MANTIS at TSAR-2022 Shared Task: Improved Unsupervised Lexical Simplification with Pretrained Encoders

TL;DR

This paper enhances an unsupervised lexical simplification system using pretrained encoders, improving candidate selection and ranking, and achieves high accuracy in the TSAR-2022 shared task.

Contribution

It introduces a RoBERTa-based candidate expansion and a novel feature weighting and filtering scheme for better lexical simplification.

Findings

Achieved 5.9% accuracy improvement over LSBert

Secured second place out of 33 solutions

Enhanced candidate selection and ranking methods

Abstract

In this paper we present our contribution to the TSAR-2022 Shared Task on Lexical Simplification of the EMNLP 2022 Workshop on Text Simplification, Accessibility, and Readability. Our approach builds on and extends the unsupervised lexical simplification system with pretrained encoders (LSBert) system in the following ways: For the subtask of simplification candidate selection, it utilizes a RoBERTa transformer language model and expands the size of the generated candidate list. For subsequent substitution ranking, it introduces a new feature weighting scheme and adopts a candidate filtering method based on textual entailment to maximize semantic similarity between the target word and its simplification. Our best-performing system improves LSBert by 5.9% accuracy and achieves second place out of 33 ranked solutions.