_version_ 1866917250066284544
author Naseem, Usman
Geislinger, Robert
Ren, Juan
Kohail, Sarah
Veliz, Rudy Garrido
Sahil, P Sam
Zhang, Yiran
Stranisci, Marco Antonio
Abdulmumin, Idris
Alacam, Özge
Acartürk, Cengiz
Jabr, Aisha
Anwar, Saba
Ayele, Abinew Ali
Frenda, Simona
Cignarella, Alessandra Teresa
Tutubalina, Elena
Rogov, Oleg
Htet, Aung Kyaw
Wang, Xintong
Thapa, Surendrabikram
Rauniyar, Kritesh
Chakraborty, Tanmoy
Zeeshan, Arfeen
Kodati, Dheeraj
Keerthi, Satya
Moradizeyveh, Sahar
Alam, Firoj
Hasan, Arid
Ahmed, Syed Ishtiaque
Thu, Ye Kyaw
Parida, Shantipriya
Qazi, Ihsan Ayyub
Wanzare, Lilian
Onyango, Nelson Odhiambo
Siro, Clemencia
Kimani, Jane Wanjiru
Ahmad, Ibrahim Said
Ali, Adem Chanie
Semmann, Martin
Biemann, Chris
Muhammad, Shamsuddeen Hassan
Yimam, Seid Muhie
author_facet Naseem, Usman
Geislinger, Robert
Ren, Juan
Kohail, Sarah
Veliz, Rudy Garrido
Sahil, P Sam
Zhang, Yiran
Stranisci, Marco Antonio
Abdulmumin, Idris
Alacam, Özge
Acartürk, Cengiz
Jabr, Aisha
Anwar, Saba
Ayele, Abinew Ali
Frenda, Simona
Cignarella, Alessandra Teresa
Tutubalina, Elena
Rogov, Oleg
Htet, Aung Kyaw
Wang, Xintong
Thapa, Surendrabikram
Rauniyar, Kritesh
Chakraborty, Tanmoy
Zeeshan, Arfeen
Kodati, Dheeraj
Keerthi, Satya
Moradizeyveh, Sahar
Alam, Firoj
Hasan, Arid
Ahmed, Syed Ishtiaque
Thu, Ye Kyaw
Parida, Shantipriya
Qazi, Ihsan Ayyub
Wanzare, Lilian
Onyango, Nelson Odhiambo
Siro, Clemencia
Kimani, Jane Wanjiru
Ahmad, Ibrahim Said
Ali, Adem Chanie
Semmann, Martin
Biemann, Chris
Muhammad, Shamsuddeen Hassan
Yimam, Seid Muhie
contents 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. All resources will be released to support further research and effective mitigation of digital polarization globally.
format Preprint
id arxiv_https___arxiv_org_abs_2505_20624
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle POLAR: A Benchmark for Multilingual, Multicultural, and Multi-Event Online Polarization
Naseem, Usman
Geislinger, Robert
Ren, Juan
Kohail, Sarah
Veliz, Rudy Garrido
Sahil, P Sam
Zhang, Yiran
Stranisci, Marco Antonio
Abdulmumin, Idris
Alacam, Özge
Acartürk, Cengiz
Jabr, Aisha
Anwar, Saba
Ayele, Abinew Ali
Frenda, Simona
Cignarella, Alessandra Teresa
Tutubalina, Elena
Rogov, Oleg
Htet, Aung Kyaw
Wang, Xintong
Thapa, Surendrabikram
Rauniyar, Kritesh
Chakraborty, Tanmoy
Zeeshan, Arfeen
Kodati, Dheeraj
Keerthi, Satya
Moradizeyveh, Sahar
Alam, Firoj
Hasan, Arid
Ahmed, Syed Ishtiaque
Thu, Ye Kyaw
Parida, Shantipriya
Qazi, Ihsan Ayyub
Wanzare, Lilian
Onyango, Nelson Odhiambo
Siro, Clemencia
Kimani, Jane Wanjiru
Ahmad, Ibrahim Said
Ali, Adem Chanie
Semmann, Martin
Biemann, Chris
Muhammad, Shamsuddeen Hassan
Yimam, Seid Muhie
Computation and Language
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. All resources will be released to support further research and effective mitigation of digital polarization globally.
title POLAR: A Benchmark for Multilingual, Multicultural, and Multi-Event Online Polarization
topic Computation and Language
url https://arxiv.org/abs/2505.20624