Unsupervised Sentence Simplification Using Deep Semantics

TL;DR

This paper introduces an unsupervised method for sentence simplification that leverages deep semantic structures, eliminating the need for annotated data and improving sentence splitting effectiveness.

Contribution

It proposes a novel unsupervised framework that uses deep semantic analysis for sentence simplification, outperforming some supervised systems.

Findings

Competitive with state-of-the-art supervised systems

Effective handling of sentence splitting through semantic analysis

No need for annotated training data

Abstract

We present a novel approach to sentence simplification which departs from previous work in two main ways. First, it requires neither hand written rules nor a training corpus of aligned standard and simplified sentences. Second, sentence splitting operates on deep semantic structure. We show (i) that the unsupervised framework we propose is competitive with four state-of-the-art supervised systems and (ii) that our semantic based approach allows for a principled and effective handling of sentence splitting.