Transformer network-based word embeddings approach for autonomous cyberbullying detection

Abstract

PurposeNowadays people are connected by social media like Facebook, Instagram, Twitter, YouTube and much more. Bullies take advantage of these social networks to share their comments. Cyberbullying is one typical kind of harassment by making aggressive comments, abuses to hurt the netizens. Social media is one of the areas where bullying happens extensively. Hence, it is necessary to develop an efficient and autonomous cyberbullying detection technique.Design/methodology/approachIn this paper, the authors proposed a transformer network-based word embeddings approach for cyberbullying detection. RoBERTa is used to generate word embeddings and Light Gradient Boosting Machine is used as a classifier.FindingsThe proposed approach outperforms machine learning algorithms such as logistic regression, support vector machine and deep learning models such as word-level convolutional neural networks (word CNN) and character convolutional neural networks with short cuts (char CNNS) in terms of precision, recall, F1-score.Originality/valueOne of the limitations of traditional word embeddings methods is context-independent. In this work, only text data are utilized to identify cyberbullying. This work can be extended to predict cyberbullying activities in multimedia environment like image, audio and video.

Multi‐aspects AI‐based modeling and adversarial learning for cybersecurity intelligence and robustness: A comprehensive overview

“…Cyber-attacks such as malware detection, 77 intrusion detection, 76,92 botnet attack 93 can also be detected using hybrid network models consisting of multiple deep basic learning models with better accuracy. In addition, a transformer network-based model can be used to solve security issues, such as autonomous cyberbullying detection, 94 intrusion or anomaly detection 95,96 and so forth. Overall, DL models and their variations mentioned above could play a vital part in the development of effective AI models to address cybersecurity issues, depending on their learning capabilities at various levels, the nature of the data, and the desired outcome, particularly for large datasets.…”

“…Cyber‐attacks such as malware detection, 77 intrusion detection, 76,92 botnet attack 93 can also be detected using hybrid network models consisting of multiple deep basic learning models with better accuracy. In addition, a transformer network‐based model can be used to solve security issues, such as autonomous cyberbullying detection, 94 intrusion or anomaly detection 95,96 and so forth.…”

A Review on Deep-Learning-Based Cyberbullying Detection

“…Transformers are deep learning models based on attention that have shown effectiveness in detecting cyberbullying [113][114][115]. Since its primary application is machine translation, the transformer model has been used for various natural language processing tasks, including the detection of cyberbullying.…”

Machine Learning for Intelligent Data Analysis and Automation in Cybersecurity: Current and Future Prospects

“…Hybrid network models, such as the ensemble of learning models, e.g., CNN and RNN, or others with their optimization can also be used to detect cyber-attacks, such as malware detection [120], phishing, and Botnet attack detection and mitigation [110]. In addition, authors in [111] describe a transformer network-based word embeddings approach for autonomous cyberbullying detection. A robust transformer-based intrusion detection system has been presented in [131].…”