<![CDATA[An in depth analysis of data security in text-based steganography is necessary to ensure the sustainability and security of the methodology used. The purpose of this study is to analyze the resilience of data security in text-based steganography. The analytical approach used involves identifying and assessing the vulnerabilities of text steganography methods using Indonesian sentence patterns. The initial stage of the research was to analyze previous works related to this field to understand previously identified vulnerabilities. The applied text embedding model is based on a dictionary consisting of 1,929 words grouped into seven word categories that correspond to sentence patterns in Indonesian, namely adj (adjective), adv (adverb), nom (noun), num (numeral), par (particle), pro (pronoun), and ver (verb). Each word class is arranged into a sentence structure and each has the same bit length, namely eight bits. The robustness analysis results show that single-word input data is still vulnerable to brute-force attacks or pattern analysis if the message embedding process uses a simple sentence structure. This is due to the relatively small search space, which makes it easier for attackers to guess the embedding pattern. Conversely, using sentence patterns consisting of more than two words significantly increases combinatorial complexity and expands the possibility space, making hacking attempts much more computationally difficult. Thus, the robustness of a steganographic system increases as the number of words in the sentence pattern increases, as the time and resources required to perform the attack become practically inefficient.]]>
Copyrights © 2025
