Understanding irony through explanations and background knowledge
- Author
- Aaron Maladry (UGent) , Cynthia Van Hee (UGent) , Els Lefever (UGent) and Veronique Hoste (UGent)
- Organization
- Project
- Abstract
- This article investigates the automatic explanation of irony in English tweets. The work covers the development and validation of a conceptual framework for annotating knowledge-informed explanations for figurative language as well as the training and evaluation of specialized generative models. Human judgements confirm that both fine-tuned open-source models (Llama 3) and proprietary models (GPT-4) can produce high-quality explanations, effectively incorporating relevant world knowledge. While metrics like BLUE and ROUGE do not seem to align with human judgement, we find that semantic similarity measures align well with human quality estimations. The resulting models and datasets for irony explanations, published as the iRONNIE collection, actively bridge the gap between theoretical understanding of irony and the technical innovations of the NLP domain. The models are be released to the public to facilitate a deeper linguistic analysis of world knowledge involved in understanding irony on social media in future work.
- Keywords
- explanation generation, world knowledge, irony, sarcasm
Downloads
-
2026.cas-1.2.pdf
- full text (Published version)
- |
- open access
- |
- |
- 347.30 KB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-01KM2V96NGPX7FD1JX858VN6RY
- MLA
- Maladry, Aaron, et al. “Understanding Irony through Explanations and Background Knowledge.” Proceedings of Computational Affective Science (CAS) @ LREC 2026, edited by Christopher Bagdon et al., European Language Resources Association, 2026, pp. 16–28, doi:10.63317/3sr9y3ffdu8w.
- APA
- Maladry, A., Van Hee, C., Lefever, E., & Hoste, V. (2026). Understanding irony through explanations and background knowledge. In C. Bagdon, K. Vishnubhotla, K. A. Lindquist, L. Ungar, R. Klinger, & S. M. Mohammad (Eds.), Proceedings of Computational Affective Science (CAS) @ LREC 2026 (pp. 16–28). https://doi.org/10.63317/3sr9y3ffdu8w
- Chicago author-date
- Maladry, Aaron, Cynthia Van Hee, Els Lefever, and Veronique Hoste. 2026. “Understanding Irony through Explanations and Background Knowledge.” In Proceedings of Computational Affective Science (CAS) @ LREC 2026, edited by Christopher Bagdon, Krishnapriya Vishnubhotla, Kristen A. Lindquist, Lyle Ungar, Roman Klinger, and Saif M. Mohammad, 16–28. European Language Resources Association. https://doi.org/10.63317/3sr9y3ffdu8w.
- Chicago author-date (all authors)
- Maladry, Aaron, Cynthia Van Hee, Els Lefever, and Veronique Hoste. 2026. “Understanding Irony through Explanations and Background Knowledge.” In Proceedings of Computational Affective Science (CAS) @ LREC 2026, ed by. Christopher Bagdon, Krishnapriya Vishnubhotla, Kristen A. Lindquist, Lyle Ungar, Roman Klinger, and Saif M. Mohammad, 16–28. European Language Resources Association. doi:10.63317/3sr9y3ffdu8w.
- Vancouver
- 1.Maladry A, Van Hee C, Lefever E, Hoste V. Understanding irony through explanations and background knowledge. In: Bagdon C, Vishnubhotla K, Lindquist KA, Ungar L, Klinger R, Mohammad SM, editors. Proceedings of Computational Affective Science (CAS) @ LREC 2026. European Language Resources Association; 2026. p. 16–28.
- IEEE
- [1]A. Maladry, C. Van Hee, E. Lefever, and V. Hoste, “Understanding irony through explanations and background knowledge,” in Proceedings of Computational Affective Science (CAS) @ LREC 2026, Palma, Mallorca (Spain), 2026, pp. 16–28.
@inproceedings{01KM2V96NGPX7FD1JX858VN6RY,
abstract = {{This article investigates the automatic explanation of irony in English tweets. The work covers the development and
validation of a conceptual framework for annotating knowledge-informed explanations for figurative language as well as the training and evaluation of specialized generative models. Human judgements confirm that both fine-tuned open-source models (Llama 3) and proprietary models (GPT-4) can produce high-quality explanations, effectively incorporating relevant world knowledge. While metrics like BLUE and ROUGE do not seem to align with human judgement, we find that semantic similarity measures align well with human quality estimations. The resulting models and datasets for irony explanations, published as the iRONNIE collection, actively bridge the gap between theoretical understanding of irony and the technical innovations of the NLP domain. The models are be released to the public to facilitate a deeper linguistic analysis of world knowledge involved in understanding irony on social media in future work.}},
author = {{Maladry, Aaron and Van Hee, Cynthia and Lefever, Els and Hoste, Veronique}},
booktitle = {{Proceedings of Computational Affective Science (CAS) @ LREC 2026}},
editor = {{Bagdon, Christopher and Vishnubhotla, Krishnapriya and Lindquist, Kristen A. and Ungar, Lyle and Klinger, Roman and Mohammad, Saif M.}},
keywords = {{explanation generation,world knowledge,irony,sarcasm}},
language = {{eng}},
location = {{Palma, Mallorca (Spain)}},
pages = {{16--28}},
publisher = {{European Language Resources Association}},
title = {{Understanding irony through explanations and background knowledge}},
url = {{http://doi.org/10.63317/3sr9y3ffdu8w}},
year = {{2026}},
}
- Altmetric
- View in Altmetric