Text Style Transfer

• DOI: 10.1145/3544903.3544906
• Publication Date: 2022-06-21T22:08:07Z
• Authors: Zhiqiang Hu; Roy Ka-Wei Lee; Charu C. Aggarwal; Aston Zhang
• Keywords: Benchmark (surveying)

The stylistic properties of text have intrigued computational linguistics researchers in recent years. Specifically, investigated the style transfer task (TST), which aims to change while retaining its independent content style. Over last few years, many novel TST algorithms been developed, industry has leveraged these enable exciting applications. field research developed because this symbiosis. This article provide a comprehensive review efforts on transfer. More concretely, we create taxonomy organize models, and summary state art. We existing evaluation methodologies for tasks conduct large-scale reproducibility study experimentally benchmark 19 state-of-the-art two publicly available datasets. Finally, expand current trends new perspectives developments field.