Jayson MackiePh.D. started in: 2012
Expected year of graduation: 2015
COINS consortium member: Norwegian University of Science and Technology
Supervised by: Katrin Franke
Research area: Digital Forensics
Project title: Interactive Evolutionary Computing
Project description: Interactive Evolutionary Computer is the synthesis of two problem solving methods to Big Data problems. Evolutionary computing methods are easily distributed making them a candidate for exploring very large non-localised datasets. Humans, both experts and random crowds, have been shown to make very rapid and accurate approximations of a problem's answer without always knowing explicitly why. I am looking to merge the computer's ability to manipulate datasets far larger than a human can handle with the human ability to make large leaps in reasoning while grouping or searching information. The challenge is how the computer can provide visualisations of the information in a form accessible to humans, and how the humans may then provide meaningful guidance back to the algorithm to improve performance.
The outcome of these IEC algorithms will be the creation of new tools for Computational Forensic investigation. This approach would enable forensic investigators to more easily and efficiently guide complex searches through massive evidence collections. As the search progresses input can be given to areas which should focused on, allowing small or currently incomplete areas of the evidence to be prioritised rather than be swamped by larger or stronger patterns in the data. The easily distributed nature of EC algorithms makes them inherently scalable to cluster or cloud-based evidence storage environments to further improving the speed of access to the search results, including simultaneous access and guidance from more than one investigator, with each searching for different objectives.