Development and validation of a cognitive model of human knowledge system: Toward an effective adaptation to differences in cognitive skills

Nong Ye, Purdue University

Abstract

Knowledge processing has a key effect on human performance in cognitive tasks. Based on previous studies in cognitive science and artificial intelligence, an integrated cognitive model of human knowledge system has been developed and validated. This cognitive model covers three dimensions of human knowledge system, namely knowledge structure, knowledge content and control strategy. It is suggested that human knowledge is structured as a schema-based semantic network, and that human knowledge content is organized in a five-level abstraction hierarchy. In such schema-based semantic network which spreads over the five levels of abstraction hierarchy of knowledge content, three types of control strategies, namely simple search strategy, schema-driven processing and goal-directed processing, are used to search and access knowledge for human's cognitive tasks. From the cognitive model, three hypotheses about skill differences due to human knowledge structure, knowledge content and control strategy are derived. These three hypotheses have been tested in three separate experiments. All three experiments utilized the same group of ten expert and ten novice programmers. In experiment one, subjects were required to provide pairwise relevance ratings of 23 concepts in C programming language and 21 concepts in the UNIX operating system so that their knowledge structure could be evaluated. In order to evaluate the knowledge content in experiment two, subjects were asked to complete a multiple choice test for the five levels of software development knowledge. In experiment three, subjects were required to understand three C programs. Their program understanding process was analyzed in order to derive their control strategies. The three hypotheses were supported since significant differences were present in knowledge content and control strategies between experts and novices but not in their knowledge structure. The developed cognitive model of human knowledge system has important implications in man-machine system design, knowledge system design, personal training and job design. The information gained on skill differences between experts and novices in human knowledge system can be applied to improve productivity and job satisfaction of cognitive tasks.

Degree

Ph.D.

Advisors

Salvendy, Purdue University.

Subject Area

Industrial engineering

COinS