Web-Based Expert System for Diagnosing Chicken and Duck Diseases Using the Dempster–Shafer Method

Abstract

Poultry farming, particularly chicken and duck production, makes a significant contribution to the supply of animal-based protein in Indonesia. However, the high vulnerability of poultry to infectious diseases such as Newcastle Disease, Avian Influenza, Cholera, and Pullorum often results in severe economic losses and impacts public health. The problem is further complicated because disease diagnosis in the field is still largely performed manually by farmers or veterinary personnel, which often requires considerable time, cost, and expertise, while similarities in symptoms among diseases frequently lead to misidentification. To address these issues, this study developed a web-based expert system using the Dempster-Shafer method. This method was selected because it can process uncertain data by combining evidence to generate more representative measures of belief and plausibility. The research stages included problem identification, data collection through interviews and literature studies, system design using the CodeIgniter framework based on the MVC model with PHP and MySQL, implementation of the Dempster-Shafer algorithm in the inference engine, and field testing at Ternak Hasbuan Kebun IV, Kotabumi. The developed system provides features such as a dashboard, symptom input, inference process, diagnosis results, and diagnosis history. Testing results showed that the system achieved an average compatibility rate of 88.76% compared to veterinary expert diagnoses, with system confidence values ranging from 84.6% to 92.4%. These findings demonstrate that the system can serve as a practical solution for farmers to perform independent diagnoses, accelerate disease identification, and effectively support poultry health management. The novelty of this research lies in the integration of dual-species (chicken and duck) diagnosis within a single web-based platform, validated directly through field testing.