<![CDATA[This study aims to develop an Android-based expert system to help fish farmers detect diseases in tilapia (Oreochromis niloticus) quickly and accurately. This system implements two inference methods, namely Certainty Factor (CF) and Dempster-Shafer (DS), which are then compared to assess their effectiveness and accuracy in the diagnosis process. The research was conducted in Purwonegoro Subdistrict, Banjarnegara Regency, which is one of the centers of tilapia farming in Central Java.The knowledge base in this system is compiled based on disease symptom data obtained from interviews with experts and scientific literature references. The developed Android application allows users to enter symptoms that appear on fish to get diagnosis results along with confidence levels and treatment suggestions. System testing is carried out using real case data from the field, and the diagnosis results are compared with evaluations by experts.The results show that both methods are able to provide fairly accurate diagnoses. The Certainty Factor method excels in terms of speed and simplicity in calculation, while the Dempster-Shafer method is better able to handle uncertainty from non-specific symptom combinations. The accuracy of the Dempster-Shafer method is slightly higher than the Certainty Factor, but the difference is not statistically significant.This expert system is expected to be a practical solution for fish farmers in identifying diseases early on, thus supporting the increase in productivity and efficiency of tilapia farming in the research area.]]>
Copyrights © 2025
