Fuser block technologies performance based on identity attributes metrics models

The identity of people and entities has continued to be a central theme in both the real and the cyber space. In this paper, we compare three fuser block technologies used in a multifactor authentication system based on two identity attributes metrics models. The fuser blocks are built using artificial intelligence technologies which include Artificial Neural Networks (ANN), Fuzzy Inference System (FIS) and Adaptive Neuro-Fuzzy Inference System (ANFIS). The identity attributes metrics model is developed using term weight from text mining techniques and entropy from information theory. The actual experimental prototype is developed using four authentication factors. The results show that ANFIS was not affected by the type of identity attribute metric model used while ANN showed a better performance when using entropy metric system and FIS showed a better performance when using term weight metric system. These results are significant when building robust web based authentication systems that combine biometrics, Internet terminals and other authentication factors such as the username and password/PIN number.