Detection of Android Ransomware Using Machine Learning Approach

The majority of Android smartphone users utilize handheld devices for almost everything in their life, including regular planning, data interchange, correspondence, social interaction, business execution, and financial transactions. The prevalence of cyberattacks on smartphones has drastically increased with people’s reliance on smartphone technology. Smartphone applications require permission to access several smartphone features, which might be used by hackers to conduct an attack or implant malware. The main aim of attackers in target cell phones is to obtain victims’ personal information for financial benefit. Despite this, as Android has the largest market share among smartphone operating systems, it is frequently attacked by cybercriminals. One such major infestation is Ransomware, which is mainly found on PCs and has the ability to latch onto smartphones. Ransomware encrypts the victim’s data and demands for a ransom amount for the decryption key. The present Android ransomware research is deficient in key components and relies on supervised machine-learning techniques. However, these techniques have several drawbacks and early detection and recognition of ransomware in android is required. The main aim of this paper is to examine various machine-learning algorithms used in Android ransomware. The novelty of the paper is to combine Ransom-Droid and concept drift by classifying raw data based on host, network, behaviour, and files. Each data is recognized with the help of static and dynamic analysis based on the type of ransomware detection. The proposed paper is informative with real-world applications in malware identification and categorization.