Using different techniques of steganography deducting null cipher in plain text

摘要

Encoding a message in plain text is an ancient art. Using null cipher (NC) is a great challenge because it requires extensive human efforts and intelligence. In this study, we propose a mechanism which can detect the existence of NC steganography in text without human involvement. We propose a statistical approach based on finding the probability of a ST to generate meaningful words and average word length. A threshhold value (THV) is defined to distinguish the normal data from unusual statistical data. On the basis of the probability and AWL each ST is assigned a risk factor value. Texts exceeding THV will be marked as infected and the remaining as clean text.