Rule-based versus structure-based models for explaining and generating expert behavior

ABSTRACT

Flexible representations are required in order to understand and generate expert behavior. Although production rules with quantifiers can encode experiential knowledge, they often have assumptions implicit in them, making them brittle in problem scenarios where these assumptions do not hold. Qualitative models achieve flexibility by representing the domain entities and their interrelationships explicitly. However, in problem domains where assumptions underlying such models change periodically, it is necessary to be able to synthesize and maintain qualitative models in response to the changing assumptions. In this paper we argue for a representation that contains partial model components that are synthesized into qualitative models containing entities and relationships relevant to the domain. The model components can be replaced and rearranged in response to changes in the task environment. We have found this "model constructor" to be useful in synthesizing models that explain and generate expert behavior, and have explored its ability to support decision making in the problem domain of business resource planning, where reasoning is based on models that evolve in response to changing external conditions or internal policies.

REFERENCES

• 1: Bobrow. D. Qualitative reasoning about physical systems: An introduction. Arfif. Infrll. 24, 1-3 (Dec. 1984). 1-5.
• 2: Randall Davis, Diagnostic reasoning based on structure and behavior, Artificial Intelligence, v.24 n.1-3, p.347-410, Dec. 1984 [doi>10.1016/0004-3702(84)90042-0]
• 3: Davis, R.. and King, J. An overview of production systems. In Maclriw It~fellrgr!tr-c. vol. 8. E. Elcock and D. Michie. Eds. Wiley. New York. 1977. pp. 300-332.
• 4: Johan De Kleer , John Seely Brown, A qualitative physics based on confluences, Artificial Intelligence, v.24 n.1-3, p.7-83, Dec. 1984 [doi>10.1016/0004-3702(84)90037-7]
• 5: de Kleer. 1.. Doyle. 1.. Steele. G.. and Sussman. G. AMORD: Explicit control of reasoning. In Artificial Intelligence: An MIT Perspective, vol. 1. P.H. Winston and R.H. Brown. Eds. MIT Press. Cambridge. Mass.. 1979.
• 6: Dhar. V. PLANET: An intelligent decision support system for the formulation and investigation of formal planning models. Ph.D. thesis, Graduate School of Business. Univ. df Pittsbirgh. Pittsburgh. Pa.. 1984.
• 7: Dhar. V. Using knowledge generated in heuristic search for nonchronological backtracking. /. Conrpuf. I,~tell. 2. 3 (1986).
• 8: Dhar. V.. and Quayle. C. An approach to dependency directed backtracking using domain specific knowledge. In Pmrrrdiqs of fhr 9fh Iulrrrlafiwuzl /oiuf C~~r~/wruce w! Artificial I?~fclliprcr (Los Angeles. Calif.. Aug. 18-23). lYfl5. pp. 188-191.
• 9: Doyle, ). A truth maintenance system. AI Lab. Memo 521. Massachusetts Instilute of Technology. Cambridge. 1978.
• 10: Doyle. ). Methodological simplicity in expert system construction: The case of judgements and reasoned assumptions. Tech. Rep. CMU-(:S-83.114. Dept. of Computer Science. Carnegie-Mellon Univ.. Pittsburgh. Pa.. 1983.
• 11: Duda. R.. and Shortliffe. E. Expert systems research. Srwnc? 220. 4594 (Apr. 15. 1983). 261-268.
• 12: Duda. R.O.. Hart. P.. Nilsson. N.. Reboh. R.. Slocum. 1.. and Sutherland, G. Development of a computer-based system for mineral axploration. Annu. Rep., SRI International. Oct. 1977.
• 13: Kenneth D. Forbus, Qualitative process theory, Artificial Intelligence, v.24 n.1-3, p.85-168, Dec. 1984 [doi>10.1016/0004-3702(84)90038-9]
• 14: Forgy. L. OPS5 users manual. Tech. Rep. CMU-CS-81.135. Dept. of Computer Science, Carnegie-Mellon Univ.. Pittsburgh. Pa., 1981.
• 15: Forgy. L.. and McDermott. I. OPS: A domain-independent production system. In Proceediqs of the 5fh International Conference on Arfificial If~fellrpm (Cambridge. Mass.. Aug.). 1977. pp. 933-939.
• 16: Michael R. Genesereth, The use of design descriptions in automated diagnosis, Artificial Intelligence, v.24 n.1-3, p.411-436, Dec. 1984 [doi>10.1016/0004-3702(84)90043-2]
• 17: Heckerman. D. Probabilistic interpretation for MYCIN's certainty factors. In LJucertaif!fy UI Arfifirraf Infclligewe. ).F. Lemmer and L. Kanal. Eds. North-Holland. Amsterdam. 1986. pp. 167-196.
• 18: McDermott. ). Rl: A rule-based configurer of computer systems. Arfif lute/l. 19. 1 (Sept. 1982). 39-88.
• 19: Pople. H.E. Heuristic methods for imposing structure on ill-structured problems: The structuring of medical diagnostics. In Arfifiual Infelligrncc itI Mediciw. P. Szolovits. Ed. Westview. Boulder. Cola.. 1982. pp. 119-184.
• 20: Shortliffe. E. MYCIN: Compufrr Based Medical Comultafion. Elsevier North-Holland. New York. 1976.
• 21: Simon, H. The structure of ill-structured problems. Arfif. Infell. 4. 3 (Sept. 1973). 181-201.
• 22: Stallman. R.. and Sussman. G. Forward reasoning and dependencydirected backtracking in a system for computer-aided circuit analysis. Arfif. Iuf?ll. 9. 2 (Oct. 1977). 135-196.
• 23: Mark Jeffrey Stefik, Planning with constraints, 1980

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.4 Knowledge Representation Formalisms and Methods
          Subjects: Representations (procedural and rule-based)

Additional Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE

General Terms:
Theory