面向大数据的移动数字图书馆情境化推荐系统研究  

Research on Mobile Digital Library Situational Recommendation System Oriented to Big Data

作　　者：刘海鸥[1] 陈晶[2] 孙晶晶[2] 张亚明[1] Liu Hai＇ou;Chen Jing;Sun Jingjing;Zhang Yarning

机构地区：[1]燕山大学经济管理学院 [2]燕山大学

出　　处：《图书馆工作与研究》2018年第9期58-64,共7页Library Work and Study

基　　金：国家社会科学基金项目"基于大数据深度融合的移动图书馆用户画像情境化推荐模型研究"(项目编号:18BTQ033)研究成果之一

摘　　要：大数据环境下,推荐系统项目评分的稀疏性问题愈加突出。如何从大数据中挖掘出与用户精准匹配的服务信息并及时推荐给相应用户,成为大数据时代移动数字图书馆个性化服务亟待解决的问题。本文在系统回顾相关研究进展的基础上,提出了面向大数据移动数字图书馆情境化推荐系统,通过融合用户的情境信息进行协同过滤推荐(CF),基于此缓解数据稀疏性导致的推荐性能下降问题;同时采用MapReduce的并行处理方式,以此提高大数据的并行挖掘性能。In the hig data environment,the proiect score sparsity of recommendation system becomes more important, how to excavate accurately service information from big data and timely recommend to the users becomes problems to be solved of mobile digital library personalized service in the era of big data. Based on a system review of related research progress,this paper puts forward the situational recommendtion systems of mobile digital library oriented to big data, conducts collaborative filtering recommendation through fusing users＇ contextual information, relieves the problems of a decrease of recommended performance caused by sparsity, and uses the parallel processing mode of MapReduce, in order to improve the parallel excavating performance of the big data.

关 键 词：大数据 移动数字图书馆 协同过滤 情境化推荐 个性化服务 

分 类 号：G250.76[文化科学—图书馆学]