关键词: 学者影响力, 结构方程模型, 天际线算法, 综合评价, 基因编辑

Abstract: With the development of science and technology, scholars have become the main part of scientific research. Evaluating the influence of scholars is conducive to promoting the enthusiasm of scholar’s research, and can also provide reference for superiors to introduce outstanding talents. The influence of scholars reflects the academic level of scholars, and the screening of scholars with excellent comprehensive influence is helpful to promote the development of scientific research. At present, there are many indicators to measure academic influence, so it is very important to distinguish the applicable indicators. Structural equation model is used to select reasonable, significant and authoritative indicators from a variety of indicators. Meanwhile, it can measure the impact of indicators on various levels and identify the core indicators at different levels. Through the construction of structural equation model between academic literature influence, academic cooperation influence, academic citation influence, social influence and network community influence of scholars in the field of gene editing, it is found that academic track is the main measure indicator in the level of academic literature influence, which integrates the quantity of publications and the citation frequency, combines the traditional indicators, embodies the depth of scholars’ research from the aspects of quantity and quality, horizontal and vertical, and can be used to evaluate scholars’ performance. It is a new perspective to measure influence. It reflects scholars’ ability in many aspects. The impact factor more represents the level of journals in which scholars publish their papers, and this indicator has year limitations on citation statistics. Therefore, the representativeness is lower than other indicators.
At the level of academic cooperation influence, citation intensity and degree centrality are the core indicators. Centrality is the most direct measurement indicator, indicating the status of nodes, citation intensity includes the quality of cooperation papers on the basis of cooperation frequency, revealing the cooperation influence of scholars from two perspectives, with strong representativeness. In the dimension of academic citation influence, PageRank value and degree centrality are important indicators. We can see that degree centrality occupies the main position in social network, and PageRank value takes into account both the quantity and quality of citations, which makes up for the drawbacks of some indicators and has certain authority. 
At the level of social influence, Altmetrics-h index can identify high-level scholars more deeply. On the basis of Altmetric Score, the concept of h-index is synthesized to reflect the public’s recognition, belonging to the core indicators. In the influence of network community, the role of Researchgate_reads and Researchgate_follower is slightly higher, and more intuitive reflect the ability of online interaction. RG Score is also a important indicator, because it considers contribution, communication, reputation and other factors from multiple perspectives, and at the same time, combined with the degree of recognition by others, it represents the overall strength of scholars. Then the skyline algorithm is used to realize multi-dimensional comprehensive evaluation, identify talents that other scholars cannot surpass in many aspects. In the social media environment, there are many platforms that can reflect the influence of the network community. Researchgate platform has been rising in the academic circles and gradually gained the recognition of researchers, providing a more conducive channel for online communication. Although the Researchgate platform is well-known, there are some controversies in academic circles about the Researchgate platform. In order to determine whether the network community impact indicators represented by Researchgate are recognized and applicable to the measurement of scholars’ influence. 
This paper evaluates the comprehensive influence of scholars in the field of gene editing by skyline algorithms from coverage and non-coverage of the influence of network community indicators two aspects, and analyses its effectiveness. The study found that 8 and 14 scholars were screened out from the coverage and non-coverage Researchgate indicators. The two situations are basically consistent, and some scholars have great turning points. The results are consistent with the actual promoters of the gene editing field research. Zhang Feng, Horvath Philippe, Mali Prashant, Koonin Eugene V, Van der Oost John, Jinek Martin are excellent scholars who have jointly measured. Zhang Feng, Horvath Philippe, Mali Prashant are all in the leading position except for the influence of the network community, which shows that they mainly focus on the performance of the literature level. 
In addition, the evaluation system that covers the Researchgate index can help to select more outstanding scholars. In the model containing the Researchgate indicator, most scholars rank the top in the influence of the network community, which shows that the network communication ability is strong, and get the attention of multi-level people. Among them, Raoult Didier scholars are the most representative, and the influence of network community ranks first, but the other dimensions are relatively backward. In recent years, Raoult Didier has published thousands of different types of academic performance mainly in the Researchgate community, with strong dissemination ability, and constantly shifting from the traditional platform to the Researchgate community, reflecting the academic status from the perspective of network media. Other outstanding scholars selected include Church George M, Chylinski Krzysztof, Jinek Martin, Holmes Michael C, Gootenberg Jonathan S, Muramatsu Masamichi, Nemazee David and Sprengl Rolf, which have promoted the development of gene editing to some extent. It further proves that the Researchgate indicator is an important factor affecting scholars’ academic ability.
It is useful for academic evaluation as a supplementary indicator, and the feasibility of establishing a multi-dimensional comprehensive evaluation system. In the environment of Web 2.0, social media platforms are becoming more and more popular. Research is no longer limited to traditional indicators, but gradually popularizes Altmetrics indicator, focusing on network interaction activities. This paper puts forward a new research idea through the integration of various categories of indicators to measure the influence of scholars from multi-dimension, which can be further popularized. To provide some reference for the expansion of scientific and technological evaluation theory and application practice, and promote the development of scientific research.

Key words: scholar influence, structural equation model, skyline algorithm, comprehensive evaluation, gene editing