Aims and objectives
Case-Based Reasoning (CBR) is a technique for solving new problems by analogy with old ones. Problem solving episodes are stored in memory as cases and later retrieved in response to a new problem situation. The best match is then adapted to fit this situation, and the cycle is repeated.
The aim of this course is to introduce students to the key concepts in this fast growing field - in particular the representation, retrieval, adaptation, repair and storage of cases. These issues will be illustrated with reference to specific case studies (CHEF, JULIA, CASEY).
Textbook
Janet Kolodner: Case-Based Reasoning, Morgan Kaufmann 1993.
Syllabus
Introduction:
What is Case-Based Reasoning? CBR as a methodology for building knowledge-based systems, and as a cognitive model. Case-based versus traditional rule-based systems.
Case Representation:
The role of cases. Individuals and generalized episodes. Case libraries. The dynamic memory model. The category & exemplar model. Indexing and storage. Case acquisition.
Retrieval:
Index-based retrieval methods. Reminding. Selecting the best case - similarity metrics.
Adaptation:
Structural and derivational adaptation techniques. Object critics. Evaluating the proposed solution. Explanation and repair.