Design and implementation of a rule-based mobile expert system for plant disease diagnosis using forward chaining method

Abstract

Plant diseases remain a significant challenge to global food security, particularly in regions with limited access to agricultural experts and diagnostic resources. This study aimed to design and implement a rule-based mobile expert system using the Forward Chaining Method to diagnose common plant diseases in crops such as Rice, Wheat, Tomato, Potato, and Maize. The system was developed using Android Studio, Java, and MySQL for database management, with a knowledge base constructed from expert inputs and agricultural datasets. The performance of the system was evaluated based on Accuracy, Precision, Recall, and F1-Score, along with user satisfaction surveys conducted with 150 farmers. The system achieved an overall accuracy of 92.4%, with the highest accuracy observed in Tomato (95%) and the lowest in Wheat (89%). Diagnostic response time averaged 3.2 seconds, significantly outperforming traditional methods (120 minutes), and the cost per diagnostic case was reduced from 15.5 USD to 1.2 USD. User satisfaction scores were consistently high across all parameters, with a strong positive correlation (r = 0.82) between diagnostic accuracy and satisfaction. This study demonstrates that mobile expert systems can provide real-time, accurate, and cost-effective diagnostic solutions, empowering farmers and enhancing sustainable agricultural practices. Future research should focus on expanding the knowledge base, enabling offline functionality, and integrating hybrid diagnostic technologies.