Artificial Intelligence based Algorithm for Detecting Android Obfuscated Applications

As technology continues to advance, so does the landscape of Android; based on its open-source nature which renders it vulnerable to various risks. Therefore, the developers need to deploy and employ obfuscation techniques in their newly developed android applications. In this paper , we present an investigation into Android obfuscation detection. Our work encompasses a comprehensive examination of Android obfuscation techniques and an exploration of their intersection with machine learning. We conducted extensive experiments involving various machine learning models to detect obfuscation. Among these models , The results show that Random Forest is the one with the most promising results with accuracy 99.5% in detecting Android Obfuscation. The dataset utilized in the experiments encompasses a diverse range of samples, including both malicious and benign samples. This diversity allows for a robust evaluation of the effectiveness of obfuscation detection across different scenarios and highlights the challenges posed by varying obfuscation techniques.