User profiling for computer security

Systems and control theory was largely responsible for realizing the industrial revolution. We believe that systems and control theory also plays a key role in the information revolution underway. In particular, systems and control theory should have much to offer network security. This paper explores the use of queueing theory and logistic regression modeling methods for profiling computer users based on simple temporal aspects of their behavior. The goal is to develop profiles for very specialized groups of users, because of the nature of their work would be expected to use their computers in a very similar and regular way, e.g., bank tellers, insurance adjusters, etc. Encouraging proof-of-concept results suggest the potential of both the idea of user profiling and the use of simple temporal features for grouping users.