Unpaired Sentiment-to-Sentiment Translation: A Cycled Reinforcement Learning Approach

TL;DR

This paper introduces a cycled reinforcement learning method for unpaired sentiment-to-sentiment translation, effectively changing sentiment while preserving content and outperforming existing systems.

Contribution

It proposes a novel cycled reinforcement learning approach with neutralization and emotionalization modules for unpaired sentiment transfer.

Findings

Significant improvement in BLEU scores on Yelp and Amazon datasets

Outperforms state-of-the-art sentiment translation systems

Enhances content preservation in sentiment transfer

Abstract

The goal of sentiment-to-sentiment "translation" is to change the underlying sentiment of a sentence while keeping its content. The main challenge is the lack of parallel data. To solve this problem, we propose a cycled reinforcement learning method that enables training on unpaired data by collaboration between a neutralization module and an emotionalization module. We evaluate our approach on two review datasets, Yelp and Amazon. Experimental results show that our approach significantly outperforms the state-of-the-art systems. Especially, the proposed method substantially improves the content preservation performance. The BLEU score is improved from 1.64 to 22.46 and from 0.56 to 14.06 on the two datasets, respectively.