---
title: "Generating Senses and RoLes: An End-to-End Model for Dependency- and Span-based Semantic Role Labeling"
authors:
  - "Rexhina Blloshmi"
  - "Simone Conia"
  - "Rocco Tripodi"
  - "Roberto Navigli"
doi: "10.24963/ijcai.2021/521"
url: "https://ijcai.org/proceedings/2021/521"
pdf: "https://ijcai.org/proceedings/2021/0521.pdf"
venue: "Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence"
year: 2021
pages: "3786-3793"
track: "Main Track"
keywords:
  - "Natural Language Processing: Natural Language Semantics"
  - "Natural Language Processing: Natural Language Processing"
  - "Natural Language Processing: Natural Language Generation"
---

# Generating Senses and RoLes: An End-to-End Model for Dependency- and Span-based Semantic Role Labeling

Rexhina Blloshmi, Simone Conia, Rocco Tripodi, Roberto Navigli

*Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence*, pages 3786-3793. DOI: [10.24963/ijcai.2021/521](https://doi.org/10.24963/ijcai.2021/521)

## Abstract

Despite the recent great success of the sequence-to-sequence paradigm in Natural Language Processing, the majority of current studies in Semantic Role Labeling (SRL) still frame the problem as a sequence labeling task.
In this paper we go against the flow and propose GSRL (Generating Senses and RoLes), the first sequence-to-sequence model for end-to-end SRL.
Our approach benefits from recently-proposed decoder-side pretraining techniques to generate both sense and role labels for all the predicates in an input sentence at once, in an end-to-end fashion.
Evaluated on standard gold benchmarks, GSRL achieves state-of-the-art results in both dependency- and span-based English SRL, proving empirically that our simple generation-based model can learn to produce complex predicate-argument structures.
Finally, we propose a framework for evaluating the robustness of an SRL model in a variety of synthetic low-resource scenarios which can aid human annotators in the creation of better, more diverse, and more challenging gold datasets.
We release GSRL at github.com/SapienzaNLP/gsrl.