摘要
传统行为分析方法存在预测能力不足、分析片面化,获取的行为特征数据规律性较差等问题,导致得到的分析结果与实际不符。基于此,提出大数据挖掘技术的图书馆移动用户行为分析方法。该方法将获取的大数据预先清洗、筛选,并利用转换算法集成特征数据;使用BP神经网络适应度函数构建评估预测模型,挖掘行为特征规律,根据预测结果将特征相互信息值排序;通过聚类算法捕捉具有关联的数据,利用交叉分析法分析用户行为内在性质,实现全面的图书馆移动用户行为的全面挖掘。实验结果表明,与传统方法相比,所提分析方法挖掘用户行为特征数据的能力更强,分析结果准确度更高,可应用于现阶段图书馆移动用户行为分析。
As the traditional behavior analysis method has some problems,such as insufficient prediction ability,one⁃sided analysis and poor regularity of the obtained behavior characteristic data,which leads to the fact that the analysis results are not consistent with the actual situation.On this basis,a library mobile users′behavior analysis based on big data mining technology is proposed.In this method,the acquired big data is cleaned and screened in advance,and the feature data is integrated by means of the transformation algorithm.The fitness function of BP neural network is used to construct the evaluation prediction model,excavate the behavior feature law,and sort the feature mutual information values according to the prediction results.The related data are captured by clustering algorithm,the inherent properties of user behavior are analyzed by means of cross analysis method,and the comprehensive mining of library mobile users′behavior is realized.The experimental results show that,in comparison with the traditional methods,the proposed analysis method has stronger ability to mining users′behavior features,and higher accuracy of analysis results.It can be applied to the analysis of library mobile users′behavior at present.
作者
孙慧
SUN Hui(Changchun Normal University,Changchun 130032,China)
出处
《现代电子技术》
北大核心
2020年第18期164-167,171,共5页
Modern Electronics Technique
基金
国家自然科学基金项目(61773009)
吉林省教育厅“十三五”社会科学项目(JJKH20170668sk)。
关键词
图书馆移动用户
行为分析
大数据挖掘技术
数据获取
预测建模
交叉分析
library mobile users
behavior analysis
big data mining technology
data acquisition
prediction modeling
cross analysis
