Disaster Information System Using Natural Language Processing

In this paper, we analyze incident response management work using free-formed information that is important for incident response communication. We develop a standard operating procedure for the work and study a support method based on a language processing technology to prevent missing or overlapping necessary tasks and optimize management of progress and integration management of related tasks. A prototype system was developed and evaluated in a workshop held by a local government, or in an exercise. As a result, it is confirmed that the use of language processing technology can make incident response management work efficient.