ParaGuide: Guided Diffusion Paraphrasers for Plug-and-Play Textual Style Transfer

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Horvitz, Zachary, Patel, Ajay, Callison-Burch, Chris, Yu, Zhou, McKeown, Kathleen
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916134776733696
author Horvitz, Zachary
Patel, Ajay
Callison-Burch, Chris
Yu, Zhou
McKeown, Kathleen
author_facet Horvitz, Zachary
Patel, Ajay
Callison-Burch, Chris
Yu, Zhou
McKeown, Kathleen
contents Textual style transfer is the task of transforming stylistic properties of text while preserving meaning. Target "styles" can be defined in numerous ways, ranging from single attributes (e.g, formality) to authorship (e.g, Shakespeare). Previous unsupervised style-transfer approaches generally rely on significant amounts of labeled data for only a fixed set of styles or require large language models. In contrast, we introduce a novel diffusion-based framework for general-purpose style transfer that can be flexibly adapted to arbitrary target styles at inference time. Our parameter-efficient approach, ParaGuide, leverages paraphrase-conditioned diffusion models alongside gradient-based guidance from both off-the-shelf classifiers and strong existing style embedders to transform the style of text while preserving semantic information. We validate the method on the Enron Email Corpus, with both human and automatic evaluations, and find that it outperforms strong baselines on formality, sentiment, and even authorship style transfer.
format Preprint
id arxiv_https___arxiv_org_abs_2308_15459
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle ParaGuide: Guided Diffusion Paraphrasers for Plug-and-Play Textual Style Transfer
Horvitz, Zachary
Patel, Ajay
Callison-Burch, Chris
Yu, Zhou
McKeown, Kathleen
Computation and Language
Artificial Intelligence
Textual style transfer is the task of transforming stylistic properties of text while preserving meaning. Target "styles" can be defined in numerous ways, ranging from single attributes (e.g, formality) to authorship (e.g, Shakespeare). Previous unsupervised style-transfer approaches generally rely on significant amounts of labeled data for only a fixed set of styles or require large language models. In contrast, we introduce a novel diffusion-based framework for general-purpose style transfer that can be flexibly adapted to arbitrary target styles at inference time. Our parameter-efficient approach, ParaGuide, leverages paraphrase-conditioned diffusion models alongside gradient-based guidance from both off-the-shelf classifiers and strong existing style embedders to transform the style of text while preserving semantic information. We validate the method on the Enron Email Corpus, with both human and automatic evaluations, and find that it outperforms strong baselines on formality, sentiment, and even authorship style transfer.
title ParaGuide: Guided Diffusion Paraphrasers for Plug-and-Play Textual Style Transfer
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2308.15459