Improving the Architecture of Agricultural Knowledge Processing Systems using Design Patterns

Abstract

Use of Knowledge Processing in agriculture has continuously increased, since the first era of knowledge based systems. Such software systems are used for support in detailed tasks, such as determining the amount of fertilizer in precision farming, as well as on high-level decision support, such as ‘what to plant in the next growing season’. Unfortunately, these software systems often have shortcomings in software quality. Applying design patterns is a recognized means to achieve better systems in terms of efficiency, flexibility, and quality. In this paper, several software design patterns are mapped to the context of knowledge processing systems in agriculture. Furthermore, additional patterns are identified and described. The need for patterns with focus on particular aspects of knowledge processing in agriculture is addressed, and an implementation is introduced as a proof of concept.