Two-Stage Data Mining Based Vehicle Navigation Algorithm in Urban Traffic Network

Along with the development of Intelligent Transportation System, traffic detectors collect numerous transportation state data in information databases and accumulate. Such data is greatly meaningful to the vehicle navigation. In this paper, we propose a noble two-stage algorithm about vehicle navigation by using data mining methods on the historical and current transportation dataset. This algorithm begins with picking sensitive data about start and end point in an urban traffic network, and data from related (or nearest) road fragments. Referring to current time and season, the algorithm gives an evaluation to every related road fragments and outputs a most reasonable route between start and end point. The experimental and theoretical analyzes show that this algorithm can form an efficient and effective route in reasonable time.