Applications of nonmonotonic logic to diagnosis | The Knowledge Engineering Review | Cambridge Core (original) (raw)
Abstract
This paper attempts to assess the practical utility of nonmonotonic logic in diagnostic problem solving. We begin with a brief review of the main assumptions which motivate work in this area, and discuss two logic-based approaches which involve nonmonotonic arguments. Then we consider two recent proposals for the application of default logic to diagnosis, as well as a proposal based on counterfactual logic. In conclusion, we briefly compare these methods with other diagnostic reasoning paradigms found in the Artificial Intelligence literature.
Information
Type
Research Article
Copyright
Copyright © Cambridge University Press 1989
Access options
Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)
Article purchase
Temporarily unavailable
References
Buchanan, BG and Shortliffe, EH eds, 1984. Rule-based expert systems, Reading, Massachusetts: Addison-Wesley [A collection of seminal papers on expert systems by researchers at Stanford University]Google Scholar
Bylander, T, Allemang, D, Tanner, MC and Josephson, JR, 1989. Some results concerning the computational complexity of abduction” In: Proceedings of the 1st International Conference on Principles of Knowledge Representation and Reasoning, pp 44–54, Morgan Kaufmann [A useful and clearly presented paper that outlines the conditions under which abductive inference is tractable]Google Scholar
Clancey, WJ, 1985. “Heuristic classification” Artificial Intelligence, 27 289–350 [A theoretical paper that attempts to reconstruct the problem solving paradigm employed by many rule-based expert systems]CrossRefGoogle Scholar
Console, L, Dupré, DT and Torasso, P, in press, “A theory of diagnosis for incomplete causal models” To appear in: Proceedings of the 11th International Joint Conference on Artificial Intelligence,Detroit Michigan,August 1989 [A formal treatment of causal reasoning with incomplete knowledge that makes interesting connections with nonmonotonic logic]Google Scholar
Davis, R, 1984. “Diagnostic reasoning based on structure and behavior” Artificial Intelligence 24 347–410 [An influential paper on electronic troubleshooting from first principles. It is not reviewed here because it makes no explicit connections with nonmonotonic logic. It is well worth reading, nonetheless]CrossRefGoogle Scholar
de Kleer, J, 1986. “An assumption-based TMS” Artificial Intelliegence 28 127–162 [The main paper on ATMS. It is neither clear not concise, but it is important]CrossRefGoogle Scholar
de Kleer, J and Williams, BC, 1987. “Diagnosing multiple faults” Artificial Intelligence 32 97–130 [A difficult but important paper that repays study. If you read the version in Ginsberg (1987), note that the columns of text on page 382 are the wrong order!]CrossRefGoogle Scholar
Eshelman, L, 1988. “MOLE: A knowledge acquisition tool for cover-and-differentiate systems” Chapter 3 of Marcus, S ed., Automating knowledge acquisition for expert systems, Boston, Massachusetts: Kluwer Academic [Describes a knowledge acquisition tool for systems that use a form of heuristic classification. MOLE is interesting because it reasons explicitly about the space of possible explanations]Google Scholar
Genesereth, MR, 1984. “The use of design descriptions in automated diagnosis” Artificial Intelligence 24 411–436 [Describes one of the first logic-based diagnosis programs]CrossRefGoogle Scholar
Ginsberg, M, 1986. “Counterfactuals” Artificial Intelligence 30 35–80 [A wide-ranging treatment of counterfactuals and their relevance to AI applications, such as diagnosis and planning]CrossRefGoogle Scholar
Ginsberg, M, 1987. Readings in nonmonotonic reasoning, Los Altos, California: Morgan Kaufmann [A good collection of important papers on nonmonotonic logic at a price you can afford]Google Scholar
Jackson, P, 1989. “Prepositional abductive logic” In: Proceedings of the 7th Conference on Artificial Intelligence and the Simulation of Behaviour, pp 89–94, London:Pitman [An attempt to provide a proof theory and semantics for abductive inference]Google Scholar
Kahn, G, 1988. “MORE: From observing knowledge engineers to automating knowledge acquisition” Chapter 2 of Marcus, S, ed., Automating knowledge acquisition for expert systems, Boston Masschusetts: Kluwer Academic [An attempt to automate the acquisition of diagnostic knowledge using causal models]Google Scholar
Laskey, K and Lehner, PE, 1988. “Belief maintenance: An integrated approach to uncertainty management” In: Proceedings of the 7th National Conference on Artificial Intelligence, pp. 210–214, American Association for Artificial Intelligence [A convincing theoretical account of the combination of ATMS and Dempster–Shafer theory]Google Scholar
McDermott, D, 1987. “A critique of pure reason” Computational Intelligence 3 151–160 [McDermott's now (in)famous attack on logic-based problem solving: still food for thought]CrossRefGoogle Scholar
Overbeek, R and Lusk, E, 1984. “The automated reasoning system ITP—user's manual” Technical Report ANL-84–27, Argonne National Laboratory [The theorem prover upon which Smith's implementation of Reiter's theory was based, chosen partly for its effective treatment of equational theories]Google Scholar
Pearl, J, 188, Probabilistic reasoning in intelligent systems: Networks of plausible inference, Los Altos, California: Morgan Kaufmann [An extended account of Bayesian belief updating, including its relation to nonmonotonic logic]Google Scholar
Peng, Y and Reggia, JA, 1986. “Plausibility of diagnostic hypotheses: The nature of simplicity” In: Proceedings of the 6th National Conference on Artificial Intelligence,140–145, American Association for Artificial Intelligence [Describes the incorporation of probabilistic reasoning into the Generalized Set Covering model of diagnosis]Google Scholar
Poole, D, 1988a. “A logical framework for default reasoning” Artificial Intelligence 36 27–47 [Proposes an account of nonmonotonic reasoning in terms of scenarios, and relates it to default logic. This forms the basís of the Theorist framework]CrossRefGoogle Scholar
Poole, D, 1988b. “Representing knowledge for logic-based diagnosis” In: Proceedings of the International Conference on Fifth Generation Computer Systems,Tokyo, Japan [Argues that the Theorist framework can be used effectively to compare abduction, diagnosis from first principles, and rule-based diagnosis]Google Scholar
Poole, D, Goebel, R and Aleliunas, R, 1987. “Theorist: A logical reasoning system for defaults and diagnosis” Chapter 13 of Cercone, N and McCalla, G, eds, The knowledge frontier, New York: Springer-Verlag [A somewhat sketchy account of the Theorist framework]Google Scholar
Pople, HE Jr, 1977. “The formation of composite hypotheses in diagnostic problem solving: An exercise in synthetic reasoning” In: Proceedings of the 5th International Joint Conference on Artificial Intelligence, pp 1030–1037, American Association for Artificial Intelligence [An early paper on abductive inference in AI; still worth reading today]Google Scholar
Provan, GM, 1988. “Solving diagnostic problems using extended truth maintenance systems” In: Proceedings of the 8th European Conference on Artificial Intelligence, pp 547–552, London:Pitman [An excellent paper that addresses the difficult problem of integrating truth maintenance and Dempster–Shafer belief functions. A more computational treatment than Laskey and Lehner]Google Scholar
Reiter, R, 1980. “A logic for default reasoning” Artificial Intelligence 31 81–132 [A thorough and well-written account of default logic that has required little or no revision over the years]CrossRefGoogle Scholar
Reiter, R, 1987a. “Nonmonotonic reasoning” Annual Reviews of Computer Science 2 147–186 [A useful overview of nonmonotonic logic]Google Scholar
Reiter, R, 1987b. “A theory of diagnosis from first principles” Artificial Intelligence 32 57–95 [A clear theoretical account of diagnoses based on the minimization of abnormality in a system description]CrossRefGoogle Scholar
Shoham, Y, 1988. Reasoning about change: Time and causation from the standpoint of artificial intelligence, Cambridge, Massachusetts: MIT Press [An attempt to integrate nonmonotonic logic and temporal logic for reasoning about change]Google Scholar
Singh, N, 1987. An artificial intelligence approach to test generation, Norwell, Massachusetts: Kluwer Academic [Describes some further work on the DART system]CrossRefGoogle Scholar
Smith, BA, 1988. “A system for the diagnosis of faults using a first principles approach” PhD thesis, Department of Computer Science, University of Missouri-Rolla [An interesting account of an implementation of Reiter's theory of diagnosis. Also contains excellent review chapters]Google Scholar
Swartout, WR, 1983. “XPLAIN: a system for creating and explaining expert consulting programs” Artificial Intelligence 21 285–325 [Describes the derivation of an expert system from a domain model by automatic programming]Google Scholar
Winslett, M, 1988. “Reasoning about action using a possible models approach” In: Proceedings of the 7th National Conference on Artificial Intelligence, pp. 89–93, American Association for Artificial Intelligence [Contains a convincing critique of Ginsberg's construction for counterfactual reasoning]Google Scholar