Research On Privacy-Perserving Search On Encrypted Data In Cloud Computing

With the rapid development of the e-business, social networking and enterprise information system, the amount of data produced by individuals and enterprises is increasing heavily, with the rise of cloud storage, enterprises and individuals stored their data in the cloud for great flexibility and economic savings. However, the sensitive data of users may be leaked. Since the researcher made a breakthrough on ciphertext search technology, the searchable encryption technology is the key to protectthesearch privacy and data confidentiality in cloud computing.The researchershad improved the existing schemes to support the search of multiple keyword by designing a novel encryption algorithm. However, most of these existing schemes can only support exact search. Once misspelled keyword occurs in the query, Results which the searcher want to retrieve may not be returned by the cloud server. The results returned from the cloud sever even is null. Therefore, the fuzzy search based on encrypted data becomes a new challenge task. In recent research results, we found that multi-keyword fuzzy search over encrypted data have acquired some achievements. But most of those schemes need to build a predefined keyword dictionary which will reduce query efficiency. In this paper we proposed a novel scheme to support the multi-keywords fuzzy search and also contain Chinese search.The specific research work of this paper is as follows:1. Studying the current ciphertext search technology. We got the conclusion that the fuzzy search based on a keyword dictionary has a low search efficiency by analyze it’s details. While we exploit a novel technology——locality-sensitive hashing to solve the fuzzy match. This technology should be combined with bloom filter to build the fuzzy search scheme.2. The Chinese fuzzy search scheme were proposed based on the scheme we proposed in previous work.This scheme can solve the query which could include misspelled Chinese keywords. In addition, we designed a ranking mechanism of query results,it used a method of custom weighting to complete results sorting.To add results sorting mechanism so that the record which at the top of the result list is the most relevant record for the user’s query.3. In order to test the feasibility and efficiency of the proposed scheme in this paper. We made experiments with selected datasource for Chinese fuzzy search and English fuzzy search, theexperimental results show that bothschemes can tackle the fuzzy searchwithin a predetermined acceptable range successfully.The experimental results also shows that the accuracy of the sequence of documents which obtained from the server is very high,it can satisfy the user’s query expectations.