Factorising Meaning and Form for Intent-Preserving Paraphrasing

TL;DR

This paper introduces a novel model that disentangles meaning and form in question paraphrasing, enabling intent-preserving rephrasing with improved semantic and syntactic control without external exemplars.

Contribution

It presents a disentangled latent space using a Vector-Quantized Variational Autoencoder for better control over paraphrase surface form while preserving intent.

Findings

Outperforms previous methods in semantic preservation and syntactic diversity.

Uses a VQ-VAE to represent surface form as discrete variables.

Human evaluation confirms improved paraphrasing quality.

Abstract

We propose a method for generating paraphrases of English questions that retain the original intent but use a different surface form. Our model combines a careful choice of training objective with a principled information bottleneck, to induce a latent encoding space that disentangles meaning and form. We train an encoder-decoder model to reconstruct a question from a paraphrase with the same meaning and an exemplar with the same surface form, leading to separated encoding spaces. We use a Vector-Quantized Variational Autoencoder to represent the surface form as a set of discrete latent variables, allowing us to use a classifier to select a different surface form at test time. Crucially, our method does not require access to an external source of target exemplars. Extensive experiments and a human evaluation show that we are able to generate paraphrases with a better tradeoff between semantic preservation and syntactic novelty compared to previous methods.