A Review On Machine Learning Algorithms In Enhancing Email Security

Abstract

Electronic mail is one of the rapid and technological ways to send important messages from one location to another place worldwide. The development of the usage of the online mail has evolved to achieve the piece of information in the pillar box, where the beneficiary gets a huge number of information, most of them create significant and diversified challenges including the challenges related to the identity of the receivers, the abruption of crucial information and destruction of Internet. These communicational bridges are so harmful that the users cannot ignore that during their usage of various types of forms including the advertisements and other electronic messages. The Trash among the messages is the process of scrutinizing those harmful messages that are harmful for noticing them and creating the communication among viewers. This particular study explains the process in improving the fraudulent activities in investigating and by providing the proposed single objective to develop the challenges in creating algorithm that uses the “Machine Learning classification algorithms” in creating the regressive in making the classifications to build the “optimal model” for accurately   categorizing the electronic messages. Cleaning the messages and selecting them for creating new features are the initial activities in modeling the processes to diminish the dimensions of “sparse text features” obtained from trash and “ham communications.” The selection of various features is used to create the best dimensions and the third stage is to discover the brouches using a “Logistic Regression classifier” with accuracy level of that.