A0316
Title: Analyzing regional suicide patterns in Japan before and after the COVID-19 pandemic and usage of generative AI for EBPM
Authors: Takafumi Kubota - Tama University (Japan) [presenting]
Abstract: The purpose is to explore regional trends in suicide deaths in Tokyo, Japan, before and after the COVID-19 pandemic. In addition, another objective is to generate evidence for policy-making by each municipality using generative AI and to examine the accuracy of such evidence. The data covered in this study are data on suicide by a municipality in Tokyo in 2016 and 2021. They are based on the ``Basic Data on Suicide in Local Communities'' of the statistics on suicide compiled by the Ministry of Health, Labour and Welfare. In addition to the suicide death rate, suicide items include age, occupation, location, etc., with 71 variables, including region names. The methodology first compares suicide data for 2016 and 2021 by visualization. Next, regions with exceptionally high rates of each item are identified. The method used to identify regions is spatial clustering of geographically referenced attributes. Moreover, the focus is placed on a single city. Here, Fuchu City in Tokyo is used as the object of comparison. The comparison targets are the annual changes in Fuchu City in 2016 and 2021 (A) and Tokyo and Fuchu City in 2021 (B). The author submits the data of 71 variables for A and B to ChatGPT (version 4), one of the generative AIs, with instructions to detect differences and have it generate a document about the results. The content of the generated documents is verified to be accurate and to provide evidence that can be used for policy-making.
