Association rules mining of vehicle routing for personalized traffic information service
Ji Li · Systems Engineering - Theory & Practice · 2013
Vehicle routing prediction is the foundation of personalized traffic information service system. In order to provide key technology support for RFID based personalized traffic service system,an improved association rules mining method is proposed to predict the vehicle routing and provide personalized recommendation.The association rules are derived from the frequent sequential patterns which are mined based on the historical routing dataset.The proposed method includes two improvements.Firstly,0-N data structure is advocated to represent the vehicle routing records.Secondly,an improved method is developed to generate frequent 2-sequence sets.Finally,the test verified the performance improvement of proposed algorithm in terms of mining time compared with GSP method.