The goal of Computer Forensics is to examine digital media with the aim of identifying, preserving, recovering, analyzing and presenting facts and opinions about the information and to detect evidence for illegal activities.
We have developed a series of different application oriented technolgies related to Computer Forensics:
- We are able to distinguish copies of some document from the original. This is reasonable, when e.g. original invoices need to be processed and instead (possibly manipulated) copies are submitted
- Using image classification mechanisms we can determine the presence of pornography on photographs - and we can even classify forbidden child pornography. Confiscated hard disks of suspects can thus be investigated with less manpower and in less time.
- Network traffic of a server can be analyzed using data mining algorithms in order to recognize denial of service attacks and thus to improve security and availability of critical services.
- Identities of persons and organizations from other cultural orignis (e.g. arabic world) whose names are transcribed into the latin alphabet, may sometimes be hard (or even impossible) to find with established database systems doing a sharp match. DFKI technology has been developed to perform a fuzzy matching on customizable phonetic normalforms of names and even accepting partial mismatch of some name parts.