面向云计算的隐私查询保护技术研究

发布时间：2018-11-27 14:55

【摘要】：随着电子商务、社交网络、企业信息化的快速发展,个人及企业产生的数据量越来越多,而云存储的兴起使得企业和个人将大量数据上传至云端,为了保护其中的用户的敏感数据不被泄露,可行的做法是将数据外包至云服务器之前对数据先进行加密处理。如何利用密文搜索技术对加密后的数据进行查询就是隐私查询保护技术研究的问题。在目前的研究成果中,一些学者通过设计优秀的索引及加密算法实现了从单个关键词到多关键词检索的突破,处理效率也得到较大的提升。然而,这些已有的模式大多数只能进行精确匹配,一旦在用户的查询关键词中出现拼写错误,云服务器返回的结果可能不是用户想要得到的查询结果,甚至返回结果可能为空。因此,基于密文的模糊搜索成为了一个新的挑战问题。在最近的研究成果中,已经有部分模式开始支持密文的模糊搜索,但是这些模式几乎都需要构建关键词词典,导致查询效率不够理想。基于此,本文提出了一种新的多关键词模糊搜索方案,该方案同时支持中英文关键词的模糊搜索。本文具体研究工作如下：1.研究了当前的密文可搜索技术,针对基于关键词字典的密文模糊搜索模式,通过分析其查询效率的不足及原因,提出一种新的搜索模式,即采用局部敏感哈希技术和布隆过滤器技术来处理模糊搜索,从而顺利构建模糊搜索模式。2.在前面提出的模式基础之上,进一步提出了针对中文的模糊搜索模式,该模式不仅支持用户输入英文查询,亦可输入中文查询。此外,本文还设计了查询结果排序机制,根据用户输入的关键词,采用自定义加权方法对查询结果排序。该排序机制使得模式能够返回有序的结果列表,位于前列的记录即是与用户查询内容最相关的。3.为了测试本文提出的模式的可行性和搜索效率。针对中、英文的模糊搜索模式,在选取的数据源上进行反复实验,验证了各模式的可行性、高效性、安全性及较高的搜索准确性。实验结果还表明返回给用户的结果集的排列顺序的准确率较高,满足用户的查询期望。
[Abstract]:With the rapid development of electronic commerce, social network and enterprise information, the amount of data generated by individuals and enterprises is increasing, and the rise of cloud storage makes enterprises and individuals upload a lot of data to the cloud. In order to protect the sensitive data of the user from being leaked, it is feasible to encrypt the data before outsourcing it to the cloud server. How to use ciphertext search technology to query encrypted data is the problem of privacy query protection technology. In the current research results, some scholars through the design of excellent indexing and encryption algorithms to achieve a breakthrough from a single keyword to multiple keyword retrieval, processing efficiency has also been greatly improved. However, most of these existing patterns can only be accurately matched. Once a spelling error occurs in the user's query keywords, the result returned by the cloud server may not be the query result that the user wants, or even the result may be empty. Therefore, the fuzzy search based on ciphertext becomes a new challenge. In recent research results, some patterns have begun to support the fuzzy search of ciphertext, but almost all of these patterns need to build keyword dictionaries, resulting in poor query efficiency. Based on this, a new multi-keyword fuzzy search scheme is proposed, which supports both Chinese and English keyword fuzzy search. The specific research work of this paper is as follows: 1. In this paper, the current searchable ciphertext technology is studied. According to the fuzzy search mode of ciphertext based on keyword dictionary, a new search mode is proposed by analyzing the inadequacies and reasons of its query efficiency. That is to say, local sensitive hashing technique and Bron filter technique are used to deal with fuzzy search, so that the fuzzy search pattern can be constructed smoothly. 2. Based on the proposed model, a fuzzy search model for Chinese is proposed, which not only supports users to input English queries, but also inputs Chinese queries. In addition, a query result sorting mechanism is designed in this paper. According to the keywords entered by the user, the query results are sorted by a custom weighting method. The sorting mechanism enables the schema to return an ordered list of results, with the front-row records being the most relevant. 3. 3 to the content of the user query. In order to test the feasibility of the proposed model and search efficiency. In view of the fuzzy search patterns in Chinese and English, repeated experiments are carried out on the selected data sources to verify the feasibility, efficiency, security and high search accuracy of each model. The experimental results also show that the accuracy of the order of the result set returned to the user is high, which meets the query expectation of the user.
【学位授予单位】：东华大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP309

【相似文献】