A Machine Learning-Based Framework for Mobile Forensics

With the development of smart mobile devices, many portable electronic equipment not only keep the connected for us but also can act as an office and amusement tool. Mobile devices and processor technology enable person to handle their works conveniently and many applications are also installed in the equipment for various purposes. In the meantime, mobile services may be employed by criminals to commit malicious activities, and the criminal acts is hard to track because the crime detail information would be investigated difficultly because of the diverse form files stored in the mobile equipment such as smartphones. In order to efficient investigation the various types of crime files that collect from mobile devices, we propose a framework that combine kernel k-means and multi SVM machine learning methods to analyze documents and images for forensics investigation.