2017.06.05
Aiming to Establish Training Guidelines for Disaster Information Analysis Using Artificial Intelligence
Keio Research Institute at SFC, Keio University
National Institute of Information and Communications Technology
National Research Institute for Earth Science and Disaster Resilience
● The Shingo Yamaguchi Laboratory at the Faculty of Environment and Information Studies, Keio University, the National Institute of Information and Communications Technology (NICT), and the National Research Institute for Earth Science and Disaster Resilience (NIED) have established a joint research council to introduce and promote the use of advanced artificial intelligence technologies in the field of disaster prevention and mitigation .
● The joint research council aims to create and publish guidelines for the effective implementation of disaster drills during normal times , concerning information analysis conducted by local governments and other organizations using artificial intelligence technology during disasters.
● A public symposium related to this theme will be held on Friday, August 4 .
● Furthermore, these training guidelines for introducing artificial intelligence technologies (such as natural language processing) for disaster prevention and mitigation are essential for a country advanced in disaster management and represent a world-first initiative .
1. Background and Significance
In recent years, in the field of disaster prevention and mitigation, there has been a growing number of cases where local governments utilize social networking services (SNS) for disseminating information during disasters. Currently, approximately 54% of local governments use SNS for disaster response (*1). In addition to such information dissemination, some local governments also use SNS as a means of "information gathering" during disasters, and this number is on the rise.
In recent years, technological innovations in artificial intelligence (AI), combined with the latest technologies such as the Internet of Things (IoT), big data, and robotics, are significantly transforming the processes of knowledge and value creation in socioeconomic activities. Among these, natural language processing technology, which enables computers to process human language, is gaining attention as a crucial artificial intelligence technology. It provides a mechanism that allows for the organization, retrieval, and analysis of information beyond human limits, even when information is convoluted during disasters or emergencies.
Against this backdrop, the active introduction of the latest information and communication technologies into the field of disaster prevention and mitigation is being positioned as a key national policy. For instance, the government published the "Guidebook on the Use of SNS in Disaster Response" (*2) this March. In April, the national Central Disaster Management Council revised the Basic Disaster Management Plan, for the first time stipulating the obligation to endeavor to use the latest technologies such as artificial intelligence (*3).
On the other hand, when the national government, local governments, designated public corporations (*4), and other related organizations attempt to use artificial intelligence technology for disaster countermeasures, it is necessary not only to introduce information systems for data collection and analysis but also to become proficient in their use through regular practice by incorporating special scenarios into disaster drills. However, the introduction of artificial intelligence in the field of disaster prevention and mitigation is still a cutting-edge endeavor, and with no appropriate policies or guidelines in place, the implementation of such disaster drills remains a process of trial and error for local governments and other bodies.
Furthermore, state-of-the-art artificial intelligence technologies are expected to contribute broadly to ensuring the safety and security of the nation and its citizens, not only in natural disaster countermeasures but also in areas such as large-scale accident response, crowd control, pandemic countermeasures, and crime prevention and security.
For this reason, we will contribute to creating innovation in the field of disaster prevention and mitigation by establishing a joint research council aimed at introducing advanced artificial intelligence technologies (such as natural language processing) for disaster prevention and mitigation, and by promoting research toward the development of guidelines for the preparation and implementation of disaster drills using artificial intelligence technology.
*1 As of 2016. Survey by the Information and Communications Technology (IT) Comprehensive Strategy Office, Cabinet Secretariat.
*2 Disaster Prevention and Mitigation Team, Information and Communications Technology (IT) Comprehensive Strategy Office, Cabinet Secretariat http://www.kantei.go.jp/jp/singi/it2/sns_guidebook.html
*3 Revision of the Basic Disaster Management Plan at the 37th Central Disaster Management Council (April 11, 2017) ("The national and local governments, etc., shall endeavor to introduce the latest information and communication-related technologies in order to swiftly and accurately analyze, organize, summarize, and search damage information and information on emergency response activities conducted by related organizations.") http://www.bousai.go.jp/kaigirep/chuobou/37/index.html
*4 Based on the Disaster Countermeasures Basic Act, these are the Japanese Red Cross Society, Japan Broadcasting Corporation (NHK), and other public institutions, as well as corporations engaged in public utility services such as electricity, gas, transportation, and communications, designated by the Prime Minister. http://www.bousai.go.jp/taisaku/soshiki/s_koukyou.html
2. Topics for Discussion at the Joint Research Council (Tentative)
l Training guidelines for disaster information analysis using artificial intelligence technologies (such as natural language processing)
l The ideal form of a standard open-source dataset (natural language) for use in disaster drills
l Measures to promote training support services for local governments, etc.
3. Future Schedule
Friday, August 4, 2017 Public Symposium
September 2017–c. January 2018 Joint research council meetings (several times)
*Information on how to participate in the joint research council will be announced at the public symposium.
c. February 2018 (Tentative) Creation and publication of disaster drill guidelines
-----------------------------------------------------------------------------------------------------------------------------
For details, please see Aiming to Establish Training Guidelines for Disaster Information Analysis Using Artificial Intelligence.pdf . (PDF)
[Inquiries regarding this matter and the joint research council]
Shingo Yamaguchi Laboratory, Faculty of Environment and Information Studies, Keio University
E-mail: shingo5 [ at ] sfc.keio.ac.jp
*When sending an email, please replace [ at ] with @.
[Distributed by]
Office of Research Development and Sponsored Projects at Shonan Fujisawa Campus, Keio University