Extracting map information from trajectory and social media data

Existing surveying methods are either labor intensive or highly costly and have a long updating cycle, which hinders the timely update of maps. In view of these problems, this paper proposes a framework of extracting digital map information from raw geospatial big data. The framework consists of four steps: data preprocessing, mathematical modeling, information extraction and map post-processing. Extracting map information based on the proposed framework is low-cost and has a short update cycle. A case study is illustrated to show the effectiveness of the framework.