The knowledge acquisition activity matrix: a systems engineering conceptual framework

An overview is presented of the knowledge acquisition challenges posed by operational artificial intelligence (AI) system development, and limitations in current knowledge acquisition approaches are identified. The authors present a systems engineering conceptual framework that views knowledge acquisition as consisting of unique knowledge acquisition steps in such system engineering phases as problem definition, requirements analysis, functional specification, system design, system development, test and evaluation, and system maintenance. The proposed conceptual framework presents a systematic and structured approach for the design, at the beginning of a knowledge-based system development project, of knowledge engineering activities. This framework allows the AI system designer to scope and prescribe knowledge acquisition activities more efficiently and realistically than many current ad hoc methods. The focus is on the goals, constraints, and results of knowledge acquisition activities as they pertain to system development phases. >