Research On Academic Influence Evaluation Of Scholars Based On Cooperation And Citation Network

Nowadays,with the continuous output of scientific research results in various research fields,a large amount of academic data has also followed.The academic subjects involved include scholars,essays,periodicals,research institutions,etc.,effectively evaluating the academic influence of these subjects.Hot issues in current academic research.Among them,the evaluation of academic influence of scholars is an important research direction in research work.It can provide reference for talent introduction,fund and project approval,and performance evaluation of researchers in universities and research institutions.Traditional scholars' influence evaluation methods include peer review method and bibliometric method.In recent years,a large number of studies use scientific research academic networks for analysis and evaluation,which are mainly divided into evaluation methods based on single academic network and multi-layer academic network.The evaluation results based on a single academic network are relatively one-sided,and the research method based on multilayer academic network evaluation is an effective supplement.However,current research based on multi-layer networks tends to simplify the attributes of nodes in real networks,the strength of relationships between nodes,and the contribution of scholars in the literature.In view of the shortcomings of the current research,this paper mainly carried out the following work: Firstly,the network model of this paper is constructed,which is a scholar-document network composed of scholars' cooperation network and literature citation network,considering the cooperation influence and citation influence of scholars..Secondly,the PageRank algorithm is applied to the above two networks respectively,and the importance of the scholars in the cooperative network and the literature in the citation network is obtained.Among them,when calculating the importance value of the literature,the publication time,the cited time,and the influence factor of the published journal are introduced into the PageRank algorithm,and the weighted PageRank algorithm is proposed.In addition,when the influence is transmitted to scholars according to the importance of the literature,the weights are assigned according to the signature order of the scholars in the literature,and it is more reasonable to consider their contribution in different documents.Finally,combined with the importance of scholars in the cooperative network and from the citation network,that is,the scholars' cooperative influence and citation influence,the scholars' academic influence evaluation method(PR_T)based on cooperation and citation network is proposed.In the experimental part,the data used in this paper is the Aminer dataset,which distributes the proportion of cooperation and citation influence in the total academic influence;and compares the PR_T method with multiple evaluation methods,and the results prove that Compared with other evaluation methods,the method has obvious advantages and can evaluate the influence of scholars more comprehensively.At the same time,this paper also analyzes the characteristics of scholars with high academic influence.