【正文】 Midrun explanation easily in expert systems General configuration of expert systems Knowledge Acquisition Knowledge Processing (Inference Engine) Knowledge Representation (Bases) Interpretation amp。 Inquiry Expert User 2. Representative Example of Expert Systems DENDRAL, E. A. Feigenbaum et al, Stanford U, 1968 Spectroscopic Analysis Molecular Structure MACSMA, C. Engleman et al, MIT, 1974 Mathematical Calculus MYCIN, E. H. Shortliffe et al, Stanford U, 1974 Bacterial Infection Diagnosing AM, D. B. Lenat, et al, Stanford, Inductive Inference PROSPECTOR, R. O. Duda et al, Stanford, Geologist HEARSAY, CarnegieMellon U, 1960s, NLP XCON, DEC amp。 CarnegieMellon, 1980, Computer Configuration … 3. Example: MYCIN First large expert system to perform at human expert level, first expert system to solve realworld problems instead of “toy problems”, and passed Turing Test. (1) General Description Functions: (1) Report the user is a bacterial infection patient。 (2) Give the infection hypothesis。 (3) Remend antibiotic therapy。 (4) Give the proper prescription. Knowledge Acquisition: (1) Built in by knowledge engineers。 (2) Learnt during manmachine interaction Knowledge Representation: Production System (2) Knowledge Bases: (1) Static knowledge base which includes (a) dictionary base has 800 English words, used for dialog between MYCIN and Doctor/Patient (b) index base simplifies the knowledge stored (c) knowledge categories (2) Dynamic Knowledge Base which includes (a) new data given by the patient (b) intermediate conclusions (c) some records during operation (3) Rule Base Rule base includes 200 production rules, ., If (i) the stain of the anism is grampositive, and (ii) the morphology of the anism is coccus, and (iii) the growth conformation of the anism is clumps Then there is suggestive evidence (.7) that the identity of the anism is staphylococcus. (4) Knowledge Processing MYCIN reasons backward from its top level goal ( of determining that there are diseasecausing anism that must be treated) to clinical observations. To solve the top level diagnostic goal, it looks for rules whose right sides suggest disease. It then uses the left side of those rules to set up subgoals whose success would enable the rules to be invoked. These subgoals are again matched against rules, and their left sides are used to set up additional subgoals. Whenever a left side describe a specific ptece of clinical evidence, MYCIN uses that evidence if it already has access to it. Otherwise, it asks the user to provide the information. (5) Uncertain Reasoning How does MYCIN bines the estimates of certainty in each of the rules to produce a final estimate of the certainty of its conclusion ? All assertions being considered have associated with them two numbers: MB (measure of belief) and MD ( measure of disbelief) of a hypothesis h given evidence e max(p(h|e), p(h)) p(h) Max(1, 0) p(h) MB (h,e) = 1 if p(h) = 1 otherwise MD (h, e) = 1 if p(h) = 0 Min (p(h|e), p(h)) p(h) Min(1, 0) p(h) otherwise The overall estimate of the confidence of the system in its belief about the hypothesis, certainty factor (CF), is given by CF(h,e) = MB(h,e) MD() The measures of belief and disbelief of a hypothesis given two observations s1 and s2 are puted by 0 if MD(h, s1amp。s2) =1 MB(h, s1amp。s2) = MB(h,s1) + MB(h,s2)[1MB(h,s1)], otherwise 0 if MB(h, s1amp。s2) = 1 MD(h, s1amp。s2) = MD(h,s1) + MD(h,s2) [1MD(h,s1)] otherwise Certainty factor of a bination of hypothesis: MB(h1amp。h2, e) = min (MB(h1,e), MB(h2,e)) MB(h1orh2, e) = max (MB(h1,e), MB(h2,e)) (6) Interaction Is the patient’s illness with ORGANISM1 a hospital acquired infection ? Why ? (Doctor asked) [] This will aid in determining the category of ORGANISM1. It has already been established that [] the gram stain of ORGANISM1 is gramneg, [] the morphology of ORGANISM1 is rod, and [] the aerobicity of ORGANISM1 is facultative therefore, if [] the infection with ORGANISM1 was not acquired while the patent was hospitalized, then there is strongly suggestive evidence (.8) that caregory of ORGANISM1 is enterobacteriacese. [also: there is weakly suggestive evidence (.1) that the identity of ORGNISM1 is pseudomonusaeruginosa] [Rule 037] How [] The following were used: [] Rule 027 indicated there is strongly suggestive evidence (.8) that the aerobicity of ORGANISM1 is facultative. [] Rule 033 indicated there is weakly suggestive evidence (.2) that the aerobicity of ORGANISM1 is not facultative. Since this give a cumulative CF of (.6) for facultative, the evidence indicates that the aerubicity of ORGANISM1 is facultative.