A Review of Bayes Machine Learning for Spam Filtering Applications
Zehao Song · Applied and Computational Engineering · 2025
The Naive Bayes algorithm uses the theorem of Bayes to filter spam emails, achieving good filtering results. The improved Bayes algorithm addresses the assumption of "feature independence given the class" in Naive Bayes algorithm, allowing for a broader application range. This paper reviews the main content and representative achievements of both the Naive Bayes algorithm and the improved Bayes algorithm, and analyzes the advantages and disadvantages of each method. This study finds that the Naive Bayes algorithm has a limited application range due to the assumption of "feature independence given the class" while the improved Bayes algorithm effectively solves this problem and it has better applicability. This paper aims to help researchers engaged in spam filtering better understand and leverage the potential of the theorem of Bayes in spam filtering, providing a summary reference to promote technological innovation in related fields and better problem-solving, as well as facilitating the understanding of other readers and the application of Bayes filtering methods.