Structured Content Preservation for Unsupervised Text Style Transfer

TL;DR

This paper introduces a structured content preserving model for unsupervised text style transfer that leverages linguistic information to better retain content while changing style, outperforming existing methods.

Contribution

The paper proposes a novel model that uses structured fine-grained supervision and linguistic features to improve content preservation in style transfer tasks.

Findings

Significant improvement in content preservation over existing models.

Enhanced style transfer quality demonstrated through automatic and human evaluations.

Effective use of linguistic information and language models for style transfer.

Abstract

Text style transfer aims to modify the style of a sentence while keeping its content unchanged. Recent style transfer systems often fail to faithfully preserve the content after changing the style. This paper proposes a structured content preserving model that leverages linguistic information in the structured fine-grained supervisions to better preserve the style-independent content during style transfer. In particular, we achieve the goal by devising rich model objectives based on both the sentence's lexical information and a language model that conditions on content. The resulting model therefore is encouraged to retain the semantic meaning of the target sentences. We perform extensive experiments that compare our model to other existing approaches in the tasks of sentiment and political slant transfer. Our model achieves significant improvement in terms of both content preservation and style transfer in automatic and human evaluation.