디지털 헬스 분야 지식생산기관 식별 정확도 제고 방안 연구

Despite the important roles of institutions and their collaboration in producing knowledge for innovation, the lack of accurate methods for identifying such knowledge-producing institutions has restricted empirical research on the role of institutions in innovation. This study explores methods to enhance the accuracy of identifying institutions involved in innovation process. To this end, we propose ways to improve accuracy in both aspects of information - data and algorithms - using bibliographic information in the digital health field. Specifically, in the data processing stage before applying algorithms, we address contextual inaccuracies of bibliographic information; in the algorithm application stage, we propose methods to improve the ambiguity of institution names (IND). When compared with the PKG dataset, which is publicly available datasets based on the same bibliographic information, our methods doubled the number of cases available for subsequent analysis. We also discovered that the contribution of Korean institutions in the digital health field is either underestimated or overestimated. The method presented in this study is expected to contribute to empirically researching the role of knowledge-producing institutions in innovation process and ecosystem.