Abstract
The innovation structure was studied for the core cities in the northern part of Kyushu, aiming at promoting regional innovation.
Elements to support the regional innovation were divided into direct elements and indirect elements. By means of principal component analysis we found that they could be represented by the following six components: "city scale", "industrial accumulation", "motivation of public sector", "function, charm and diversity of city", "infrastructure", and "safety, relief and residence". Multiple regression analysis was then applied to estimate the innovation index of each city, taking the score of each component as dependent variable. We found the following two variables were generally the most significant: "function, charm and diversity of city", and "industrial accumulation".
On the basis of these findings, we classified the regional innovation structure into three categories by cluster analysis. Some suggestions on city promotion policy are also discussed.
