Exploiting the Ordering of Observed Problem-Solving Steps for Knowledge Base Refinement: An Apprenticeship Approach

Book One

Volume

Proceedings of the AAAI Conference on Artificial Intelligence, 12

Student Abstracts

Apprenticeship is a powerful method of learning among humans in which a student refines his knowledge by observing and analyzing the problem-solving steps of an expert. In this paper we focus on knowledge base (KB) refinement for classification problems and examine how the ordering of the intermediate steps of an observed expert can be used to yield leverage in KB refinement. In the classical classification problem, the problem-solver is given an example consisting of a set of attributes and their corresponding values, and it must put the example in one of a pre-enumerated set of classes.