SEAMLESS SOFTWARE DEVELOPMENT FOR SYSTEMS BASED ON BAYESIAN NETWORKS - An Agricultural Pest Control System Example

Abstract

This work presents a specific solution for the development of software systems that embed functionalities based and not based on knowledge, concerning the decision support process and the information management processes, respectively. When constructing a knowledge model, the processes to be performed are mainly focus on the description of the steps necessary to build it. Usually, all approaches concentrate on adapting the software engineering lifecycle to develop a knowledge model and forget the problem of integrating it in the final software system. We propose a process model that allows developing software systems that use a Bayesian network as knowledge model. In order to show how to apply our software process model, we have included a partial view of the development process of a knowledge-based system, related to decision making in an agricultural domain, specifically with pest control in a given crop.