A systematic method to build a knowledge base to be used in a human reliability analysis model

Abstract

A human’s knowledge base is a key component for the development of a mechanistic model of human response to be used for human reliability analysis. This paper proposes a new method for constructing this knowledge base. The proposed method is comprised of three steps: (1) systematic literature review, which is used to collect data pertinent to the subject under study; (2) summarization, the goal of which is to extract key points that are expressed in the literature; (3) qualitative coding, a process in which codes closely related to the topic are derived and the relationships between these codes are expressed. As a case study, the proposed method is being applied to construct an operator’s knowledge base concerning severe accident phenomenology in a nuclear power plant. Part of this application is explored in this paper. With the proposed method and the resulting knowledge base, it is expected that an individual’s response when presented with a specific context can be modeled in more detail.