Developing Hybrid Knowledge-Based System for the Diagnosis and Treatment of Banana Disease

Abstract

 Banana production in Ethiopia is widely affected by disease and attacked by several insect pests. The banana disease needs sufficient and knowledgeable agricultural experts to identify the disease and describe the methods of treatment and protection at an early stage of infestation. However, Agricultural experts’ assistance may not always be available and accessible to every farmer when the need arises for their help. Therefore, this study presents a hybrid knowledge-based system for the diagnosis and treatment of banana disease to identify the disease timely and apply the control measure effectively. The system aims to provide a guide for research centers and development agents to facilitate the diagnostic process of mango disease. To develop the proposed method, data and knowledge are acquired from documented and non-documented sources. The acquired knowledge is modeled decision tree structure that represents concepts and procedures involved in the diagnosis of banana disease. For the rule-based reasoning module production rule is used as knowledge representation and for the Case-Based reasoning module, the cases are prepared from the collected dataset using jCOLIBRI studio. Finally, the researcher uses a rule dominant approach for the integration of RBR and CBR Module. The system is developed using SWI Prolog programing language and Java Net Beans and JPL Library is used for the integration of GUI and production rules. The system has been evaluated to ensure the performance of the system is accurate and is the system usable by the researcher and development agent. The system has registered an overall performance of 80.90% accuracy of user acceptance testing. Hence, this study concludes that the integration of rule-based and case-based reasoning approaches achieve better performance concerning the performance of individual reasoning approaches in the identification, recommending first-line treatment, and prevention of Mango infection. The finding of this study can be used as a supportive tool for agricultural extension workers, farmers, and farmworkers to help in the diagnosis and treatment of mango disease.