Identifying Android Banking Malware Through Measurement of User Interface Complexity

Trojan Android banking malware can steal login credentials from users. This type of malware enables threat actors to gain unauthorized access to online banking accounts and steal funds. Such infostealers, when purpose-built to impersonate a legitimate online banking application, often lack the attention to detail observed in legitimate benign Android applications. This disparity is anecdotally observable but also empirically measurable. This research shows that a design science artifact can extract significant user interface features from APK files, and a binary logistic regression approach can quantitatively analyze and classify mobile banking mal ware on these features with a 94.1 % accuracy.