User-adaptive interface based on symbol matching for in-vehicle information systems

Drivers are distracted by the large amount of information displayed, which raises the driver´s visual and cognitive workload. Conventional hierarchical searching, which invariably requires time even for those familiar with systems, is another factor. Our proposed user-adaptive interface consists of user modeling using a hidden Markov model (HMM) and symbol matching which is the reconfiguration of a preferred navigation sequence. The interface learns a user´s navigation patterns based on accumulated navigation sequences. The navigation sequence is reconfigured by the user model and a clue. The symbol matching is performed in compliance with the user´s selection and the navigation sequence is recommended to the user. Experimental results showed that reconfiguring the navigation sequence and switches reduced navigation reaction time.