Research on a resilience structure model using the text mining technique

After the 911 tragedy, the American society began to build a mechanism that can prevent, mitigate, or recover quickly the damage caused by external shock. This has been achieved is by creating a risk management system called Business Continuity Planning(BCP) and the capability of this risk management is " resilience ". On other hand, European countries have tried to cope with extremely large-scale shock by integrating nationwide systems. Finland's seven shocks test is a pioneer project for increasing national resilience, and the project's achievement provided a set of robust portfolio action plans to the Finnish government. However, the definition of resilience is different in different documents, and components of resilience are different in each reports and research papers, even they deal with same nationwide problem, or organizational problem. In Japan, the term resilience is translated into the word "DAN RYOKU SEI " in Japanese, which could be inversely translated to" elasticity"in English; and this causes confusion when further researching resilience and its structure. In this research, we aim to subtract a common component of resilience from Japanese documents and construct a resilience structure model for the purpose of providing guidance or a clue for enhancing the resiliency of an organization. The text mining technique is used to analyze ten Japanese language documents, and we concluded that the term resilience consists of three major components( i.e., diversity, redundancy and robustness). We further analyzed each component and found that: (1) the diversity consists of scale, functions, design, and strategy of a system (or organization), (2) the redundancy consists of facility duplication, system duplication, scale and strategy of a system, and (3) the robustness consists of earthquake-proofing, facility toughening, security software, system environment, BCP capability, BCP training and risk management. According to an analysis of the text mining data, we constructed a two-layer structure model of resilience. The model can be applied or provide clue to risk management of a company, as well as resilience evaluation of a company.