Style Transfer as Unsupervised Machine Translation
This addresses the problem of generating stylistically varied text without parallel data for researchers and practitioners in natural language processing, representing an incremental improvement.
The paper tackled the lack of parallel data for language style transfer by adapting unsupervised machine translation methods, resulting in a model that outperformed previous state-of-the-art models in accuracy and quality on benchmark datasets.
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.