Abstract: One of the greatest methods of communication convenient involves using e-mail for personal messages or commercial objective. Being one of the strongest and quick ways of communication, emails publicity has led to increased undesirable spam email. Email spam is one of the main problems of the Internet today and bringing financial damage to companies and individual users. Spam mails can be harmful as they may contain malware and links to phishing Web sites. So, necessary to separate spam from mail messages into a separate folder. Filter classification can be classified in two techniques-learning method based on machine learning techniques and non-machine techniques. Most popular machine learning techniques due to the high accuracy and athletic support. Machine learning techniques include Naive Bayes and support vector machine learning and decision tree, etc. while non-machine learning techniques, black and white list, signatures and verify email address and mail header checking, etc. In this study utilize one of mechanism learning techniques is Naive Bayes algorithm and for extract features from dataset used Term Frequency Invers Term Frequency (TFIDF) method. For reduce dimensionality of feature space use Information Gain (IG) method.
Shahad Suhail Najam and Karim Hashim AL-Saedi, 2019. Spam Classification by using Naive Bayes Algorithm Based on Segmentation. Research Journal of Applied Sciences, 14: 437-447.