In this paper we give a critical view on current rule oriented expert systems and show that there are arguments against the rule oriented strategy from the practical as well as the theoretical point of view. Instead of this strategy we propose a formalization of analogical reasoning, which leads for second order expert systems to a case oriented strategy. In some areas of artificial intelligence certain approaches to the formalization of analogy exist. But a unified mathematical theory of analogy is needed. After a short description of a possible general mathematical approach to analogical reasoning we will describe some applications and implications of this new method. The potential applications for analogical reasoning have one difficulty in common: the used knowledge is to represent in a structural way for detecting similarities and analogies on a certain level.
CITATION STYLE
Poetschke, D. (1989). Analogical reasoning for second generation expert systems. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 397 LNAI, pp. 264–276). Springer Verlag. https://doi.org/10.1007/3-540-51734-0_67
Mendeley helps you to discover research relevant for your work.