首页 | 本学科首页   官方微博 | 高级检索  
     

基于链路预测的iSchools联盟院校URL共引潜在关联识别研究
引用本文:袁国廷,岳增慧,刘星. 基于链路预测的iSchools联盟院校URL共引潜在关联识别研究[J]. 中华医学图书情报杂志, 2020, 29(7): 1-13
作者姓名:袁国廷  岳增慧  刘星
作者单位:济宁医学院国际教育学院(外国语学院),山东 日照 276826;济宁医学院医学信息工程学院,山东 日照276826
基金项目:国家自然科学基金青年科学基金项目“学科知识扩散规律及动力学机制研究”(71704063);济宁医学院大学生创新训练计划项目“基于相似性的iSchools联盟院校URL共现网络链路预测研究”(cx2019008)
摘    要:目的:探索iSchools联盟院校关联特征及潜在演进态势,为网络时代背景下iSchools联盟院校间交互结构性能的优化、互联互通引导机制的健全、国际交流与合作策略的完善以及协同创新与发展战略的制定提供可资借鉴的理论和实践参考。方法:以iSchools联盟院校的URL共引网络结构信息为基础,采用10项基于局部信息的相似性指标分别对无权和加权URL共引网络进行链路预测分析,对比各指标的预测性能。引入权重调节系数,剖析强弱连接对预测精度的影响。利用无权PA指标对iSchools联盟院校在网络空间中的潜在关联进行预测识别。结果:不同链路预测指标在无权和加权iSchools联盟院校URL共引网络中的适用性存在一定差异,iSchools联盟院校URL共引链路预测过程中存在一定程度的强弱连接效应。结论:我国高校信息学院与国际院校的联系将日益密切,在iSchools联盟网络中的地位具有较大的提升空间。

关 键 词:iSchools;URL共引;局部信息相似性指标;链路预测;网络空间
收稿时间:2020-06-12

Link prediction-based recognition of potential associations in co-citations of iSchool URL
YUAN Guo-ting,YUE Zeng-hui,LIU Xing. Link prediction-based recognition of potential associations in co-citations of iSchool URL[J]. Chinese Journal of Medical Library and Information Science, 2020, 29(7): 1-13
Authors:YUAN Guo-ting  YUE Zeng-hui  LIU Xing
Affiliation:School of International Education or School of Foreign languages, Jining Medical College, Rizhao 276826, Shandong Province, China;School of Medical Information Engineering, Jining Medical College, Rizhao 276826, Shandong Province, China
Abstract:Objective To provide the theories and practical experiences for iSchool URL to optimize its interactive structure performance, to perfect its guiding mechanism of interlink and intercommunication, to improve its strategies for international communication and cooperation, and to formulate its strategies for cooperative innovation and development under the background of network era by studying the association characteristics and potential evolution trend of iSchool URL. Methods The unweighted and weighted URL co-citation networks were analyzed by link prediction analysis using the 10 local information-based similarity indexes, and the prediction performances of 10 local information-based similarity indexes were compared. The effects of strong link and weak link on the prediction accuracy were analyzed by introducing the weight adjustment coefficient. The potential associations in iSchool URL were predicted and recognized using the unweighted PA index. Results The applicability of different link prediction indexes was different in co-citations of iSchool URL. The effects of strong link and weak link on the prediction accuracy were detected in co-citation link prediction of iSchool URL. Conclusion The contact between domestic information schools and foreign universities and colleges will become increasingly close and there is a rather large room for domestic information schools to improve their position in iSchool URL networks.
Keywords:iSchool   URL co-citation   Local information similarity index   Link prediction   Network space
点击此处可从《中华医学图书情报杂志》浏览原始摘要信息
点击此处可从《中华医学图书情报杂志》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号