分类评价视角下学者研究优势识别与评价研究OA

[Purpose/Significance]Against the backdrop of the reform of the national science and technology innova-tion system,this study proposes a novel talent evaluation method integrating research topics,scholarly contributions,and the z-index to address the demands for classified,multidimensional,and comprehensive talent evaluation in scientific and technological fields.[Method/Process]First,the Word2vec model and the k-means algorithm were employed to conduct topic clustering of keywords in the field of discipline,and different research topics were identified.Then,a three-dimensional relationship data of topic-scholar-citation was constructed.Finally,the average weight of keywords per article and the co-authorship weight were incorporated into the z-index to construct the improved ZAK index.[Result/Conclusion]Through empirical analysis,the ZAK index significantly enhances evaluation differentiation and fairness by incorporating per-article keyword weighting and author collaboration weighting while preserving the consistency of traditional metrics.This method can not only effectively identify outstanding talents in specific fields but also discover excellent scholars engaged in small-scale or emerging topic research,providing a scientific and objective evaluation method for the identifica-tion of outstanding talents,serving the reform of the national science and technology system,and promoting academic exchanges and knowledge innovation.