Published on February 20, 2014
Cursing in English on Twitter Wenbo Wang Lu Chen T.K. Prasad Amit Sheth firstname.lastname@example.org email@example.com firstname.lastname@example.org email@example.com Ohio Center of Excellence in Knowledge-enabled Computing (Kno.e.sis) Wright State University, Dayton, OH, USA Kno.e.sis Center 1 CSCW 2014
Kno.e.sis Center Photo credit: Wikimedia Cursing, also called swearing, profanity, or bad language, is the use of certain words and phrases that are considered by some2 to be CSCW 2014
Cursing: academic point of view • Not uncommon during conversations in physical world – Rates: curse words (0.5%) vs. first person plural pronouns (1.0%, we, us, our) [Mehl and Pennebaker 2003] • “A rich emotional, psychological and sociocultural phenomenon” [Jay 2009] Mehl, M. R., & Pennebaker, J. W. (2003). The sounds of social life: a psychometric analysis of students' daily social environments and natural conversations. Journal of personality and social psychology, 84(4), 857. Kno.e.sis Center Jay, T. (2009). The utility and ubiquity of taboo words. Perspectives on Psychological Science, 4(2), 153-161. 3 CSCW 2014
Physical World Twitter Private to participants Face to face Kno.e.sis Center Public to the world Don’t see each other Photo credit: 紫流, Wikimedia1, Wikimedia2, Tony Alter 4 CSCW 2014
How do people use curse words on Twitter? Kno.e.sis Center 5 CSCW 2014
Research questions • Ubiquity: How often do people use curse words on Twitter? • Utility: Why do people use curse words on Twitter? • Contextual Variables: when, where or communication type? • People: who says curse words to whom on Twitter? Kno.e.sis Center 6 CSCW 2014
Data Collection • Twitter sample API, between March 11th 2013 and April 7th 2013 • Retain only English tweets • 51 Million tweets sent by 14 Million users Kno.e.sis Center Photo credit: Geograph UK 7 CSCW 2014
Cursing Lexicon Coding • Collected four lists of curse words from Internet • Annotated by two native English speakers – 1: mostly used for cursing – 2: can be used for both cursing and non-cursing purposes – 3: usually is not used for cursing – 4: don’t know its meaning – Cohen’s Kappa: 0.5582 • Kept only 788 words that got label 1 from both annotators – ass, a55, @$$, …. Kno.e.sis Center 8 CSCW 2014
Cursing: Ubiquity Kno.e.sis Center Phto credit: Wikimedia 9 CSCW 2014
How often do people use curse words on Twitter? People curse more on Twitter! [Mehl and Pennebaker 2003] [Subrahmanyam and Smahel 2006] Our work 52 undergraduates 1,150 chatroom users 14 million Twitter users 4 days’ tape recording 12,258 utterances 51 million tweets 0.5% of all words 3% of all utterances 0.80% of all words 7.73% of all tweets Kno.e.sis Center Mehl, M. R., & Pennebaker, J. W. (2003). The sounds of social life: a psychometric analysis of students' daily social environments and natural conversations. Journal of personality and social psychology, 84(4), 857. Subrahmanyam, K., Smahel, D., & Greenfield, P. (2006). Connecting developmental constructions to the internet: identity presentation 10 and sexual exploration in online teen chat rooms. Developmental psychology, 42(3), 395. CSCW 2014
What are the most frequently used curse words? 2M 90.40% count 1.5M 1M 0.5M 0 retard blowjob cock cum cunt fag damn tit puta slut pussy piss dick whore hell nigga bitch ass shit fuck curse words Top seven curse words cover 90.40% of all the curse word occurrences. Kno.e.sis Center 11 CSCW 2014
Cursing: Utility Kno.e.sis Center Photo credit: Dan Thompson 12 CSCW 2014
Emotion identification from tweets • Collect self-labeled emotion tweets (Wang et. al. 2012) – “And all I need is one fuckin sheet stamped! #rage” • Evaluation on detecting five emotions from 250K tweets – Precision: 76.17%; recall: 46.07% Kno.e.sis Center Wang, W., Chen, L., Thirunarayan, K., & Sheth, A. P. (2012, September). Harnessing twitter" big data" for automatic emotion identification. In Privacy, Security, Risk and Trust (PASSAT), 2012 International Conference on and 2012 International Confernece on 13 Social Computing (SocialCom) (pp. 587-592). IEEE. CSCW 2014
Emotion distributions in both cursing and non-cursing tweets Curse words are used to vent out negative emotions! Kno.e.sis Center 14 CSCW 2014
Cursing: Contextual variables Kno.e.sis Center Photo credit: Wikipedia 15 CSCW 2014
In which place, are cursing words most frequently used? A. College academic places B. High schools C. Residence Kno.e.sis Center 16 CSCW 2014
Where? Users are usually in a good mood while out in the nature? TWO exceptions! Kno.e.sis Center The pattern of more swearing in more relaxed environment still holds 17 CSCW 2014
During a day, when are cursing words most frequently used? A. B. C. D. Kno.e.sis Center Morning: 6am-11am Afternoon: 3pm-6pm Evening: 6pm-11pm Late night: 12pm-1:30am 18 CSCW 2014
When: time of day People curse more and more as they go through the day! humans’ diurnal activity schedule 3pm-1:30am 6am-11am Kno.e.sis Center 19 CSCW 2014
During a week, when are cursing words most frequently used? A. B. C. D. Kno.e.sis Center Monday Wednesday Friday Sunday 20 CSCW 2014
When: Day of week Users start with relatively high cursing ratios on Mondays, Tuesdays and Wednesdays Keep decreasing on the following three days Kno.e.sis Center 21 CSCW 2014
Message type Users perform self-censorship to some extent when they talk to others directly The number of times a tweet is retweeted has a negligible correlation with whether a tweet has curse words: r=0.00154, p<2.2e16 They pay less attention to the use of curse words in mentions and updates Kno.e.sis Center 22 CSCW 2014
Cursing: People Kno.e.sis Center Photo credit: Wikimedia 23 CSCW 2014
Gender: cursing ratio Female-to-Male has the lowest cursing ratio! Male-to-Male communication has the highest cursing ratio! People are more likely to use curse words within the samegender context. Kno.e.sis Center Photo credit: Wikipedia 24 CSCW 2014
Gender: word choice Some words are used significantly more often within specific groups. Kno.e.sis Center Photo credit: Wikipedia 25 CSCW 2014
Limitations • Users in the data may NOT mirror overall population in the physical world • Rely on self-reported data • Mostly descriptive, providing insights on the “what” aspects Kno.e.sis Center 26 CSCW 2014
Thanks! • Original CSCW paper: link • More info at knoesis.org • Questions? firstname.lastname@example.org • Special thanks to: – NSF – Anonymous reviewers Kno.e.sis Center 27 CSCW 2014
... Krishnaprasad Thirunarayan and Amit P. Sheth. Cursing in English on Twitter. ... (CSCW 2014), 2014. It primary corrects two programming errors.
Cursing in English on Twitter. Added by. Amit Sheth. www.slideshare.net/knoesis/cursing-in-english-on-twitter-at-cscw-2014
... 2014. Cursing in English on Twitter project page can be ... knoesis/cursing-in-english-on-twitter-at-cscw-2014. Abstract. Cursing is not uncommon ...
... Thirunarayan, K., & Sheth, A. P. (2014). Cursing in English on Twitter. ... with circles stands for the ratio of cursing tweets to overall CSCW 2014 ...
Publication » Cursing in English on Twitter. ... (CSCW 2014), At Baltimore ... We first identified English cursing tweets in the collection,
Cursing in English on twitter. Full Text: PDF: ... 2012-2014: Publication count: 2: ... As CSCW creates and investigates technologies for social ...
The latest Tweets from ACM CSCW ... English UK; Español; Filipino; Français; Italiano; ... Twitter stores that location.