A Dual Reinforcement Learning Framework for Unsupervised Text Style Transfer

Abstract

Unsupervised text style transfer aims to transfer the underlying style of text but keep its main content unchanged without parallel data. Most existing methods typically follow two steps: first separating the content from the original style, and then fusing the content with the desired style. However, the separation in the first step is challenging because the content and style interact in subtle ways in natural language. Therefore, in this paper, we propose a dual reinforcement learning framework to directly transfer the style of the text via a one-step mapping model, without any separation of content and style. Specifically, we consider the learning of the source-to-target and target-to-source mappings as a dual task, and two rewards are designed based on such a dual structure to reflect the style accuracy and content preservation, respectively. In this way, the two one-step mapping models can be trained via reinforcement learning, without any use of parallel data. Automatic evaluations show that our model outperforms the state-of-the-art systems by a large margin, especially with more than 8 BLEU points improvement averaged on two benchmark datasets. Human evaluations also validate the effectiveness of our model in terms of style accuracy, content preservation and fluency. Our code and data, including outputs of all baselines and our model are available at https://github.com/luofuli/DualLanST.

Accepted by IJCAI 2019

References in corpus (4)

• A Deep Reinforced Model for Abstractive Summarization
• Dual Learning for Machine Translation
• Style Transfer as Unsupervised Machine Translation
• LiveBot: Generating Live Video Comments Based on Visual and Textual Contexts

Cited by in corpus (11)

• Unsupervised Text Generation by Learning from Search
• From Theories on Styles to their Transfer in Text: Bridging the Gap with a Hierarchical Survey
• Cycle-Consistent Adversarial Autoencoders for Unsupervised Text Style Transfer
• XL-Editor: Post-editing Sentences with XLNet
• A Neural Approach to Irony Generation
• Unsupervised Dual Paraphrasing for Two-stage Semantic Parsing
• ST${}^2$: Small-data Text Style Transfer via Multi-task Meta-Learning
• Exploring Contextual Word-level Style Relevance for Unsupervised Style Transfer
• XFORMAL: A Benchmark for Multilingual Formality Style Transfer
• GTAE: Graph-Transformer based Auto-Encoders for Linguistic-Constrained Text Style Transfer
• SE-DAE: Style-Enhanced Denoising Auto-Encoder for Unsupervised Text Style Transfer