Abstract
With globalization, the notion of "creative city" has become a core concept of many cities in the world development policies, with real properties being upgraded or used to change, renewal is being conducted, and creative industries are emerging. This trend has reached its peak in the past decade, with different forms and scales gathering global development momentum among the creative communities to promote the development of creative economies. In recent years, however, there was still skepticism about the sustainability of the current creative communities. Many scholars have pointed out that signs of unsustainability have begun to appear in many creative communities. To overcome these obstacles, the development of rational and highly effective improvement strategy requires a dynamic thinking process. Therefore, this study employs the DEMATEL-based ANP with modified VIKOR (D-DANP-mV) model in presenting an assessment framework for the sustainability of creative communities. This system is used to assess the sustainability of current creative communities and determine how to solve their problems. Thus, continuous and systemic improvement strategies can be developed to achieve the aim of sustainable development. Two creative communities in Taiwan, Taichung Cultural and Creative Industries Park (TCCIP), and Shen-Ji New Village (SJNV), are used as case studies in this study. Based on the concept of systematic improvement from fundamental issues, the results indicate that the improvement priorities can be determined by applying the D-DANP-mV model. This approach is different from those found by a conventional method with the hypothesis of independent criteria (e.g., diversification of creative talents in TCCIP), and cannot use for performance improvement (e.g., only can be used for ranking and selection among alternatives). Considering these points, unreasonable premises, biased errors, and lack of some real application functions in the process of resource allocation could be more efficient improvement strategies generated in this proposed model.