No abstract

A user identification system which matches the keystroke dynamics of the users with the Cognitive RAM (CogRAM) weightless neural net is discussed in this paper. The keystroke patterns are made up of a common password for all users. While there are several common approaches to represent the users' keystroke patterns, the approach adopted here is based on the force applied to each key. Effectively, they will then constitute a fixed length passkey. In addition, the system was developed based on an 8-bit AVR enhanced, RISC microcontroller. From the experimental results obtained, it can be seen that the identity of the users can be successfully matched just from their keystroke biometric patterns alone.

No abstract