Formal Concept Analysis in Investigation of Normal Accidents

Document Type

Article

Publication Date

10-1-2004

Publication Title

International Journal of General Systems

DOI

10.1080/03081070412331283789

ISSN

1563-5104

Abstract

Formal concept analysis (FCA) is a theory of data analysis that identifies conceptual structures among datasets. A strong feature of FCA is its capability of producing graphical visualizations of the inherent structures among data. High-risk systems have some special characteristics that result in inevitable  accidents. This research utilizes FCA to investigate vessel accidents on the lower Mississippi river. FCA identifies the concepts that are vital indicators of a special type of accident, a collision. It is possible to analyze navigation environments and thereby gain a better understanding of these accidents. Such accidents may not be completely eliminated, though it is beneficial to identify when they were most likely to occur and take appropriate control measures. Furthermore, identification of the most dangerous environments might suggest specific safety regimens and regulatory control designs. The FCA suggests three groups of policy decisions, representing a series of increasingly dangerous situations for vessel traffic in a treacherous inland waterway. The three groups are made of: two sets of “cautionary” conditions, three sets of “serious” conditions, and one set for “prohibitive” action, respectively. The FCA does not make decisions; it clearly identifies distinct environments that require guidelines (i.e. regimens) for safe navigation to lessen or eliminate normal or system accidents.

Share

COinS