Elamur Research⌕
☰
home / papers / 2015 / Humor Recognition and Humor Anchor Extraction

Humor Recognition and Humor Anchor Extraction

Paper · 2015 · Diyi Yang, Alon Lavie, Chris Dyer et al.

Summary

This EMNLP 2015 paper addresses two tasks: deciding whether a short sentence is humorous, and locating the words or phrases that trigger the joke, which the authors call humor anchors. They frame recognition as binary classification and build feature groups around four proposed sources of humor: incongruity, ambiguity, interpersonal effect and phonetic style. Positive examples come from Pun of the Day and the 16,000 One-Liner dataset. Negative sentences are drawn from AP News, New York Times, Yahoo! Answers and proverbs, then filtered to match the humorous data in vocabulary and length so that topic differences do not inflate accuracy. For anchors, they propose Maximal Decrement Extraction (MDE). It builds on the trained recognizer, but its full procedure falls in a part of the text that was unavailable for this summary. Compared with three human annotators, MDE clearly beats random and POS-based baselines. The work matters because it is presented as the first attempt to extract humor anchors from general sentences, and because its negative-sampling design guards against misleading results.

Key points

  • Datasets:

    • Pun of the Day: 2423 positive and 2403 negative sentences.
    • 16,000 One-Liners: 16,000 positive and 16,002 negative sentences.
    • The proverb negatives (654) were extracted manually.

  • Negative filtering: every word of a negative sentence must appear in the vocabulary of the positive set, and sentences are limited to roughly 10–30 words.

  • Incongruity features: pretrained generic Word2Vec vectors give two features over content-word pairs. Disconnection is the maximum distance between a pair, and Repetition is the minimum.

  • Ambiguity feature: Sense Combination is the log of the product of WordNet sense counts for nouns, verbs, adjectives and adverbs.

  • Anchor results on Pun of the Day (F1):

    MethodExact matchingAt-Least-One matching
    MDE0.4380.756
    POS baseline0.1650.360
    Random baseline0.1160.285

    On One-Liners, MDE reaches 0.288 (Exact) and 0.616 (At-Least-One).

Interesting ideas

  • I1. Domain-matched negative sampling: restricting negatives to the positive set's vocabulary and length range stops a classifier from simply learning topic, such as news versus everyday life. This is a useful check for any task where positives and negatives come from different sources.
  • I2. Theory-driven feature groups: tying each feature set to a specific hypothesis about humor makes the model easier to interpret, and lets you test which mechanism contributes. The same approach could suit sarcasm or figurative-language detection.
  • I3. Cheap proxies for abstract phenomena: the paper approximates incongruity with embedding distances between word pairs and ambiguity with WordNet sense counts. These are simple, reusable signals wherever wordnets and embeddings exist, including for Ukrainian.
  • I4. Strict and lenient span matching: reporting both Exact and At-Least-One matching shows how good the extracted spans are when boundaries are fuzzy. It also helps for keyword or rationale extraction evaluated against several annotators, whose scores can be averaged.
  • I5. Explanation from an existing classifier: MDE derives anchors from the trained recognizer instead of training a separate span model. This makes anchor extraction largely independent of which classifier is used underneath.

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

 Institutions

🎓Carnegie Mellon UniversityPittsburgh · United States365 works · 158,048 citations

📚 Shelves 6 Computer Science · 9,086 works Artificial Intelligence · 8,889 works Arts and Humanities · 9,296 works Language and Linguistics · 9,206 works Psychology · 798 works Social Psychology · 129 works

Cite

Cite as Diyi Yang, Alon Lavie, Chris Dyer and Eduard Hovy. 2015. Humor Recognition and Humor Anchor Extraction. Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/d15-1284

@inproceedings{yang2015humor,
  title = {{Humor Recognition and Humor Anchor Extraction}},
  author = {Diyi Yang, Alon Lavie, Chris Dyer and Eduard Hovy},
  booktitle = {Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing},
  year = {2015},
  doi = {10.18653/v1/d15-1284},
  url = {https://doi.org/10.18653/v1/d15-1284}
}

References

28 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Salvatore Attardo. 1994. Linguistic theories of humor, volume 1. Walter de Gruyter. find
  2. Tristan A Bekinschtein, Matthew H Davis, Jennifer M Rodd, and Adrian M Owen. 2011. Why clowns taste funny: the relationship between humor and semantic ambiguity. The Journal of Neuroscience, 31(26):9665–9671. find
  3. Kim Binsted and Graeme Ritchie. 1997. Computational rules for generating punning riddles. Humor: International Journal of Humor Research. find
  4. Chiara Bucaria. 2004. Lexical and syntactic ambiguity as a source of humor: The case of newspaper headlines. Humor, 17(3):279–310. find
  5. Christiane Fellbaum. 1998. WordNet. Wiley Online Library. find
  6. Bryan Anthony Hong and Ethel Ong. 2009. Automatically extracting word relationships as templates for pun generation. In Proceedings of the Workshop on Computational Approaches to Linguistic Creativity, CALC ’09, pages 24–31, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  7. Chloe Kiddon and Yuriy Brun. 2011. That’s what she said: double entendre identification. In Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies: short papers-Volume 2, pages 89–94. Association for Computational Linguistics. find
  8. Sujay Kumar Jauhar, Chris Dyer, and Eduard Hovy. 2015. Ontologically grounded multi-sense representation learning for semantic vector space models. In The 2015 Conference of the North American Chapter of the Association for Computational Linguistics. find
  9. Igor Labutov and Hod Lipson. 2012. Humor as circuits in semantic networks. In Proceedings of the 50th Annual Meeting of the Association for Computational Linguistics: Short Papers-Volume 2, pages 150–155. Association for Computational Linguistics. find
  10. Herbert M Lefcourt. 2001. Humor: The psychology of living buoyantly. Springer Science & Business Media. find
  11. Rada Mihalcea and Carlo Strapparava. 2005. Making computers laugh: Investigations in automatic humor recognition. In Proceedings of the Conference on Human Language Technology and Empirical Methods in Natural Language Processing, pages 531– 538. Association for Computational Linguistics. our card
  12. Tomas Mikolov, Ilya Sutskever, Kai Chen, Greg S Corrado, and Jeff Dean. 2013. Distributed representations of words and phrases and their compositionality. In Advances in Neural Information Processing Systems, pages 3111–3119. our card
  13. Tristan Miller and Iryna Gurevych. 2015. Automatic disambiguation of english puns. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 719–729, Beijing, Chi… find
  14. Gözde Ozbal and Carlo Strapparava. 2012. Computational humour for creative naming. find
  15. Computational Humor 2012, page 15. find
  16. John Allen Paulos. 2008. Mathematics and humor: A study of the logic of humor. University of Chicago Press. find
  17. Amruta Purandare and Diane Litman. 2006. Humor: Prosody analysis and automatic recognition for f*r*i*e*n*d*s*. In Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing, EMNLP ’06, pages 208–215, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  18. Victor Raskin. 1985. Semantic mechanisms of humor, volume 24. Springer. find
  19. Yishay Raz. 2012. Automatic humor classification on twitter. In Proceedings of the 2012 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Student Research Workshop, pages 66–70. Association for Computational Linguistics. find
  20. Graeme Ritchie. 2005. Computational mechanisms for pun generation. In Proceedings of the 10th European Natural Language Generation Workshop, pages 125–132. Citeseer. find
  21. Oliviero Stock and Carlo Strapparava. 2003. Getting serious about the development of computational humor. In IJCAI, volume 3, pages 59–64. find
  22. Oliviero Stock and Carlo Strapparava. 2005. Hahacronym: A computational humor system. In Proceedings of the ACL 2005 on Interactive poster and demonstration sessions, pages 113–116. Association for Computational Linguistics. find
  23. J Taylor and L Mazlack. 2004. Computationally recognizing wordplay in jokes. Proceedings of CogSci 2004. find
  24. Julia M Taylor. 2009. Computational detection of humor: A dream or a nightmare? the ontological semantics approach. In Proceedings of the 2009 IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology-Volume 03, pages 429–432. IEEE Computer Society. find
  25. Kristina Toutanova, Dan Klein, Christopher D. Manning, and Yoram Singer. 2003. Feature-rich part-of-speech tagging with a cyclic dependency network. In Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology - Vol… our card
  26. Janyce Wiebe and Rada Mihalcea. 2006. Word sense and subjectivity. In Proceedings of the 21st International Conference on Computational Linguistics and the 44th annual meeting of the Association for Computational Linguistics, pages 1065–1072. Association for Computational Linguistics. find
  27. Theresa Wilson, Janyce Wiebe, and Paul Hoffmann. 2005. Recognizing contextual polarity in phrase-level sentiment analysis. In Proceedings of Human Language Technology Conference and Conference on Empirical Methods in Natural Language Processing, pages 347–354, Vancouver, British Columbia, Canada, Oc… our card
  28. Renxian Zhang and Naishi Liu. 2014. Recognizing humor on twitter. In Proceedings of the 23rd ACM International Conference on Conference on Information and Knowledge Management, pages 889–898. ACM. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/yang-2015-humor-recognition-humor-anchor-extraction/