most of the personal and professional communication is done using the electronic media such as E-Mailing, SMS etc. But these all services also suffer from the problem of unwanted messages or the communicating information called Spam. The Spam Message can be an email virus, charity latter, commercial advertisement etc. But it affects the user time, memory and the attention. In this paper, a DCA based improved decision tree approach is suggested to identify the spam emails over the dataset. The work is implemented in integrated weka environment. The obtained result from the system shows the effective identification of spam mails over the dataset. KeywordsEmail, DCA, Decision Tree, Secure Communication