Recognizing Humor on Twitter

Humor Recognition and Humor Anchor Extraction

“…For example, Mihalcea and Strapparava (2005) defined three types of humorspecific stylistic features: Alliteration, Antonym and Adult Slang, and trained a classifier based on these feature representations. Similarly, Zhang and Liu (2014) designed several categories of humor-related features, derived from influential humor theories, linguistic norms, and affective dimensions, and input around fifty features into the Gradient Boosting Regression Tree model for humor recognition. Taylor and Mazlack (2004) recognized wordplay jokes based on statistical language recognition techniques, where they learned statistical patterns of text in N-grams and provided a heuristic focus for a location of where wordplay may or may not occur.…”

“…That is, humor is essentially associated with sentiment (Zhang and Liu, 2014) and subjectivity (Wiebe and Mihalcea, 2006). For example, a sentence is likely to be humorous if it contains some words carrying strong sentiment, such as 'idiot' as follows.…”

HumorHawk at SemEval-2017 Task 6: Mixing Meaning and Sound for Humor Recognition

“…They noted that negative sentiment, humancenteredness and lexical centrality were their most important model features. Zhang and Liu (2014) trained a classifier using tweets that use the hashtag #Humor for positive examples. They concluded that tweet part-of-speech ratios are a major factor in humor detection.…”

“…In line with previous work (Radev et al, 2015;Zhang and Liu, 2014), we used the following features as input to the model:…”

QUB at SemEval-2017 Task 6: Cascaded Imbalanced Classification for Humor Analysis in Twitter

“…This can be extremely challenging (Attardo, 1994) because no universal definition of humor has been achieved, humor is highly contextual, and there are many different types of humor with different characteristics (Raz, 2012). Previous studies (Mihalcea and 1 http://alt.qcri.org/semeval2017/task6/ Strapparava, 2005;Yang et al, 2015;Zhang and Liu, 2014;Purandare and Litman, 2006;Bertero and Fung, 2016) dealt with the humor recognition task as a binary classification task, which was to categorize a given text as humorous or non-humorous (Li et al, 2016). Textual data consisting of comparable amounts of humorous texts and nonhumorous texts were collected, and a classification model was then built using textual features.…”