Research on Academic Impact Evaluation of Papers based on Multi-generation Citation and Knowledge Inheritance

Abstract: ［Purpose/Significance］The evaluation of academic papers plays a fundamental and radiant role in the whole scientific evaluation system. It is an important part of the evaluation system to construct the measurement index that can reflect the long-term academic influence of academic papers.［Method/Process］This paper constructs a paper academic impact evaluation method called MGR-KI method for short which can comprehensively consider the citations chain relationship between direct citations and indirect citations, and take into account the level of sustained knowledge contribution of academic achievements, then uses 7652 sets of citation relationships and 14 074 058 citation links involved in 2070 papers in the field of electromagnetic metamaterials as data sources for the empirical research and the feasibility and effectiveness of the proposed method are verified.［Result/Conclusion］The results show that the academic impact evaluation method based on MGR-KI established in this paper shows stronger differentiation compared with the impact evaluation results of papers based on citation frequency and cited h index, which takes into account the characteristics of different citation types and has fine comparison of different papers.The difference of text similarity calculation methods based on feature dimension and semantic dimension will have a significant impact on the results. More fine-grained researches on knowledge genetic relationship mining and genetic strength calculation methods should be carried out to achieve more accurate measurement and evaluation.