Title: Visualization and spatial statistics for spatial data that contribute to comprehensive suicide countermeasures in Japan
Authors: Takafumi Kubota - Tama University (Japan) [presenting]
Marina Ishikawa - JR Shikoku Information Systems (Japan)
Yoshikazu Yamamoto - Tokushima Bunri University (Japan)
Abstract: Spatial data of suicides in Japan are visualized, which contribute to the development of comprehensive suicide countermeasures. To observe the suicide risks based on area, age, and gender, small area suicide data (suicide data) are visualized. These data are derived on the premise that the place is determined based on the location at which a person commits suicide, and such locations are identified by observing the ``mobile spatial statistics''. To estimate the actual relation between industrial structures and suicide risks, the suicide data are visualized based on the principle that the place of suicide is based on the residence of the person who has committed suicide; the required information can be obtained using the RESAS (Regional Economy and Society Analyzing System) data. Suicide data are also used to conduct spatial autocorrelation and conditional autoregressive modeling. To detect spatial autocorrelation, a local Moran plot of the suicide data for visualization is created. Further, a regression model is used based on several variables, such as health, income, tax, education, and so on, to identify the risks that are associated with various regions.