NAVIGATION INFORMATION DISPLAY STRATEGY BASED ON DRIVERS' STATED PREFERENCE SURVEY AND DATA MINING TECHNIQUE

摘要

State preference survey and analysis for CNS (car navigation system) based traffic information provision has been conducted. The motivation is to better provide more individualized traffic information for diverse drivers by increasing the amount of more individualized information based upon more proper traffic information provision strategy while accounting for diverse preferences of drivers regarding travel. An Internet based survey has been conducted to collect the state preference data for dynamic individualized traffic information over the pool of potential CNS users. The survey design contained data for individuals' socio-economic data and traffic information preferences. The former included the age, driving career, income and other characteristics and the latter included preferences in face of congestion information and route guidance information. The information provision strategy for diverse groups of drivers has been constructed using a machine learning or data mining algorithm. The results could be expected to be used for both designers and drivers of CNS systems.