Enhanced Filter For Detecing And Prevanting Spam Zombies

A B S T R A C T With the rapid development, various new network services in every sector of state, society, and public services are becoming increasingly attractive targets to large scale malicious attacks by cyber criminals. Spam Zombies are emerging as effective threats and being increasingly used by cyber attackers to control millions of computers, steel information and commit various malicious activities. Compromised machines or spam zombies are one of the key security threats on the Internet. Spamming provides a key economic incentive for attackers to recruit the large number of compromised machines. The presence of these compromised machines has posed a great challenge to the internet. These compromised machines involved in the spamming activities are referred to as spam zombies. Network of spam zombies are recognized as one of the most serious security threats today. The discovery of e-mail spam using the spam detection technique from these machines is essential for the prevention of network-related attacks. This thesis focuses on the creation of enhanced filter for detection and prevention of spam zombies. When messages enter into the spot filter, it detects valid messages and invalid or spam messages. Sometimes spam messages are not harmful to our computers. So there is nessicity to differentiate good spam and bad spam. According to SPRT, spot filter will decide the incoming messages are spam messages or valid messages. If the probability of incoming messages is more then it will consider as spam messages. But it will not show that messages are good spam or bad spam. This thesis integrates spot filter approach with Gray filter detection approach to overcome drawbacks of spot filter.