基于Spark计算模型的资源供给新范式  

Park computing model-based new canonical form of resources supply

作　　者：屈艳玲[1] QU Yan-ling(Hunan Library, Changsha 410011, Hunan Province, China)

机构地区：[1]湖南图书馆,湖南长沙410011

出　　处：《中华医学图书情报杂志》2020年第3期37-42,共6页Chinese Journal of Medical Library and Information Science

基　　金：2019年度湖南省社会科学成果评审委员会课题“智能数据驱动图书馆智慧服务的研究”(XSP2019WT002)研究成果之一。

摘　　要：大数据背景下,图书馆各类信息系统、平台、数字资源总量的不断增多以及用户兴趣与关注焦点的不断变化,加速了碎片资源的产生。基于Spark计算框架,对不同分布式平台上的碎片资源进行重组、推荐。通过交互式查询和朴素贝叶斯分类算法设计的推介模型,提升了碎片资源供给时效,减轻了分别管理各种平台造成的负担,促进了图书馆碎片资源治理微生态的健康发展。The number of different information systems,platforms and digital resources in libraries and the changing interests and concerned focuses of library users have accelerated the generation of fragmented resources under the background of big data.The fragmented resources on different distributed platforms were thus reorganized and recommended.The recommendation model designed in accordance with the interactive query and naive Bayes classification algorithm can improve the effectiveness of fragmented resources supply time,reduce the burden of separate management of different platforms,and promote the healthy development of microecology by bringing the fragmented resources in library into control.

关 键 词：大数据 SPARK 碎片资源 碎片资源池 分布式计算 

分 类 号：G250.73[文化科学—图书馆学] R-058[医药卫生]