Disentangled Representation Learning for Text Style Transfer
share
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 objectives: a multi-task classification objective, and dual 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, using this approach. This disentangled latent representation learning method is applied to attribute (e.g. style) transfer on non-parallel corpora. We achieve similar content preservation scores compared to previous state-of-the-art approaches, and significantly better style-transfer strength scores. Our code is made publicly available for replicability and extension purposes.
READ FULL TEXT