An Approach to detection of privacy leakage Vulnerability in Android mobile Applications

Abstract

Malware is malicious programs that intentionally carry out destructive actions. it is categorized as software that can harm any computer, mobile device, or operating system. Over the past decade, an increasing number of malwares has been created. The exponential growth and complexity of malware pose a serious threat to computer and network security. In recent times, cybercriminals (attackers) have used malware as a weapon to carry out attacks on computer systems to fulfill their nefarious intentions. Malware is classified into several types based on its behavior and execution nature, such as worms, viruses, Trojans, rootkits, backdoors, spyware, logic bombs, adware, and ransomware. The Internet is the primary medium for carrying out malware attacks on computer systems through emails, malicious websites, and attached software code. Computer systems are compromised for many reasons such as: for financial gain, stealing confidential or private data, using the system to act as a bot, or making services unavailable to the system, etc. Therefore, detecting vulnerability is essential for recognizing and offering early warnings of potential attacks. Android operating system is particularly vulnerable to malware attacks due to its open-source nature. newlineZeus GameOver, Blackrock, Cerberus, StrandHogg, and AgentSmith are the current malware attacks active in 2019-2022. The usual approaches used for detecting Android vulnerabilities are ineffective due to the variety and many Flavors of Android malware families. Vulnerability detection heavily utilizes the field of machine learning, which can be used to overcome the shortcomings of conventional procedures. Vulnerability detectors are evaluated using machine learning on a small dataset with a limited variety of malicious files. newlineTo develop a system that detects a vulnerability, malicious samples are analyzed using techniques called static and dynamic analysis. Static analysis methods retrieve the characteristics of malware without running the malicious file.