Data Fusion of Heterogeneous Data Sources for Intelligent Transportation Systems

Intelligent Transportation Systems (ITS) provide applications to enhance road safety, minimize travel time, and detect incidents. However, providing accurate information for road status is considered as a challenging task. These days, the availability of big data from multiple sources provides valuable information to ITS applications. These data require data fusion techniques to handle and extract useful knowledge. Additionally, deep learning approaches have been applied to the ITS field and provided promising results; but it required the input data to be in large amounts of data and in a uniform representation. In this paper, we provide data fusion techniques to integrate and extract features from heterogeneous data sources to be ready for deep learning training approaches. We also generate preprocessed real-world traffic datasets can be used to solve ITS-related problems. Furthermore, we provide two applications to show the importance of our data fusion techniques. Traffic datasets analysis and visualization, where we build a data cube to provide an in-depth analysis of the dataset. Traffic flow forecasting using deep learning; we perform a comprehensive study on how different features can improve the traffic flow prediction models. The experimental results show that deep learning approaches achieve better results when extra features are considered.