Map Matching Algorithms: An Experimental Evaluation

Na Ta,Jiuqi Wang,Guoliang Li

DOI: https://doi.org/10.1007/978-3-319-96893-3_14

2018-01-01

Abstract:Map matching is an important operation of location-based services, which matches raw GPS trajectories onto real road networks, and facilitates tasks of urban computing, such as intelligent traffic systems, etc. More than ten algorithms have been proposed to address this problem in the recent decade. However, existing algorithms have not been thoroughly compared under the same experimental framework. For example, some algorithms are tested only on specific datasets. This makes it rather difficult for practitioners to decide which algorithms should be used for various scenarios. To address this problem, in this paper we provide a survey on a wide spectrum of existing map matching algorithms, classify them into different categories based on their main techniques, and compare them through extensive experiments on a variety of real-world and synthetic datasets with different characteristics. We also report comprehensive findings obtained from the experiments and provide new insights about the strengths and weaknesses of existing map matching algorithms which can guide practitioners to select appropriate algorithms for various scenarios.