基于专长吻合度、学术影响力与社会关联值的专家推荐模型研究

发布时间：2018-06-13 13:25

  本文选题：同行评议 + 专家推荐　； 参考：《情报学报》2017年04期

【摘要】：学术评审活动中,专家的学术专长和学术水准对于项目评审的科学性十分关键,专家的社会关系对于科研项目评审的公平性有重要影响。论文基于Web of Science期刊论文、中国博硕士学位论文、简历、网络新闻等组成的多源数据,首先定义了专长吻合度(S)、学术影响力(I)与社会关联值(C)三个变量;然后,基于这三个变量构建了专家遴选、回避与推荐模型R。专家的研究专长依赖于其研究成果的关键词,专家的学术影响力依赖于专家的h指数与入选人才计划(包括两院院士、国家自然科学基金杰出青年基金、中组部千人计划学者、教育部长江学者等),专家的社会关联值依赖于各种利害关系,具体包括:师承关系、同事关系与合作关系。论文以精准医疗领域的100位知名专家为例构建了专家库,测试了模型的效果,研究发现,该模型能有效运行,并能基于专长吻合度、学术影响力、社会关联值三个变量推荐专家。最后,论文提出了模型的不足之处与优化的可能。
[Abstract]:In the activities of academic review, the academic expertise and academic level of experts are very important to the scientific nature of the project review, and the social relationship of experts has an important impact on the fairness of the evaluation of scientific research projects. This paper is based on the multi-source data of Web of Science journal thesis, Chinese Ph. D. thesis, resume, network news, etc. Firstly, it defines three variables: the degree of specialty coincidence, academic influence (I) and social association value (C). Based on these three variables, a model of expert selection, avoidance and recommendation is constructed. Experts' research expertise depends on the key words of their research results, and experts' academic influence depends on experts'h index and selected talent programs (including academicians in both chambers, outstanding youth funds of the National Natural Science Foundation of China, and scholars with thousands of people in the Central Organization Department). The social relevance value of the experts depends on various interests, including: teacher-bearing relationship, co-worker relationship and cooperative relationship. This paper takes 100 well-known experts in the field of precision medicine as an example to construct the expert bank, and tests the effect of the model. The research finds that the model can work effectively, and can be based on the degree of expertise and academic influence. Social correlation value three variables recommend experts. Finally, the paper puts forward the deficiency of the model and the possibility of optimization.
【作者单位】： 浙江大学公共管理学院信息资源管理系;
【基金】：中国科技信息研究所情报工程实验室开放基金“基于多源数据的专家遴选、回避与推荐技术研究”
【分类号】：G353.1
，

本文编号：2014166