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

2026-02-26

AI Summary

Cross-lingual predicate-argument structure annotation is costly and has long been constrained by English-centric resources, lacking efficient and scalable multilingual construction methods. This work proposes a cross-lingual projection framework based on Question Answering–driven Semantic Role Labeling (QA-SRL), which transfers knowledge from an English QA-SRL parser to target languages through constrained machine translation and word alignment to automatically generate high-quality annotations. The approach achieves, for the first time, effective predicate-argument parsing in Hebrew, Russian, and French. Models trained on the generated data significantly outperform strong multilingual large language model baselines such as GPT-4o and LLaMA-Maverick, demonstrating the effectiveness of QA-SRL as a transferable semantic interface.

Technology Category

• Natural Language Processing: Question Answering
• Machine Learning: Large Multimodal Models (LMMs)
• Knowledge Representation and Reasoning: Argumentation

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Large pretrained models with web data

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.

Problem

• predicate-argument relations
• semantic annotation
• cross-lingual
• QA-SRL
• multilingual parsing

Innovation

• QA-SRL
• cross-lingual projection
• predicate-argument parsing
• semantic role labeling
• natural-language interface