MUSS: Multilingual Unsupervised Sentence Simplification by Mining Paraphrases

TL;DR

MUSS introduces a multilingual, unsupervised sentence simplification system that leverages paraphrase mining from large-scale data, achieving competitive results without requiring labeled simplification datasets.

Contribution

It proposes a novel unsupervised approach to multilingual sentence simplification using paraphrase mining, eliminating the need for labeled simplification data.

Findings

Achieves state-of-the-art results on English, French, and Spanish benchmarks.

Outperforms previous supervised methods despite not using labeled data.

Further improves performance by incorporating labeled simplification data.

Abstract

Progress in sentence simplification has been hindered by a lack of labeled parallel simplification data, particularly in languages other than English. We introduce MUSS, a Multilingual Unsupervised Sentence Simplification system that does not require labeled simplification data. MUSS uses a novel approach to sentence simplification that trains strong models using sentence-level paraphrase data instead of proper simplification data. These models leverage unsupervised pretraining and controllable generation mechanisms to flexibly adjust attributes such as length and lexical complexity at inference time. We further present a method to mine such paraphrase data in any language from Common Crawl using semantic sentence embeddings, thus removing the need for labeled data. We evaluate our approach on English, French, and Spanish simplification benchmarks and closely match or outperform the previous best supervised results, despite not using any labeled simplification data. We push the state of the art further by incorporating labeled simplification data.