用户借阅行为的可视化方法研究

发布时间：2018-04-01 02:09

本文选题：图书采购　切入点：大数据　出处：《天津工业大学》2017年硕士论文

【摘要】：天津市高校联合图书馆成立13年来,在图书馆信息化的道路上一直走在前列,但是随着时间的积累,信息化系统中的数据日益变得庞大复杂,如何利用这些数据成为了现在图书馆管理工作的一个迫切需求。基于该需求,作者设计了两个可视分析的模型,重点分析了馆藏资源数据与读者借阅记录这两种数据。针对当前高校图书馆纸质文献采购决策过程中,决策主观性较强,客观依据不足,提出图书借阅可视分析系统模型,实现对历史信息的采集、挖掘、可视分析。分析馆藏图书数据,构建图书借阅可视分析系统模型,搭建大数据平台,验证模型的可用性。实验表明,系统能够很好地对海量图书历史数据进行可视化分析,达到辅助制定图书采购决策的目的。目前的可视分析结果视图还可以进一步丰富。该系统可以对图书馆借阅历史数据进行可视化分析,为决策者进行图书采购提供科学化的数据支持。为了进一步研究读者的借阅习惯,提出基于序列模式挖掘的读者借阅行为可视分析模型,发现读者的借阅模式。在使用序列模式挖掘的过程中发现读者的序列模式比较庞大,本文提出使用模式裁剪的方式降低序列数据库中序列模式的规模,让之后的可视化结果更容易被理解。实验表明该模型能够对读者的借阅模式进行有效地挖掘,可视化的结果具有一定的实用价值。
[Abstract]:In the past 13 years since the establishment of Tianjin University United Library, it has been in the forefront of library informatization, but with the accumulation of time, the data in the information system has become increasingly huge and complex. How to make use of these data has become an urgent need of library management. Based on this demand, the author designs two visual analysis models. This paper mainly analyzes the data of library collection resources and readers' borrowing records. In view of the subjectivity of decision-making and the lack of objective basis in the decision-making process of paper literature procurement in university libraries, a visual analysis system model for library borrowing is put forward. Collection, mining, visual analysis of historical information, analysis of library data, construction of visual analysis system model of library borrowing and reading, setting up big data platform to verify the usability of the model. The system can visualize and analyze the massive historical data of books. The present visual analysis result view can be further enriched. The system can visually analyze the historical data borrowed by the library. In order to further study readers' borrowing habits, a visual analysis model of readers' borrowing behavior based on sequential pattern mining is proposed. Discover the reader's borrowing pattern. In the process of using sequential pattern mining, we find that the reader's sequential pattern is quite large. In this paper, we propose to reduce the scale of sequence pattern in sequence database by pattern clipping. The experimental results show that the model can effectively mine the readers' borrowing patterns, and the results of visualization have some practical value.
【学位授予单位】：天津工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：G250.7;TP311.13

【相似文献】