Style Transfer as Unsupervised Machine Translation

Abstract

Language style transferring rephrases text with specific stylistic attributes while preserving the original attribute-independent content. One main challenge in learning a style transfer system is a lack of parallel data where the source sentence is in one style and the target sentence in another style. With this constraint, in this paper, we adapt unsupervised machine translation methods for the task of automatic style transfer. We first take advantage of style-preference information and word embedding similarity to produce pseudo-parallel data with a statistical machine translation (SMT) framework. Then the iterative back-translation approach is employed to jointly train two neural machine translation (NMT) based transfer systems. To control the noise generated during joint training, a style classifier is introduced to guarantee the accuracy of style transfer and penalize bad candidates in the generated pseudo data. Experiments on benchmark datasets show that our proposed method outperforms previous state-of-the-art models in terms of both accuracy of style transfer and quality of input-output correspondence.

References in corpus (12)

• Neural Machine Translation by Jointly Learning to Align and Translate
• Sequence to Sequence Learning with Neural Networks
• ADADELTA: An Adaptive Learning Rate Method
• Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
• Style Transfer from Non-Parallel Text by Cross-Alignment
• Word Translation Without Parallel Data
• Phrase-Based & Neural Unsupervised Machine Translation
• A Persona-Based Neural Conversation Model
• Toward Controlled Generation of Text
• Semantically Conditioned LSTM-based Natural Language Generation for Spoken Dialogue Systems
• Sequence-Level Knowledge Distillation
• Unpaired Sentiment-to-Sentiment Translation: A Cycled Reinforcement Learning Approach

Cited by in corpus (37)

• Controllable Unsupervised Text Attribute Transfer via Editing Entangled Latent Representation
• Multiple-Attribute Text Style Transfer
• A Dual Reinforcement Learning Framework for Unsupervised Text Style Transfer
• Style Transfer for Texts: Retrain, Report Errors, Compare with Rewrites
• Unsupervised Text Generation by Learning from Search
• Break-It-Fix-It: Unsupervised Learning for Program Repair
• Neural Language Generation: Formulation, Methods, and Evaluation
• The Importance of Suppressing Domain Style in Authorship Analysis
• Decomposing Textual Information For Style Transfer
• Zero-Shot Fine-Grained Style Transfer: Leveraging Distributed Continuous Style Representations to Transfer To Unseen Styles
• Style Transformer: Unpaired Text Style Transfer without Disentangled Latent Representation
• Improving GAN Training with Probability Ratio Clipping and Sample Reweighting
• Generalized Data Augmentation for Low-Resource Translation
• Grammatical Error Correction and Style Transfer via Zero-shot Monolingual Translation
• On Synthetic Data for Back Translation
• XL-Editor: Post-editing Sentences with XLNet
• Hooks in the Headline: Learning to Generate Headlines with Controlled Styles
• Thank you BART! Rewarding Pre-Trained Models Improves Formality Style Transfer
• A Hierarchical Reinforced Sequence Operation Method for Unsupervised Text Style Transfer
• 3M: Multi-style image caption generation using Multi-modality features under Multi-UPDOWN model
• Expertise Style Transfer: A New Task Towards Better Communication between Experts and Laymen
• Multimodal Text Style Transfer for Outdoor Vision-and-Language Navigation
• IMaT: Unsupervised Text Attribute Transfer via Iterative Matching and Translation
• Fine-grained Sentiment Controlled Text Generation
• Generating Classical Chinese Poems from Vernacular Chinese
• On Learning Text Style Transfer with Direct Rewards
• Exploring Contextual Word-level Style Relevance for Unsupervised Style Transfer
• Unsupervised Neural Text Simplification
• Semi-supervised Text Style Transfer: Cross Projection in Latent Space
• Back-Training excels Self-Training at Unsupervised Domain Adaptation of Question Generation and Passage Retrieval
• SE-DAE: Style-Enhanced Denoising Auto-Encoder for Unsupervised Text Style Transfer
• NAST: A Non-Autoregressive Generator with Word Alignment for Unsupervised Text Style Transfer
• Multi-Pair Text Style Transfer on Unbalanced Data
• Generic resources are what you need: Style transfer tasks without task-specific parallel training data
• Transductive Learning for Unsupervised Text Style Transfer
• Preventing Author Profiling through Zero-Shot Multilingual Back-Translation
• GTAE: Graph-Transformer based Auto-Encoders for Linguistic-Constrained Text Style Transfer