Jointly Predicting Predicates and Arguments in Neural Semantic Role Labeling

Abstract

Recent BIO-tagging-based neural semantic role labeling models are very high performing, but assume gold predicates as part of the input and cannot incorporate span-level features. We propose an end-to-end approach for jointly predicting all predicates, arguments spans, and the relations between them. The model makes independent decisions about what relationship, if any, holds between every possible word-span pair, and learns contextualized span representations that provide rich, shared input features for each decision. Experiments demonstrate that this approach sets a new state of the art on PropBank SRL without gold predicates.

arXiv comment5 pages, ACL 2018

References in corpus (4)

• Exact solutions to the nonlinear dynamics of learning in deep linear neural networks
• End-to-End Relation Extraction using LSTMs on Sequences and Tree Structures
• Segmental Recurrent Neural Networks
• A Minimal Span-Based Neural Constituency Parser

Cited by in corpus (6)

• Natural Language Processing Advancements By Deep Learning: A Survey
• Graph Neural Networks for Natural Language Processing: A Survey
• Clinical Concept Extraction with Contextual Word Embedding
• Open-Domain Targeted Sentiment Analysis via Span-Based Extraction and Classification
• Towards Understanding and Answering Multi-Sentence Recommendation Questions on Tourism
• Retrieve, Read, Rerank: Towards End-to-End Multi-Document Reading Comprehension