Exemplar-based Knowledge Acquisition

Book excerpt: Learning from past experiences is a fundamental element of intelligent behavior. An important organizing principle of such learning is that common experiences are collected into categories or cases. Despite its simple description, this principle is surprisingly difficult to implement computationally. This book describes the design, implementation, and experimental evaluation of the Protos knowledge acquisition system, a practical tool for the construction of knowledge-based programs. Protos is a case-based learning apprentice that learns by solving problems under the guidance of an expert teacher.