基于相似树查询的隐私大数据定向检索算法  

Private Big Data Oriented Retrieval Algorithm Based on Similar Tree Query

作　　者：李盼盼[1] 赵浩[1] 林慧恩 LI Pan-pan;ZHAO Hao;LIN Hui-en(Jinshan College,Fujian Agriculture and Forestry University,Fuzhou Fujian 350002,China;Library of Fuzhou University,Fuzhou Fujian 350108,China)

机构地区：[1]福建农林大学金山学院,福建福州350002 [2]福州大学图书馆,福建福州350108

出　　处：《计算机仿真》2021年第11期429-432,437,共5页Computer Simulation

基　　金：2018年福建省中青年教师教育科研项目(JT180853);2017年福建省高等教育教学改革项目(FBJG20170323);福建省2017年创新创业教育改革项目试点专业(jx180301)。

摘　　要：网络用户文档涵盖大量隐私敏感数据,传统隐私大数据检索方法面对多种类繁杂数据存在检索顺序混乱问题,导致检索命中率不高、效率低下等问题。为解决上述问题,提出基于相似树查询的隐私大数据定向检索算法。结合物-雾-云建立适用于海量数据的隐私保护检索系统,设计基于相似树查询的兄弟叶节点查询结构。利用网格区域分割将多种类繁杂数据集划分成多个子集,完成大数据的相互匹配与滤波。结合模糊谱聚类完成数据关联属性特征提取,按照数据聚类属性识别分类特征,引入识别结果完成隐私大数据定向检索。仿真结果表明,与传统方法相比,所提隐私大数据定向检索方法检索性能具有明显应用优势,检索耗时短、精度高,能够有效维护网络用户数据安全。The privacy-sensitive data in network user documents leads to many shortcomings of the traditional privacy big data retrieval methods,such as chaotic retrieval order,low retrieval hit rate,and efficiency.In this regard,a privacy big data-oriented retrieval algorithm based on similarity tree query was reported.According to the object fog cloud,a privacy protection retrieval system suitable for massive data was founded.The query structure of sibling leaf nodes based on similarity tree query was designed.Grid region segmentation was applied to divide multiple kinds of complex data sets into multiple subsets,matching and filtering big data.According to fuzzy spectral clustering,data association attribute features were extracted.Through data clustering attributes,classification features were identified.Eventually,the recognition results were utilized to achieve the directional retrieval of private big data.The simulation results show that the retrieval performance,time-consuming,accuracy,and data security of this method are better than those of traditional methods.

关 键 词：相似树查询 隐私数据 大数据 定向检索 模糊理论 关联属性特征 

分 类 号：TP352[自动化与计算机技术—计算机系统结构]