Hierarchical Model for Email Fraud Detection Using Naïve Bayes and SVM

Email fraud detection is one of the major problems in our society with the quick rise of internet frauds and spam nowadays. Some people were using it for illegal and evil purposes like phishing and fraud. The spammers are pretending like a genuine person in the spam emails for creating a fake profile and fake account. In this paper, we propose a method for email fraud detection that combines the strengths of the Naive Bayes algorithm and Support Vector Machine (SVM) in machine learning. The Naive Bayes algorithm is one of the most popular and effective in handling large datasets, and this can be used to model their relationships between words in emails and their properties of being spam or ham. Meanwhile, SVM is able to separate classes in high-dimensional spaces, and it is also used for the accuracy classification. By integrating these two approaches, the proposed method aims to improve the detection of spam emails by using the advantages of both algorithms. This mixed-breed model offers a strong solution to the challenge of distinguishing between spam and not spam emails. Most of the existing methods have less accuracy in detecting spam. The proposed Naïve Bayes and SVM algorithms achieve higher accuracy of 97.49% for email fraud detection and fast processing speed compared to other proposed methods.