Determining green identity profiles of urban residents in city branding with self-organizing maps: Case of "Green Bursa"

Advanced technology and rapid urbanization have caused waste of natural resources and environmental damage by disconnecting human from nature. Sustainability is considered as a solution to problems arising from such activities, and local solutions are suggested for global sustainability. Urban branding, on the other hand, makes the city attractive for investors businesses and citizens, and increases the livability and welfare level of the city. Sustainability and green city branding activities should start from the "citizen", who gives the city and its brand an identity and keeps them alive. The purpose of this thesis is to discover ecological identity profiles by taking attitudes, human-nature relationship and sustainable consumption behaviors of citizens into consideration. Data mining and knowledge discovery process has been adopted and implemented throughout the study. Bursa, one of the important cities of our country, has been working on adaptation to the "European Green Capital (EGC)" system for the last ten years. However, it has not yet passed the pre-selection stage to become an EGC. In order to contribute to the process, the city of Bursa has been chosen as the application area of the study. The data were collected with three adapted measurement scales that are designed for measuring environmental concerns (attitude), human-nature relationship and environmentally friendly behavior. In the first stage, the mediating role of the nature-relatedness in the relationship between attitude and sustainable consumption behavior was investigated by Partial Least Squares Structural Equation Modeling. In the second stage, the data were clustered with Super Self-Organizing maps, Ward-linked hierarchical and k-medoid clustering algorithms. Clusters were evaluated with internal and external evaluation criteria and the most appropriate cluster number was determined to be five. Each cluster was investigated with box-whiskers plots, bar charts and decision trees in order to describe green identity profiles. Unfortunately, a completely "green" typology compatible with sustainability has not been identified. Therefore, the need for local governments to design policies and activities that will change the perspective and behavior of citizens for the sustainability of the city emerged and it was concluded that "Green Bursa" can be achieved with green citizens. Keywords: Self-Organizing Maps, Mixed-Data Clustering, Partial Least Square Structural Equation Modelling, PERMANOVA, City Branding, Sustainable Cities, Pro-Environmental Behavior