User behavior modeling for the prediction, profiling and detection of malware in mobile devices
Recent advances in Smartphone technology led to rapid development of mobile applications that closely integrate into our daily lives. The Android platform is the fastest growing market in Smartphone operating systems to date. It has become the target for security threats because of its wide spread presence among mobile operating systems to develop and publish applications into the Android market. It presents an opportunity for attackers to easily deliver malicious applications such as malware which steals sensitive information, gets root privilege and performs abuse functions such as send SMS, delete files etc., without the user's knowledge. Also, the presence of advanced sensing functions attracts malware developers to develop targeted malware. This type of malware complicates behavioral detection since it triggers malicious behavior based on the factors such as user actions, profile and/or the presence of other applications. Many dynamic detection techniques have been proposed to detect Smartphone malware. However, a real time, scalable approach to detect the targeted malware in a dynamic analysis has not been studied extensively. The main objective of this thesis is to develop User behavior models which help in detection of malware and prediction of application usage that would prevent further loss to the user by prioritizing in real time the malware testing queue and send an alert beforehand to the user to make them aware about the malicious presence. Also, grouping similar User behavior models will provide scalability and alert multiple users who are using the same malware application in the group.
Computer Engineering|Computer science
Divya Naga Devi Guntu,
"User behavior modeling for the prediction, profiling and detection of malware in mobile devices"
ETD Collection for Tennessee State University.