EKSPLORASI TUTUPAN LAHAN BERDASARKAN KONSEP SPATIAL DATA MINING STUDI KASUS: JAWA BARAT

• المؤلف الرئيسي: Fitriana Wulandari, Dini
• التنسيق: Final Project
• اللغة: Indonesia
• الوصول للمادة أونلاين: https://digilib.itb.ac.id/gdl/view/48292
• المؤسسة: Institut Teknologi Bandung

Land cover provides patterns of landscape features found on earth. Large amount of land cover data due to temporal and spatial variety causing the patterns and information found varies but increasing in complexity and quantity. Spatial data mining can be implemented to extract interesting information from large and complex spatial data, one of which is land cover data. This study uses the Association Rule method from spatial data mining to discover association information between land classes in West Java region in 2005, 2010, and 2015. The main steps done in association rules is frequency pattern mining, to form land cover combination that meets threshold value, and continued by association rules formation to make representative rules for land cover data in West Java in the specified period. A total of 2 rules from threshold 5%, 5 rules from threshold 1%, and 23 rules from threshold 0% were acquired to discover land cover patterns in West Java region in 2005, 2010, and 2015. This study also acquired trends found from rules parameters to describe the characteristics of land cover in West Java region in 2005, 2010, and 2015, which shows that at least 76.614% of land cover in West Java region remains unchanged from 2005 to 2015.