Instance-based cognitive mapping: a process for discovering a knowledge worker's tacit mental model

This paper addresses tacit–to–explicit knowledge externalization, arguably the most critical, and yet problematic, phase of Nonaka's knowledge creation theory. Specifically, we propose and describe instance–based cognitive mapping (ICM), a unique externalization process that analyzes multiple decision instances using the inductive learning algorithms of artificial intelligence to generate a polynomial representation of the knowledge worker's mental model, explicitly relating how the knowledge worker implicitly selects and weighs key factors in making decisions within a specific problem domain. After reviewing current externalization techniques, we describe the characteristics, and evaluate the advantages, of the ICM process. An exploratory test of the process suggests that inductive learning algorithms, such as the group method of data handling, can be used to discover a reasonable polynomial estimate of a knowledge worker's tacit mental model. This estimate can then be compared with other explicit models and standards, updated with new information and knowledge, and internalized by all interested knowledge workers.