An MRC Framework for Semantic Role Labeling

Abstract

Semantic Role Labeling (SRL) aims at recognizing the predicate-argument structure of a sentence and can be decomposed into two subtasks: predicate disambiguation and argument labeling. Prior work deals with these two tasks independently, which ignores the semantic connection between the two tasks. In this paper, we propose to use the machine reading comprehension (MRC) framework to bridge this gap. We formalize predicate disambiguation as multiple-choice machine reading comprehension, where the descriptions of candidate senses of a given predicate are used as options to select the correct sense. The chosen predicate sense is then used to determine the semantic roles for that predicate, and these semantic roles are used to construct the query for another MRC model for argument labeling. In this way, we are able to leverage both the predicate semantics and the semantic role semantics for argument labeling. We also propose to select a subset of all the possible semantic roles for computational efficiency. Experiments show that the proposed framework achieves state-of-the-art results on both span and dependency benchmarks. 1

Improving Aspect-Based Sentiment with End-to-End Semantic Role Labeling Model

“…Papay et al (2021) propose regular-constrained conditional random fields (CRF) decoding on top of the same model. There are many other complex deep models Wang et al, 2021) For our experiments, we need an end-to-end SRL model which encodes most of the information in the Transformer's hidden state. However, to the best of our knowledge, there is no such model.…”

A Joint Learning Model for Aspect-based Sentiment Analysis via Machine Reading Comprehension and Graph Convolutional Network

“…In recent years, machine reading comprehension (MRC) [11][12][13][14][15] has become a topic of discussion in the field of named entity recognition. Generally, the definition of an MRC model is to provide the correct answer based on the given query statement and target text.…”