Multilingual Target-Stance Extraction

2025-10-25

AI Summary

Existing target–stance extraction (TSE) research is confined to English, lacking multilingual benchmarks and a unified evaluation framework—hindering cross-lingual public opinion analysis. Method: We introduce the first multilingual TSE benchmark covering Catalan, Estonian, French, Italian, Chinese, and Spanish; propose the first language-agnostic multilingual TSE framework integrating cross-lingual representation learning with joint target identification and stance classification; and conduct ablation studies to quantify sensitivity of target formulation to F1 performance. Contribution/Results: We identify target prediction as the critical bottleneck and demonstrate that our framework achieves an F1 score of 12.78 across the six languages—confirming the task’s inherent difficulty. The benchmark and framework provide a reproducible, extensible foundation for multilingual stance analysis, enabling systematic cross-lingual comparison and future model development.

Technology Category

• Natural Language Processing: Machine Translation, Multilinguality, Cross-Lingual NLP
• Knowledge Representation and Reasoning: Argumentation
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Social media enables data-driven analysis of public opinion on contested issues. Target-Stance Extraction (TSE) is the task of identifying the target discussed in a document and the document's stance towards that target. Many works classify stance towards a given target in a multilingual setting, but all prior work in TSE is English-only. This work introduces the first multilingual TSE benchmark, spanning Catalan, Estonian, French, Italian, Mandarin, and Spanish corpora. It manages to extend the original TSE pipeline to a multilingual setting without requiring separate models for each language. Our model pipeline achieves a modest F1 score of 12.78, underscoring the increased difficulty of the multilingual task relative to English-only setups and highlighting target prediction as the primary bottleneck. We are also the first to demonstrate the sensitivity of TSE's F1 score to different target verbalizations. Together these serve as a much-needed baseline for resources, algorithms, and evaluation criteria in multilingual TSE.

Problem

• Extending target-stance extraction to multiple languages
• Creating first multilingual benchmark across six languages
• Addressing target prediction as primary performance bottleneck

Innovation

• Extends TSE pipeline to multilingual setting without separate models
• Introduces first multilingual TSE benchmark across six languages
• Demonstrates F1 score sensitivity to different target verbalizations