Methodological proposal for modelling and implementing regulation application problems in a knowledge-based system

In this paper, a methodology is proposed for designing and constructing a k nowledge-based system (KBS) in order to generate and evaluate administrativ e grants. The methodology is applied to the Common Agricultural Policy regu lations covering Spanish farmers for validation. The proposed methodology f acilitates the development of applications of this type from preliminary pr oblem analysis to application implementation and maintenance. This methodol ogy is well adapted to the characteristics of the knowledge-based software under development: open requirements and diverse computational models. It c onstitutes a complete guide for the knowledge engineer in producing and mai ntaining an automated solution to a real-world problem, as the methodology is not only declarative (i.e. indicates what to do) but also procedural (i. e. indicates how to do it). The KBS demonstrator built is a query system fo r use by specialists from the Agricultural Extension Services. The prototyp e implemented by means of the Kappa-PC V. 2.3 was developed jointly by the Departments of Artificial Intelligence, Rural Projects and Planning and Mat hematics applied to Agronomics of the Polytechnic University of Madrid and the Castile-Leon Provincial Office of Agriculture in Segovia. (C) 1999 Else vier Science Ltd. All rights reserved.