Detecting malware by data mining

The exponentially growth of malware has created number of security threats in IT industry. A large number of viruses are developed and millions of applications are infected and suffered on daily basis. Trojan is one of the fatal and deadly types of malware. But it is often said as legitimate software. They hide themselves within harmless programs. Trojan survived by going unnoticed. They look like just about anything like the computer game as downloaded from different websites. Sometimes even a popup advertisement might try to install something on our computer. Trojan can trick you into using them. In this paper, data mining technique is being proposed to detect Trojan. The technique is based on Naive Bayes-this technique is simple to put into practice and we achieve amazing results in large number of cases. But practically, dependencies exist among variables.