Affiliations: Department of Computer Engineering, Hankyong National
University, Anseong, South Korea | Department of Computer Education, Silla University,
Busan, South Korea | Department of Computer Science & Information
Systems, Texas A&M University – Commerce, TX, U.S.A
Abstract: This paper describes a Service Oriented Architecture (SOA) based on
Web services technology designed to assist producers in precision farming in
the implementation of decision tree based Cropland Suitability Evaluation (CSE)
decisions. The proposed SOA delivers a recommendation service to carry out CSE.
The recommendation service is supported by three components that perform the
knowledge discovery process by data extraction (Data Provider), refinement of
relevant dataset formats (Data Refinement Assistant) and data-driven estimation
of significances-by-attributes (Attribute Evaluator). We also suggest several
data extraction functions in knowledge discovery process, the data refinement
method transforming naïve candidate datasets into compact and unified ones and
the data-driven Attribute Significance Quantification (ASQ) algorithm for
quantifying the influence of attributes on the decision tree.
Keywords: decision support systems, service oriented architecture, data extraction, decision tree, cropland suitability evaluation
