23 results about "Private information retrieval" patented technology

The invention relates to the technical field of information encryption and decryption, can be applied to business scenes such as financial science and technology and medical health, and discloses a filter matrix keyword retrieval method, device and equipment and a medium. Receiving a query keyword and generating a hash position set, constructing an index vector and generating an encrypted query vector and query state information, executing homomorphic inner product processing by the server based on the filter matrix and the encrypted query vector to generate encrypted response data, and reconstructing and generating a query result based on the query state information. According to the method, encryption query and result reconstruction are achieved through the binary fusion filter and the homomorphic inner product mechanism, query content leakage is avoided, calculation and communication overhead is reduced through matrix processing, the keyword retrieval function is kept, meanwhile, the efficiency is close to index type PIR, and the private information retrieval practicability is improved.

The invention provides a multi-keyword private information retrieval method and system based on inverted indexes, and belongs to the field of information security and privacy computing. The method comprises the following steps that: a server splits an original database into a reverse index part of a keyword-index set and a key value pair part of an index-data value; mapping the index set into an integer by utilizing Godel coding; processing the data by adopting probability batch coding and binary random linear coding and publishing a hash function; the client maps a multi-keyword query into a bucket by using a hash function to generate a query vector, and the query vector is sent to the server after being subjected to fully homomorphic encryption; the server performs homomorphic calculation on a matching result in a ciphertext state and returns the result; and the client obtains a matching index set after decryption verification, and then initiates a second round of query to obtain final data. According to the method, symmetric privacy protection of multi-keyword non-primary key query in a single-server environment is realized, leakage of query content and database information is effectively prevented, and private information retrieval security is improved.

The invention discloses a dynamic single-server client preprocessing private information retrieval method, and belongs to the technical field of private information retrieval. The method comprises a preprocessing stage: a client generates a private prompt and stores the private prompt in a local device; a query and response stage: the client accelerates query by using a private prompt; in the refreshing stage, the client repositions the used private prompt and updates the private prompt stored locally; and a dynamic updating stage: when the server updates or deletes the original data item or adds a new data item, the client directly updates the private prompt stored locally. According to the method, the client storage time and the server online calculation time are optimally balanced, and meanwhile, efficient updating of the dynamic database is supported. And when the database of the server dynamically changes, the private prompt stored in the local equipment of the client can be updated within the expected PRP calling time, so that the private prompt updating efficiency in a dynamic database scene is remarkably improved.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for batch retrieving data are described. In one aspect, a method includes receiving, from a client device and by a first multi-party computation (MPC) server of a cluster of MPC servers, a batch request for retrieving multiple database values stored in one or more databases. The batch request includes a first byte array that includes, for each requested key of multiple requested keys, a first secret share of the requested key. Each database includes multiple data items that each include a database key and a corresponding value. The MPC server processes each database key to generate first secret shares of matching data indicating whether the database key matches at least one requested key. The MPC server generates one or more results that represent database values corresponding to each database key that matches at least one requested key.

This application proposes a verifiable private information retrieval method against RAG knowledge base poisoning attacks, comprising the following steps: The server maps each data fragment in the knowledge base to a polynomial form in a polynomial ring to obtain the server-side knowledge base; the client generates a query polynomial and a private key based on the polynomial ring, generates a private commitment based on the private key, constructs a private query, and sends it to the server; the server performs ciphertext field retrieval calculation on the polynomial ring based on the private query to generate a ciphertext polynomial for the retrieval response; the client performs decryption operation on the ciphertext polynomial for the retrieval response in the response message using the private key to obtain the retrieval result. This scheme generates a unique version identifier and knowledge base commitment for the server-side knowledge base. When the knowledge base content is legally updated, a new knowledge base commitment and version identifier are generated and published simultaneously, without the need to reconstruct the entire knowledge base index, significantly reducing the preprocessing overhead and time cost of dynamic updates to the knowledge base in private retrieval.

The application discloses a kind of private information retrieval method, equipment, storage medium and program product of supervision, it is related to cryptography and privacy computing technical field.The method comprises: by data holder, public verifiable cryptography commitment is completed in system initialization phase to database version, and database version commitment is recorded to evidence chain by supervision committee, and threshold signature system is established;By supervision committee, the compliance audit is carried out to external authorization certificate and authorization message provided by inquirer, threshold signature is generated when audit passes, and authorization event is recorded to evidence chain;By inquirer, authorization query package is assembled based on threshold signature and is submitted to data holder;In the case where data holder is verified to threshold signature and authorization query package and confirms that authorization is legal and is consistent with database version commitment binding, response ciphertext is generated.The application realizes the authorization decision and after-the-fact accountability of encrypted query, reduces the trust assumption to single-point approval subject.

The application discloses a dynamic single-server client preprocessing private information retrieval method, and belongs to the technical field of private information retrieval. The method comprises a preprocessing stage: a client generates a private hint and stores the private hint in a local device; a query and response stage: the client uses the private hint to accelerate query; a refreshing stage: the client repositions the used private hint and updates the locally stored private hint; and a dynamic updating stage: when the server updates or deletes original data items or adds new data items, the client directly updates the locally stored private hint. The application realizes the best trade-off between the client storage and the server online calculation time, and supports efficient updating of a dynamic database. When the database of the server dynamically changes, the private hint stored in the local device of the client can be updated within the expected PRP calling time, and the updating efficiency of the private hint in a dynamic database scenario is remarkably improved.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for batch retrieving data are described. In one aspect, a method includes receiving, from a client device and by a first multi-party computation (MPC) server of a cluster of MPC servers, a batch request for retrieving multiple database values stored in one or more databases. The batch request includes a first byte array that includes, for each requested key of multiple requested keys, a first secret share of the requested key. Each database includes multiple data items that each include a database key and a corresponding value. The MPC server processes each database key to generate first secret shares of matching data indicating whether the database key matches at least one requested key. The MPC server generates one or more results that represent database values corresponding to each database key that matches at least one requested key.

The application belongs to the technical field of privacy protection, and provides a DID-based authorized private information retrieval method, which comprises the following steps: a receiver and a possessor register to generate a DID and a DID document and upload them to a blockchain; the possessor encrypts a file through a symmetric key to obtain ciphertext and uploads the ciphertext to a database server; the receiver sends file information to the possessor; the possessor signs file attribute information of the corresponding file to obtain a signature value, encapsulates the signature value and the DID of the possessor into a verifiable credential, and sends the verifiable credential to the receiver; the receiver analyzes the verifiable credential to obtain the signature value, and then performs an OSBE protocol on the signature value among the receiver, the database server and the possessor; the receiver obtains the symmetric key, the database server sends the ciphertext to the receiver, and the receiver decrypts the ciphertext through the symmetric key to obtain the file. The application not only meets the query privacy requirement, but also provides a data retrieval scheme in which access control and data subject authorization mechanism are unified.

The application relates to the technical field of anonymous communication and provides an anonymous communication method and system based on blind signature and a medium, the method comprising the following steps: a sender generates a blind signature request containing public metadata and a message token, sends the blind signature request to a signature server, and the signature server performs blind signature; the sender performs de-blinding on the received blind signature result, obtains an anonymous token, and sends an authorization request to a bulletin board server; the bulletin board server verifies the validity of the anonymous token, allocates a mailbox and a corresponding writing time window to the sender; the sender obtains the writing permission of the mailbox within the time window, and then delivers a ciphertext to the mailbox; a receiver sends a private information retrieval request to the bulletin board server based on the mailbox identifier informed by the sender, and receives returned response ciphertext; and the receiver uses a PIR private key and a session key to decrypt the response ciphertext, and obtains a plaintext message. The application guarantees the strong privacy of both communication parties and realizes the safe transmission of confidential messages.

The invention belongs to the technical field of privacy protection, and provides a DID-based authorizable private information retrieval method, which comprises the following steps: a receiver and an owning party register to generate a DID and a DID document and upload the DID and the DID document to a block chain, the owning party encrypts the document through a symmetric key to obtain a ciphertext and uploads the ciphertext to a database server, the receiver sends the file information to the owning party, and the owning party sends the file information to the database server. The owner signs the file attribute information of the corresponding file to obtain a signature value, packages the signature value and the DID of the owner into a verifiable voucher, sends the verifiable voucher to the receiver, analyzes the verifiable voucher to obtain a signature value, then executes an OSBE protocol on the signature value among the receiver, the database server and the owner, the receiver obtains a symmetric key, and sends the symmetric key to the database server. And the database server sends the ciphertext to a receiver, and the receiver decrypts the ciphertext through the symmetric key to obtain the file. According to the method, the query privacy requirement is met, and a data retrieval scheme in which access control and a data subject authorization mechanism are unified is provided.

The invention discloses a large-model fine-tuning intelligent disaster recovery big data early warning analysis method and system, and the method comprises the steps: obtaining disaster index data from each distributed data center, carrying out the clustering cleaning of the data through a clustering algorithm, and forming a standardized clustering data set; mapping the data of different disaster levels into computing power tasks, and quantifying the computing power demand of each task in combination with a computing power efficiency value; performing three-stage fine tuning on the basic large model based on the standardized data set and the computing power efficiency value, wherein the three-stage fine tuning comprises ridge regression preliminary training, difficult point data intensified training and special small model generation, and joint fine tuning according to disaster grades; and a longitudinal joint prediction algorithm is adopted, homomorphic encryption and private information retrieval technologies are combined, and the occurrence probability of each disaster grade and a final early warning result are output. The system comprises a data preprocessing module, a calculation power conversion module, a large model fine tuning module and a joint prediction and early warning module. According to the invention, the accuracy, real-time performance and privacy protection capability of disaster recovery early warning are improved.

A system for performing private information retrieval may include a data requesting computer that is configured to receive a query string, homomorphically encrypt the query string to generate an encrypted query, transmit the encrypted query to a data repository computer, and receive resultant encrypted datasets. The data repository computer may be configured to access a database comprising a plurality of datasets, receive the encrypted query, and perform a search of the plurality of datasets to identify the resultant encrypted datasets that include data that matches the encrypted query for provision to the data requesting computer. The search may be performed such that encryption of the encrypted query is maintained while performing computational operations of the search.

The present application belongs to the technical field of password leakage detection, and provides a password leakage detection method and system based on private information retrieval, which performs password leakage service detection, obtains a username-password pair, encodes the username-password pair, interacts with a server, obtains an unintentional key-value storage data structure stored by the server, decodes the encoded username-password pair by using the unintentional key-value storage data structure, obtains a specific location index of a data bucket in which the username-password pair exists, performs private information retrieval between the unintentional key-value storage data structure and the server, obtains a data set in which the username-password pair to be detected exists, searches the obtained data set for whether there is an encoded value matching the encoded value of the username-password pair to be detected, and respectively prompts according to the result. The present application can solve the problems of low efficiency, insufficient security, and insufficient integration effect of existing password leakage detection, and improve the efficiency of password leakage detection.

The invention discloses a client non-inductive distributed private information retrieval method supporting metadata protection, which relates to the technical field of communication security and comprises the following steps: initializing a system to determine a public parameter set; the system distributes a message matrix slot position for the online client and completes the establishment of communication connection between the two communication clients; the online client encrypts a transmission message and transmits the encrypted transmission message to the server, and a corresponding matrix slot of the client is filled with ciphertext information at the server side to obtain a message matrix; the online client sends a private information retrieval request according to the matrix slot position of the communication opposite side; performing client login and distributed private information retrieval calculation under a challenge-response mechanism; the server performs final private information retrieval request calculation and returns a final response to the requesting client, and the client recovers communication plaintext data by using a private key; according to the invention, the calculation pressure of the server side is effectively reduced; and meanwhile, metadata such as communication time and communication objects are protected.

Disclosed is a server device. The device includes: a communication interface configured to communicate with an external device; a memory storing a plurality of reference data; and a processor. Here, the processor is configured to obtain an inner product operation result based on the plurality of reference data and a homomorphic ciphertext if the homomorphic ciphertext generated using a coefficient encoding method is received from the external device through the communication interface, and transmit at least one of private information retrieval (PIR) result data, private membership test (PMT) result data, or private set intersection (PSI) result data, obtained based on the inner product operation result, to the external device through the communication interface.

The application relates to the technical field of information security and relates to a high-throughput private information retrieval method suitable for a dynamic database.The method comprises the following steps: S1, server preprocessing: inputting security parameters, a public seed and a database, and outputting offline hints containing the first dimension of the database; S2, client generating query information: inputting the security parameters, the public seed and an index (i, j) to be queried, and outputting query information qu and a private key sk; S3, server online responding to the query: inputting the database, the query information qu and the offline hints, and outputting response information; and S4, client decrypting the response information: inputting the private key sk and the response information, and outputting retrieved data items.The application further improves the throughput, eliminates the need for downloading offline hints by the client, ensures that the server only returns the information required by the user, greatly shortens the time for private information retrieval, and significantly improves the user experience and data retrieval efficiency.

The invention discloses a discrete asymmetric private information retrieval coding method based on a random feedforward neural network. The method is composed of three parts of neural networks: 1, a mutual information estimation neural network which estimates mutual information according to database responses and original messages; 2, a KL (Kullback-Leibler) divergence estimation neural network is used for estimating the KL divergence of a database response and quantifying the KL divergence into privacy constraints; and 3, coding the neural network, generating a response according to the database message and binarizing the response into discrete output. According to the method, low-packetization coding can be realized in a discrete asymmetric private information retrieval scene under the condition of any database and message number.

The invention relates to the technical field of anonymous communication, and provides an anonymous communication method and system based on blind signature and a medium, and the method comprises the steps that a sender generates a blind signature request containing public metadata and a message token, the blind signature request is sent to a signature server, and the signature server carries out blind signature; the sender carries out blind removal on the received blind signature result, obtains an anonymous token and sends an authorization request to the bulletin board server, and the bulletin board server distributes a mailbox and a corresponding write-in time window for the sender after verifying the validity of the anonymous token; the sender delivers a ciphertext to the mailbox after obtaining a write-in permission to the mailbox in the time window; the receiver sends a privacy information retrieval request to the bulletin board server based on the mailbox identifier notified by the sender, and receives a returned response ciphertext; and the receiver decrypts the response ciphertext by using the PIR private key and the session key to obtain a plaintext message. According to the invention, strong privacy of both communication parties is ensured, and secure transmission of confidential messages is realized.

The invention provides a data query method and device, equipment, a medium and a program product, which are applied to the technical field of information security, and the method comprises the following steps: in the case of receiving a query request sent by a client, responding to the query request, and based on a bucket index in the query request and a query mask in the query request, querying the bucket index and the query mask in the query request; querying in the B hash buckets to obtain a query result; and sending the query result to the client. According to the method, on the basis of realizing privacy information retrieval, an original data set in a server is distributed into B hash buckets, when data are queried, the whole original data set does not need to be traversed, only data related to the hash buckets corresponding to bucket indexes need to be processed in each query, the calculation amount and communication traffic are remarkably reduced, and the data query efficiency is improved. Compared with the prior art, the method has the advantages that calculation, storage and communication expenses are reduced, the amount of data needing to be processed in each query is reduced, query efficiency and response speed can be improved, and both high efficiency and safety of data query are considered.

The application discloses a batch privacy information retrieval method and device, and the method comprises the following steps: a client and a server agree on a hash function, the serial number index in an original database is calculated through a cuckoo hash technology, the result of each serial number index for each hash function is reserved, and a first hash bucket containing a plurality of baskets is obtained; the client fills the index of batch query into a second hash bucket with the same length as the first hash bucket through the cuckoo hash algorithm, and the filling result is encoded in a manner of using a fixed weight code for coding; the client encrypts the obtained coding result by using a homomorphic encryption technology as a batch query request; the server compares the batch query request with the first hash bucket by using the homomorphic encryption technology, obtains a corresponding intermediate result, performs corresponding calculation on the intermediate result and the original database, and obtains corresponding query result ciphertext; and the client decrypts the query result ciphertext, so that the query result of the batch query request is obtained.

The application discloses a kind of extremely small storage verifiable private information retrieval method, device and system, belong to information security and cryptography field. Offline pre-processing stage, client sends challenge ciphertext matrix to server, server calculates proof ciphertext matrix and returns to client, and client is compressed into extremely small persistent state data by random linear combination with challenge matrix with the proof matrix obtained by decryption;Online query stage, client sends error learning query ciphertext with homomorphic private key ciphertext, server calculates response ciphertext and outsourcing ciphertext;Client completes linear verification and decryption using persistent state data. In addition, the present application also introduces public commitment and cheating detection mechanism, provides covert security protection for stateless client. The present application completely decouples client storage overhead and database size, realizes thousands of times storage efficiency improvement, and is suitable for deploying high security privacy retrieval on mobile devices and Internet of Things terminals.

The invention belongs to the technical field of artificial intelligence, information retrieval and data encryption, and discloses a privacy protection approximate nearest neighbor search method and system of a retrieval enhancement generation system, the method realizes approximate nearest neighbor index by using LSH, greatly reduces the quantity of vectors to be compared on the premise of ensuring privacy, and realizes quick query. Experiments prove that the scheme adopts the ciphertext inner product calculation of the ANN index, and as the scale of a corpus is increased, the retrieval delay degradation degree is far lower than that of methods based on homomorphic k neighbor query, such as RemoteRAG and the like. Particularly, when the number of documents is large, the advantages of ShieldRAG are more obvious, query can be completed within millisecond-level time, and due to the fact that homomorphic cosine similarity calculation and an additional private information retrieval mechanism are introduced into Remote RAG, delay is obviously higher.