Using Data Mining for Predicting Cultivable
Uncultivated Regions in the Middle East
Ahsan Abdullah1, Ahmed
Bakhashwain2, Abdullah Basuhail3, and Ahtisham Aslam4
1Department
of Information Technology, King Abdulaziz University, Saudi Arabia
2Department
of Arid Regions Agriculture, King Abdulaziz University, Saudi Arabia
3Department
of Computer Science, King Abdulaziz University, Saudi Arabia
4Department of Information Systems, King Abdulaziz
University, Saudi Arabia
Abstract: Middle-East region is
mostly characterized by a hot and dry climate, vast deserts and long
coastlines. Deserts cover large areas, while agricultural lands are described
as small areas of arable land under perennial grass pastures or crops. In view
of the harsh climate and falling ground-water level, it is critical to identify
which agriculture produce to grow, and where to grow it? The traditional
methods used for this purpose are expensive, complex, prone to subjectivity,
risky and are time-consuming; this points to the need of exploring novel IT
techniques using Geographic Information Systems (GIS). In this paper, we
present a data-driven stand-alone flexible analysis environment i.e., Spatial
Prediction and Overlay Tool (SPOT). SPOT is predictive spatial data mining GIS
tool designed to facilitate decision support by processing and analysing
agro-meteorological and socio-economic thematic maps and generating crop
cultivation geo-referenced prediction maps by predicative data mining. In this
paper, we present a case study of Saudi Arabia by using decade old wheat
cultivation data, and compare the historically uncultivated regions predicted
by SPOT with their current cultivation status. The prediction results were
found to be promising after verification in time and space using latest satellite
imagery followed by on-site physical ground verification using GPS.
Keywords: Data mining, image processing, GIS, prediction,
wheat, alfalfa.