Spam Email and Malware Elimination employing various Classification Techniques
M. S. Swetha, Gaurav Sarraf · 2019
Machine Learning (ML) is a subsection of Artificial Intelligence (AI) which is a scientific approach of using statistical models, that computer systems can apply and carry out specific task without human intervention. Since its inception email has been one of the most important and prominent means of communication, as with all other modes of communication email has also fallen victim to spam (Spiced-Ham). Studies show about 59.56% of all emails sent in 2018 where spam. Many of these spam emails also have attachments, which might have hidden malicious code which may execute on victims' computer when opened locally. The malicious code can help attackers gain sensitive information of the victim, leading to theft of money or worse identity theft. To solve the spam and malicious file problems we propose to use a subsection of ML known as supervised learning classification which performs binary signature analysis. In this paper we compare over ten different classification techniques like k-Nearest Neighbors (kNN), Support Vector Machine (SVM), Naïve Bayes and Decision Tree. These algorithms are trained on previously labeled data and accuracy of the classifier is computed based on unseen data.