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

Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, pages 2701–2711, Hong Kong, China, November 3–7, 2019. c⃝2019 Association for Computational Linguistics 2701 Keep Calm and Switch On! Preserving Sentiment and Fluency in Semantic Text Exchange Steven Y. Feng∗ Aaron W. Li∗ David R. Cheriton School of Computer Science University of Waterloo Waterloo, Ontario, Canada {sy2feng, w89li, jhoey}@uwaterloo.ca Jesse Hoey 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 se- mantic correction of text generated by chat- bots and virtual assistants. We introduce a pipeline called SMERTI that combines entity replacement, similarity masking, and text in- filling. 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 con- trol the amount of semantic change in the text. Our experiments demonstrate that SMERTI can outperform baseline models on Yelp re- views, Amazon reviews, and news headlines. 1 Introduction There has been significant research on style transfer, with the goal of changing the style of text while preserving its semantic content. The alternative where semantics are adjusted while keeping style intact, which we call semantic text exchange (STE), has not been investi- gated to the best of our knowledge. Consider the following example, where the replace- ment entity defines the new semantic context: Original Text: It is sunny outside! Ugh, that means I must wear sunscreen. I hate being sweaty and sticky all over. Replacement Entity: weather = rainy Desired Text: It is rainy outside! Ugh, that means I must bring an umbrella. I hate being wet and having to carry it around. The weather within the original text is sunny, ∗Authors contributed equally whereas the actual weather may be rainy. Not only is the word sunny replaced with rainy, but the rest of the text’s content is changed while preserving its negative sentiment and fluency. With the rise of natural language processing (NLP) has come an increased demand for mas- sive amounts of text data. Manually collecting and scraping data requires a significant amount of time and effort, and data augmentation techniques for NLP are limited compared to fields such as com- puter vision. STE can be used for text data aug- mentation by producing various modifications of a piece of text that differ in semantic content. Another use of STE is in building emotionally aligned chatbots and virtual assistants. This is useful for reasons such as marketing, overall en- joyment of interaction, and mental health therapy. However, due to limited data with emotional con- tent in specific semantic contexts, the generated text may contain incorrect semantic content. STE can adjust text semantics (e.g. to align with reality or a …

Content truncated to 500 words