Effective QA-driven Annotation of Predicate-Argument Relations Across Languages

TL;DR

This paper introduces a cross-linguistic method for automatically annotating predicate-argument relations using QA-SRL, enabling semantic analysis in multiple languages with high-quality data and outperforming large multilingual models.

Contribution

It presents a novel cross-linguistic projection approach that adapts English QA-SRL parsers for multiple languages, reducing annotation costs and expanding semantic analysis capabilities.

Findings

High-quality predicate-argument annotations for Hebrew, Russian, and French

Language-specific parsers outperform multilingual LLM baselines

Efficient transfer of semantic annotation across languages

Abstract

Explicit representations of predicate-argument relations form the basis of interpretable semantic analysis, supporting reasoning, generation, and evaluation. However, attaining such semantic structures requires costly annotation efforts and has remained largely confined to English. We leverage the Question-Answer driven Semantic Role Labeling (QA-SRL) framework -- a natural-language formulation of predicate-argument relations -- as the foundation for extending semantic annotation to new languages. To this end, we introduce a cross-linguistic projection approach that reuses an English QA-SRL parser within a constrained translation and word-alignment pipeline to automatically generate question-answer annotations aligned with target-language predicates. Applied to Hebrew, Russian, and French -- spanning diverse language families -- the method yields high-quality training data and fine-tuned, language-specific parsers that outperform strong multilingual LLM baselines (GPT-4o, LLaMA-Maverick). By leveraging QA-SRL as a transferable natural-language interface for semantics, our approach enables efficient and broadly accessible predicate-argument parsing across languages.