Congestion Prediction for Urban Areas by Spatiotemporal Data Mining

Traffic congestion is a fairly serious problem in many cities over the world. Thus, congestion prediction is a significant research problem for society. However, traffic congestion is affected by several factors and vary with spatial and temporal environments, which make it difficult to define and predict. This paper proposes a spatiotemporal traffic prediction mining model based on extending association rule mining algorithm. This model can predict the traffic congestion by extracting the frequent congestion areas and spatiotemporal congestion propagation rules. The feasibility and effectiveness of the proposed model have been evaluated through the analysis of congestion areas using the real floating car data in Beijing, China.