A Similarity Metric for Retrieval of Cases
Imperfectly Explained *

Carlos Bento and Ernesto Costa

Laboratrio de Informtica e Sistemas - Univ. de Coimbra
Quinta da Boavista, lote 1, 1
3000 Coimbra - PORTUGAL
E-mail: bento@alma.uc.pt ernesto@moebius.uc.pt



Abstract. This paper describes a new quantitative similarity metric for
cases comprising an episode represented by a problem/solution pair and
an interpretation in the form of a set of imperfect explanations.
Our explanation-based metric is an alternative to other approaches that
use cases explanations for selection and retrieval. We analyse this metric
and present some examples.
We describe CLASH, a Case-Based Reasoning Expert System imple-
mented in PROLOG, which applies this metric. The results provided by
CLASH are compared with another system based on a different explanation-based 
metric for case retrieval.
References

1.	Cain, Timothy, Pazzani, M. 3. and Silverstein, Glenn, Using Domain Knowledge to
Influence Similarity Judgments, in Proceedings of a Case-Based Reasoning Work-
shop, Morgan-Kaufmann, 1991.
2.	Duda, R., Hart, P., Pattern Classification and Scene Analysis, New York: Wiley,
1973.
3.	Kolodner, J., Retrieval and Organizational Strategies in Conceptual Memory: A
Computer Model, Hillsdale, NJ.: Lawrence Eribaum Associates, 1984.
4.	Koton, Phyllis, Using Experience in Learning and Problem Solving, Massachusets
Institute of Technology, Laboratory of Computer Science (PhD diss., October 1988),
MIT/LCS/TR-441, 1989.
5.	Lebowitz, M., Concept Learning in an Rich Input Domain: Generalization-Based
Memory, in Michalski, R., Carbonell, 3., and Mitchell T. (Ed.), Machine Learning,
Vol. 2, Los Altos, Ca.: Morgan Kaufmann Publishers, 1986.
6.	Riesbeck, C., and R. Schank, Inside Case-Based Reasoning, Lawrence Erlbaum Associates, 
Hillsdale, NJ, 1989.
7.	Veloso, M., Learning by Analogical Reasoning in General Problem Solving, Ph D
thesis, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA,
August 1992.
