Abstract
While pure case based classification systems can take little advantage of heuristic expert knowledge, heuristic classification systems have much difficulty to adapt themselves with case knowledge. There are three integration strategies: Combine two independent problem solvers, learn rules from cases, or incorporate heuristic knowledge into the case based approach. Pursuing the last alternative, we present a case comparison language (CCL) designed for incremental refinement with knowledge from domain experts. Nevertheless CCL needs considerably less knowledge than is usually necessary for probabilistic rules in heuristic classification. The key features of CCL are combining static and dynamic weighting of symptoms, providing several schemes for computation of partial similarities between symptoms, and a categorical rule language with abstraction layers for inferring the domain terminology from the raw symptoms. Its simplicity, adaptability, and explainability, its strong dependence on domain terminology, its high potential for parallelism, and its incremental approach to knowledge acquisition seem to make it a candidate for a model of human classification problem solving in experience-rich domains. CCL is implemented as a component named CCC+ of the diagnostic expert system shell D3.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Althoff, K.-D., de la Ossa, A., Maurer, F., Stadler, M. and WeB, S.: Case based Reasoning for Real World Applications, Internal Report, Kaiserslautern University, 1990.
Ashley, K. and Rissland, E.: Waiting on Weighting: A Symbolic Least Commitment Approach, AAAI-88, 239–244, 1988.
Bamberger, S., Gappa, U., Goos, K., Meinl, A., Poeck, K. and Puppe, F.: The Diagnostic Expert System Shell D3 (in German), Handbook, Version 1.0, Karlsruhe University, 1991.
Clancey, W.: Heuristic Classification, AI-Journal 27, 289–350, 1985.
Hammond, K.: Case-Based-Planning: Viewing Planning as a Memory Task, Academic Press, 1989.
Hestermann, C.: CCC: Design and Implementation of an Expert System Component for Case Comparison and its Use for Similarity Searching of Prismatic Workpieces (in German), Diploma Thesis, Karlsruhe University, 1990.
Kolodner, J.: Capitalizing on Failure through Case-Based-Inference, in: Proc. of the Conference on the Cognitive Science Society, 1987.
Koton, P.: Using Experience in Learning and Problem Solving, PhD-Thesis, Department of Electrical Engineering and Computer Science, MIT, 1988.
Mooney, R., Shavlik, J., Towell, G., and Gove, A.: An Experimental Comparison of Symbolic and Connectionist Learning Algorithms, ICJAI-89, 775–780, 1989.
Miller, R., Pople, H., and Myers, J.: INTERNIST1 , an Experimental Computer-Based Diagnostic Consultant for General Internal Medicine, New England Journal of Medicine 307, 468–476, 1982.
Patil, R.: A Case Study on Evolution of System Building Expertise: Medical Diagnosis, in Grimson, E. and Patil, R. (eds.): AI in the 1980s and Beyond, MIT Press, 75–108, 1987.
Porter, B., Bareiss, R., and Holte, R.: Concept Learning and Heuristic Classification in Weak-Theory Domains, AI-Journal 45, 229–263, 1990.
Puppe, F.: Requirements of a Classification Shell and Their Realization in MED2, Applied Artificial Intelligence 1, 163–171, 1987.
Puppe, F.: Problem Solving Methods in Expert Systems (in German, translation to English in preparation), Springer, 1990.
Puppe, F., Legleitner, T., and Huber, K.: DAX I MED2 - A Diagnostic Expert System for Quality Assurance of an Automatic Transmission Control Unit, to appear in Zarri, G. (ed.): Operational Expert Systems in Europe, Pergamon Press, 1991.
Rissland, E. and Skalak, D.: Combining Case-Based and Rule-Based Reasoning: A Heuristic Approach, IJCAI-89, 524–530, 1989.
Stanfill, G. and Waltz, D.: Toward Memory-Based Reasoning, CACM 29, 1213–1228, 1986.
Weiss, S. and Kapoulas, I.: An Empirical Comparison of Pattern Recognition, Neural Nets and Machine Learning Classification Methods, UCAI-89, 781–787, 1989.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1991 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Puppe, F., Goos, K. (1991). Improving Case Based Classification with Expert Knowledge. In: Christaller, T. (eds) GWAI-91 15. Fachtagung für Künstliche Intelligenz. Informatik-Fachberichte, vol 285. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-662-02711-0_21
Download citation
DOI: https://doi.org/10.1007/978-3-662-02711-0_21
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-54558-3
Online ISBN: 978-3-662-02711-0
eBook Packages: Springer Book Archive