Study Of The Effectiveness of Expert Sistems to Diagnose Diseases In Cattle

Abstract

Cattle often face the risk of diseases that can lead to significant losses for farmers. Limited knowledge about diseases in cattle often makes farmers rely on livestock experts or seek assistance from veterinarians. Therefore, this research aims to compare the effectiveness of two expert system methods, namely Euclidean Probability and Bayes' Theorem, in detecting diseases in cattle. Euclidean Probability is used as a case-based approach technique to measure the likelihood or certainty of conclusions based on the causes that occur. On the other hand, Bayes' Theorem is a method for calculating the probability of hypotheses based on previous data. Both methods have similar goals, which are to determine the presentation of diseases based on the symptoms experienced by cattle, with the main difference lying in the calculation processes they employ. The application of the expert system resulting from this research can assist clinic personnel in detecting diseases in cattle. The effectiveness of the method from the program trial for six different diseases resulted in an accuracy of 83% for the Euclidean Probability method, while the Bayes' Theorem method resulted in 50% accuracy. This concludes that the Euclidean Probability method is more effective than the Bayes' Theorem method in diagnosing diseases in cattle.