A Study on The Establishment of Knowledage Base for a Herd Mastitis Prevention and diagnosis Expert System

• Main Authors: Kuo, Wen-yi, 郭文益
• Other Authors: Wan Ye-nu
• Format: Others
• Language: zh-TW
• Published: 1996
• Online Access: http://ndltd.ncl.edu.tw/handle/64479116638111105091

碩士 === 國立中興大學 === 農業機械工程學系 === 84 === ABSTRACT Dairy herd mastitis not only reduce dairy farmers' income but also disturb the management of the dairy herd. In Taiwan, mastitis may be caused during the process of breeding by milking hygienic, milking procedure, milking machine performance, and milking machine hygienic which are interact with each other. Therefore, it needs long time development and training to make dairy farmer and counselorhave the knowledge and enough experience to identify sources causing mastitis. An expert system is available to effectively keep and combine domain experts' knowledge and experiences to be the reference for diagnosis and decision. In this study, expert system technologies are applied to identify the milking problems causing mastitis and theirrelated possible solutions. The rule base method usually used on an expert system is applied to classify knowledge of main factors causing mastitis into environment hygienic, milking hygienic, milking procedure, milking machine performance, and milking machine hygienic five categories. There are 66 rules and 11 tables are conducted in the rule-base knowledge system to describe relations. The certainty factor of each rule is also given. Investigation forms are used to collect fired examples to verify the correctness of the rules is on the expert system and the driving procedure of the rules. Verbal description and table look up are used to solve the uncertainly and ambiguity of diagnosis knowledge. The certainty factors (CF) method is applied to calculate the belief of the rule conclusion for the diagnosis reliance.The results from the filed examples are similar to the conclusion driving from the system, which prove the classification of knowledge and the inference of the rule are appropriate. It is also encourage the using CF method to deal with the uncertainty knowledge of herd mastitis prevention and diagnosis expert system.