Scholarlink

Email-based cyberbullying is the practice of harassing, threatening, or hurting others which covers things like distributing misleading information, starting online arguments, and sending menacing comments. Cyberbullying in email spam detection entails using a variety of methods to recognize and filter out offensive or threatening messages. Detecting cyberbullying using machine learning algorithms involves training models to recognize patterns and characteristics associated with harmful online behavior. This research paper offers a thorough analysis of the use of several machine learning algorithms to categorize emails including cyberbullying. The efficacy of decision trees, Naive Bayes, random forest, and XGBoost algorithms in detecting and classifying emails containing dangerous content is investigated by the author. Random forest classifier gives the best accuracy of 96% in cyberbullying datasets. This makes it the best choice in identifying and combating cyberbullying effectively.

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

Implementation of Machine Learning Algorithms for Precise Detection and Classification of Email-Based Cyberbullying

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Janhvi Juyal,
  • Vriddhi Mittal,
  • Prisha Virmani,
  • Samarth Bhutani,
  • Sonakshi Vij

摘要

Email-based cyberbullying is the practice of harassing, threatening, or hurting others which covers things like distributing misleading information, starting online arguments, and sending menacing comments. Cyberbullying in email spam detection entails using a variety of methods to recognize and filter out offensive or threatening messages. Detecting cyberbullying using machine learning algorithms involves training models to recognize patterns and characteristics associated with harmful online behavior. This research paper offers a thorough analysis of the use of several machine learning algorithms to categorize emails including cyberbullying. The efficacy of decision trees, Naive Bayes, random forest, and XGBoost algorithms in detecting and classifying emails containing dangerous content is investigated by the author. Random forest classifier gives the best accuracy of 96% in cyberbullying datasets. This makes it the best choice in identifying and combating cyberbullying effectively.