大数据背景下的图书馆读者借阅行为分析

Library readers′ borrowing behavior analysis under background of big data

图书馆读者的数量大、借阅行为复杂多变,而传统图书馆读者借阅行为分析方法精度不高,且工作效率极低,无法满足现代图书馆管理的要求。为了更好地刻画图书馆读者借阅行为,提高图书馆读者借阅行为分析的准确性,设计了大数据背景下的图书馆读者借阅行为分析方法。首先,分析图书馆读者借阅行为的研究现状,对图书馆读者借阅行为分析数据进行聚类分析,并提取图书馆读者借阅行为分析特征;然后,采用大数据分析技术--最小二乘支持向量机对图书馆读者借阅行为变化特点进行拟合,构建图书馆读者借阅行为分析模型;最后进行图书馆读者借阅行为分析实例验证。结果表明,大数据背景下的图书馆读者借阅行为分析精度超过93%,而当前其他分析方法的精度均小于90%,同时,减少了图书馆读者借阅行为分析时间,所提方法分析速度明显加快,具有显著的优势。

Due to the large number of library readers and the various and complex library readers ′ borrowing behavior,the precision of traditional analysis methods of library readers′ borrowing behavior is not high,and the working efficiency of the methods is extremely low. Therefore,they cannot meet the requirements of modern library management. In order to satisfactorily describe the library readers′ borrowing behavior and improve the precision of library readers′ borrowing behavior analysis,an analysis method of the library readers′ borrowing behavior under the background of big data is designed. The research status of the library readers′ borrowing behavior is studied,and then,the data of the library readers′ borrowing behavior analysis is subjected to clustering analysis,and the features of library readers′ borrowing behavior analysis are extracted. After that,the big data analysis technology named least squares support vector machine(LS-SVM)is used to fit the change characteristics of the library readers′ borrowing behavior and construct the analysis model of library readers′ borrowing behavior. Finally,examples are given to verify the library readers′ borrowing behavior analysis. The results show that the precision of the library readers′ borrowing behavior analysis under the background of big data exceeds 93%,while the precision obtained with other analysis methods is below 90%. In addition,the duration of the library readers′ borrowing behavior analysis using the proposed method is reduced and the analysis speed is obviously accelerated. Therefore,the proposed method has significantly advantages.