A Comparative Analysis and Discussion of Email Spam Classification Methods Using Machine Learning Techniques

Abstract

Email communication is by far the most easy, ubiquitous, and convenient medium ever used by humankind for communicating with people. Free email service providers such as Gmail, Yahoo! Mail, Hotmail, and Outlook are used by millions of people worldwide. Emails as a form of communication became popular when Internet service providers such as AOL, Prodigy, and CompuServe provided each user with an email account by default in the 1990s. Machine learning is currently one of the most-discussed fields in computer science. With the rise in the adoption of data analysis and machine learning fueled by Google open sourcing, most email classifiers started using some variation of this method. Machine learning is one of the most-discussed fields in computer science. Multiple machine learning models, supervised or unsupervised, can be trained and tested on the proposed model. The proposed system architecture focuses on retrieving emails through available third-party APIs and processes them appropriately for machine learning.