Abstract: Nowadays more and more enterprise and individual users store a large amount of data in the cloud.To protect data privacy,important data has to be encrypted before being stored in the cloud,which brings a severe challenge to the retrieval of data.Traditional retrieval schemes based on plaintext have not been applicable,and existing retrieval schemes based on ciphertext have many shortcomings,some of which do not support fuzzy search or multi-keyword search, poor efficiency or large space overhead,or do not return ranking results.Therefore,researching secure and efficient retrieval scheme on ciphertext is of great significance.A new fuzzy multi-keyword retrieval scheme over encrypted data in cloud computing was proposed,which can retrieve the ciphertext containing multiple keywords from cloud ser-ver,support fuzzy keyword search,and will not leak any plaintext information of data and retrieval to cloud server and other attackers.In the scheme,counting bloom filter and MinHash algorithm are used to construct index vectors and query vectors,which makes the process of building index and querying more efficient,and the ranking results more accurate.The security analysis and performance evaluation show that our scheme has high security,reliability,retrieval efficiency and accuracy.

Key words: Cloud computing,Fuzzy search,Multi-keyword search,Encrypted data,Secure index