TY - GEN
T1 - From royals to vegans
T2 - 41st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2018
AU - Guy, Ido
AU - Shapira, Bracha
N1 - Publisher Copyright: © 2018 ACM.
PY - 2018/6/27
Y1 - 2018/6/27
N2 - The phenomenon of trolling has emerged as a widespread form of abuse on news sites, online social networks, and other types of social media. In this paper, we study a particular type of trolling, performed by asking a provocative question on a community question-answering website. By combining user reports with subsequent moderator deletions, we identify a set of over 400,000 troll questions on Yahoo Answers, i.e., questions aimed to inflame, upset, and draw attention from others on the community. This set of troll questions spans a lengthy period of time and a diverse set of topical categories. Our analysis reveals unique characteristics of troll questions when compared to "regular" questions, with regards to their metadata, text, and askers. A classifier built upon these features reaches an accuracy of 85% over a balanced dataset. The answers' text and metadata, reflecting the community's response to the question, are found particularly productive for the classification task.
AB - The phenomenon of trolling has emerged as a widespread form of abuse on news sites, online social networks, and other types of social media. In this paper, we study a particular type of trolling, performed by asking a provocative question on a community question-answering website. By combining user reports with subsequent moderator deletions, we identify a set of over 400,000 troll questions on Yahoo Answers, i.e., questions aimed to inflame, upset, and draw attention from others on the community. This set of troll questions spans a lengthy period of time and a diverse set of topical categories. Our analysis reveals unique characteristics of troll questions when compared to "regular" questions, with regards to their metadata, text, and askers. A classifier built upon these features reaches an accuracy of 85% over a balanced dataset. The answers' text and metadata, reflecting the community's response to the question, are found particularly productive for the classification task.
KW - Abusive behavior
KW - Antisocial behavior
KW - Community question answering
KW - Content abuse
KW - Question trolling
KW - Trolling
UR - http://www.scopus.com/inward/record.url?scp=85051462863&partnerID=8YFLogxK
U2 - 10.1145/3209978.3210058
DO - 10.1145/3209978.3210058
M3 - Conference contribution
T3 - 41st International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2018
SP - 835
EP - 844
BT - 41st International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2018
Y2 - 8 July 2018 through 12 July 2018
ER -
