Scholarlink

Global digital communication has been noticed as the highest priority because of the rapid growth in population and their hectic schedules. Email communication is considered primarily professional commitment in several public and private organizations such educational institutes, healthcare, and business. Because of the growing rush of digital communication, only email communication is recommended safely as a most professional recognition among these organizations. Raising spam email volume creates a security risk for national and international organizations. It requires high attention when AI-generated email attempts cyber-attack by virus dissemination and finally destroys the complete system. Targeting the contemporary scope against security threats and intricacy refers to meeting the real time solution of email spam detection. This article demonstrated a comparative as well as isolated study on two publicly available datasets Enron and Spambase. The model also evaluated a hybrid model by the combination of Artificial Neural Network (ANN), Convolutional Neural Network (CNN), and Bidirectional Encoder Transformer (BERT) showing their respective performance 93.7%, 95.2%, and 97.15 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 Hybrid Emerging Framework for Spam Email Classification with ANN CNN and BERT Models

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Dawoud Sulaiman Al-Aamri,
  • Salim Saud Al-Sarhni,
  • Hamdan Abdullah AL balushi,
  • Naresh Kumar

摘要

Global digital communication has been noticed as the highest priority because of the rapid growth in population and their hectic schedules. Email communication is considered primarily professional commitment in several public and private organizations such educational institutes, healthcare, and business. Because of the growing rush of digital communication, only email communication is recommended safely as a most professional recognition among these organizations. Raising spam email volume creates a security risk for national and international organizations. It requires high attention when AI-generated email attempts cyber-attack by virus dissemination and finally destroys the complete system. Targeting the contemporary scope against security threats and intricacy refers to meeting the real time solution of email spam detection. This article demonstrated a comparative as well as isolated study on two publicly available datasets Enron and Spambase. The model also evaluated a hybrid model by the combination of Artificial Neural Network (ANN), Convolutional Neural Network (CNN), and Bidirectional Encoder Transformer (BERT) showing their respective performance 93.7%, 95.2%, and 97.15 respectively.