Controllable Lexical Simplification for English

TL;DR

This paper introduces ConLS, a controllable lexical simplification system based on T5, which achieves comparable or better performance than current state-of-the-art models across multiple datasets, with insights into control token effectiveness.

Contribution

The paper presents the first application of Transformer fine-tuning for lexical simplification, introducing controllability and demonstrating competitive results.

Findings

ConLS performs comparably to LSBert on three datasets.

Control tokens significantly influence model outputs.

ConLS outperforms LSBert in some evaluation cases.

Abstract

Fine-tuning Transformer-based approaches have recently shown exciting results on sentence simplification task. However, so far, no research has applied similar approaches to the Lexical Simplification (LS) task. In this paper, we present ConLS, a Controllable Lexical Simplification system fine-tuned with T5 (a Transformer-based model pre-trained with a BERT-style approach and several other tasks). The evaluation results on three datasets (LexMTurk, BenchLS, and NNSeval) have shown that our model performs comparable to LSBert (the current state-of-the-art) and even outperforms it in some cases. We also conducted a detailed comparison on the effectiveness of control tokens to give a clear view of how each token contributes to the model.