Al and Similarity.

The article discusses topics on similarity-driven reasoning, analogy, learning and explanation that are important in the realization of a truly robust artificial intelligence. Case-based reasoning (CBR) is a process of solving new problems in reference to old cases with its two types, the interpretive CBR and the problem-solving CBR. Further understanding of similarity based-reasoning includes methods for knowledge representation and techniques in dealing with concepts, contexts and purpose. Intelligent systems require reasoning with explanation that is important in learning, decision-making and giving advice. Several concepts such as messy, open-textured, nonstationary and nonconvex are discussed. INSET: Al's Abiding Interest in Messy Concepts..