Email Spam: A New Strategy of Screening Spam Emails using Natural Language Processing

Due to its cost-effectiveness, email has become the primary method of commercial and general communication. These emails can be accessed by everyone quickly with an internet connection. For these reasons, despite the existence of other communication tools, the usage of email keeps growing. The rapid increase in technology has caused many applications to be automated. Considering this example: Email spamming increased as a result of the automation of email classification. This has become a significant problem for email communication. Before artificial intelligence and machine learning techniques, various solutions included manual detection by identifying the specific keywords in the email but this method is not possible to solve the current problem. Since the number of emails generated nowadays is vast, then the picture of artificial intelligence and machine learning algorithms comes into focus. Emails can be classified into various categories very easily based on the requirements of a particular user or a company. This email spam classification model is implemented by using Python & classification algorithms like Logistic Regression, Naive Bayes, and Support Vector Machine.

I. Introduction

In this modern age, where most of the work is happening on the internet, email has become the most convenient and efficient mode of communication. The email system is ingrained in both the business and academic worlds, to the point where it handles all significant conversations and day-to-day tasks. Due to its popularity, cost-effectiveness, and efficiency, its usage has increased rapidly.[1] Most of the official and some of the unofficial communication is done through email. According to a recent survey, noting that access to the internet is expanding globally, the estimated daily average count for emails sent and received by 2025 is 376.4 billion. Of these numbers, about 85% of the emails sent or received are spam.[7] One of the biggest issues with the Internet today is email spam, which costs the company time and money as well as irritates individual users. This demonstrates how these spams waste the time and resources of email users while producing no beneficial results.[2] It consistently makes it difficult for the network administrator and security professional to efficiently avoid or reduce it. Basically, the email contains two types of messages one is ham and the other one is spam. "Ham" refers to the use and wanted by the recipient in digital communication. "Ham" means messages which do not spam, and "Spam" refers to any unwanted, uninvited digital communication transmitted in bulk.[11]