris, Stranisci, Marco Antonio, Alacam, Özge, Acarturk, Cengiz, Jabr, Aisha, Anwar, Saba, Ayele, Abinew Ali, Frenda, Simona, Cignarella, Alessandra Teresa, Tutubalina, Elena, Rogov, Oleg, Htet, Aung Kyaw, … Yimam, Seid Muhie. (2026). POLAR: A benchmark for multilingual, multicultural, and multi-event online polarization. Findings of the Association for Computational Linguistics (2026), 28699–28720.https://doi.org/10.18653/v1/2026.findings-acl.1433 |
|