Semi-supervised Text Style Transfer: Cross Projection in Latent Space

TL;DR

This paper introduces a semi-supervised approach for text style transfer that leverages both parallel and nonparallel data, utilizing a projection function in latent space to improve style transfer quality.

Contribution

It proposes a novel semi-supervised model with a projection function in latent space and introduces a new dataset for style transfer between ancient and modern Chinese.

Findings

Effective style transfer with limited parallel data

Improved content preservation and style accuracy

New dataset for Chinese style transfer

Abstract

Text style transfer task requires the model to transfer a sentence of one style to another style while retaining its original content meaning, which is a challenging problem that has long suffered from the shortage of parallel data. In this paper, we first propose a semi-supervised text style transfer model that combines the small-scale parallel data with the large-scale nonparallel data. With these two types of training data, we introduce a projection function between the latent space of different styles and design two constraints to train it. We also introduce two other simple but effective semi-supervised methods to compare with. To evaluate the performance of the proposed methods, we build and release a novel style transfer dataset that alters sentences between the style of ancient Chinese poem and the modern Chinese.