Implementation of Secure Internet Email System Through Outgoing Email Filtering Based on Semantic Approach

Modem society is highly dependent on the smooth and safe flow of information over communication and computer networks. One of the methods to exchange information is email. Email, which is now a dominating medium of communication, has become the target of abuse. Mass posting of unsolicited email messages called spam is increasingly prevalent. Spam is a huge problem for all email users who lose time deleting all the junk emails before reading the legitimate emails. But the more dangerous problem is that email can be used to treat information from user because it can contain a virus or worm. The virus or warm can be attached in the junk email, and the virus tries to collect confidential information from an email user. Therefore, many email service companies spend much money trying to protect treat of information by virus. Many scientists and companies are trying to avoid the unwanted email to protect virus. One of their efforts is an email filtering method. Emails are filtered before reaching users; therefore, the users can be safe about information leakage by virus. However, even though we can check and filter all incoming email, we cannot protect outgoing email if someone tries to leak confidential information. Because email filtering is focused only on incoming email, this thesis suggests an email filtering for outgoing email based on the combination of currently existing email incoming filtering and a new semantic approach. This outgoing email-filtering method detects a virus before sending email. The outgoing server checks for spam, email virus, or worm when email is sent from sender, filters the email having some confidential information, passes the email to the incoming server, and finally, the receiver gets the email from incoming server