抄録
This study aims to develop the comprehensive GIS-based traffic accident database system through the integration of hospital-based data, police data and the road inventory data. To determine how to integrate the data from three data sources, data taxonomy is utilized. Available data are hierarchically classified based on their share characteristic. Grouping data in this way is useful for understanding, designing, and building integrated data system. Data warehouse, a common data storage approach to integration, is utilized for the data integration. GIS is an enabling technology for the integration as well. The scope of data integration is established by identifying the model of target data (or integrated data) and identifying the disparate data that would be mapped to the target data. During the physical data integration process, data from the three data sources are extracted, transformed, cleaned and finally loaded into an integrated data source, a data mart or data warehouse.
