Triage in-Lab

Abstract

Since it exist a huge backlog of cases and few digital forensic specialists in the Justice System, usually there is not possible to move them to contribute directly into the digital crime scene. On the other side, the law enforcement has a lack of skilled forensic staff available to perform forensic triage. Moreover, the reviews on the fly are taking significant time delays, under pressure, technical restrictions and time framed. At this point, when a suspect target system and data are found, it leads to be seized and moved to a dedicated forensic laboratory where the expert can perform the analysis of their content. Under some circumstances, all that may be required is to quickly and efficiently review a number of target systems to establish if they are likely to contain material of interest to an investigation. However, when the digital evidence comes to the specialist, he has a little knowledge of the previous stage, and it is difficult to make decisions about the priorities or activities on the sized devices. Such reviews are often referred to as "forensic triage" reviews and must be performed using forensically acceptable methods in order that any evidence that is identified during the forensic triage process is not damaged, modified or contaminated, literally or from a legal perspective, by the process of acquiring and reviewing the evidence. We have developed a novel triage tool, which tries to catch a criminal profile with an automated predictive classifier focused on child pornography and intellectual property theft. This software detects few critical attributes into the digital evidence and they are compared with other vectors of characteristics extracted from a digital data corpus based on devices of past cases. As a result of this automated process, a criminal profile prediction is done. This tool will assist to computer forensic experts, in order to make decisions about priorities to make full analysis of suspect devices or discard them with low probabilities of losing digital evidence. Our approach should be useful to mitigate the backlog of computer forensics laboratories.