Elamur ResearchβŒ•
☰
home / papers / 2021 / Chandler: An Explainable Sarcastic Response Generator

Chandler: An Explainable Sarcastic Response Generator

Summary

Earlier sarcasm generators assume that sarcasm simply means the opposite of what is said. The authors argue this assumption is neither necessary nor sufficient for sarcasm. Instead, they build Chandler on Utsumi's Implicit Display Theory (IDT). IDT treats an utterance as sarcastic exactly when three conditions hold. There must be an ironic environment, meaning a speaker's expectation failed and they feel negatively about it. The utterance must also implicitly display that environment: it alludes to the failed expectation, is pragmatically insincere, and hints at the negative attitude.

Chandler first infers the ironic environment from an input utterance, then produces a sarcastic reply. Because each step follows the theory, the system can also explain why each reply counts as sarcastic. The authors see this transparency as important, since sarcasm can easily offend or be misread.

Crowdworkers on Prolific rated outputs for 250 inputs balanced by sentiment. Chandler was more sarcastic than DialoGPT-based systems and slightly less sarcastic than the template-based SarcasmBot. However, it was far more diverse and usually more specific to the input. This makes it a more realistic basis for sarcastic conversational agents.

Key points

  • The system is grounded in Implicit Display Theory (Utsumi, 1996), not the traditional "opposite meaning" view. The paper also reviews Gricean, echoic and pretense theories and finds them lacking.
  • There are four Chandler variants, each alluding to the failed expectation through a different relation type: Ch-xAttr, Ch-xNeed, Ch-xReact and Ch-xEffect.
  • The evaluation compared the variants with DialoGPT, DialoGPT+R3 and SarcasmBot. It used 250 inputs, 3 annotations per input, and 0–4 scales for sarcasm, humour, coherence and specificity.
  • Ch-xAttr was the most sarcastic variant (2.1), against 2.5 for SarcasmBot and 0.6 for plain DialoGPT. Diversity, measured as the share of unique responses, was 0.80 for Ch-xAttr and Ch-xNeed versus 0.14 for SarcasmBot, which produced only 28 distinct replies.
  • A live demo is available, and the authors planned to release the code, model checkpoints, inputs and responses under CC BY-NC.

Our summary of the paper, in our own words β€” not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Cited by

7 papers of our library cite it (of the 140673 whose bibliographies we read); 3 of them have a page here, 4 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2021.emnlp-demo.38

@inproceedings{oprea2021chandler,
  title = {{Chandler: An Explainable Sarcastic Response Generator}},
  author = {Oprea, Silviu and Wilson, Steven and Magdy, Walid},
  booktitle = {Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing: System Demonstrations},
  year = {2021},
  doi = {10.18653/v1/2021.emnlp-demo.38},
  url = {https://doi.org/10.18653/v1/2021.emnlp-demo.38}
}

References

31 references, 12 of them with a card of ours. β†— leads to the work, Β«findΒ» searches for one we have no id for yet.

  1. Silvio Amir, Byron C. Wallace, Hao Lyu, Paula Carvalho, and Mario J. Silva. 2016. Modelling context with user embeddings for sarcasm detection in social media. In CoNLL, pages 167–177. ACL. β†— our card
  2. David Bamman and Noah A. Smith. 2015. Contextualized sarcasm detection on twitter. In ICWSM, pages 574–577. AAAI Press. find
  3. Francesco Barbieri, Jose Camacho-Collados, Luis Espinosa Anke, and Leonardo Neves. 2020. TweetEval: Unified benchmark and comparative evaluation for tweet classification. In Findings of the Association for Computational Linguistics: EMNLP 2020, pages 1644–1650, Online. ACL. find our card
  4. Antoine Bosselut, Hannah Rashkin, Maarten Sap, Chaitanya Malaviya, Asli Celikyilmaz, and Yejin Choi. 2019. COMET: Commonsense transformers for automatic knowledge graph construction. In ACL, pages 4762–4779. ACL. find our card
  5. Richard Bruntsch and Willibald Ruch. 2017. Studying irony detection beyond ironic criticism: Let’s include ironic praise. Frontiers in Psychology, 8:606. β†—
  6. Tuhin Chakrabarty, Debanjan Ghosh, Smaranda Muresan, and Nanyun Peng. 2020. RΛ†3: Reverse, retrieve, and rank for sarcasm generation with commonsense knowledge. In ACL, pages 7976–7986. ACL. find our card
  7. Herbert H. Clark. 1996. Using Language. Cambridge University Press. find
  8. Aditya Joshi, Vaibhav Tripathi, Kevin Patel, Pushpak Bhattacharyya, and Mark Carman. 2016. Are word embedding-based features useful for sarcasm detection? In EMNLP, pages 1006–1011. ACL. β†— our card
  9. Mikhail Khodak, Nikunj Saunshi, and Kiran Vodrahalli. 2018. A large self-annotated corpus for sarcasm. In LREC. ELRA. find our card
  10. Roger J. Kreuz and Sam Glucksberg. 1989. How to Be Sarcastic: The Echoic Reminder Theory of Verbal Irony. Journal of Experimental Psychology: General, 118(4):374–386. β†—
  11. Sachi Kumon-Nakamura, Sam Glucksberg, and Mary Brown. 1995. How about another piece of pie: The allusional pretense theory of discourse irony. β†—
  12. Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692. β†—
  13. William C. Mann and Sandra A. Thompson. 1987. Rhetorical Structure Theory: Description and Construction of Text Structures, pages 85–95. Springer Netherlands, Dordrecht. find
  14. George A. Miller. 1995. Wordnet: A lexical database for english. Communications of the ACM, 38(11):39–41. β†— our card
  15. Abhijit Mishra, Tarun Tater, and Karthik Sankaranarayanan. 2019. A modular architecture for unsupervised sarcasm generation. In EMNLP-IJCNLP, pages 6144–6154. ACL. find our card
  16. Silviu Oprea and Walid Magdy. 2019. Exploring author context for detecting intended vs perceived sarcasm. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 2854–2859, Florence, Italy. Association for Computational Linguistics. β†— our card
  17. Herbert H. Clark and Richard J. Gerrig. 1984. On the pretense theory of irony. Journal of Experimental Psychology: General, 113(1):121–126. β†—
  18. Rachel Giora. 1995. On irony and negation. Discourse Processes, 19(2):239–264. β†—
  19. H. P. Grice. 1975. Logic and conversation. In Peter Cole and Jerry L. Morgan, editors, Syntax and Semantics: Vol. 3: Speech Acts, pages 41–58. Academic Press, Cambridge, UK. β†—
  20. Devamanyu Hazarika, Soujanya Poria, Sruthi Gorantla, Erik Cambria, Roger Zimmermann, and Rada Mihalcea. 2018. Cascade: Contextual sarcasm detection in online discussion forums. In COLING, pages 1837–1848. ACL. find our card
  21. Julia Jorgensen. 1996. The functions of sarcastic irony in speech. Journal of Pragmatics, 26(5):613–634. β†—
  22. Aditya Joshi, Anoop Kunchukuttan, Mark James Carman, and Pushpak Bhattacharyya. 2015. Sarcasmbot: An open-source sarcasm-generation module for chatbots. In WISDOM at KDD. ACM. find
  23. Penny M. Pexman and Meghan T. Zvaigzne. 2004. Does irony go better with friends? Metaphor and Symbol, 19(2):143–163. β†—
  24. Ashwin Rajadesingan, Reza Zafarani, and Huan Liu. 2015. Sarcasm detection on twitter: A behavioral modeling approach. In WSDM, pages 97–106. ACM. β†—
  25. Ellen Riloff, Ashequl Qadir, Prafulla Surve, Lalindra De Silva, Nathan Gilbert, and Ruihong Huang. 2013. Sarcasm as contrast between a positive sentiment and negative situation. In EMNLP, pages 704–714. ACL. find our card
  26. Sara Rosenthal, Noura Farra, and Preslav Nakov. 2017. SemEval-2017 task 4: Sentiment analysis in Twitter. In Proceedings of the 11th International Workshop on Semantic Evaluation (SemEval-2017), pages 502–518, Vancouver, Canada. ACL. find our card
  27. Maarten Sap, Ronan Le Bras, Emily Allaway, Chandra Bhagavatula, Nicholas Lourie, Hannah Rashkin, Brendan Roof, Noah A. Smith, and Yejin Choi. 2019. Atomic: An atlas of machine commonsense for if-then reasoning. Proceedings of the AAAI Conference on Artificial Intelligence, 33(01):3027–3035. β†—
  28. John R Searle and John Rogers Searle. 1969. Speech acts: An essay in the philosophy of language, volume 626. Cambridge University Press, Cambridge, UK. find
  29. Dan Sperber and Deirdre Wilson. 1981. Irony and the use-mention distinction. Philosophy, 3:143–184. find
  30. Dan Sperber and Deirdre Wilson. 1998. Irony and relevance: A reply to seto, hamamoto and yamanashi. In R. Carston and S. Uchida, editors, Relevance theory: Applications and implications, pages 289–293. Benjamins, Amsterdam. find
  31. Martin Sykora, Suzanne Elayan, and Thomas W Jackson. 2020. A qualitative analysis of sarcasm, irony and related #hashtags on twitter. Big Data & Society, 7(2):2053951720972735. β†—

↑ back to top

This page in Markdown β†’ https://md.research.elamur.ai/free/papers/oprea-2021-chandler-explainable-sarcastic-response-generator/