Does Commonsense help in detecting Sarcasm?

TL;DR

This paper investigates whether integrating commonsense knowledge improves sarcasm detection in NLP, finding that it does not outperform baseline models and analyzing the contexts where it helps or hinders.

Contribution

The study introduces a method to incorporate commonsense knowledge into sarcasm detection using graph convolution networks and thoroughly analyzes its impact.

Findings

Commonsense knowledge integration does not outperform baseline models.

Analysis reveals specific scenarios where commonsense helps or hurts detection.

Public implementation available for further research.

Abstract

Sarcasm detection is important for several NLP tasks such as sentiment identification in product reviews, user feedback, and online forums. It is a challenging task requiring a deep understanding of language, context, and world knowledge. In this paper, we investigate whether incorporating commonsense knowledge helps in sarcasm detection. For this, we incorporate commonsense knowledge into the prediction process using a graph convolution network with pre-trained language model embeddings as input. Our experiments with three sarcasm detection datasets indicate that the approach does not outperform the baseline model. We perform an exhaustive set of experiments to analyze where commonsense support adds value and where it hurts classification. Our implementation is publicly available at: https://github.com/brcsomnath/commonsense-sarcasm.