Download PDFOpen PDF in browserProfiling Irony and Stereotype Spreaders on Twitter Using Multi-View LearningEasyChair Preprint 83065 pages•Date: June 19, 2022AbstractWith the rise of social media, millions of people are using it every day. They may publish content about everything. In addition to maintaining freedom of speech, social media executives must restrict the spread of harassing speech. For this purpose, the Sheykhlan team developed a system with multi-view learning in combination with an SVM, to identify malicious users. The proposed approach achieved 94.63% accuracy on 5-fold cross-validation on the English dataset. Keyphrases: Irony and Stereotype Spreaders, RoBERTa, SVM, TF-IDF, Twitter, deep learning, machine learning
|