Disentangled Representation Learning for Non-Parallel Text Style Transfer
It addresses the problem of style transfer without parallel data for natural language processing applications, but is incremental as it builds on existing disentanglement techniques.
The paper tackled disentangling style and content in language models for non-parallel text style transfer, achieving substantially better results in transfer accuracy, content preservation, and fluency compared to previous state-of-the-art methods.
This paper tackles the problem of disentangling the latent variables of style and content in language models. We propose a simple yet effective approach, which incorporates auxiliary multi-task and adversarial objectives, for label prediction and bag-of-words prediction, respectively. We show, both qualitatively and quantitatively, that the style and content are indeed disentangled in the latent space. This disentangled latent representation learning method is applied to style transfer on non-parallel corpora. We achieve substantially better results in terms of transfer accuracy, content preservation and language fluency, in comparison to previous state-of-the-art approaches.