关键词: 空间科学计量学, 科学合作, 科研产出, 知识流动, 可视化

Abstract: [Purpose/Significance] As an important research field of scientometrics, spatial scientometrics has been specifically proposed by scholars to discuss the spatial aspects of scientific research activities. At present, scholars at home and abroad have carried out quantitative research of spatial scientometrics, covering almost all aspects of scientometrics; however, there is little systematic review of the concept and research progress of spatial scientometrics. This article systematically reviews, analyzes, and summarizes the related research work of spatial scientometrics, with an aim to provide a reference for scholars to carry out spatial scientometrics research. [Method/Process] Based on the introduction of the concept of scientometrics and spatial scientometrics, this paper uses the literature research method to expound from three aspects including research theme, research method and visualization practice. [Results/Conclusions] The research finds that the current spatial scientometrics research is mainly based on the geographical location attributes of scientific information, and comprehensively uses bibliometric methods, network analysis methods, model construction and statistical methods, and multi-indicators construction methods to examine scientific collaboration, scientific research output and knowledge flow. This research also reveals the remaining problems and challenges in spatial scientometrics. First, in terms of the theoretical research system, spatial scientometrics essentially serves a specific field of scientometrics research, drawing on technical methods from the fields of computer science, network science, and economic geography, and has not formed a relatively independent theoretical system. Second, in terms of method applicability and utility research, its method selection is also arbitrary. Regarding the same research problems, researchers usually choose their own and familiar methods and tools based on experience. Thus, it is difficult to accurately evaluate the scientificity and validity of their analysis results, and even different research methods and tools reveal opposed so-called "characteristics and laws". Third, in terms of data analysis sources, the existing research is mostly around the data collected from Web of Science and Scopus databases. However, for a comprehensive and objective analysis of the spatial structure of scientific activities, other forms of data should be considered such as reports, the fund-funded projects and policy documents. Last, for data acquisition, when using the literature database to conduct quantitative research, since many well-known institutions and companies are not centrally distributed organizations, it is difficult to guarantee the author's real address. Due to the different scope and definition criteria of cities in different countries, an analysis of city-level data would affect the accuracy of the analysis results.

Key words: spatial scientometrics, scientific collaboration, scientific output, knowledge flow, visualization