Trends in Research and Artificial Intelligence Methods for Steganalysis - A Systematic Literature Review (SLR)

The advancement of steganography techniques in the modern era has reached the point where malware can be embedded within an object or medium, known as stegomalware. Consequently, the need for developing new techniques and methods in steganalysis that can detect cover objects, steganograms, and stegomalware is increasing. Artificial Intelligence (AI) is a technology capable of mimicking human cognitive functions by recognizing its surroundings, capturing patterns, making decisions, and learning. AI functions as a solution to various problems more quickly and intelligently. Therefore, modern steganalysis techniques must address these challenges using AI. This study aims to review research trends and methods used in the application of AI for steganalysis over the past five years (2018–2023), which have not been well identified, using the Systematic Literature Review (SLR) method. By employing the SLR method, research will be identified using keywords for searches in digital libraries, specifically IEEE Xplore. Based on predetermined inclusion and exclusion criteria, 13 journals were selected for investigation. The literature results indicate that research trends in this field continue to evolve, with most cases being addressed in stegoimage, comprising 77% of the research. Ten methods were identified, with the most used method being Convolutional Neural Network (CNN). The results of this study can serve as a guide for researchers and other readers to observe current research trends.