With the increasing number of interactions, social media users have been vulnerable to intentional aggressive acts and cyberbullying instances. In this paper, first, we carry out a message-level cyberbullying annotation on an Instagram dataset. Second, we use the correlations on the Instagram dataset annotated with emotion, sentiment and bullying labels. Third, we build a message-level emotion classifier automatically predicting emotion labels for each comment in the Vine bullying dataset. Fourth, we build a session-based bullying classifier with the use of n-grams, emotion, sentiment and concept-level features. For both emotion and bullying classifiers, we use Linear Support Vector Classification. Our results show that “anger” and “negative” labels have a positive correlation with the presence of bullying. Concept-level features, emotion and sentiment features in different levels contribute to the bullying classifier, especially to the bullying class. Our best performing bullying classifier with n-grams and concept-level features (e.g., polarity, averaged polarity intensity, moodtags and semantics features) reaches to an F1-score of 0.65 for bullying class and a macro average F1-score of 0.7520.
Pinar Arslan, M.C. (2019). Overwhelmed by negative emotions?: maybe you are being cyber-bullied! [10.1145/3297280.3297573].
Overwhelmed by negative emotions?: maybe you are being cyber-bullied!
Michele Corazza;
2019
Abstract
With the increasing number of interactions, social media users have been vulnerable to intentional aggressive acts and cyberbullying instances. In this paper, first, we carry out a message-level cyberbullying annotation on an Instagram dataset. Second, we use the correlations on the Instagram dataset annotated with emotion, sentiment and bullying labels. Third, we build a message-level emotion classifier automatically predicting emotion labels for each comment in the Vine bullying dataset. Fourth, we build a session-based bullying classifier with the use of n-grams, emotion, sentiment and concept-level features. For both emotion and bullying classifiers, we use Linear Support Vector Classification. Our results show that “anger” and “negative” labels have a positive correlation with the presence of bullying. Concept-level features, emotion and sentiment features in different levels contribute to the bullying classifier, especially to the bullying class. Our best performing bullying classifier with n-grams and concept-level features (e.g., polarity, averaged polarity intensity, moodtags and semantics features) reaches to an F1-score of 0.65 for bullying class and a macro average F1-score of 0.7520.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.