Revisiting Semantic Role Labeling: Efficient Structured Inference with Dependency-Informed Analysis

2026-05-04

AI Summary

This work addresses the absence of explicit predicate-argument structure representations in current large language models and the incompatibility of traditional semantic role labeling (SRL) approaches with modern encoder architectures and efficient inference requirements. The authors propose a novel structured SRL framework built upon pretrained encoders—such as BERT, RoBERTa, and DeBERTa—that preserves explicit predicate-argument structures while incorporating a dependency-informed inference mechanism to enhance structural stability. The method supports multilingual SRL projection and achieves up to a tenfold speedup in inference compared to existing systems, while matching or surpassing their F1 performance. These results underscore the critical role of syntactic dependency cues in structured semantic parsing.

Abstract

Semantic Role Labeling (SRL) provides an explicit representation of predicate-argument structure, capturing linguistically grounded relations such as who did what to whom. While recent NLP progress has been dominated by large language models (LLMs), these systems often rely on implicit semantic representations, often lacking explicit structural constraints and systematic explanatory mechanisms. Traditionally, SRL systems have often relied on AllenNLP; however, the framework entered maintenance mode in December 2022, limiting compatibility with evolving encoder architectures and modern inference requirements. We revisit structured SRL modeling, introducing a modernized encoder-based framework that preserves explicit predicate-argument structure while enabling inference 10 times faster. Using BERT-base, the model attains comparable predictive performance, and RoBERTa and DeBERTa further improve F1 performance within the same framework. We adopt a dependency-informed diagnostic methodology to characterize span-level inconsistencies and conduct a representation-level analysis of LLM behavior under dependency-informed structural signals. Results indicate that dependency cues primarily improve structural stability. Finally, we illustrate how the framework's explicit predicate-argument structure can support multilingual SRL projection as a downstream application.

Problem

• Semantic Role Labeling
• large language models
• structured inference
• explicit predicate-argument structure
• dependency-informed analysis

Innovation

• Semantic Role Labeling
• dependency-informed analysis
• structured inference
• efficient encoder framework
• multilingual projection