Art transaction privacy protection and data security encryption method

Abstract

The present application relates to the technical field of computer data security and access control, in particular to an art transaction privacy protection and data security encryption method. The method establishes a state sequence according to an art transaction process, generates a state description object containing a state identifier, a participant principal digest, a data visibility level and a previous state commitment value; divides transaction data into data units, forms core ciphertext using a content key and an authentication encryption algorithm, and derives a state packaging key according to the state description object, the data unit identifier, the transaction random number and the previous state commitment value; when the transaction state migrates, one-way proxy re-encryption is performed according to the state migration voucher, and the threshold key share, the one-time capability token and the access closure are combined to complete authorization verification and content key recovery. The method synchronously converts the transaction permission change into the ciphertext decryption condition change, blocks the decryption of the combination of the old token, the old packaging ciphertext and the cross-state key share, and reduces the plaintext exposure caused by full re-encryption.

Description

Technical Field

[0001] This invention relates to the field of computer data security and access control technology, specifically to methods for privacy protection and data security encryption in art transactions. Background Technology

[0002] Art trading platforms typically handle a wide range of data, including basic artwork information, high-resolution images, ownership certificates, authentication reports, inquiry records, quotation records, transaction vouchers, payment information, and delivery documents. Current technologies often employ a combination of account authentication, role-based access control, and database encryption for protection. When data is written, the platform generates a content key for the file or field, uses a symmetric encryption algorithm to form ciphertext, and then encapsulates the content key using the platform's master key, the user's public key, or a key from a key management service. When a user accesses the platform, the server determines the access scope based on the user's role, transaction participation relationship, and access control table. After successful verification, the content key is decapsulated, and the server either decrypts the data or returns decryptable data to the terminal. Changes during the transaction phase are usually controlled by modifying the access control table, adjusting role markers, shortening the validity period of access tokens, or removing download access points; the ciphertext itself and its key encapsulation relationship generally remain unchanged. For data that has already been downloaded, cached, or copied, the platform can typically only stop subsequent API calls and cannot directly change the decryption conditions of the existing ciphertext.

[0003] It also includes processing methods such as envelope encryption, key rotation, proxy re-encryption, and threshold key recovery. Envelope encryption separates the data ciphertext from the content key, re-encapsulating the content key when permissions change; key rotation restricts subsequent access by changing the master key or version key; proxy re-encryption allows conversion nodes to convert ciphertext intended for the original authorized subject into ciphertext intended for the new authorized subject without obtaining the plaintext; threshold key recovery splits the content key into multiple shares and restores the content key after a preset combination condition is met. Existing implementations typically use user identity, role, or key version as the basis for conversion, and transaction status is only used as a business field in the permission system, without forming a unified cryptographic constraint with ciphertext version, previous status, data visibility level, and the set of participating subjects. Therefore, access control after a state switch mainly relies on server-side records, while the distributed key encapsulation, token, and cached copy still retain the original decryption association.

[0004] When an art transaction status changes, the change in access permissions cannot be synchronously translated into a change in the conditions for decryption. Statuses such as preview, inquiry, bidding, transaction, settlement, and resale correspond to different participants and data visibility ranges. However, the core ciphertext, content key encapsulation, and access tokens in conventional encryption schemes lack a continuous binding relationship with specific transaction statuses. Even after the permission table has been updated, key encapsulations, access tokens, or cached ciphertext generated in the original state can still be used to recover the content key through combinations of old versions. Server-side revocation records cannot directly render distributed data undecryptable. If all transaction data is re-decrypted and re-encrypted to eliminate old authorizations, plaintext needs to be restored at the centralized processing node, and ciphertext needs to be recalculated for a large number of images and files, creating a plaintext exposure range. Therefore, it is necessary to address the problem of the lack of synchronous, continuous, and irreversible constraints between transaction status migration, changes in the authorization set, and the decryptable range of ciphertext. Summary of the Invention

[0005] The purpose of this invention is to provide a method for protecting privacy and encrypting data in art transactions, which can solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: The method for protecting privacy and encrypting data in art transactions includes: establishing a sequence of transaction states with sequential relationships based on the art transaction process, and generating a state description object for each transaction state that includes a state identifier, a summary of the participating entities, a data visibility level, and a commitment value of the preceding state. Art transaction data is divided into multiple data units, each generating a content key and using an authentication encryption algorithm to form the core ciphertext. A state authorization vector is generated based on the participant entity summary and the data visibility level. A state packaging key is derived based on the state description object, data unit identifier, transaction random number and the preceding state commitment value. A state packaging ciphertext is formed by combining the content key and the state authorization vector. During transaction state transition, the state package ciphertext is re-encrypted along the predetermined transition direction based on the state transition certificate to generate the target state package ciphertext. Upon receiving an access request, the status identifier, ciphertext version, and authorization information are verified, the content key is recovered, and the core ciphertext is decrypted.

[0007] Preferably, generating the state description object includes: performing irreversible digest processing on the participant identifiers in the current transaction state, generating a subject authorization bitmap based on the access scope of each participant for different data units, and mapping the subject authorization bitmap and the data visibility level to the state authorization vector; A joint commitment is made to the state description object of the previous transaction state, the encrypted version of the state package, and the summary of the transition event to obtain the commitment value of the preceding state. The current state identifier, the main authorization bitmap, the state authorization vector, the previous state commitment value, and the state effective interval are encoded into normalized state data. The state commitment value of the normalized state data is calculated, and the state commitment value is written into the state description object of the next transaction state.

[0008] Preferably, the state transition certificate consists of a state commitment value of the previous transaction state, a state commitment value of the target transaction state, a set of data units that are allowed to be migrated, a migration direction identifier, and a migration authorization signature; Performing the proxy re-encryption includes: generating a one-way conversion parameter based on the state packaging key corresponding to the previous transaction state and the state packaging key corresponding to the target transaction state; binding the one-way conversion parameter to the state migration certificate; and converting the state packaging ciphertext whose version number, data unit identifier, and preceding state commitment value are all consistent with the state migration certificate. After the conversion is completed, a migration record containing a source version summary and a target version summary is generated, and the migration record is incorporated into the state commitment value calculation of the target transaction state.

[0009] Preferably, for a preset highly sensitive data unit, the corresponding content key is split into multiple key shares according to the set of participating entities in the current transaction state, and the share combination required for recovery is determined based on the state authorization vector; Generate a share-encapsulated ciphertext for each key share, which is bound to the participant digest, the data unit identifier, the ciphertext version, and the state commitment value; When an access request is received, the allowed key shares are extracted based on the status authorization vector corresponding to the requester, the encapsulation binding relationship and version consistency of each key share are verified, and the content key is reconstructed when the share combination is satisfied. When the set of participating entities changes, the content key is re-splitled based on the target transaction status, and a corresponding share-encapsulated ciphertext is generated.

[0010] Preferably, after dividing the art transaction data into multiple data units, a confidentiality dependency graph is constructed based on the reference relationships, generation relationships and sensitivity levels between the data units, and a set of parent data units and a set of affected child data units are recorded for each data unit; When generating the state package ciphertext, the package ciphertext version digest and the current state commitment value corresponding to the parent data unit are incorporated into the state package key derivation input of the child data unit; When the state authorization vector, ciphertext version, or state commitment value of any parent data unit changes, the affected subgraph is determined along the confidentiality dependency graph, the state ciphertext is regenerated for the data units in the affected subgraph, and the original state ciphertext of the unaffected data units is retained.

[0011] Preferably, generating the one-way conversion parameters includes: generating independent conversion factors from the authorization control terminal of the previous transaction state and the authorization control terminal of the target transaction state, respectively, and combining the two conversion factors after commitment exchange into a joint conversion parameter bound to the migration direction identifier; The joint transformation parameters include source state commitment value constraints, target state commitment value constraints, and data unit permission set constraints. Before performing the state-packaged ciphertext transformation, the consistency of the commitments of the two transformation factors and the migration authorization signature are verified. After the conversion is complete, the source state-side conversion factor is deleted, and the reverse conversion factor required for converting the target state package key to the source state package key is omitted from the joint conversion parameters.

[0012] Preferably, determining the share combination includes: generating a subject category matrix based on the authorization category of the participating subject in the current transaction state, using the key shares corresponding to different subject categories for at least two independent share sets, and setting a combination constraint for each share set determined by the state authorization vector; When reconstructing the content key, the subject category, state commitment value and ciphertext version of the submitted shares in each share set are verified respectively, and reconstruction is performed when multiple share sets simultaneously satisfy the corresponding combination constraints; When the target transaction status revokes the access rights of any subject category, update factors with complementary relationships are generated, the update factors are allocated to the key shares of the unrevoked subjects, and the remaining update factors are used to generate compensation shares and corresponding share encapsulated ciphertext.

[0013] Preferably, the access request also carries a one-time capability token, which is generated by the requester's subject digest, the target data unit set, the current state commitment value, the target encrypted version, the token sequence value, and the valid range. Upon receiving an access request, the token sequence value is verified not to be recorded in the used token set, and the current state commitment value is verified to be consistent with the target state package ciphertext. Based on the target data unit set and the subject category matrix, the share package ciphertext that can be submitted is limited. After the content key reconstruction is completed, the token sequence value and access result digest are written into the used token set, and the set commitment value bound to the current state commitment value is updated.

[0014] Preferably, when the same source transaction state is concurrently migrated to multiple candidate transaction states, a branch state commitment value, a branch migration record, and independent joint transformation parameters are generated for each candidate transaction state. After receiving a branch merge request, the common migration data unit and the branch-limited data unit are determined according to the source version summary and target version summary in each branch migration record, and all the branch status commitment values ​​are aggregated into a merge status commitment value. For the common migration data unit, a merge transformation parameter bound to the merge state commitment value is generated. For the branch-limited data unit, the joint transformation parameter of the corresponding branch is retained. The branch identifier, the data unit permission set constraint item, and the merge state commitment value are written together into the state migration certificate of the target transaction state.

[0015] Preferably, before generating the one-time capability token, an access closure is generated based on the direct and indirect reference relationships between the target data units. The encrypted version digest and state commitment value of each data unit in the access closure are aggregated into a closure commitment value according to the data unit identifier order, and the closure commitment value is incorporated into the one-time capability token. When verifying the access request, the access closure is regenerated based on the current transaction status and the current closure commitment value is calculated. The current closure commitment value is verified to be consistent with the closure commitment value in the token and the token sequence value is not recorded in the invalid token set. The ciphertext is encapsulated according to the access closure and the subject category matrix to limit the share that can participate in the content key reconstruction. When any data unit in the access closure undergoes a version migration or state commitment update, the token sequence value containing the original closure commitment value is written into the invalid token set.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The transaction status identifier, participant summary, data visibility level, and prior status commitment value are written into the status description object. The status description object, data unit identifier, transaction random number, and prior status commitment value are used as derived inputs to the status packaging key, binding the encapsulation result of the content key to the current transaction status. During transaction status migration, a one-way proxy re-encryption is performed on the status packaging ciphertext based on the status migration certificate containing the source status commitment value, target status commitment value, migration direction, and permitted data unit set. During the access phase, the status identifier, ciphertext version, authorization information, and status commitment continuity are simultaneously verified. Tokens, packaging ciphertext, or key shares in the original status cannot be combined with data in the target status to recover the content key. Permission revocation can be directly converted into the invalidation of the old ciphertext encapsulation relationship, avoiding the disconnect between status control and ciphertext control caused by relying solely on the server permission table. Status commitments are written continuously along the transaction status sequence, making it impossible to pass subsequent status verification when rolling back to a historical status or replacing an intermediate status.

[0017] 2. After separating the core ciphertext from the state-packaged ciphertext, changes in transaction state only require converting the ciphertext corresponding to the content key. This eliminates the need for full decryption and re-encryption of the artwork's high-resolution image, authentication documents, and delivery documents, reducing the plaintext processing steps during state transitions. The confidentiality dependency graph incorporates the ciphertext version digest and state commitment value of the parent data unit into the key derivation input of the child data unit. When any associated data undergoes a version or authorization change, only the affected subgraph's state-packaged ciphertext is regenerated, while the original core ciphertext of unaffected data units is preserved. The content key corresponding to highly sensitive data is split into multiple shares according to the participating entity category, and protected by ciphertext encapsulated by shares bound to the entity digest, data unit identifier, ciphertext version, and state commitment value. One-time capability tokens are further bound to the access closure, target data unit set, current state commitment value, ciphertext version, and token sequence value, restricting key recovery under conditions of old token replay, cross-state share concatenation, and inconsistent associated data versions. Attached Figure Description

[0018] Figure 1 A flowchart illustrating the overall process of encrypting and decrypting art transaction data status coupling; Figure 2 Flowchart for generating transaction state transition credentials and re-encrypting state-packaged ciphertext via one-way proxy; Figure 3 Flowchart for threshold key allocation, one-time capability token verification, and content key recovery for highly sensitive transaction data; Figure 4 Flowchart for ensuring transaction data confidentiality through propagation, updating affected subgraphs, and merging concurrent transaction state branches. Detailed Implementation

[0019] In one embodiment, please refer to Figure 1The privacy protection and data security encryption method for art transactions includes: establishing a sequence of transaction states with sequential relationships based on the art transaction process, and generating a state description object for each transaction state, containing a state identifier, participant summary, data visibility level, and previous state commitment value; dividing the art transaction data into multiple data units, generating content keys for each unit, and using an authentication encryption algorithm to form core ciphertext; generating a state authorization vector based on the participant summary and data visibility level, and deriving a state packaging key based on the state description object, data unit identifier, transaction random number, and previous state commitment value, and applying this key to the content key and state authorization. Vectors form state-packaged ciphertext; during transaction state transitions, the state-packaged ciphertext is re-encrypted along the predetermined transition direction based on the state transition certificate to generate the target state-packaged ciphertext; upon receiving an access request, the state identifier, ciphertext version, and authorization information are verified, the content key is recovered, and the core ciphertext is decrypted. The transaction state sequence can at least represent the actual state during data registration, restricted display, inquiry, bidding, transaction confirmation, payment verification, delivery, and re-transfer. States form a directed state chain according to the allowed transition relationships. Each state only references the confirmed preceding state, and it is prohibited to directly generate subsequent states from unconfirmed states.

[0020] Specifically, the system establishes a unique transaction identifier for the same artwork transaction and divides the work name, author identifier, creation year, material description, size information, high-resolution image, partial detail image, ownership certificate, appraisal report, restoration record, quotation record, transaction certificate, payment certificate, and delivery documents into data units according to read boundaries. The read boundaries are determined by whether the data requires independent authorization, independent updates, and can be verified independently. Each data unit is configured with a stable data unit identifier and an incrementing ciphertext version number. The content key is independently generated by a cryptographically secure random number generator. The core ciphertext consists of plaintext, random numbers, associated data, and authentication tags. The associated data includes at least the transaction identifier, data unit identifier, and ciphertext version number. Therefore, even if the ciphertext content has not changed, it cannot pass authentication verification if the ciphertext version or the transaction to which it belongs is replaced. After the core ciphertext is generated, it will not be repeatedly decrypted with the normal state migration. State changes only affect the state-packaged ciphertext containing the content key.

[0021] In this embodiment, the state description object is encoded using a deterministic field order. The state identifier indicates the position of the current state in the state sequence. The participant digest is formed by irreversibly digesting the identifiers of the subjects currently authorized to participate in the state. The data visibility level indicates whether each type of data unit is allowed to be read completely, partially, only verified, or prohibited from reading in the current state. The preceding state commitment value is used to fix the description content and encrypted version relationship of the previous state. The state commitment value is calculated according to the following formula: ; in, This represents the state commitment value for the t-th transaction state. This represents the collision-resistant summary operation. Indicates the status flag. This indicates a summary of the participating entities. Indicates the data visibility level code. This represents the commitment value of the preceding state. Indicates the effective period of the state. This indicates that the fields are concatenated in a predetermined order. For example, when the third state is the bidding state, the bidding state identifier, the set of bidder entity summaries, the visibility level of the bid data, the commitment value of the second inquiry state, and the bidding effective range are concatenated before calculation. The replacement of any field will result in a different state commitment value.

[0022] The state packaging key is not directly generated from the transaction master key, but is derived from the transaction random number, data unit identifier, current state field, and previous state commitment value. The calculation relationship is as follows: ; in, Representing data unit The state-packaged key in state t. This represents deterministic derivation operations based on key materials. Represents the random number of the transaction. This indicates the data unit identifier. For example, the same appraisal report may have the same core encrypted text in both the inquiry and transaction confirmation states, but because... , , and Different states result in different state packaging keys. The state packaging ciphertext consists of a content key, a state authorization vector, a current state commitment value, a ciphertext version, and an authentication tag. When desealing, the same associated fields as those used during encapsulation must be provided.

[0023] In a preferred embodiment, generating a state description object includes: performing irreversible digest processing on the participant identifiers in the current transaction state; generating a subject authorization bitmap based on the access scope of each participant for different data units; mapping the subject authorization bitmap to the data visibility level as a state authorization vector; jointly committing to the state description object of the previous transaction state, the state package ciphertext version, and the migration event digest to obtain the preceding state commitment value; encoding the current state identifier, subject authorization bitmap, state authorization vector, preceding state commitment value, and state effective interval into normalized state data; calculating the state commitment value of the normalized state data; and writing the state commitment value into the state description object of the next transaction state. The subject authorization bitmap is sorted according to the data unit identifier, with each bit corresponding to one of the authorization codes: complete read, restricted read, verification only, and prohibited access. A random salt is added to the participant identifier before entering the digest operation to avoid the same subject in different transactions forming directly comparable digest results.

[0024] The state authorization vector is generated along three dimensions: subject, data unit, and transaction state. The specific calculation relationship is as follows: ; in, This represents the final authorization bit of subject u to data unit i in state t. This represents the basic authorization bit of the corresponding data unit in the main authorization bitmap. This indicates whether the subject belongs to the set of participating subjects in its current state. This indicates whether the data visibility level of the data unit in the current state allows the requested access type. This represents a logical AND operation. For example, if a bidder has basic access authorization to a high-definition image and is a participant in the bidding process, but the high-definition image is currently only allowed to be read under limited access, then the bidder must fully read the corresponding image. Taking 0 restricts the reading of the corresponding value. If set to 1, the state authorization vector is formed by concatenating all valid authorization bits in the order of data unit identifiers, and serves as the protected content of the state package ciphertext.

[0025] Furthermore, the normalized state data adopts a fixed field order, a fixed-length field header, and a deterministic set sorting rule. The participating entity summary set is arranged in ascending order of summary value, and the entity authorization bitmap is arranged in a dual order of participating entity summary and data unit identifier. The state effective interval is encoded with a unified time base. The migration event summary at least covers the migration initiator summary, the migration target state identifier, the affected data unit set summary, and the source ciphertext version summary. When the system receives the state description object, it re-executes normalization encoding and calculates the state commitment value. If the received result is inconsistent with the commitment value carried by the state description object, no state packaging ciphertext is generated, and subsequent state migrations are not accepted, so that the participating entity set, authorization scope, data version, and state order enter the same verifiable data chain.

[0026] In another preferred embodiment, please refer to Figure 2 The state transition certificate consists of the state commitment value of the previous transaction state, the state commitment value of the target transaction state, the set of data units allowed to be migrated, the migration direction identifier, and the migration authorization signature. The proxy re-encryption process includes: generating one-way transformation parameters based on the state packaging key corresponding to the previous transaction state and the state packaging key corresponding to the target transaction state; binding the one-way transformation parameters to the state transition certificate; and transforming the state packaging ciphertext, whose version number, data unit identifier, and preceding state commitment value are all consistent with the state transition certificate. After the transformation, a migration record containing the source version digest and the target version digest is generated, and the migration record is incorporated into the calculation of the state commitment value of the target transaction state. The migration direction identifier is formed by encoding the source state identifier and the target state identifier according to a directed state chain. The migration authorization signature covers all fields of the certificate, and the set of data units allowed to be migrated is represented by a sorted data unit identifier digest.

[0027] During proxy re-encryption, the content key in the source state package ciphertext remains encrypted. The transformation node reads the source state commitment value, target state commitment value, data unit identifier, source version, and target version, and uses one-way transformation parameters to transform the source package component into the target package component. The transformation node cannot calculate the source state package key or the target state package key from the one-way transformation parameters. The associated data of the target state package ciphertext is rewritten as the target state commitment value, target version, and target authorization vector. The transformation node also generates a migration record, which includes the source package ciphertext digest, the target package ciphertext digest, the migration credential digest, the migration time code, and the transformation result authentication tag. When the target state is established, all migration records are sorted according to the data unit identifier to form a migration record digest, and the migration record digest is incorporated into the extended field of the target state commitment value.

[0028] Preferably, generating unidirectional conversion parameters includes: generating independent conversion factors from the authorization control end of the previous transaction state and the authorization control end of the target transaction state respectively; combining the two conversion factors after commitment exchange into a joint conversion parameter bound to the migration direction identifier; the joint conversion parameter includes source state commitment value constraint items, target state commitment value constraint items, and data unit permission set constraint items; before performing state packaging ciphertext conversion, verifying the commitment consistency of the two conversion factors and the migration authorization signature; after conversion, deleting the source state side conversion factor, and omitting the reverse conversion factor required for conversion from the target state packaging key to the source state packaging key in the joint conversion parameter; each of the two authorization control ends generates a random conversion factor and submits the factor commitment value first; after both parties' commitment values ​​are fixed, they exchange the encrypted conversion factors; the joint conversion parameter is generated only when the commitment value verification passes.

[0029] Specifically, the source state commitment value constraint is used to limit the convertible package ciphertext to carry the specified source state commitment value. The target state commitment value constraint is used to fix the target state that the conversion result can enter. The data unit permission set constraint is used to limit the conversion scope to the data unit set registered in the migration certificate. The joint conversion parameter also carries the migration direction identifier and the effective usage count. Each time the conversion node completes a data unit conversion, it writes the corresponding migration record and increments the count. After all permitted data unit conversions are completed, the joint conversion parameter is closed. If any of the data unit identifier, source version, previous state commitment value, or migration direction of a state package ciphertext is inconsistent, the conversion node refuses to output the target package component. The source state side conversion factor is deleted from the runtime context after the conversion is completed. The remaining content only includes the factor commitment value and the migration record summary.

[0030] In a preferred embodiment, please refer to Figure 3 For preset highly sensitive data units, the corresponding content key is split into multiple key shares according to the set of participating entities in the current transaction state, and the share combination required for recovery is determined based on the state authorization vector. For each key share, a share-encapsulated ciphertext is generated, which is bound to the participating entity digest, data unit identifier, ciphertext version, and state commitment value. When an access request is received, the allowed key shares are extracted according to the state authorization vector corresponding to the requester, and the encapsulation binding relationship and version consistency of each key share are verified. When the share combination is satisfied, the content key is reconstructed. When the set of participating entities changes, the content key is split again based on the target transaction state and the corresponding share-encapsulated ciphertext is generated. The highly sensitive data unit can be an undisclosed transaction price, payment voucher, complete ownership identity, and undisclosed authentication documents. The specific scope is determined by the data visibility level code.

[0031] Key shares are generated using a polynomial allocation method over a finite field. The content key is reversibly encoded and mapped to elements of a finite field. A polynomial with a degree lower than the recovery threshold is randomly generated, and share values ​​are calculated using different non-zero indices. When the share combination is satisfied, the content key is recovered according to the following formula: ; in, This represents the finite field element corresponding to the content key of data unit i. This represents the set of indexes of shares currently participating in the recovery. This represents the key share value corresponding to index j, m represents other share indices in the same set, and p represents the modulus of a large prime number. For example, when the recovery threshold is 3 and the qualified share indices are 1, 4, and 7, then... By taking 1, 4, and 7 and substituting them into the three share values, the problem can be restored. Any two shares cannot determine the polynomial constant term. After restoration, the finite field elements are inversely mapped to the content key byte sequence.

[0032] In this embodiment, the associated data of each share encapsulated ciphertext includes a participant summary, data unit identifier, ciphertext version, state commitment value, share index, and subject category. The share value is authenticated and encrypted using a share packaging key jointly determined by the corresponding participant and the current state. The server only stores the share encapsulated ciphertext and share index, and does not store the plaintext shares that can directly recover the content key. After an access request arrives, the system parses the subject category and data unit range that the requester can submit from the state authorization vector, uncapsulates the qualified shares, and verifies the authentication tag. Only shares with consistent ciphertext version, state commitment value, and data unit identifier are included in the interpolation calculation. Shares from the old state, other data units, or other ciphertext versions will not be combined.

[0033] Further, determining the share combination includes: generating a subject category matrix based on the authorization category of the participating subject in the current transaction state; using the key shares corresponding to different subject categories for at least two independent share sets; and setting combination constraints determined by the state authorization vector for each share set; when reconstructing the content key, verifying the subject category, state commitment value, and ciphertext version of the submitted shares in each share set, and performing reconstruction when multiple share sets simultaneously satisfy the corresponding combination constraints; when the target transaction state revokes the access qualification of any subject category, generating update factors with complementary relationships, allocating the update factors to the key shares of the unrevoked subjects, and using the remaining update factors to generate compensation shares and corresponding share-encapsulated ciphertexts. The rows of the subject category matrix correspond to the participating subject categories, the columns correspond to the data unit types, and the matrix elements record whether the corresponding category can participate in the recovery and the share set to which it belongs.

[0034] For example, the platform custodian's share and the shares of both parties in the transaction are assigned to the first share set, and the authenticator's share and the ownership confirmation's share are assigned to the second share set. The content key recovery condition is set to the first share set reaching the internal limit and the second share set reaching the internal limit. The system performs share legality verification on the two sets respectively, and then combines the intermediate recovery quantities output by the two sets into the content key according to the predetermined combination relationship. When an authenticator is revoked in the target state, the system does not decrypt the core ciphertext, but generates an update factor group with a total of 0 based on the share commitment corresponding to the original content key. The non-zero update factor is added to the share value of the unrevoked subject and repackaged. The remaining complementary factor is written into the share of the new authenticator or compensation subject, so that the old authenticator's share and the updated other shares cannot form a consistent interpolation set.

[0035] In a further embodiment, please refer to Figure 4 After dividing the art transaction data into multiple data units, a confidentiality dependency graph is constructed based on the reference relationships, generation relationships, and sensitivity levels between the data units. For each data unit, a set of parent data units and a set of affected child data units are recorded. When generating state-packaged ciphertext, the digest of the ciphertext version corresponding to the parent data unit and the current state commitment value are incorporated into the state-packaged key derivation input of the child data unit. When the state authorization vector, ciphertext version, or state commitment value of any parent data unit changes, the affected subgraph is determined along the confidentiality dependency graph. State-packaged ciphertext is regenerated for the data units within the affected subgraph, while the original state-packaged ciphertext of unaffected data units is retained. Reference relationships indicate that a child data unit needs to reference a parent data unit semantically or for verification. Generation relationships indicate that a child data unit is formed by processing by a parent data unit. Sensitivity level relationships indicate that the authorization scope of the parent data unit imposes an upper limit constraint on downstream data units.

[0036] The affected subgraphs are determined according to directed reachability relations, and their set expression is: ; in, This represents the set of subgraph nodes affected when the parent data unit x changes. This indicates that there exists a directed path from x to y in the confidentiality dependency graph. This indicates whether the dependency type on this path requires inheritance of the wrapped ciphertext version or state commitment. A value of 1 means that node y will be included in the update scope, while a value of 0 means it will not. For example, if the ownership certificate is the parent data unit and the transaction certificate references the ownership certificate digest, a change in the ownership certificate version will affect the corresponding transaction certificate. The value is 1, while the value for directory files, which are only related to the display order, can be 0. The system only considers sets. The data unit in the middle is used to re-derive the state packaging key.

[0037] The confidential dependency graph is stored using both an adjacency list and a reverse index. Each directed edge records the dependency type, inherited field, and edge version. The packaged ciphertext version digest of the parent data unit is formed by connecting the data unit identifier, ciphertext version, state commitment value, and packaged ciphertext digest. When a child data unit derives its state package key, all parent data unit version digests are sorted according to the parent data unit identifier and then incorporated into the derived input. After a parent data unit changes, the system performs a breadth-first traversal using the reverse index. When an edge is encountered where the inherited field does not contain the changed field, the propagation along that edge stops. After the affected subgraph is determined, the state package ciphertext is updated in topological order, ensuring that the parent data unit version digest referenced by each child data unit during the update is the current version. Data units that have not entered the affected subgraph retain their original package ciphertext and original ciphertext version.

[0038] In a further preferred embodiment, the access request also carries a one-time capability token, which is generated by the requester's subject digest, the target data unit set, the current state commitment value, the target ciphertext version, the token sequence value, and the valid range. After receiving the access request, it is verified that the token sequence value is not recorded in the used token set, and the current state commitment value is verified to be consistent with the target state packaged ciphertext. The ciphertext is packaged according to the target data unit set and the subject category matrix to limit the share that can be submitted. After the content key reconstruction is completed, the token sequence value and the access result digest are written into the used token set, and the set commitment value bound to the current state commitment value is updated. The one-time capability token is protected by a signature and carries a token version. The token sequence value is generated by the transaction identifier, the request random number, and the monotonic count value.

[0039] The used token set is stored in a set structure partitioned by state commitment value. Each partition records the token sequence value digest, access result digest, target data unit set digest, and write order. When an access request is received, the token signature and valid range are verified first. Then, the corresponding partition is located according to the current state commitment value, and the token sequence value digest is retrieved. If the search result exists, the share is refused to be unsealed. If the search result does not exist, the token sequence value is locked and the share verification is initiated. After the content key is recovered, the locked record is converted into a used record. If the access process terminates abnormally, the locked record is converted into a non-reusable record. The set commitment value is formed by aggregating the records in the partition according to the write order and is written into the access audit data of the current state, so that the same token cannot be accepted repeatedly due to concurrent requests.

[0040] Furthermore, before generating a one-time capability token, an access closure is generated based on the direct and indirect reference relationships between target data units. The ciphertext version digest and state commitment value of each data unit within the access closure are aggregated into a closure commitment value according to the data unit identifier order, and the closure commitment value is incorporated into the one-time capability token. When verifying an access request, the access closure is regenerated based on the current transaction state and the current closure commitment value is calculated. It is verified that the current closure commitment value is consistent with the closure commitment value in the token and that the token sequence value is not recorded in the invalid token set. The ciphertext is encapsulated according to the access closure and the subject category matrix to limit the share that can participate in the content key reconstruction. When any data unit in the access closure undergoes version migration or state commitment update, the token sequence value containing the original closure commitment value is written into the invalid token set.

[0041] The closure commitment value is generated according to the following formula: ; in, This represents the access closure commitment value. This represents the ordered set of data units obtained by tracing upwards from the target data unit along direct and indirect reference relationships. Indicates the identifier of the data unit within the set. This represents the digest of the corresponding ciphertext version. This indicates the current associated state commitment value of the data unit. This indicates that the data units are connected consecutively in the order of their identifiers. For example, when accessing a transaction document, if the document references an appraisal report and a certificate of ownership, then... It contains three data units, and a version migration of any of the referenced data units will cause... Change, carrying the original One-time capability tokens are added to the expiration token set.

[0042] In a concurrent transaction state processing embodiment, when the same source transaction state is concurrently migrated to multiple candidate transaction states, branch state commitment values, branch migration records, and independent joint conversion parameters are generated for each candidate transaction state. After receiving a branch merging request, common migration data units and branch-limited data units are determined according to the source version summary and target version summary in each branch migration record. All branch state commitment values ​​are aggregated into a merged state commitment value. Merging conversion parameters bound to the merged state commitment value are generated for the common migration data units. For the branch-limited data units, the joint conversion parameters of the corresponding branches are retained. The branch identifier, data unit permission set constraint items, and merged state commitment value are jointly written into the state migration certificate of the target transaction state. Concurrent branches can correspond to the payment verification branch and the delivery document verification branch that are performed simultaneously in the transaction confirmation state. Each branch is generated from the same source state commitment value but has a different data unit permission set.

[0043] During branch merging, the system first verifies that each branch migration record references the same source state commitment value. Then, it compares the source version summary and target version summary of each data unit in each branch. Data units with the same source and target versions are included in the common migration set, while data units that appear only in some branches or have different target versions are included in the branch-limited set. The merge state commitment value is generated according to the following formula: ; in, This represents the commitment value for the merged status. to This represents the branch status commitment value after sorting by branch identifier. Represents a summary of the set of shared migration data units. This represents a set of data units that define a branch and a summary of the branch's affiliation. For example, two branches form... and The common migration set contains basic information about the work, and the branch-limited sets contain payment vouchers and delivery documents, respectively. Substituting the corresponding abstracts into these sets yields a unique result. .

[0044] The merge transformation parameters only apply to the state package ciphertext in the common migration set. Its source constraint item references the consistent source version digest in each branch migration record, and its target constraint item references the merge state commitment value. The branch-limited data unit continues to use the original branch joint transformation parameters and retains the branch identifier in the state package ciphertext. When the target transaction state reads the common migration data unit, it verifies the merge state commitment value. When reading the branch-limited data unit, it verifies both the merge state commitment value and the corresponding branch state commitment value. If two branches form different target versions for the same data unit, the system does not automatically select either version, but marks the data unit as a pending branch-limited data unit and suspends the generation of the corresponding target state package ciphertext until a branch merge request containing a clear version selection is received.

[0045] In the access processing phase of this embodiment, the accessing party submits an art transaction identifier, a requester's subject summary, a target data unit set, a one-time capability token, and an available share encapsulated ciphertext. The system reads the current state description object and recalculates the state commitment value, verifying that the preceding state commitment value in the state description object is consistent with the state chain record. Subsequently, it verifies the current state commitment value, closure commitment value, target ciphertext version, and valid range in the one-time capability token. After the token is approved, the system determines the set of shares that can be used from the subject category matrix, and then verifies the subject summary, data unit identifier, ciphertext version, state commitment value, share index, and subject category of each share encapsulated ciphertext. When the constraints of each share set combination are satisfied, the content key is restored, and decryption is performed using the associated data and authentication tag carried in the core ciphertext. If any verification is inconsistent, the intermediate shares and temporary content key of this process are cleared.

[0046] In the specific application of the transaction status transition from confirmation to settlement, the system uses the confirmation status commitment value as the source status commitment value to generate a target status description object containing a summary of the participants in the settlement status, the visibility level of the settlement document, and the restriction level of the payment voucher. The payment voucher and the settlement document form data units respectively. The content key corresponding to the payment voucher is split according to the subject categories of the two parties to the transaction, the platform custodian, and the payment verifier. The settlement document references the ownership certificate and the transaction certificate and enters the confidentiality dependency graph. The state transition certificate only allows the data units that have entered the settlement status to be converted. The proxy re-encryption node converts the corresponding state package ciphertext into the target package ciphertext. The bidding bid records that have not entered the permission set retain the original package ciphertext and do not enter the settlement status status authorization vector. When the accessing party reads the settlement document, it generates an access closure containing the ownership certificate, the transaction certificate, and the settlement document, and binds a one-time capability token with the closure commitment value.

[0047] In specific applications where the transaction status transitions from the settlement state to the re-transfer state, the system regenerates the participant summary and the subject authorization bitmap based on the new set of participants. The original buyer is converted into the current ownership entity, the original bidder is removed from the participant set, and a status authorization vector is generated based on the new data visibility level. The original state package ciphertext is transformed into the target package ciphertext through joint transformation parameters along the direction from the settlement state to the re-transfer state. Data units involving the original payment entity's identity and the undisclosed transaction price are re-splitted into content key shares and generate share-encapsulated ciphertexts. The original state one-time capability token enters the invalid token set due to changes in the state commitment value and closure commitment value. Data units referencing the original ownership proof in the confidentiality dependency graph update the state package ciphertext along the affected subgraph. Data units whose core ciphertext content remains unchanged continue to retain the original core ciphertext.

[0048] In practice, all state description objects, migration credentials, migration records, state packaging ciphertext, share encapsulation ciphertext, one-time capability tokens, and set commitment records use a unified versioned encoding format. Each encoded object includes object type, encoding version, transaction identifier, field length, and field content. Set fields are deterministically sorted before encoding, and the digest field clearly defines the corresponding original encoding range. The associated data used for authentication encryption is consistent with the immutable fields in the object encoding. The parser stops processing when it encounters an unknown encoding version, duplicate fields, inconsistent field order, or out-of-bounds length. State commitments, migration signatures, packaging ciphertext authentication tags, and share encapsulation authentication tags are verified separately. If any verification fails, subsequent key recovery or proxy re-encryption processing will not proceed.

[0049] In anomaly recovery, the system uses three types of persistent records for state transitions: preparation records, transition records, and acknowledgment records. Preparation records fix the source state commitment value, the target state commitment value, and the set of data units allowed to migrate. Transition records save the source package ciphertext digest and the target package ciphertext digest item by item. Acknowledgment records save the aggregate digest of all transition records and the target state commitment value. During system recovery, only the target state with a complete acknowledgment record is set to an accessible state. When a preparation record exists but an acknowledgment record is missing, the already generated target package ciphertext remains inaccessible and is cleaned up according to the transition record, while the source state package ciphertext remains available. When an acknowledgment record exists but the access index has not been updated, the target state access index is rebuilt according to the acknowledgment record to avoid mixed decryption conditions formed by some data units entering the new state while the rest of the data units remain in the old state.

[0050] When deleting temporary key materials, the system sets limited lifecycles for state-packaged keys, source state-side transition factors, unsealed key shares, intermediate recovery quantities, and content keys. Key materials only exist in the corresponding encapsulation, transformation, or decryption context. Upon completion of processing or abnormal exit, they are overwritten and released by unified cleanup logic. Long-term stored data includes only core ciphertext, state-packaged ciphertext, share-packaged ciphertext, commitment values, signatures, migration records, and token set records. Access audit data does not record plaintext content keys, plaintext shares, or decrypted art transaction data. Audit correlation is completed through transaction identifier digests, data unit identifiers, ciphertext versions, state commitment values, and access result digests.

[0051] Through the above implementation methods, art transaction data is bound to transaction identifiers and data unit identifiers from the generation of core ciphertext. The state description object incorporates the set of participating entities, data visibility level, and prior state commitments into a continuous state chain. The state packaging ciphertext fixes the recoverability conditions of the content key in the current state. One-way proxy re-encryption limits state transitions to a predetermined direction and a permitted set of data units. Threshold shares, entity category matrices, one-time capability tokens, and access closures jointly limit the recovery conditions of highly sensitive data. The confidentiality dependency graph and concurrent branch merging mechanism handle related data updates and parallel state transitions. All types of verification fields are carried by clearly coded objects and cryptographic operations, which can be directly implemented according to data structure, operation relationship, and processing order.

Claims

1. A method for privacy protection and data security encryption in art transactions, characterized in that, include: Based on the art transaction process, a sequence of transaction states with sequential relationships is established, and a state description object containing a state identifier, a summary of the participating entities, a data visibility level, and a commitment value of the preceding state is generated for each transaction state. Art transaction data is divided into multiple data units, each generating a content key and using an authentication encryption algorithm to form the core ciphertext. A state authorization vector is generated based on the participant entity summary and the data visibility level. A state packaging key is derived based on the state description object, data unit identifier, transaction random number and the preceding state commitment value. A state packaging ciphertext is formed by combining the content key and the state authorization vector. During transaction state transition, the state package ciphertext is re-encrypted along the predetermined transition direction based on the state transition certificate to generate the target state package ciphertext. Upon receiving an access request, the status identifier, ciphertext version, and authorization information are verified, the content key is recovered, and the core ciphertext is decrypted.

2. The method for protecting privacy and encrypting data in art transactions according to claim 1, characterized in that, Generating the state description object includes: performing irreversible digest processing on the participant identifiers in the current transaction state, generating a subject authorization bitmap based on the access scope of each participant for different data units, and mapping the subject authorization bitmap and the data visibility level to the state authorization vector; A joint commitment is made to the state description object of the previous transaction state, the encrypted version of the state package, and the summary of the transition event to obtain the commitment value of the preceding state. The current state identifier, the main authorization bitmap, the state authorization vector, the previous state commitment value, and the state effective interval are encoded into normalized state data. The state commitment value of the normalized state data is calculated, and the state commitment value is written into the state description object of the next transaction state.

3. The method for protecting privacy and encrypting data in art transactions according to claim 1, characterized in that, The state transition certificate consists of the state commitment value of the previous transaction state, the state commitment value of the target transaction state, the set of data units that can be migrated, the migration direction identifier, and the migration authorization signature; Performing the proxy re-encryption includes: generating a one-way conversion parameter based on the state packaging key corresponding to the previous transaction state and the state packaging key corresponding to the target transaction state; binding the one-way conversion parameter to the state migration certificate; and converting the state packaging ciphertext whose version number, data unit identifier, and preceding state commitment value are all consistent with the state migration certificate. After the conversion is completed, a migration record containing a source version summary and a target version summary is generated, and the migration record is incorporated into the state commitment value calculation of the target transaction state.

4. The method for protecting privacy and encrypting data in art transactions according to claim 2, characterized in that, For the preset highly sensitive data unit, the corresponding content key is split into multiple key shares according to the set of participating entities in the current transaction state, and the share combination required for recovery is determined according to the state authorization vector; Generate a share-encapsulated ciphertext for each key share, which is bound to the participant digest, the data unit identifier, the ciphertext version, and the state commitment value; When an access request is received, the allowed key shares are extracted based on the status authorization vector corresponding to the requester, the encapsulation binding relationship and version consistency of each key share are verified, and the content key is reconstructed when the share combination is satisfied. When the set of participating entities changes, the content key is re-splitled based on the target transaction status, and a corresponding share-encapsulated ciphertext is generated.

5. The method for protecting privacy and encrypting data in art transactions according to claim 1, characterized in that, After dividing the art transaction data into multiple data units, a confidentiality dependency graph is constructed based on the reference relationships, generation relationships and sensitivity levels between the data units, and a set of parent data units and a set of affected child data units are recorded for each data unit. When generating the state package ciphertext, the package ciphertext version digest and the current state commitment value corresponding to the parent data unit are incorporated into the state package key derivation input of the child data unit; When the state authorization vector, ciphertext version, or state commitment value of any parent data unit changes, the affected subgraph is determined along the confidentiality dependency graph, the state ciphertext is regenerated for the data units in the affected subgraph, and the original state ciphertext of the unaffected data units is retained.

6. The method for protecting privacy and encrypting data in art transactions according to claim 3, characterized in that, Generating the one-way conversion parameters includes: generating independent conversion factors from the authorization control terminal of the previous transaction state and the authorization control terminal of the target transaction state, and combining the two conversion factors after commitment exchange into a joint conversion parameter bound to the migration direction identifier; The joint transformation parameters include source state commitment value constraints, target state commitment value constraints, and data unit permission set constraints. Before performing the state-packaged ciphertext transformation, the consistency of the commitments of the two transformation factors and the migration authorization signature are verified. After the conversion is complete, the source state-side conversion factor is deleted, and the reverse conversion factor required for converting the target state package key to the source state package key is omitted from the joint conversion parameters.

7. The method for protecting privacy and encrypting data in art transactions according to claim 4, characterized in that, Determining the share combination includes: generating a subject category matrix based on the authorization category of the participating subject in the current transaction state, using the key shares corresponding to different subject categories for at least two independent share sets, and setting a combination constraint for each share set determined by the state authorization vector; When reconstructing the content key, the subject category, state commitment value and ciphertext version of the submitted shares in each share set are verified respectively, and reconstruction is performed when multiple share sets simultaneously satisfy the corresponding combination constraints; When the target transaction status revokes the access rights of any subject category, update factors with complementary relationships are generated, the update factors are allocated to the key shares of the unrevoked subjects, and the remaining update factors are used to generate compensation shares and corresponding share encapsulated ciphertext.

8. The method for protecting privacy and encrypting data in art transactions according to claim 7, characterized in that, The access request also carries a one-time capability token, which is generated by the requester's subject digest, the target data unit set, the current state commitment value, the target encrypted version, the token sequence value, and the valid range. Upon receiving an access request, the token sequence value is verified not to be recorded in the used token set, and the current state commitment value is verified to be consistent with the target state package ciphertext. Based on the target data unit set and the subject category matrix, the share package ciphertext that can be submitted is limited. After the content key reconstruction is completed, the token sequence value and access result digest are written into the used token set, and the set commitment value bound to the current state commitment value is updated.

9. The method for protecting privacy and encrypting data in art transactions according to claim 6, characterized in that, When the same source transaction state is concurrently migrated to multiple candidate transaction states, a branch state commitment value, a branch migration record, and independent joint transformation parameters are generated for each candidate transaction state. After receiving a branch merge request, the common migration data unit and the branch-limited data unit are determined according to the source version summary and target version summary in each branch migration record, and all the branch status commitment values ​​are aggregated into a merge status commitment value. For the common migration data unit, a merge transformation parameter bound to the merge state commitment value is generated. For the branch-limited data unit, the joint transformation parameter of the corresponding branch is retained. The branch identifier, the data unit permission set constraint item, and the merge state commitment value are written together into the state migration certificate of the target transaction state.

10. The method for protecting privacy and encrypting data in art transactions according to claim 8, characterized in that, Before generating the one-time capability token, an access closure is generated based on the direct and indirect reference relationships between the target data units. The encrypted version digest and state commitment value of each data unit in the access closure are aggregated into a closure commitment value according to the data unit identifier order, and the closure commitment value is incorporated into the one-time capability token. When verifying the access request, the access closure is regenerated based on the current transaction status and the current closure commitment value is calculated. The current closure commitment value is verified to be consistent with the closure commitment value in the token and the token sequence value is not recorded in the invalid token set. The ciphertext is encapsulated according to the access closure and the subject category matrix to limit the share that can participate in the content key reconstruction. When any data unit in the access closure undergoes a version migration or state commitment update, the token sequence value containing the original closure commitment value is written into the invalid token set.

Images