Modelling and Mining of Networked Information Spaces

Project Title: Attack
Participants: Gunes Kayacik, Patrick LaRoche, Dr. Malcolm Heywood, Dr. Nur Zincir-Heywood
Project Description: The principal objective of this project is to automate the process of building instances of a generic class of attack. To do so we recognize that (1) non operational code is necessary for masking the operational or real intent of an attack. (2) Solutions have been widely observed to have functionality that is distributed across the length of the individual. Thus, the basic objective of an automated process for generating attacks will be to build code sequences, which have the same function as the original generic attack, but camouflaged in non-operational but syntactically correct code. To this end a machine learning framework will be developed for developing a family of attacks from a generic objective (denial of service, stack/heap overflow etc.).