Modeling Individual Differences and Stressors Using the SAMPLE Cognitive Architecture

摘要

Building on previous work by Hudlicka on the Methodology for Analysis and Modeling of Individual Differences (MAMID), we have developed a generic architecture for modeling the effects of behavior moderators on human decision-making performance. Our approach to Modeling Individual Differences and Stressors (MINDS) employs the generic, parameterized approach of MAMID while applying it to the SAMPLE cognitive architecture. In doing so, we hope to make the behavior moderating capabilities of MAMID accessible to the wider user community of SAMPLE, as well as explore alternative methods of modeling combined moderator effects. Additionally, SAMPLE's cognitive architecture supports multitasking, which provides another opportunity to apply the effects of behavior moderators to behavior models. In this paper we discuss the similarities and differences between the MAMID and SAMPLE cognitive architectures as well as our approach to modeling combined moderator effects in SAMPLE's three principal computational technologies: fuzzy logic, belief networks and rule-based expert systems.