How can the human dimensions of disaster impacts be more accurately captured and represented in the analysis, modeling and simulation of disasters?
In order to improve the disaster resilience of society; residents must possess and improve the three powers of prediction; prevention; and response individually. In particular; knowledge of science and engineering is necessary for prediction and prevention; while for response; we must learn lessons from past disasters and learn appropriate response methods from successful cases; from the aspects of social informatics. In these ten years; much information on disaster prevention and mitigation has been shared on the web. Natural language processing using AI is common for analyzing this information. However; the importance of information and wisdom to be known differ depending on a person's position; role and the environment. There is no correct answer for disaster response; and it is difficult to solve problems with centralized AI technology alone. Therefore; I thought that it would be possible to extract wisdom that would serve as a model for response through harmonious collaboration among AIs and humans. In other words; the appropriate method is to incorporate the information processing process into the process of people working on disaster prevention and mitigation; and to extract the necessary wisdom. Specifically; after setting a purpose; a group of people as an aggregate attaches tags to reference sites and blocks of information; extracts and accumulates them. By focusing on common information from these information groups and comparing it with actions to be taken; we believe that it is possible to generate a collection of wisdom that can be used as a reference for each decision-making timing.
What type of data and supporting research infrastructure would be necessary to enable novel, transdisciplinary approaches to answering these and other human-centered disaster questions?
I believe that we need a knowledge-based information infrastructure that allows humans to centrally collect and reuse information scattered on the web in their decision making. A website is a convenient platform for easily disseminating various information. As a result; various people will post what they have noticed; and after a disaster; a large amount of information about their experiences will be posted. However; in the past research; we found that about 30% of information has been lost in Japan in about 15 years. An archive like Google's is conceivable to overcome this problem; however the amount would be huge and unrealistic. It would be good if there was a mechanism to appropriately manage only the information useful for sustainable resilience. There is a need for a mechanism for appropriately archiving information that has been evaluated by people and has a track record of being used by someone. In other words; there is a need for an information infrastructure that allows users who browse websites to record the process of using the information; archive the information; and reuse it. In addition; to systematically organize and manage human behavior and disaster responses that are the background of using information. We believe that we need a unified information infrastructure that integrates these; systematically manages actions; decisions; and the information groups necessary for them; and evaluates those values by human beings. No such infrastructure currently exists. This could be a concrete platform of “Online Synthesis System for Sustainable Resilience (OSS-SR)”.
In what ways can US-Japan collaborations advance these questions in new and important ways?
In recent years; with the development of ICT; the technology of automatic translation in natural language processing has improved dramatically. It is expected that this can be used to remove the language barrier. Although information understanding in Japanese is basic in Japan; the amount of information is limited when viewed from the world. On the other hand; the United States also has a wealth of experience with disasters and many advanced initiatives. By using automatic translation technology; it is assumed that mutual use of information content will become possible. Therefore; by comparing and verifying the difference in how information is valued according to the attributes of each user in the United States and Japan; it is possible to classify the knowledge into common wisdom and wisdom that each of them should possess independently. In particular; by condensing common wisdom; we believe that we can derive wisdom that the world should have in common from the knowledge and experience of the developed countries; the United States and Japan. To that end; we will develop a platform called OSS-SR under a common framework in Japan and the United States; and encourage the use and application of it by involving local communities in both countries. Furthermore; we will establish a new community of researchers from the United States and Japan; review the stored knowledge in OSS-SR from both countries; and promote the extraction of wisdom. Involving researchers from various fields; we aim to establish a common global governance for disaster prevention.