The perfect solution for detecting sarcasm in tweets #not

To avoid a sarcastic message being understood in its unintended literal meaning, in microtexts such as messages on Twitter.com sarcasm is often explicitly marked with the hashtag ‘#sarcasm’. We collected a training corpus of about 78 thousand Dutch tweets with this hashtag. Assuming that the human labeling is correct (annotation of a sample indicates that about 85% of these tweets are indeed sarcastic), we train a machine learning classifier on the harvested examples, and apply it to a test set of a day’s stream of 3.3 million Dutch tweets. Of the 135 explicitly marked tweets on this day, we detect 101 (75%) when we remove the hashtag. We annotate the top of the ranked list of tweets most likely to be sarcastic that do not have the explicit hashtag. 30% of the top-250 ranked tweets are indeed sarcastic. Analysis shows that sarcasm is often signalled by hyperbole, using intensifiers and exclamations; in contrast, non-hyperbolic sarcastic messages often receive an explicit marker. We hypothesize that explicit markers such as hashtags are the digital extralinguistic equivalent of nonverbal expressions that people employ in live interaction when conveying sarcasm.