Keep Calm and Switch On! Preserving Sentiment and Fluency in Semantic Text Exchange

TL;DR

This paper introduces SMERTI, a pipeline that effectively adjusts text semantics while maintaining sentiment and fluency, aiding data augmentation and correction in conversational AI.

Contribution

The paper presents a novel semantic text exchange method combining entity replacement, similarity masking, and text infilling, with a new evaluation metric called STES.

Findings

SMERTI outperforms baseline models on multiple datasets

The masking rate threshold effectively controls semantic change

SMERTI preserves sentiment and fluency during semantic adjustments

Abstract

In this paper, we present a novel method for measurably adjusting the semantics of text while preserving its sentiment and fluency, a task we call semantic text exchange. This is useful for text data augmentation and the semantic correction of text generated by chatbots and virtual assistants. We introduce a pipeline called SMERTI that combines entity replacement, similarity masking, and text infilling. We measure our pipeline's success by its Semantic Text Exchange Score (STES): the ability to preserve the original text's sentiment and fluency while adjusting semantic content. We propose to use masking (replacement) rate threshold as an adjustable parameter to control the amount of semantic change in the text. Our experiments demonstrate that SMERTI can outperform baseline models on Yelp reviews, Amazon reviews, and news headlines.