Research On Keyword Searchable Encryption Based On Cloud Data Privacy Protection

The widespread adoption and use of smart mobile devices has led to a rapid increase in the amount of data and the speed of data growth.Such a large amount of data makes it difficult for users to store,analyze,and manage them.The capabilities of traditional databases are now far behind.As a new mode of data services,cloud computing can provide users with services of storage and computing through the Internet,providing effective support for the management and analysis of massive data.In cloud computing,data owners usually upload their data to the cloud server,which can reduce the users' storage costs and provide users with quick access to services.But this kind of situation needs to consider the privacy of the data,so these data will be encrypted before uploading to the cloud server.At the same time we also need offer the service of searching for the user,so searchable encryption technology arises at the historic moment.Through the research on the existing problems of searchable encryption technology,such as the relatively simple support of logical operations ?the exponential increase in the cost of searching as the number of documents increases exponentially?insufficient search accuracy,etc.So we intend to increase the accuracy of search results,enrich functions,and improve search efficiency.So our paper builds two searchable encryption schemes: Support Fine-grained Multi-keyword Ranking Searchable Encryption Scheme and Multi-keyword Ranking Searchable Encryption Scheme Based on Cluster Index.The specific research content is divided into the following two aspects:(1)In this thesis,we propose a searchable encryption scheme that supports fine-grained multi-keyword ranking search.By quantifying the search user's preference,we generate preference factor.Then the cloud server comprehensively considers the weight of the keyword and the user's preference factor when performing the search.The weight of the keywords can make the cloud server return more accurate results.The definition of the preference factor can allow the user to define the importance of each keyword.Through the combination of the two parts,the user can get the result that meets their needs.At the same time,in this program,we also support mixed logical queries for multiple keywords,such as AND,OR,and NOT.We continuously narrow the search space by layer-by-layer screening to quickly find the documents the user needs.So the proposed scheme can support more logical operations and improve the functionality of our scheme.(2)In this thesis,we propose a multi-keyword ranking searchable encryption scheme based on cluster index.The program will divide the documents in the document collection into different clusters,and then create indexes for each cluster and documents in the cluster.In the search process,the cluster with the highest correlation will be matched first.Then the documents will be taken out from this cluster and calculate the correlation score.Finally the documents with high correlation will return to the user.This method saves a lot of computational overhead and improves the retrieval efficiency.At the same time,the TF-IDF weight calculation method is used to calculate the relevance score,which improves the accuracy of the search.So this scheme satisfies the high-efficiency search and improves the accuracy of the search results at the same time.