Scholarlink

As per the study, at every single second, new 8.9 users is getting connected to the internet. Users may be the attacker or non-attacker type. Nowadays, attackers are also using emotional factors to create social media messages to steal user data which makes it important to identify authentic social media posts from phishing messages. The paper proposes an approach with self-regularization for processing cyber-attacks that includes textual threats, and vulnerable data. The self-attention approach finds contextual relationships among words in a text input by assessing each word’s specific arrangement in the sentence. Our approach is performed over a continuous interacting process that goes through machine learning. The research work is validated by its results where we have achieved an accuracy of 99.99, 99.73 and 99.23% for Random Forest, Decision Tree, and Naive Bayes Respectively.

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

A Novel Approach for Self-Regularization for Malware Prediction in Online Social Media

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Varsha Mittal,
  • Anupama Mishra,
  • Kwok Tai Chui

摘要

As per the study, at every single second, new 8.9 users is getting connected to the internet. Users may be the attacker or non-attacker type. Nowadays, attackers are also using emotional factors to create social media messages to steal user data which makes it important to identify authentic social media posts from phishing messages. The paper proposes an approach with self-regularization for processing cyber-attacks that includes textual threats, and vulnerable data. The self-attention approach finds contextual relationships among words in a text input by assessing each word’s specific arrangement in the sentence. Our approach is performed over a continuous interacting process that goes through machine learning. The research work is validated by its results where we have achieved an accuracy of 99.99, 99.73 and 99.23% for Random Forest, Decision Tree, and Naive Bayes Respectively.