Searchable encryption method based on semantic sorting of keywords

Abstract

The invention discloses a searchable encryption method based on semantic sorting of keywords, includes such steps as generating encryption key by data owner and sending it to authorized user, extracting keywords from plaintext document set, constructing document mark vector according to keywords and sending it to private cloud server, creating secure index tree and sending it to public cloud server; The authorized user sends the query mark vector generated by the search keyword to the private cloud server, and sends the generated query trap to the public cloud server. The private cloud servermatches the query tag vector with the document tag vector, and sends the set of security index identifiers that may satisfy the user's query request to the public cloud server. The public cloud servercalculates document similarity, sorts the calculation results, and returns the set of ciphertext documents with the highest similarity to the authorized users. By realizing the multi-keyword semanticsorting method, the invention not only ensures the data privacy safety, but also effectively improves the retrieval efficiency.

Description

technical field [0001] The invention relates to the field of cloud security, in particular to a searchable encryption method based on keyword semantic ranking. Background technique [0002] With the rapid development of cloud technology, more and more users outsource their data to cloud servers. However, while cloud servers provide users with high-quality data storage services, they also have a great impact on users' data privacy. threaten. In order to ensure that the user's sensitive data is not leaked, most data owners use encryption technology to encrypt and store the data before uploading the data to the cloud server. Therefore, some search technologies for encrypted data have emerged as the times require. After the data is encrypted and stored in the cloud server, neither the cloud server administrator nor external attackers can obtain the real content of the data, which protects the privacy of users. But this has brought great challenges to the information retrieval ...

AI Technical Summary

Problems solved by technology

[0003] When the searchable encryption technology was first proposed, it used a symmetric key to encrypt each word in the document with a double-layer structure, but the full-text search method was too inefficient, and then a keyword-based public key searchable encryption system was proposed , so that any user with a public key can store data to the server, and only users with a private key can search for keywords, but this bilinear pairing-based structure makes retrieval efficiency very low, and is not suitable for large-scale data cloud computing environment

[0004] In order to better meet the search request, a keyword fuzzy search scheme was proposed. By constructing a keyword fuzzy set, it can tolerate spelling mistakes and inconsistent formats when users search, but the disadvantage is that they cannot search for documents related to the semantics of keywords.

[0005] Although some encryption methods in the existing research methods support retrieval methods such as semantic fuzzy search, multi-keyword search, parallel search, and similarity retrieval, they cannot take semantic similarity into document scoring.

Images