Meta-Cases: Explaining Case-Based Reasoning

AI research on case-based reasoning has led to the development of many laboratory case-based systems. As we move towards introducing these systems into work environments, explaining the processes of case-based reasoning is becoming an increasingly important issue. In this paper we describe the notion of a meta-case for illustrating, explaining and justifying case-based reasoning. A meta-case contains a trace of the processing in a problem-solving episode, and provides an explanation of the problem-solving decisions and a (partial) justiication for the solution. The language for representing the problem-solving trace depends on the model of problem solving. We describe a task-method-knowledge (TMK) model of problem-solving and describe the representation of meta-cases in the TMK language. We illustrate this explanatory scheme with examples from Interactive Kritik, a computer-based design and learning environment presently under development. One goal of AI research on case-based reasoning is to develop theories for designing useful and usable interactive case-based environments. In an interactive case-based environment, a human may acquire knowledge by navigating and browsing a case library, address a problem in cooperation with a case-based system , or learn about problem solving by observing the problem solving in the case-based system. The goal of designing case-based interactive systems that are both useful and usable raises the issue of explaining the reasoning of the case-based system. This issue is especially important in moving laboratory case-based systems into real work environments. Explanation of reasoning is a recurrent theme in AI research. Consider, for example, the history of AI research on knowledge systems. Starting with MYCIN (Shortliie 1976), which probably was the rst useful and usable knowledge system, explanation became an increasingly important issue. In the context of MYCIN, for example, AI researchers rst built an explanatory interface called GUIDON for tutoring medical students (Clancey 1987). Explanations in GUIDON initially were expressed in the language of goals, production rules, and