Leveraging Machine Learning for Mobile Forensics
摘要
Mobile digital investigations have assumed great importance in the past years because of the enhancement of the number of devices, especially mobile ones, in everyday life. Today, phones and tablet forms have become the means of communication, Internet, business, and social interaction. Therefore, these apparatuses bear extensive amounts of personal and business communications, such as text messages, emails, phone call records, GPS tracks, photographs, films, and information on the applications that are installed. Due to the increasing usage of mobile devices in everyday life activities and communications, they become the most valuable source of evidence during criminal investigations, cybersecurity incidents, or civil lawsuits, namely, regarding activity or communication tracks left by a suspect in a criminal investigation, cyberattack, or involved party in civil proceedings. However, this is an area of concern because the storage capacity on mobile devices is very large, thereby creating a challenge for forensic investigators. Current digital investigation techniques that involve mechanically copying or transferring data to a new environment are insufficient. Given the fact that thousands of mobiles churn out gigabytes of data each day, the process of knowledge extraction mandates a lot of time and can lead to the generation of numerous errors. However, mobile platforms, including Android and iOS, are revised more often, implying that new techniques of encryption, security elements, and operating system complications complicate the extraction and analysis of data. Tools that can be used for analysis of the cell phones affect the necessity of inventing more progressive technologies for analysis by forensic investigators due to constant development of mobile technologies and the growth of investigative complications.