Leveraging LaBSE with Progressive Curriculum Learning for Multicultural Polarization

2026-06-19

AI Summary

This study addresses the challenge of limited performance in polarization detection for low-resource languages within multilingual and multicultural contexts. To this end, the authors propose a cross-lingual polarization detection framework that integrates LaBSE embeddings with progressive curriculum learning, and systematically evaluate the performance of Qwen-series encoders within a retrieval-based prompting architecture. Notably, the work innovatively repurposes LaBSE—originally designed for semantic retrieval—for polarization detection, effectively mitigating the data scarcity inherent to low-resource languages. Experimental results demonstrate that the proposed approach achieves up to a 0.2 improvement in macro F1 score on low-resource languages, significantly outperforming baseline methods. This advancement offers an efficient and scalable solution for multilingual polarization analysis.

Abstract

Detecting online polarization remains a critical challenge, particularly in multilingual and multicultural contexts where intergroup hostility is prevalent. The problem is particularly challenging due to the data scarcity for these tasks in the low-resource languages. Identifying such phenomena has become an active area of research and is addressed in SemEval-2026 Task 9: Multilingual, Multicultural Online Polarization Detection. To address this problem we propose an architecture that leverages LaBSE embeddings - an unconventional choice typically reserved for retrieval tasks, to obtain strong cross-lingual learning which enhances scores in low-resource language by a score up to 0.2 macro F1. Furthermore, we provide a comprehensive ablation study evaluating the performance of diverse encoder models in the Qwen model family within a retrieval-based prompting framework. Our code will be soon available at https://github.com/carrycurious/PolarMind.

Problem

• online polarization
• multilingual
• multicultural
• low-resource languages
• intergroup hostility

Innovation

• LaBSE
• progressive curriculum learning
• multilingual polarization detection
• low-resource languages
• retrieval-based prompting