<![CDATA[A web-based Geographic Information System (GIS) has been built by previous researchers to visualize hotspots data in Indonesia. That GIS still has not contained a hotspot analysis module. Data mining method can be used to analyze hotspot data. This research aims to develop and to integrate a clustering module of hotspot in GIS which has been developed in the previous research. The clustering module for grouping hotspot data was built using the DBSCAN algorithm with PHP programming language. Clustering hotspot data was performed based on year, month, and province. Clustering parameters are epsilon and minimum points (MinPts). Epsilon value that used ranged from 0.01 to 0.1 while MinPts ranges from 1 to 6. The clustering results are shown in form of table which consists of the attribute Province, Regency, Latitude, Longitude and Cluster. Cluster column is the final result of clustering using algorithm DBSCAN. The attribute cluster represents clusters are visualized using the map of Indonesia that was built using MapServer. Visualization can help parties involved in making effective and efficient decisions to prevent forest fires.Key words: clustering, DBSCAN algorithm, geographic information system, hotspot]]>
Copyrights © 2016
