Abstract
Expertise consists of rapid selection and application of compiled experience. Robust reasoning, however, requires adaptation to new contingencies and intelligent modification of past experience. And novel or creative reasoning, by its real nature, necessitates general problem-solving abilities unconstrained by past behavior. This article presents a comprehensive computational model of analogical (case-based) reasoning that transitions smoothly between case replay, case adaptation, and general problem solving, exploiting and modifying past experience when available and resorting to general problem-solving methods when required. Learning occurs by accumulation of new cases, especially in situations that required extensive problem solving, and by tuning the indexing structure of the memory model to retrieve progressively more appropriate cases. The derivational replay mechanism is discussed in some detail, and extensive results of the first full implementation are presented. These results show up to a large performance improvement in a simple transportation domain for structurally similar problems, and smaller improvements when less strict similarity metrics are used for problems that share partial structure in a process-job planning domain and in an extended version of the strips robot domain.
Article PDF
Similar content being viewed by others
References
Allen, J., amp; Langley, P. (1990). Integrating memory and search in planning. In Proceedings of theDARPA Workshop on Innovative Approaches to Planning, Scheduling, and Control (pp. 301–312). SanDiego: CA: Morgan Kaufmann.
Bareiss, R., amp; King, J.A. (1989). Similarity assessment in case-based reasoning. In Proceedings of theSecond Workshop on Case-Based Reasoning (pp. 67–71). Pensacola, FL: Morgan Kaufmann.
Carbonell, J.G. (1983). Learning by analogy: Formulating and generalizing plans from past experience. InR.S. Michalski, J.G. Carbonell, amp; T.M. Mitchell (Eds.), Machine learning, an artificial intelligenceapproach (Vol. 1), pp. 137–162. Palo Alto, CA: Tioga Press.
Carbonell, J.G. (1986). Derivational analogy: A theory of reconstructive problem solving and expertiseacquisition. In R.S. Michalski, J.G. Carbonell, amp; T.M. Mitchell (Eds.), Machine learning, anartificial intelligence approach (Vol. 2), pp. 371–392. San Mateo, CA: Morgan Kaufmann.
Carbonell, J.G., amp; Gil, Y. (1990). Learning by experimentation: The operator refinement method. InR.S. Michalski amp; Y. Kodratoff (Eds.), Machine learning: An artificial intelligence approach, (Vol.3), pp. 191–213. Palo Alto, CA: Morgan Kaufmann.
Carbonell, J.G., amp; Veloso, M.M. (1988). Integrating derivational analogy into a general problem solvingarchitecture. In Proceedings of the First Workshop on Case-Based Reasoning (pp. 104–124).Tampa, FL: Morgan Kaufmann.
Cheng, P.W., amp; Carbonell, J.G. (1986). Inducing iterative rules from experience: The FERMI system. In Proceedings of AAAI-86 (pp. 490–495). Philadelphia, PA.
DeJong, G.F., amp; Mooney, R. (1986). Explanation-based learning: An alternative view. MachineLearning, 1 (2), 145–176.
Doyle, J. (1984). Expert systems without computers. AI Magazine, 5 (2), 59–63.
Drummond, M., amp; Currie, K. (1989). Goal ordering in partially ordered plans. In Proceedings ofIJCAI-89 (pp. 960–965). Detroit, MI.
Etzioni, O. (1990a). A structural theory of explanation-based learning (Technical Report CMU-CS-90-185). Ph.D. thesis, School ofComputer Science, Carnegie Mellon University, Pittsburgh, PA.
Etzioni, O. (1990b). Why Prodigy/EBL works. In Proceedings of AAAI-90 (pp. 916–922).Boston, MA.
Fikes, R.E., amp; Nilsson, N.J. (1971). Strips: A new approach to the application of theorem proving to problemsolving. Artificial Intelligence, 2, 189–208.
Hammond, K. (1986). Case-based planning: An integrated theory of planning, learning and memory. Ph.D.thesis, Department of Computer Science, Yale University, New Haven, CT.
Harandi, M.T., amp; Bhansali, S. (1989). Program derivation using analogy. In Proceedings of IJCAI-89 (pp. 389–394). Detroit, MI.
Hickman, A.K., amp; Larkin, J.H. (1990). Internal analogy: A model of transfer within problems. In The 12th Annual Conference of TheCognitive Science Society (pp. 53–60). Hillsdale, NJ: LawrenceErlbaum Associates.
Hickman, A.K., Shell, P., amp; Carbonell, J.G. (1990). Internal analogy: Reducing search during problemsolving. In C. Copetas (Ed.), The Computer Science Research Review 1990. The School ofComputer Science, Carnegie Mellon University, Pittsburgh, PA.
Joseph, R.L. (1989). Graphical knowledge acquisition. In Proceedings of the 4th Knowledge AcquisitionFor Knowledge-Based Systems Workshop, Banff, Canada.
Kambhampati, S. (1989). Flexible reuse and modification in hierarchical planning: A validation structurebased approach. Ph.D. thesis, Computer Vision Laboratory, Center for Automation Research, University ofMaryland, College Park, MD.
Kedar-Cabelli, S. (1985). Purpose-directed analogy. In Proceedings of the Seventh Annual Conference ofthe Cognitive Science Society (pp. 150–159).
Knoblock, C.A. (1991). Automatically generating abstractions for problem solving. (Technical ReportCMU-CS-91-120). Ph.D. thesis, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA.
Kolodner, J. (1989). Judging which is the “best” case for a case-based reasoner. In Proceedings ofthe Second Workshop on Case-Based Reasoning (pp. 77–81). Pensacola, FL: Morgan Kaufmann.
Kolodner, J.L. (1984). Retrieval and organization strategies in conceptual memory. Hillsdale, NJ: Lawrence Erlbaum Associates.
Korf, R.E. (1985). Macro-operators: A weak method for learning. Artificial Intelligence, 26,35–77.
Laird, J.E., Rosenbloom, P.S., amp; Newell, A. (1986). Chunking in soar: The anatomy of a general learningmechanism. Machine Learning, 1, 1–46.
Minton, S. (1985). Selectively generalizing plans for problem solving. In Proceedings of AAAI-85 (pp.596–599).
Minton, S. (1988). Learning effective search control knowledge: An explanation-based approach.(Technical Report CMU-CS-88-133). Ph.D. thesis, Computer Science Department, Carnegie MellonUniversity, Pittsburgh, PA.
Minton, A., Knoblock, C.A., Kuokka, D.R., Gil, Y., Joseph, R.L., amp; Carbonell, J.G. (1989). {tiprodigy 2.0: The manual and tutorial}(Technical Report CMU-CS-89-146). School of ComputerScience, Carnegie Mellon University, Pittsburgh, PA.
Mitchell, T.M., Keller, R.M., amp; Kedar-Cabelli, S.T. (1986). Explanation-based generalization: A unifyingview. Machine Learning, 1 (1), 47–80.
Mitchell, T.M., Utgoff, P.E., amp; Banerji, R. (1983). Learning by experimentation: Acquiring and refiningproblemsolving heuristics. In R.S. Michalski, J.G., J.G. Carbonell, amp; T.M. Mitchell (Eds.), Machine learning, an artificial intelligence approach (pp. 163–190), Palo Alto, CA: Tioga Press.
Newell, A. (1980). Physical symbol systems. Cognitive Science, 4 (2), 135–184.
Pazzani, M. (1990). Creating a memory of causal relationships: An integration of empirical and explanation-basedlearning methods. Hillsdale, NJ: Lawrence Erlbaum Associates.
Polya, G. (1945). How to solve it. Princeton, NJ: Princeton University Press.
Porter, B., Bareiss, R., amp; Holte, R. (1989). Knowledge acquisition and heuristic classification in weak-theorydomains (Technical Report AI-TR-88-96). Department of Computer Science, University of Texasat Austin.
Riesbeck, C.K., amp; Schank, R.C. (1989). Inside case-based reasoning. Hillsdale, NJ: LawrenceErlbaum Associates.
Sacerdoti, E.D. (1975). The nonlinear nature of plans. In Proceedings of IJCAl-75 (pp. 206–213). Tbilisi, USSR.
Schank, R.C. (1982). Dynamic memory. Cambridge: Cambridge University Press.
Schank, R.C. (1983). The current state of AI: One man's opinion. Artificial Intelligence Magazine, 4 (1), 1–8.
Shell, P., amp; Carbonell, J.G. (1989). FRuleKit: A frame-based production system. User's manual. Internalreport, Department of Computer Science, Carnegie Mellon University, Pittsburgh, PA.
Simpson, R.L. (1985). A computer model of case-based reasoning in problem solving: An investigation in thedomain of dispute mediation. Ph.D. thesis, School of Information and Computer Science, Georgia Institute ofTechnology, Atlanta, GA.
Sussman, G.J. (1973). A computational model of skill acquisition (Technical Report AI-TR-297).Artificial Intelligence Laboratory, MIT, Cambridge, MA.
Sycara, E.P. (1987). Resolving adversarial conflicts: An approach to integrating case-based and analyticmethods. Ph.D. thesis, School of Information and Computer Science, Georgia Institute of Technology, Atlanta,GA.
Veloso, M.M. (1989). Nonlinear problem solving using intelligent casual-commitment (Technical ReportCMU-CS-89-210). School of Computer Science, Carnegie Mellon University, Pittsburgh, PA.
Veloso, M.M. (1991). Replaying multiple cases in analogical problem solving (Technical Report forthcoming).School of Computer Science, Carnegie Mellon University, Pittsburgh, PA.
Veloso, M.M., amp; Carbonell, J.G. (1989). Learning analogies by analogy—The closed loop of memoryorganization and problem solving. In Proceedings of the Second Workshop on Case-Based Reasoning (pp.153–158). Pensacola, FL: Morgan Kaufmann.
Veloso, M.M., amp; Carbonell, J.G. (1990). Integrating analogy into a general problem-solving architecture. InM. Zemankova amp; Z. Ras (Eds.), Intelligent systems (pp. 29–51). Chicester, England: Ellis Horwood.
Veloso, M.M., amp; Carbonell, J.G. (1991a). Learning by analogical replay In prodigy: First results. In Proceedings of the European Working Session on Learning (pp. 375–390). Porto, Portugal: Springer-Verlag.
Veloso, M.M., amp; Carbonell, J.G. (1991b). Variable-precision case retrieval in analogical problem solving. In Proceedings of the 1991 DARPA Workshop on Case-Based Reasoning (pp. 93–106). Washington,DC: Morgan Kaufmann.
Veloso, M.M., Pérez, M.A., amp; Carbonell, J.G. (1990). Nonlinear planning with parallel resource allocation.In Proceedings of the DARPA Workshop on Innovative Approaches to Planning, Scheduling, and Control (pp.207–212). San Diego, CA: Morgan Kaufmann.
Waldinger, R. (1981). Achieving several goals simultaneously. In B.L. Webber amp; N.J. Nilsson (Eds.),Readings in Artificial Intelligence (pp. 250–271). Los Altos, CA: Morgan Kaufmann.
Wilkins, D.E. (1989). Can AI planners solve practical problems? (Technical Note 468R). SRIInternational. Stanford, CA.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Veloso, M.M., Carbonell, J.G. Derivational Analogy in PRODIGY: Automating Case Acquisition, Storage, and Utilization. Machine Learning 10, 249–278 (1993). https://doi.org/10.1023/A:1022686910523
Issue Date:
DOI: https://doi.org/10.1023/A:1022686910523