Analysis of chronological records by text mining: From medical activities in Fukushima Prefecture Typhoon No. 19 in 2019

Abstract

【Purpose】Focusing on the Disaster Medical Assistance Team (DMAT) activities during Typhoon No. 19 in 2019 in Fukushima Prefecture, we conducted an objective evaluation of this topic and its changes during the disaster by analyzing chronological records. 【Results】The analysis reveals a notable progression in both number of words and sentences, transitioning from the hyperacute phase to the acute period, followed by a marked decline during the subacute phase. Furthermore, the frequency of words in the chronological records changed with period. 【Discussion】During the hyperacute phase, there was an active engagement by both hospitals and DMAT in collecting information from medical institutions. In the acute phase, frequent mentions of evacuation centers and district names suggested a transition in DMAT activities toward local community support. As the situation moved into the subacute phase, an increase in the usage of words related to logistics operations indicated a shift in other work. 【Conclusion】 In conclusion, text mining on chronological records offers a promising and objective approach to evaluate topics and their changes during disasters.