Spatial data mining for poor village characterization at West Java and Banten- Indonesia

Abstract

Data mining, also popularly referred to as Knowledge discovery in databases (KDD), is the automated or convenient extraction of pattern representing knowledge implicitly stored in farge databases, data warehouses, and other massive information repositories. Spatial data mining deal with knowledge discovery in spatial databases. One of spatial data mining task is characterization - to find a description of the spatial and non-spatial properties which are typical for the target objects but not for the whole database. In this research we developed a system prototype which can be implement spatial data mining technique to discover poor village characteristics at West Java and Banten. The prototype consist of four components: pre-processing module, spatial data mining module, database module and visualization module. Spatial data mining module consist of neighbourhood graph, predicate filter, neighbourhood index and spatial characterization algorithm. From this research we discovered several factor related with village poverty i.e. road type, the existence of newspaper subscribers, distance form health facility and type of waste disposal facility.