A Knowledge-Engineering Facility for Building Scientific Expert Systems

---

2

Downloaded by UNIV OF CALIFORNIA SANTA BARBARA on March 5, 2018 | https://pubs.acs.org Publication Date: April 30, 1986 | doi: 10.1021/bk-1986-0306.ch002

A Knowledge-Engineering Facility for Building Scientific Expert Systems Charles E . Riese and J . D. Stuart Radian Corporation, Austin, T X 78766-0948

RuleMaster is a general-purpose software package for building and delivering expert systems. Its features include 1) knowledge acquisition by inductive learning, 2) specialized a r t i f i c i a l intelligence programming s k i l l s are not required, and 3) it runs on a wide range of micro-computers and mini-computers. RuleMaster was developed to enable scientists and engineers to incorporate human-like decision making as part of their computer applications. One such application is TOGA, an expert system to diagnose faults in large transformers based on gas chromatographic analysis of the insulating oil. An e x p e r t system i s a computer program w h i c h c o n t a i n s t h e c a p t u r e d knowledge o f an e x p e r t i n some s p e c i f i c domain. The program i s a b l e t o g i v e a d v i c e w i t h i n t h e d o m a i n i n much t h e same manner as t h e human e x p e r t w o u l d , a s k i n g f o r i n f o r m a t i o n as i t i s n e e d e d , v o l u n t e e r i n g p a r t i a l diagnoses as t h e y a r e r e a c h e d , and f u n c t i o n i n g w i t h incomplete or p o s s i b l y erroneous i n f o r m a t i o n . The e x p e r t s y s t e m i s a b l e t o p r o v i d e an e x p l a n a t i o n o f t h e l i n e o f r e a s o n i n g upon demand. U n t i l r e c e n t l y , most e x p e r t system b u i l d i n g t o o k p l a c e i n t h e r e s e a r c h departments o f u n i v e r s i t i e s and a few major c o r p o r a t i o n s . The p r i m a r y e m p h a s i s was i n v e s t i g a t i o n o f a r t i f i c i a l i n t e l l i g e n c e p r i n c i p l e s , a n d t h e a p p l i c a t i o n was o f s e c o n d a r y i m p o r t a n c e . The e x p e r t systems t o o l s used r e f l e c t t h i s i n t e r e s t . They a r e t y p i c a l l y s t a n d - a l o n e A I computer systems, u s i n g s p e c i a l hardware and s o f t w a r e environments ( u s u a l l y L i s p ^ b a s e d ) not commonly found i n s c i e n t i f i c and e n g i n e e r i n g o r g a n i z a t i o n s . But a p p l i c a t i o n s u s u a l l y n e e d a d i f f e r e n t t y p e o f computing environment. The r e a s o n i n g t a s k , a c c o m p l i s h e d b y A I t e c h n i q u e s , o f t e n c o n s t i t u t e s t e n p e r c e n t o r l e s s o f t h e code o f an a p p l i c a t i o n . The m a j o r i t y o f t h e code i s f o r c o n v e n t i o n a l programming t a s k s , such as d a t a a c q u i s i t i o n , d a t a base a c c e s s , n u m e r i c a l c a l c u l a t i o n s , and graphics. I n each a p p l i c a t i o n domain, computer hardware and s o f t w a r e has b e e n s e l e c t e d t o m a t c h t h e n e e d s o f i t s t a s k s . In 0097-6156/86/0306-0018$06.00/0 © 1986 American Chemical Society

Pierce and Hohne; Artificial Intelligence Applications in Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1986.

Downloaded by UNIV OF CALIFORNIA SANTA BARBARA on March 5, 2018 | https://pubs.acs.org Publication Date: April 30, 1986 | doi: 10.1021/bk-1986-0306.ch002

2.

RIESE A N D STUART

A Knowledge-Engineering Facility

19

e s t a b l i s h e d f i e l d s l i k e c h e m i s t r y , computer s o l u t i o n s have been i m p l e m e n t e d and i n use f o r y e a r s . I t i s n o t r e a s o n a b l e f o r t h e A I component, a r e l a t i v e l y s m a l l a d d i t i o n t o t h e t o t a l s y s t e m , t o d i c t a t e major changes t o t h e computing environment. W h i l e t h e o r i g i n a l e x p e r t system approaches were s u i t a b l e f o r A I r e s e a r c h , s e v e r a l t y p e s o f p r o b l e m s a r e e n c o u n t e r e d when t h e emphasis i s s h i f t e d t o s c i e n t i f i c e x p e r t system a p p l i c a t i o n s . In t h e o r i g i n a l a p p r o a c h e s , e x p e r t s y s t e m b u i l d i n g i s s l o w and e x p e n s i v e due t o t h e amount o f e x p e r t a n d k n o w l e d g e e n g i n e e r t i m e r e q u i r e d t o e x p r e s s and t e s t r u l e s . The c o s t o f A I h a r d w a r e a n d s p e c i a l A I p r o g r a m m e r s makes s m a l l a p p l i c a t i o n s p r o h i b i t i v e l y expensive. The e x p e r t systems a r e s t a n d - a l o n e programs, and i t i s d i f f i c u l t or i m p o s s i b l e t o i n t e g r a t e t h e i r reasoning w i t h e x i s t i n g s c i e n t i f i c software. Sometimes, f i n i s h e d e x p e r t systems can not be used i n the f i e l d because they are too s l o w , or r e q u i r e i n a p p r o p r i a t e l y e x p e n s i v e hardware. B e c a u s e o f t h e c u r r e n t h i g h demand f o r e x p e r t s y s t e m a p p l i c a t i o n s , s o f t w a r e packages w h i c h a r e o p t i m i z e d f o r a p p l i c a t i o n building, rather than for AI technique research, h a v e been developed. One o f t h e s e i s R u l e M a s t e r (l) which i s designed to e x t r a c t e x p e r t r e a s o n i n g and t o i n c o r p o r a t e i t i n t o a w i d e range o f s c i e n t i f i c a n d e n g i n e e r i n g a p p l i c a t i o n s . I n c o n t r a s t w i t h many o t h e r A I approaches, R u l e M a s t e r i s based on contemporary s t r u c t u r e d programming p r i n c i p l e s . C o n v e n t i o n a l m i c r o - and m i n i - c o m p u t e r s may be u s e d b y a n y c o m p u t e r p r o f e s s i o n a l t o b u i l d e x p e r t s y s t e m s i n t e g r a t e d w i t h e x i s t i n g computer programs. A knowledge a c q u i s i t i o n system based on i n d u c t i v e l e a r n i n g speeds up t h e r u l e g e n e r a t i o n and t e s t i n g process. A p r o c e d u r a l r e p r e s e n t a t i o n o f the r u l e base i s a u t o m a t i c a l l y g e n e r a t e d , p r o v i d i n g c o n s i s t e n c y and c o m p l e t e n e s s c h e c k i n g and e f f i c i e n t r u n - t i m e b e h a v i o r . Embedding e x p e r t system r e a s o n i n g i n t o e x i s t i n g systems i s s u p p o r t e d by two f e a t u r e s : a c c e s s t o e x t e r n a l u s e r programs from t h e R u l e M a s t e r r u l e l a n g u a g e , and t h e a u t o m a t i c g e n e r a t i o n o f a C c o d e r e p r e s e n t a t i o n o f t h e expert system. 9

RuleMaster D e s c r i p t i o n History. R a d i a n C o r p o r a t i o n i s a t e c h n i c a l c o n s u l t i n g company, e m p l o y i n g about 1000 p e o p l e . About h a l f o f R a d i a n ' s b u s i n e s s i s i n t h e c h e m i s t r y and c h e m i c a l e n g i n e e r i n g f i e l d s . I n 1981, Radian management r e a l i z e d t h a t e x p e r t systems c a p a b i l i t y c o u l d enhance and complement e x i s t i n g c o n s u l t i n g a c t i v i t i e s . R a d i a n e n t e r e d i n t o an agreement w i t h D o n a l d M i c h i e , o f E d i n b u r g h U n i v e r s i t y and I n t e l l i g e n t T e r m i n a l s L i m i t e d (ITL). F o r a number o f y e a r s , he had done r e s e a r c h i n i n d u c t i v e l e a r n i n g a n d i n o t h e r e x p e r t s y s t e m t e c h n i q u e s , and o f t e n used c o n v e n t i o n a l s t r u c t u r e d programming languages l i k e P a s c a l . He n o t e d t h a t t h e s p e c i a l A I environments were p r i m a r i l y u s e f u l f o r r e s e a r c h i n t o A I t e c h n i q u e s , and were n o t n e c e s s a r y f o r an e x p e r t systems package o r i e n t e d toward b u i l d i n g applications. R u l e M a s t e r was d e s i g n e d a n d d e v e l o p e d by I T L a n d R a d i a n d u r i n g 1982 and 1 9 8 3 . S i n c e t h e n , b o t h companies have c o n t i n u e d e n h a n c i n g R u l e M a s t e r , and s e v e r a l d o z e n e x p e r t s y s t e m a p p l i c a t i o n s a r e under c o n s t r u c t i o n o r c o m p l e t e d .

Pierce and Hohne; Artificial Intelligence Applications in Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1986.

20

ARTIFICIAL INTELLIGENCE APPLICATIONS IN CHEMISTRY

Downloaded by UNIV OF CALIFORNIA SANTA BARBARA on March 5, 2018 | https://pubs.acs.org Publication Date: April 30, 1986 | doi: 10.1021/bk-1986-0306.ch002

Components.

The two p r i n c i p l e components o f R u l e M a s t e r a r e :

Radial:

a procedural, block structured e x p r e s s i n g d e c i s i o n r u l e s , and

language

for

RuleMaker:

t h e knowledge a c q u i s i t i o n s y s t e m ; induces d e c i s i o n t r e e s from examples o f e x p e r t d e c i s i o n - m a k i n g , and e x p r e s s e s t h e s e d e c i s i o n s t r e e s as executable R a d i a l code,

R u l e M a s t e r e x p e r t s y s t e m s a r e r e p r e s e n t e d as R a d i a l programs. To b u i l d an e x p e r t system, domain knowledge i s n o r m a l l y e n t e r e d i n two parts: a m o d u l e s t r u c t u r e and t h e b o d i e s o f t h e m o d u l e s . The s t r u c t u r e d e f i n e s t h e h i e r a r c h i c a l o r g a n i z a t i o n o f d e c i s i o n s used t o s o l v e t h e p r o b l e m . The code w i t h i n each module d e f i n e s t h e d e t a i l s o f one o f t h e s e d e c i s i o n s . R u l e M a k e r i s a knowledge e x t r a c t i o n u t i l i t y f o r b u i l d i n g and t e s t i n g the d e c i s i o n l o g i c contained w i t h i n R a d i a l modules. The l o g i c i s s p e c i f i e d as a t a b l e o f e x a m p l e s o f c o r r e c t e x p e r t d e c i s i o n s f o r each module. R u l e M a k e r t r a n s f o r m s each example s e t i n t o an e q u i v a l e n t d e c i s i o n t r e e , and a u t o m a t i c a l l y generates t h e body o f t h e module i n t h e form o f R a d i a l code. System b u i l d e r s may a l s o choose t o e n t e r R a d i a l code d i r e c t l y , a l t h o u g h t h e y u s u a l l y p r e f e r t o work w i t h example t a b l e s . C o n s u l t a t i o n o f an e x p e r t system i s a c c o m p l i s h e d by u s i n g i t s R a d i a l code r e p r e s e n t a t i o n as i n p u t t o t h e R a d i a l i n t e r p r e t e r . The i n t e r p r e t e r f i r s t performs completeness and c o n s i s t e n c y c h e c k s , and then provides i n t e r a c t i v e run-time support. I n d u c t i v e L e a r n i n g ( R u l e M a k e r ) . Experts are best a b l e t o e x p l a i n complex concepts t o human a p p r e n t i c e s i m p l i c i t l y by u s i n g examples o f t h e e x p e r t ' s d e c i s i o n - m a k i n g , r a t h e r t h a n by e x p l i c i t l y s t a t i n g fundamental t h e o r e t i c a l p r i n c i p l e s . The a p p r e n t i c e quickly g e n e r a l i z e s t h e s e example d e c i s i o n s t o form w o r k i n g r u l e s , w h i c h he a p p l i e s when s i m i l a r s i t u a t i o n s a r e e n c o u n t e r e d . R u l e M a s t e r s knowledge a c q u i s i t i o n t o o l , R u l e M a k e r , employs a l e a r n i n g process s i m i l a r to that o f the apprentice. To t e a c h a concept t o R u l e M a k e r , t h e e x p e r t p r o v i d e s a s e t o f examples ( c a l l e d a t r a i n i n g s e t ) o f c o r r e c t d e c i s i o n s w i t h i n some c o n t e x t . Each t r a i n i n g set contains a l i s t o f the a t t r i b u t e s which are factors for d e t e r m i n i n g t h e c h o i c e o f a c t i o n . Each example c o n t a i n s a v a l u e f o r e a c h o f t h e a t t r i b u t e s , t o g e t h e r w i t h t h e s p e c i f i e d a c t i o n s t o be t a k e n when t h a t c o m b i n a t i o n o f a t t r i b u t e v a l u e s i s encountered. The R u l e M a k e r u t i l i t y c h e c k s e a c h t r a i n i n g s e t f o r c o m p l e t e n e s s and c o n s i s t e n c y , and t h e n g e n e r a t e s a p r o c e d u r a l r e p r e s e n t a t i o n o f t h e knowledge embodied i n t h e example. To i l l u s t r a t e t h i s , t h e e x a m p l e s e t o f F i g u r e 1 shows how a s i m p l e corona d e t e c t i o n d e c i s i o n ( l i k e l y , p o s s i b l e , or u n l i k e l y ) i n TOGA (Transformer O i l Gas A n a l y s i s ) might be s p e c i f i e d . TOGA i s an e x p e r t system t h a t diagnoses f a u l t s i n l a r g e e l e c t r i c a l t r a n s f o r m e r s a n d w i l l be d e s c r i b e d i n d e t a i l l a t e r i n t h i s p a p e r . The c o r o n a d e c i s i o n i s b a s e d on f o u r a t t r i b u t e s : H 2 , t h e r m a l , H 2 / C 2 H 2 , and temperature. The a t t r i b u t e " H 2 i s t h e c o n c e n t r a t i o n o f h y d r o g e n g a s ; i t may be l o w , medium, o r h i g h , a c c o r d i n g t o n u m e r i c a l r a n g e s f

I!

Pierce and Hohne; Artificial Intelligence Applications in Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1986.

Downloaded by UNIV OF CALIFORNIA SANTA BARBARA on March 5, 2018 | https://pubs.acs.org Publication Date: April 30, 1986 | doi: 10.1021/bk-1986-0306.ch002

2.

RIESE A N D STUART

A Know ledge-Engineering Facility

21

s e t by t h e e x p e r t i n a n o t h e r R a d i a l m o d u l e . " T h e r m a l " r e f e r s t o t h e r m a l l y generated hydrocarbon g a s e s , w h i c h may be a b s e n t , s l i g h t , or d e f i n i t e l y p r e s e n t . The o t h e r two a t t r i b u t e s a r e t h e h y d r o g e n t o - a c e t y l e n e r a t i o and t h e e s t i m a t e o f t h e t e m p e r a t u r e a t w h i c h t h e h y d r o c a r b o n gases were generated. A h i e r a r c h y o f r u l e s s u p p l i e d by t h e e x p e r t determines t h e v a l u e o f each o f t h e s e a t t r i b u t e s , based e v e n t u a l l y on t h e n u m e r i c a l c o n c e n t r a t i o n s r e c e i v e d from t h e gas chromatograph. The d e c i s i o n f o r each example i s e x p r e s s e d as an " a c t i o n - n e x t state" pair. The " a c t i o n " i s a r e f e r e n c e t o e x e c u t a b l e R a d i a l code, w h i c h c o n s i s t s o f a sequence o f R a d i a l s t a t e m e n t s . These s t a t e m e n t s may c o n t a i n r e f e r e n c e s t o e x t e r n a l p r o g r a m s i n v a r i o u s l a n g u a g e s ( t h i s w i l l be d i s c u s s e d f u r t h e r l a t e r ) . The "next s t a t e " d e s c r i b e s the c o n t e x t t o which c o n t r o l i s t o pass a f t e r the a c t i o n i s c o m p l e t e d . F o r d i a g n o s t i c e x p e r t s y s t e m s , s u c h as TOGA, t h e n e x t s t a t e w i l l u s u a l l y be t h e " g o a l " s t a t e o f t h e module. T h i s passes c o n t r o l back t o t h e c a l l i n g module. F o r p r o c e d u r a l e x p e r t systems, s u c h as r o b o t i c s a n d i n s t r u m e n t a t i o n c o n t r o l a p p l i c a t i o n s , t h e c o n t r o l w i l l be t r a n s f e r r e d between s e v e r a l s t a t e s w i t h i n a module t o implement l o o p i n g . The d e c i s i o n t r e e f o r t h e t r a i n i n g s e t o f F i g u r e 1, as g e n e r a t e d b y R u l e M a k e r , i s shown i n F i g u r e 2. The g e n e r a t e d t r e e a g r e e s w i t h a l l d e c i s i o n s r e p r e s e n t e d i n t h e example s e t , and g e n e r a l i z e s to reach decisions for u n s p e c i f i e d portions of the space. The r u l e i n d u c t i o n a l g o r i t h m , c a l l e d ID3 ( 2 ) , uses i n f o r m a t i o n t h e o r e t i c t e c h n i q u e s t o r e d u c e t h e number o f d e c i s i o n nodes i n t h e g e n e r a t e d t r e e . R u l e Language ( R a d i a l ) . R u l e M a s t e r e x p e r t systems a r e e x p r e s s e d i n R a d i a l , a b l o c k s t r u c t u r e d i n t e r p r e t e d language w i t h a syntax s i m i l a r t o P a s c a l a n d ADA. R a d i a l i s a s i m p l e , e a s y - t o - l e a r n language which supports the f u l l range of expert system capabilities. The b u i l d i n g b l o c k o f R a d i a l , c o r r e s p o n d i n g t o t h e P a s c a l p r o c e d u r e , i s c a l l e d a "module". The s y n t a x w i t h i n each module i s based on f i n i t e automata t h e o r y , t o p r o v i d e t h e c o n t r o l s t r u c t u r e s needed t o s u p p o r t b o t h d i a g n o s t i c and p l a n n i n g a s p e c t s o f e x p e r t systems a p p l i c a t i o n s . Other language features include recursive r o u t i n e c a l l s , argument p a s s i n g , s c o p e d v a r i a b l e and f u n c t i o n s , a b s t r a c t d a t a t y p e s , and u s e r - d e f i n e d o v e r l o a d e d o p e r a t o r s . Builti n d a t a t y p e s i n c l u d e s t r i n g , i n t e g e r , f l o a t i n g p o i n t , and b o o l e a n . The R a d i a l c o d e f o r t h e d e c i s i o n t r e e o f F i g u r e 2 i s shown i n F i g u r e 3. T h i s c o d e was g e n e r a t e d b y R u l e M a k e r . Experts have d i f f i c u l t y c o r r e c t l y g e n e r a t i n g a d e e p l y n e s t e d c o n d i t i o n a l phrase l i k e t h i s , but t h e y are a b l e t o i n s p e c t i t f o r p o s s i b l e e r r o r s or omissions. TOGA u s e s t h e b u i l t - i n n u m e r i c a l c a p a b i l i t i e s o f R a d i a l t o compute f u n c t i o n s o f c o n c e n t r a t i o n v a l u e s , w h i c h are used e x t e n s i v e l y i n the r u l e s . The r a t i o o f h y d r o g e n t o a c e t y l e n e c o n c e n t r a t i o n i n t h e corona r u l e i s a s i m p l e example o f t h i s . Userd e f i n e d compound d a t a t y p e s a r e used t o h a n d l e b l o c k s o f d a t a as a s i n g l e named s t r u c t u r e . These f e a t u r e s a r e i n v a l u a b l e i n b u i l d i n g p r a c t i c a l e x p e r t systems, but a r e not a v a i l a b l e w i t h a l l packages. Most R a d i a l code i s c o n s t r u c t e d by R u l e M a k e r from t r a i n i n g s e t s

Pierce and Hohne; Artificial Intelligence Applications in Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1986.

ARTIFICIAL INTELLIGENCE APPLICATIONS IN CHEMISTRY

22

H2 high med

Downloaded by UNIV OF CALIFORNIA SANTA BARBARA on March 5, 2018 | https://pubs.acs.org Publication Date: April 30, 1986 | doi: 10.1021/bk-1986-0306.ch002

high med high med med med low

-

thermal

H2/C2H2

absent

absent

absent present slight

action

next state GOAL) GOAL)

high high

low low

=> =>

( likely, I: l i k e l y ,

high high

moderate moderate

=> =>

< possible, ( possible,

high high

high high moderate moderate

=> => => => => =>

( ( ( ( ( (

-low

-— Figure

temperature

-—

unlikely, unlikely, unlikely, unlikely, unlikely, unlikely,

Example s e t f o r corona r u l e .

1.

unlikely

unlikely

marf unlikely

likely

likely

unlikely

( thermal )

absent possible F i g u r e 2.

Decision

possible