Elamur Research⌕
☰
home / papers / 2014 / Modelling Irony in Twitter

Modelling Irony in Twitter

Summary

Barbieri and Saggion treat irony detection on Twitter as binary classification. They separate ironic tweets from non-ironic ones using hand-crafted lexical features rather than word identity. They used the corpus built by Reyes et al. (2013), which contains 40,000 tweets split evenly across four hashtag-defined topics: #irony, #education, #humour and #politics. The hashtags were stripped before experiments. The authors designed seven feature groups, many intended to capture the "unexpectedness" associated with irony: Frequency, Written-Spoken, Intensity, Structure, Sentiments, Synonyms and Ambiguity. Word frequencies come from the American National Corpus (ANC) Frequency Data. They trained Decision Tree and Random Forest classifiers in Weka and compared them with a bag-of-words baseline and with Reyes et al. (2013). Bag-of-words wins within a single topic, which suggests it mostly learns topic cues. When training and test topics differ, however, the feature-based model is significantly better. It also matches or slightly exceeds the prior state of the art, with more balanced precision and recall. The features are cheap to compute, so the approach is easy to replicate.

Key points

  • Data: 40,000 tweets from Reyes et al. (2013), in four classes of 10,000 each. Irony is the positive class; Education, Humour and Politics are the negative domains.
  • Novel feature groups: the authors present Frequency, Written-Spoken, Intensity and Synonyms as new for this task. These features rely on ANC frequency data, WordNet tools and the vinhkhuc Twitter Tokenizer.
  • Comparison with Reyes et al. (2013), Decision Tree F1:
    • Education: 0.73 vs 0.70
    • Politics: 0.75 vs 0.73
    • Humour: 0.75 vs 0.76
  • Cross-domain results: the advantage over bag-of-words was statistically significant at 99% confidence. Random Forest beats Decision Tree in cross-validation, but the two are comparable across domains.
  • Most informative features (by information gain):
    • The top features are rarest word frequency, synonym-based features and punctuation, with four features scoring above 0.1.
    • Ironic tweets use few emoticons, but about one in five of them is a wink. In the other corpora the rate is roughly one in 18 to 30.

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

Fields: Linguistics & languages

Cited by

18 papers of our library cite it (of the 140673 whose bibliographies we read); 4 of them have a page here, 14 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.3115/v1/e14-3007

@inproceedings{barbieri2014modelling,
  title = {{Modelling Irony in Twitter}},
  author = {Barbieri, Francesco and Saggion, Horacio},
  booktitle = {Proceedings of the Student Research Workshop at the 14th Conference of the European Chapter of the Association for Computational Linguistics},
  year = {2014},
  doi = {10.3115/v1/e14-3007},
  url = {https://doi.org/10.3115/v1/e14-3007}
}

References

27 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Kalina Bontcheva, Leon Derczynski, Adam Funk, Mark A. Greenwood, Diana Maynard, and Niraj Aswani. 2013. TwitIE: An Open-Source Information Extraction Pipeline for Microblog Text. In Proceedings of Recent Advances in Natural Language Processing Conferemce. ↗
  2. Paula Carvalho, Luı́s Sarmento, Mário J Silva, and Eugénio de Oliveira. 2009. Clues for detecting irony in user-generated contents: oh...!! it’s so easy;-). In Proceedings of the 1st international CIKM workshop on Topic-sentiment analysis for mass opinion, pages 53–56. ACM. find
  3. Kushal Dave, Steve Lawrence, and David M. Pennock. 2003. Mining the peanut gallery: Opinion extraction and semantic classification of product reviews. In Proceedings of the 12th International Conference on World Wide Web, WWW ’03, pages 519–528, New York, NY, USA. ACM. find
  4. Dmitry Davidov, Oren Tsur, and Ari Rappoport. 2010. Semi-supervised recognition of sarcastic sentences in twitter and amazon. In Proceedings of the Fourteenth Conference on Computational Natural Language Learning, pages 107–116. Association for Computational Linguistics. find our card
  5. Leon Derczynski, Alan Ritter, Sam Clark, and Kalina Bontcheva. 2013. Twitter part-of-speech tagging for all: Overcoming sparse and noisy data. In Proceedings of the International Conference on Recent Advances in Natural Language Processing. find
  6. Andrea Esuli and Fabrizio Sebastiani. 2006. Sentiwordnet: A publicly available lexical resource for opinion mining. In Proceedings of Language Resources and Evaluation Conference, volume 6, pages 417–422. find our card
  7. Elena Filatova. 2012. Irony and sarcasm: Corpus generation and analysis using crowdsourcing. In Proceedings of Language Resources and Evaluation Conference, pages 392–398. find our card
  8. Rachel Giora. 1995. On irony and negation. Discourse processes, 19(2):239–264. ↗
  9. Roberto González-Ibáñez, Smaranda Muresan, and Nina Wacholder. 2011. Identifying sarcasm in twitter: A closer look. In ACL (Short Papers), pages 581–586. Citeseer. find our card
  10. H Paul Grice. 1975. Logic and conversation. 1975, pages 41–58. ↗
  11. Daniel C Howe. 2009. Rita wordnet. java based api to access wordnet. find
  12. Nancy Ide and Keith Suderman. 2004. The American National Corpus First Release. In Proceedings of the Language Resources and Evaluation Conference. find
  13. Christine Liebrecht, Florian Kunneman, and Antal van den Bosch. 2013. The perfect solution for detecting sarcasm in tweets# not. WASSA 2013, page 29. find our card
  14. Joan Lucariello. 1994. Situational irony: A concept of events gone awry. Journal of Experimental Psychology: General, 123(2):129. ↗
  15. George A Miller. 1995. Wordnet: a lexical database for english. Communications of the ACM, 38(11):39–41. ↗ our card
  16. Bo Pang and Lillian Lee. 2008. Opinion Mining and Sentiment Analysis. Found. Trends Inf. Retr., 2(1-2):1–135, January. ↗
  17. Bo Pang, Lillian Lee, and Shivakumar Vaithyanathan. 2002. Thumbs up?: Sentiment classification using machine learning techniques. In Proceedings of the ACL-02 Conference on Empirical Methods in Natural Language Processing - Volume 10, EMNLP ’02, pages 79–86, Stroudsburg, PA, USA. Association for Computational Linguistics. ↗
  18. Christopher Potts. 2011. Developing adjective scales from user-supplied textual metadata. NSF Workshop on Restructuring Adjectives in WordNet. Arlington,VA. find
  19. Quintilien and Harold Edgeworth Butler. 1953. The Institutio Oratoria of Quintilian. With an English Translation by HE Butler. W. Heinemann. find
  20. Antonio Reyes, Paolo Rosso, and Tony Veale. 2013. A multidimensional approach for detecting irony in twitter. Language Resources and Evaluation, pages 1–30. ↗
  21. Brett Spell. 2009. Java api for wordnet searching (jaws). find
  22. Akira Utsumi. 2000. Verbal irony as implicit display of ironic environment: Distinguishing ironic utterances from nonirony. Journal of Pragmatics, 32(12):1777–1806. ↗
  23. Tony Veale and Yanfen Hao. 2010a. Detecting ironic intent in creative comparisons. In ECAI, volume 215, pages 765–770. ↗
  24. Tony Veale and Yanfen Hao. 2010b. An ironic fist in a velvet glove: Creative mis-representation in the construction of ironic similes. Minds and Machines, 20(4):635–650. ↗
  25. Tony Veale. 2004. The challenge of creative information retrieval. In Computational Linguistics and Intelligent Text Processing, pages 457–467. Springer. find
  26. Deirdre Wilson and Dan Sperber. 2002. Relevance theory. Handbook of pragmatics. find
  27. Ian H Witten and Eibe Frank. 2005. Data Mining: Practical machine learning tools and techniques. Morgan Kaufmann. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/barbieri-2014-modelling-irony-twitter/