Characterizing Japan's technological regions using patent data

Abstract

This study examines the spatial distribution of knowledge creation and invention in Japan using patent data. It applies a methodology tailored for the Japanese patent database, combining the spatial Gini Index and Principal Component Analysis (PCA) to analyze inventive activities at the municipal level. The research uncovers a significant concentration of data in a few municipalities and a consistent spatial distribution of technology domains that align with the locations of industrial activities. The study reveals the spatial concentration of inventions and their technological proximity, utilizing a spatial Gini Index and PCA for assessing technology concentration and colocation, as well as a Local Indicator of Spatial Association (LISA) for spatial autocorrelation. The findings indicate high levels of concentration and spatial autocorrelation in technology fields, with a notable correlation between urban areas and their industrial activities. The study highlights the importance of accurate inventor location data and underscores the significance of urban areas in patented activities, suggesting that inventive activity concentration often mirrors the geography of industrial locations in Japan.