Performance Analysis on Text Steganalysis Method Using A Computational Intelligence Approach

Roshidi Din, Shafiz Affendi Mohd Yusof, Angela Amphawan, Hanizan Shaker Hussain, Hanafizah Yaacob, Nazuha Jamaludin, Azman Samsudin


In this paper, a critical view of the utilization ofcomputational intelligence approach from the text steganalysisperspective is presented. This paper proposes a formalization ofgenetic algorithm method in order to detect hidden message on ananalyzed text. Five metric parameters such as running time, fitnessvalue, average mean probability, variance probability, and standarddeviation probability were used to measure the detection performancebetween statistical methods and genetic algorithm methods.Experiments conducted using both methods showed that geneticalgorithm method performs much better than statistical method,especially in detecting short analyzed texts. Thus, the findings showedthat the genetic algorithm method on analyzed stego text is verypromising. For future work, several significant factors such as datasetenvironment, searching process and types of fitness values throughother intelligent methods of computational intelligence should beinvestigated.

Full Text: PDF


  • There are currently no refbacks.