| Summary: | 碩士 === 國立臺北科技大學 === 商業自動化與管理研究所 === 89 === “Data mining” is being extensively used in sales, finance, banking, and telecommunication. Nonetheless, Data mining often fails to perform well because of different reasons, such as invalid data, excessive patterns, and time and space factors.
The theme of this study is “Spatial Data Mining,” or “Knowledge Discovery in Spatial Database.” 50 to 85% of business data is spatial. Therefore, this study focuses on analyzing the autocorrelation of database and discovering the knowledge, spatial relations, and patterns of the database.
Data for this research come from the R.O.C Monthly Population Report of Taiwan and the Outer Islands, the R.O.C Quarterly Population Report of Taiwan and the Outer Islands, and the R.O.C Report of Household Income in Taiwan and the Outer Islands. The Geographic Information System, moreover, forms a spatial database by connecting the data mentioned above with those of the 25 cities and counties in Taiwan. The spatial database includes various information such as the basic data of every township, the population density, and types of occupations. Based on the spatial database, this research involves the analytical technology of spatial data-- Moran’s I Index and Geary’s C Index for the analysis of the global spatial autocorrelation and Moran’s I Index and Getis Gi* Index for the analysis of the local spatial autocorrelation. The results of the analysis contribute to the spatial database, come to the public through data visualization with the business geographic information system, and provide analysts with more information for further study.
The model of data analysis in this research integrates various technologies-- Spatial Autocorrelation, Business Geographic Information System, and Data Visualization─to conduct the analysis of spatial data. The result indicates a significant spatial autocorrelation. Moreover, data visualization through the geographic information system is very beneficial for data presentation and analysis.
|
|---|
