Discrepancy in real-time traveler info websites traffic condition estimation based on rough set theory

More and more real-time traveler information Websites have been set up to provide real-time traffic condition information. They derive traffic conditions with different algorithms based on different data sources. We have found that there is a discrepancy among those Websites. This certainly confuses users because they provide inconsistent traffic condition information. As known, people draw conclusions of traffic conditions in a vague way. Estimating traffic conditions is not necessary with high resolution, while a rough classification might be enough. This paper presents a rough set model to address the problem of traffic condition estimation. A densely used urban arterial in Beijing was selected to test the performance of this model. Also, the existing models used in practice are used for comparison. The results of the comparisons indicate that this proposed model is capable of classifying traffic conditions with satisfying accuracy.