Fast and Accurate End-to-End Span-based Semantic Role Labeling as Word-based Graph Parsing

TL;DR

This paper introduces a novel end-to-end span-based semantic role labeling method by formulating it as a word-based graph parsing task, achieving superior accuracy and speed on benchmark datasets.

Contribution

It proposes four graph representation schemata for span encoding and a constrained Viterbi decoding to ensure valid SRL structures, significantly improving performance and efficiency.

Findings

Outperforms previous SRL methods on CoNLL datasets

Achieves high parsing speed of 669/252 sentences per second

Demonstrates the effectiveness of BES schema for span representation

Abstract

This paper proposes to cast end-to-end span-based SRL as a word-based graph parsing task. The major challenge is how to represent spans at the word level. Borrowing ideas from research on Chinese word segmentation and named entity recognition, we propose and compare four different schemata of graph representation, i.e., BES, BE, BIES, and BII, among which we find that the BES schema performs the best. We further gain interesting insights through detailed analysis. Moreover, we propose a simple constrained Viterbi procedure to ensure the legality of the output graph according to the constraints of the SRL structure. We conduct experiments on two widely used benchmark datasets, i.e., CoNLL05 and CoNLL12. Results show that our word-based graph parsing approach achieves consistently better performance than previous results, under all settings of end-to-end and predicate-given, without and with pre-trained language models (PLMs). More importantly, our model can parse 669/252 sentences per second, without and with PLMs respectively.