Applying spatial data mining to develop a predicative model for illegal land exploitaion

A large-scale land use/land cover change detection system has been developed to detect illegal land practice for the land management administrations in Taiwan since 2001. Through long-term and regular monitoring of land change, the system has accumulated a mass investigation data of illegal land practice. The main objective of this study is to build a predictive model with the integration of data mining technology and Geographic Information Systems (GIS). The model incorporates spatial attributes with land change investigation data to gain insight into the relationship between illegal land exploitation and spatial data. Consequently, the predictive model can serve as a reference for authorities to identify critical factors of illegal land use and conduct land management policies for Taiwan in a more effective way.