基于k-mean聚类算法的加密云数据排序搜索仿真

Simulation of Sorting and Searching Encrypted Cloud Data Based on K-Mean Clustering Algorithm

密钥加密索引向量生成加密数据索引时,会产生数据维度升高问题,导致加密云数据排序搜索时间较长,搜索结果不准确,资源利用率低。为此提出基于k-mean聚类算法的加密云数据排序搜索方法。采用k-mean聚类算法对加密云数据进行预处理,选取质心点,并对数据点进行重新聚类,计算出最新的质心点;根据获取的数据聚类结果,选取文档中具有高代表性的关键词作为文档的索引,同时引入加密搜索插件,对索引向量进行维度缩减,使其在检索时形成和用户检索相关的文档摘要,从而能够发送摘要供用户选择目标文档,得到排序搜索结果,实现加密云数据排序搜索。仿真结果表明,所提方法的搜索结果准确率较高,能够有效提升资源利用率,缩短搜索时间。

When the index vector of key encryption generates the encrypted data index,the data dimension will be increased.In addition,the time of sorting and searching time of the encrypted cloud data is long,the search results are not accurate,and the resource utilization is low.Therefore,a sorting and searching method for encrypted cloud data based on K-mean clustering algorithm was proposed.At first,K-mean clustering algorithm was used to prepro⁃cess the encrypted cloud data and select the center of mass.Meanwhile,the data points were clustered again.Then,the latest center of mass was calculated.According to the data clustering result,the highly representative keywords in document were selected as the index of document.Moreover,the encrypted search plug-in was introduced to reduce the dimension of index vector,so that document summarization related to user retrieval was formed,and the summary could be sent to users to select the target documents.Finally,the sorting search result was obtained.Thus,the sorting search of encrypted cloud data was achieved.Simulation results show that the proposed method has high accuracy of search results,so it can effectively improve the resource utilization and shorten the search time.