Scholarlink

In the era of digital evolution and global connectivity, social networking sites have emerged as significant platforms for virtual interaction. As user engagement on these platforms increases, so does the prevalence of cyber-crimes, including cyber-bullying. Cyber-bullying, a pervasive threat that can originate from both strangers and acquaintances, presents substantial challenges in detection and identification. This paper reviews key advancements in countering cyber-bullying and employs a dataset of 47,000 Twitter tweets to investigate the prevalence of such incidents. Our approach involves expanding this textual dataset and applying multi-class classification techniques to categorize tweets based on their relevance to cyber-bullying. We utilize machine learning classifiers to develop and train the classification model. A comparative analysis of classifier performance demonstrates that the XGBoost model outperforms other classifiers in terms of efficiency and accuracy.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Classification of Tweets to Identify Cyber-Bullying Attack Based on Multi-class Classification Using Machine Learning

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Sarah Fatima,
  • Aqeel Khalique,
  • Farheen Siddiqui,
  • Mohd Abdul Ahad,
  • Mohammad Faisal Siddiqi,
  • Nameer Khan

摘要

In the era of digital evolution and global connectivity, social networking sites have emerged as significant platforms for virtual interaction. As user engagement on these platforms increases, so does the prevalence of cyber-crimes, including cyber-bullying. Cyber-bullying, a pervasive threat that can originate from both strangers and acquaintances, presents substantial challenges in detection and identification. This paper reviews key advancements in countering cyber-bullying and employs a dataset of 47,000 Twitter tweets to investigate the prevalence of such incidents. Our approach involves expanding this textual dataset and applying multi-class classification techniques to categorize tweets based on their relevance to cyber-bullying. We utilize machine learning classifiers to develop and train the classification model. A comparative analysis of classifier performance demonstrates that the XGBoost model outperforms other classifiers in terms of efficiency and accuracy.