Advancing Sarcasm Detection: The Case for Naturalistic Conversational Datasets

View Full Paper

Ask AI

Bookmark

Authors

Discussion

This review analyzes sarcasm detection datasets, highlighting the importance of multimodal resources for enhancing NLP applications.

Sarcasm detection can improve NLP applications, and diverse datasets are crucial for accurate model training.
The review categorizes 46 papers and identifies three types of sarcasm detection datasets: text-based, visual-textual, and audio-visual.
Naturalistic conversational data is needed to capture emotional tone and speaker variability in sarcasm detection.
Publicly available datasets often lack critical elements, limiting the effectiveness of sarcasm detection in real-world scenarios.

Awasthi et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark