Context: Irony in language is when a statement is produced with one meaning but the intended meaning is exactly the opposite. For instance, someone who has burned toast might serve it and say ironically “it’s a little underdone”. Automatically detecting when language is ironic is an especially difficult task in Natural Language Processing. Content: This dataset contains 1950 comments, which have been labeled as ironic (1) or not ironic (-1) by human annotators. The text was taken from Reddit comments. Acknowledgements: This dataset and analysis of it is presented in the following paper. Wallace, B. C., Do Kook Choe, L. K., Kertz, L., & Charniak, E. (2014, April). Humans Require Context to Infer Ironic Intent (so Computers Probably do, too). In ACL (2) (pp. 512-516). Url: http://www.byronwallace.com/static/articles/wallace-irony-acl-2014.pdf Made possible by support from the Army Research Office (ARO), grant 64481-MA / W9111F-13-1-0406 "Sociolinguistically Informed Natural Language Processing: Automating Irony Detection" Inspiration: * Is irony more likely when discussing certain topics? * Does ironic text tend to have more positive or more negative sentiment? * What novel features can you develop to help detect irony?