Query driven spatial pattern analysis and visualization through GIS

Earth features are arranged following one or many rules called spatial patterns. Around the world many features share common patterns. To discover those patterns and analyzing them through some data visualization technique can lead to a good decision supporting system. In this paper analysis and visualization model is designed for the Uttarakhand (India) region. Data like district wise population, schools and hospitals is collected from different sources. To store GIS data PostgreSQL and PostGIS is providing a favorable support and powerful database. GIS maps are created using tool QGIS (Open source) for the visualization purpose. To classify common pattern from the given dataset, clusters are formed from population field. To cluster data OPTICS clustering method is used. Then decision tree by information gain method is used for the supervised learning. To display the result open source software Geoserver is used. Geoserver handle the user queries and display the result on the dynamic maps.