项目名称: 科学合作网络的不连通问题研究
项目编号: No.71473237
项目类型: 面上项目
立项/批准年度: 2015
项目学科: 管理科学
项目作者: 韩红旗
作者单位: 中国科学技术信息研究所
项目金额: 50万元
中文摘要: 科学合作网络作为一种强关系网络,在揭示科研工作者的关系和科研社区、评价研究者的学术能力、进行学术推荐和科研合作推荐、服务科研论文和项目评审等有着重要的应用。不少研究者观察到构建的科学合作网络是由很多不连通的子网络构成的,但并没有就这一问题开展研究。科学合作网络的不连通问题会给一些实际应用带来错误的结果,影响其应用效果。本项目拟对不连通问题的原因、不连通问题对实际应用的影响进行分析,探索提高科学合作网络连通性的方法。基于不连通是因为数据不全面的假设,对中英双语言体系下的跨数据库姓名消歧问题进行研究,在此基础上,对采用关系扩展方法提高科学合作网络连通性的方法进行研究。基于关系扩展法成本高、效率低等考虑,对采用链接预测算法提高科学合作网络连通性的方法进行研究。通过对科学合作网络不连通问题的研究,探索提高连通性的方法,希望为科学合作网络的深入研究和应用提供新的理论或方法的指导。
中文关键词: 科学合作网络;不连通问题;姓名消歧;链接预测
英文摘要: As a strong-tie social network,scientific collaboration networks (SCN) have many important applications, including uncovering relationships between researchers and scientific communities, evaluating academic influence of researchers, recommending academic publications and collaborative teams, selecting review experts for scientifc projects and papers. Many researcher have observed that SCN is composed of a great deal of disconnected sub-networks, yet they didn't study the problem. The disconnected problem of SCN will bring wrong results for practical application, and thus influence its effectiveness. The study will analysis the disconnected reasons and influence in practical application, and explore the methods to improve connectivity of SCN. Under the assumption of incomplete data, we will study author name disambiguation methods under the conditions of Chinese-English dual language and cross-database, and then we will study relation expansion methods to improve connectivity of SCN.Considering high cost and low effiency of the relation expansion methods, we will study link prediction methods to imporove connectivity of SCN. We hope the study will bring guidance to new theory or methods for SCN.
英文关键词: scientific collaboration network;disconnected problem;name disambiguation;link prediction