Abstract
Real-time traffic information is essential in intelligent transportation system (ITS) service for users to estimate arrival time to destinations. Statistical data, generated from historical data, is a useful compensation for real time data to improve the quality of traffic information. Therefore, the generation of high quality statistical data becomes very important.The paper introduces an irregular data cleansing method, which is a key issue to produce fluent data for statistical data generation. Evaluation results based on two months historical data in Beijing indicate that the proposed method can efficiently cleanse irregular data.