Expert System For Identification Of Symptoms And Diseases In Lobsters Using The Backward Chaining Method

Lobster is a high-value fishery commodity widely cultivated, including in Watorumbe Bata Village. However, diseases attacking lobsters are often difficult for farmers to identify early, leading to economic losses due to delayed treatment. This study aims to develop an expert system to identify lobster symptoms and diseases using the backward chaining method. This method enables the system to reason logically and systematically from disease hypotheses to symptom facts. Data collection was conducted through observation, interviews with lobster experts, and literature study. The system development followed the Waterfall model, comprising analysis, design, coding, and testing phases. The implementation results show that the system can diagnose diseases based on symptom inputs and provide information including disease name, cause, solution, and likelihood level. Black-box testing confirmed that all system functions operated properly, while accuracy testing using 20 sample data showed a system accuracy rate of 90%. These results indicate that the expert system using the backward chaining method is effective in assisting farmers to identify lobster diseases more quickly and accurately, thus supporting the sustainability and productivity of lobster farming.