Cipher-Text Classification with Data Mining

Cryptography plays an important role in information technology. With the explosive growth of Internet, the importance of information and network security is increased. The principle goal of cryptography is information hiding and block-cipher algorithms achieve this goal through information recoding. However, various kinds of information have distinctive features, which are employed in classification tasks. If these features are transformed to the cipher-text, cipher-texts may be classified to their original classes. In this paper, a classification strategy is proposed for cipher-text classification. The proposed strategy is evaluated through simulations. Different classifiers (SVM, KNN, and LDA) are used for this purpose.