A Crowd-Annotated Spanish Corpus for Humor Analysis

Abstract

Computational Humor involves several tasks, such as humor recognition, humor generation, and humor scoring, for which it is useful to have human-curated data. In this work we present a corpus of 27,000 tweets written in Spanish and crowd-annotated by their humor value and funniness score, with about four annotations per tweet, tagged by 1,300 people over the Internet. It is equally divided between tweets coming from humorous and non-humorous accounts. The interannotator agreement Krippendorff's alpha value is 0.5710. The dataset is available for general use and can serve as a basis for humor detection and as a first step to tackle subjectivity.

CMU-MOSEAS: A Multimodal Language Dataset for Spanish, Portuguese, German and French

“…Polarity classification tasks based on Twitter data have also been collected in Portuguese (Brum and das Graças Volpe Nunes, 2017) (TweetSentBR), German (Cieliebak et al, 2017;Flender and Gips, 2017) (SB10k), and French (Rhouati et al, 2018). Another line of related work aims to predict humor from text in multiple languages (Castro et al, 2016(Castro et al, , 2017. Table 1 demonstrates that CMU-MOSEAS is a unique resource for the languages of Spanish, Portuguese, German and French.…”

Large Dataset and Language Model Fun-Tuning for Humor Recognition

“…The latter one contains about 4,000 contexts, 71% of which are puns, annotated with WordNet senses. Most of humor recognition research deals with English; exceptions are studies working with Italian (Reyes et al, 2009), Russian (Ermilov et al, 2018), and Spanish (Castro et al, 2018).…”

Telling the Whole Story: A Manually Annotated Chinese Dataset for the Analysis of Humor in Jokes

“…• Level: We weighed the fine-grained annotation of humor (Bressler and Balshine, 2006;Castro et al, 2018;Deckers and Devine, 1981) and presented a 5 point-scale for humor rating, 1 signifying the non-humor, 2-5 signifying the gradually increasing degree of humor. (Gan, 2015;Stein, 1998;Tsakona, 2009) • DataSource: Table 1 presents the source of jokes.…”

“…Our annotation not only contains the degree of funniness, but also key words that trigger humor, as well as character relationship, scene, and humor categories. Specially, we have improved on the study of Castro et al (2018) by providing evidence of what causes humor and explaining how humor arises in text. Furthermore, our data have come from a range of sources in numerous domains rather than only from Twitter.…”

“…The humor level annotation is based on a 5-point scale, 1 signifying the lowest level and 5 signifying the highest level. Castro et al (2018) revised the Spanish Twitter corpus with crowd notes and presented a 27,000 tweets dataset in total. Specially, the authors used 5 different emojis to represent the 5 degrees of humor instead of using the 5-point annotation.…”