A knowledge-based System for diagnosis in Veterinary Medicine

Abstract

An Expert System is here presented for diagnosis of goat diseases. The principal aim is to build a simple and reliable software construction, giving a field assistance, taking into account a knowledge representation as it is used in the domain of Veterinary Medicine. The system consists of about 50 rules, one for each disease; each rule contains from 15 to 25 elements of description with four valued certainty factors (never-no, sometimes, mostly, yes-always). A first software product was carried out in Ada language with a Vax 4000/300, and a smaller program has been written in C language, to be used with a PC on any farming concern. The elements of a rule can be related to any aspect of diagnosis in the field of Veterinary Medicine, as for instance general symptoms, laboratory tests, visual characteristics of a specific lesion, location of the farming concern, and so on ...