RT Article
T1 Automatic detection of cyberbullying: racism and sexism on Twitter
JF Cybersecurity in the age of smart societies
SP 105
OP 122
A1 Wang, Linfeng
A2 Islam, Tasmina
LA English
YR 2023
UL https://krimdok.uni-tuebingen.de/Record/1918680892
AB With the increasing number of people more people utilising social media platforms, the production of aggressive language online such as attacks, abuse, and denigration increase. However, the constantly changing and different forms of online language provide difficulties in detecting violent language. Not only is this a difficult undertaking, but it is also an area for research and growth, considering the harm caused by cyber violence to children, women, and victims of racial prejudice, as well as the severity of cyberbullying's consequences. This paper identifies some violent terms and proposes a model for detecting racism and sexism on social media (twitter) based on TextCNN and Word2Vec sentiment analysis achieving 96.9% and 98.4% accuracy.
NO Literaturverzeichnis: Seite 122
SN 9783031201592
K1 Cyberbullying detection
K1 Online social media
K1 Racism
K1 Sexism
K1 Convolutional neural network