Multilingual Target-Stance Extraction

TL;DR

This paper introduces the first multilingual Target-Stance Extraction benchmark covering six languages, extending the task beyond English and highlighting the challenges and sensitivities involved.

Contribution

It presents a new multilingual TSE benchmark and pipeline that works across multiple languages without separate models, establishing a baseline for future research.

Findings

Achieved a modest F1 score of 12.78 in multilingual TSE

Identified target prediction as the main bottleneck

Demonstrated the impact of target verbalizations on F1 score

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.