EXPERT SYSTEM FOR DETERMINING POULTRY FEED NUTRITION USING THE BACKWARD CHAINING METHOD

Nutritional fulfillment in poultry feed is a critical factor in supporting optimal livestock growth, yet many farmers still face significant challenges due to limited knowledge regarding appropriate feed compositions for various growth phases. To address this gap, this study develops a web-based expert system designed to provide precise nutritional recommendations for various poultry types, including laying hens, ducks, quail, and broiler chickens. The system utilizes the Backward Chaining method as its inference engine, operating through a knowledge base structured with 13 nutritional goals and 8 distinct symptom/phase indicators. The system was subjected to rigorous evaluation through Black Box testing to ensure full functional integrity. Furthermore, diagnostic accuracy was assessed using 15 comparative test cases between the system's output and expert assessments. The results demonstrate that the expert system achieves an 80% accuracy rate in determining the optimal nutritional composition of poultry feed. This research concludes that the application of a backward chaining-based expert system can effectively assist both novice and experienced farmers in optimizing livestock productivity and reducing the risk of nutritional errors.