Investigation of methods of emails spam detection

Abstract
     E-mail service has created a convenient communication platform for users to exchange information, and according to the latest statistics from lifewire site in January 2019, each employee typically receives 121 emails and sends 40 emails, who work in the field of business economics believe that the spam economy is not something that will be forgotten anytime soon, and the only thing that can be guessed about the fate of spam is its change and completion at the same time with the change and completion and development of technology because spams know as a chronic security problem for computers and in case of hacking can be very dangerous and costly. With the advent of social media and other online information exchange sites, dependence on email communications has increased over the years, leading to an urgent need to improve spam factors. Although many spam filters have been created to prevent users from entering the inbox, there is a lack of research to improve text algorithms. The Naïve Bayes algorithm is currently one of the most popular methods of classifying spam because of its simplicity and efficiency, which is why it is used in many methods. Given that spammers are looking for new ways to target these networks, there are ongoing efforts to identify such malicious emails every day, but it is nonetheless noteworthy that They have increased by about 5.32% compared to in 2018. In this article, eight methods of spam filtering are compared, and in the end, it is concluded that the use of an artificial intelligence library is more efficient than other methods due to the combination of several algorithms. 
Keywords: Learning Machine, Naïve Bayes ،SVM ،Spam