Block chain-based evaluation data evidence storage method and system, product and equipment

Abstract

The embodiment of the invention relates to an evaluation data evidence storage method and system based on a block chain, a computer program product and electronic equipment, and relates to the technical field of block chains, and the method comprises the steps: carrying out the model evaluation processing of original data according to an evaluation rule generated based on an evaluation data evidence storage protocol, and obtaining the evaluation data of the original data; and constructing a multi-layer hash chain, and in response to an evaluation evidence storage request, storing the evaluation data to a block chain through the constructed multi-layer hash chain based on a Merkel tree. According to the invention, the reliability of evaluation data evidence storage can be improved.

Description

Background Art

[0002] With the rapid development of artificial intelligence technology, large language models are increasingly used in various fields. The performance evaluation of these models has become a key link to ensure their reliability and effectiveness. However, existing evaluation methods mainly focus on the design of evaluation schemes and the formulation of evaluation standards, lacking standardized management of the evaluation process and a credible guarantee mechanism.

[0003] The traditional evaluation data management method has problems such as inconsistent data standards, opaque evaluation process, difficult to verify results, and low reliability of evaluation data storage. Summary of the invention

[0004] The purpose of the present disclosure is to provide a blockchain-based evaluation data notarization method, a blockchain-based evaluation data notarization system, a computer program product, and an electronic device, thereby at least to a certain extent overcoming the problem of low reliability of evaluation data notarization caused by the limitations and defects of related technologies.

[0005] According to one aspect of the present disclosure, a blockchain-based model evaluation data notarization method is provided, including: performing model evaluation processing on original data according to evaluation rules generated based on an evaluation data notarization protocol to obtain evaluation data of the original data; constructing a multi-layer hash chain, and in response to an evaluation notarization request, storing the evaluation data in the blockchain through the constructed multi-layer hash chain based on a Merkle tree.

[0006] In an exemplary embodiment of the present disclosure, the evaluation data evidence protocol includes an evaluation rule protocol; the model evaluation processing is performed on the original data according to the evaluation rules generated based on the evaluation data evidence protocol to obtain the evaluation data of the original data, including: configuring the evaluation rules according to the evaluation rule protocol and initializing the evaluation environment; using the original data as the data to be evaluated, and performing model evaluation processing on the data to be evaluated in the initialized evaluation environment based on the evaluation rules to generate evaluation data corresponding to the original data.

[0007] In an exemplary embodiment of the present disclosure, the constructing of a multi-layer hash chain includes: constructing the multi-layer hash chain based on a data layer, a process layer, and a result layer; each layer in the multi-layer hash chain is associated with the previous layer through a hash value.

[0008] In an exemplary embodiment of the present disclosure, the evaluation data is stored in the blockchain through a multi-layer hash chain constructed based on a Merkle tree, including: obtaining a hash value of a data layer, a hash value of a process layer, and a hash value of a result layer corresponding to the evaluation data; unifying the hash value of the data layer, the hash value of the process layer, and the hash value of the result layer into a target hash value through a Merkle tree, and storing the target hash value in the blockchain through a blockchain evidence storage protocol.

[0009] In an exemplary embodiment of the present disclosure, the hash value of the data layer is the hash value corresponding to the original data and metadata, the hash value of the process layer is the hash value of the key node information of the model evaluation processing, and the hash value of the result layer is the hash value of the evaluation data and verification data.

[0010] In an exemplary embodiment of the present disclosure, the target hash value is stored in the blockchain through the blockchain evidence protocol, including: verifying the data format of the evaluation data corresponding to the target hash value, and packaging the evaluation data corresponding to the target hash value into blocks to generate an evidence voucher.

[0011] In an exemplary embodiment of the present disclosure, the method further includes: verifying the evaluation data to obtain a verification result, and storing the verification result in the blockchain.

[0012] According to one aspect of the present disclosure, there is provided an evaluation data storage system based on blockchain, including: an evaluation data generation module, used to perform model evaluation processing on original data according to evaluation rules generated based on an evaluation data storage protocol, and obtain evaluation data of the original data; an evaluation data storage module, used to construct a multi-layer hash chain, and in response to an evaluation storage request, store the evaluation data to the blockchain through the constructed multi-layer hash chain based on a Merkle tree.

[0013] According to one aspect of the present disclosure, a computer program product is provided, including a computer program, wherein when the computer program is executed by a processor, the evaluation data storage method based on blockchain described in any one of the above is implemented.

[0014] According to one aspect of the present disclosure, there is provided an electronic device, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute any one of the above-mentioned blockchain-based evaluation data storage methods by executing the executable instructions.

[0015] In the technical solution provided in the embodiments of the present disclosure, on the one hand, a multi-layer hash chain is combined with the Merkle tree technology to build a complete evidence storage and certification system. The multi-layer hash chain ensures the traceability of the entire evaluation process, and the Merkle tree structure optimizes the verification efficiency of large-scale data. This combination mechanism ensures the reliability of evidence storage, while improving system performance and reducing storage costs. On the other hand, by configuring the evaluation data evidence storage protocol, the standard format and processing specifications of the evaluation data are defined, and standardized management of the evaluation data is achieved.

[0016] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the specification are used to explain the principles of the present disclosure. Obviously, the accompanying drawings described below are only some embodiments of the present disclosure, and for ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without creative work.

[0018] Figure 1 The following schematically shows a flowchart of a blockchain-based evaluation data storage method in an embodiment of the present disclosure.

[0019] Figure 2 The overall architecture diagram of the evaluation data storage system in the embodiment of the present disclosure is schematically shown.

[0020] Figure 3 The specific flow chart of evaluation data storage in the embodiment of the present disclosure is schematically shown.

[0021] Figure 4 The block diagram of the evaluation data storage system based on blockchain in the embodiment of the present disclosure is schematically shown.

[0022] Figure 5 The block diagram of the electronic device according to the embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0023] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as being limited to the examples set forth herein; on the contrary, these embodiments are provided so that the present disclosure will be more comprehensive and complete, and the concepts of the example embodiments are fully conveyed to those skilled in the art. The described features, structures, or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced while omitting one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other cases, known technical solutions are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0024] In addition, the accompanying drawings are only schematic illustrations of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0025] In some embodiments, traditional evaluation data management methods have problems such as inconsistent data standards, opaque evaluation processes, and difficult to verify results. Although some evaluation platforms have introduced blockchain technology to store evaluation data, they only stay at the simple data storage level and fail to build a complete evaluation storage system, which cannot meet the industry's needs for the credibility of the evaluation process and the verifiability of the results.

[0026] In order to solve the above technical problems, the present disclosure provides a blockchain-based model evaluation data storage method, which can be applied to the application scenario of blockchain-based model evaluation data storage for any type of model. Figure 1 As shown in , the blockchain-based model evaluation data storage method mainly includes the following steps:

[0027] In step S110, the original data is subjected to model evaluation processing according to the evaluation rules generated based on the evaluation data storage protocol to obtain evaluation data of the original data;

[0028] In step S120, a multi-layer hash chain is constructed, and in response to the evaluation evidence request, the evaluation data is stored in the blockchain through the constructed multi-layer hash chain based on the Merkle tree.

[0029] The technical solution provided in the embodiments of the present disclosure, on the one hand, combines multi-layer hash chains with Merkle tree technology to build a complete evidence storage and certification system. The multi-layer hash chain ensures the traceability of the entire evaluation process, and the Merkle tree structure optimizes the verification efficiency of large-scale data. This combination mechanism ensures the reliability of evidence storage, while improving system performance and reducing storage costs. On the other hand, by configuring the evaluation data evidence storage protocol, the standard format and processing specifications of the evaluation data are defined, and standardized management of the evaluation data is achieved.

[0030] Next, refer to Figure 1 The blockchain-based evaluation data storage method in the embodiment of the present disclosure is specifically described in detail.

[0031] In step S110, model evaluation processing is performed on the original data according to the evaluation rules generated based on the large evaluation data storage protocol to obtain evaluation data of the original data.

[0032] In the embodiment of the present disclosure, the raw data in the raw data set can first be submitted to the data preparation stage, and the raw data in the raw data set can be processed in the data preparation stage. After the raw data is processed, the model evaluation process is performed on the raw data in the evaluation execution stage to obtain the evaluation data corresponding to the raw data. In some embodiments, before the raw data is evaluated, data standardization processing and raw data encryption can be performed in the data preparation stage to process the raw data in the raw data set and generate data fingerprints. Among them, the data standardization process can include format conversion, quality verification and metadata extraction. Metadata can be data describing the raw data in the raw data set, mainly information describing the attributes of the raw data, used to support functions such as indicating storage location, historical data, resource search, file records, etc. Exemplarily, when the raw data is processed, data standardization processing can be performed specifically according to the model evaluation data packet protocol. The evaluation data attestation protocol (MEDAP, Model Evaluation Data Attestation Protocol) refers to a data attestation standard protocol specifically for large model evaluation scenarios. The evaluation data attestation protocol defines the format specification, storage structure and verification mechanism of the evaluation data to ensure the standardization and credibility of the evaluation process. The evaluation data notarization protocol includes the data set package protocol, the evaluation rule protocol, the evaluation result protocol, and the blockchain notarization protocol. When the original data is subjected to data standardization, it can be standardized according to the data set package protocol in the evaluation data notarization protocol. The data set package protocol in the evaluation data notarization protocol can be as shown in Table 1. The data set package protocol includes the basic information layer, the encryption information layer, and the metadata layer, and each layer has specific field definitions and verification rules. The protocol supports serialization and deserialization of data to ensure consistent understanding of the data.

[0033] Table 1

[0034]

[0035]

[0036] When encrypting the original data, the AES-256 algorithm can be used to encrypt the encrypted data, and the SHA-256 algorithm can be used to generate a data fingerprint for the encrypted data.

[0037] Among them, the AES-256 algorithm is an Advanced Encryption Standard that uses a 256-bit key length. It is a symmetric encryption algorithm. Its encryption process may include: Key expansion: Generate a series of round keys based on the input key. Initial round: Perform XOR operation on the plaintext and the first round key. Main round: Use multiple rounds of substitution, row shift, column confusion and round key addition operations to confuse and diffuse the data. Final round: In the last round, no column confusion operation is performed, only round substitution, row shift and round key addition operations are performed.

[0038] The SHA-256 algorithm is an algorithm that converts any length of message into a shorter, fixed-length message digest. For any length of message, SHA256 will generate a 256-bit hash value, called a message digest. Based on this, encrypted data can be converted into a data fingerprint to ensure the security and integrity of the data.

[0039] Data fingerprint refers to the unique identifier generated after cryptographic processing of the original data set. The data fingerprint can contain data content hash, metadata information, etc., which is used to ensure the uniqueness and integrity of the data and is an important part of the evidence storage process. The data fingerprint is used to verify the integrity of the data and whether it is the same original data.

[0040] Figure 2 The overall system architecture of the evaluation data storage system is schematically shown in Figure 2 As shown in , the overall architecture of the system may include a data layer, an evaluation layer, a proof layer, and a verification layer. Based on this, data standardization processing and encryption of raw data in the original data set can be performed in the data layer to generate a data fingerprint.

[0041] After the data fingerprint is generated, the original data can be subjected to model evaluation processing to generate evaluation data of the original data. The evaluation data here refers to the evaluation results. Model evaluation processing refers to large model evaluation. Large model evaluation is used to represent the systematic testing and evaluation process of the performance, effect and capability of the model. In some embodiments, large model evaluation refers to the evaluation process performed while protecting data privacy. Among them, the model can be any type of large-scale artificial intelligence model, for example, a large language model or other machine learning model.

[0042] In the disclosed embodiment, when the original data is subjected to model evaluation processing to obtain evaluation data, the evaluation rules can be configured and the evaluation environment can be initialized. Further, the original data can be used as the data to be evaluated, and the model evaluation processing can be performed on the data to be evaluated in the initialized evaluation environment based on the configured evaluation rules to generate evaluation data. Among them, the evaluation rules can be configured according to the evaluation rule protocol in the evaluation data storage protocol, and the evaluation rule protocol can include evaluation dimensions and operation configurations. The evaluation dimensions can include weight configurations, scoring rules, and threshold definitions. The evaluation rule protocol can refer to that shown in Table 2.

[0043] Table 2

[0044]

[0045] When executing the model evaluation process, resource allocation can be performed, and parallel model evaluation processing can be performed based on the allocated resources to obtain evaluation data of the original data. The content of the evaluation data is determined according to the evaluation data protocol in the evaluation data storage protocol, and the evaluation data protocol refers to the evaluation result protocol. As shown in Reference Table 3, the evaluation data protocol may include basic information and result data. Among them, the basic information may include evaluation identification, data set association, model information, and execution timestamp; the result data may include original responses, dimension scores, and statistical indicators. The evaluation data packet protocol includes a multi-layer structure represented by basic information, encrypted information, and metadata.

[0046] Table 3

[0047]

[0048] refer to Figure 2 As shown in the overall system architecture of the evaluation data storage system, the model evaluation processing can be performed in the evaluation layer, which is mainly used to generate evaluation data, encrypt and store the evaluation data, generate evaluation logs, etc.

[0049] Exemplarily, the input of the model evaluation process can be encrypted original data and configured evaluation rules. The steps of the model encryption process may include: verifying the compliance of the evaluation rules; configuring the evaluation environment parameters; executing the model evaluation task; collecting the original evaluation results; encrypting the evaluation data; and recording the evaluation execution log. The output of the model evaluation process is the encrypted evaluation data and the execution log.

[0050] In the disclosed embodiment, by configuring the measurement data evidence storage protocol (MEDAP), the standard format and processing specifications of the measurement data are defined, including the three-layer structure of basic information, encrypted information and metadata, so as to realize the standardized management of the measurement data. Through the unified data interface and verification mechanism, the problems of inconsistent measurement data format and chaotic management are effectively solved, and the standardization and efficiency of the measurement process are improved.

[0051] In step S120, a multi-layer hash chain is constructed, and in response to the evaluation evidence request, the evaluation data is stored in the blockchain through the constructed multi-layer hash chain based on the Merkle tree.

[0052] In the disclosed embodiment, after generating the evaluation data, an evaluation evidence request may be submitted to the blockchain service. The evaluation evidence request is used to store the evaluation data. After receiving the evaluation evidence request, a multi-layer hash chain may be constructed. The multi-layer hash chain refers to a hierarchical data integrity protection mechanism, in which the hash value of each evaluation data may be calculated. The hash values ​​of two evaluation data are combined to calculate the hash value, and then combined with the hash value of the next evaluation data to calculate the hash value. And so on, finally the hash value of the end of the chain is obtained.

[0053] In some embodiments, when constructing a multi-layer hash chain, for a certain task scenario, a multi-layer hash chain can be constructed based on the original data, the key node information of the task execution, and the task processing result. For example, for the application scenario of model evaluation processing, a multi-layer hash chain can be constructed based on the original data, the key node information of the model evaluation processing, and the evaluation results. Exemplarily, for the application scenario of model evaluation processing, the constructed multi-layer hash chain includes a data layer, an evaluation layer, and a result layer, and the data layer, the evaluation layer, and the result layer all contain their respective hash values. By constructing hash associations based on hash values ​​at different levels such as the data layer, the evaluation layer, and the result layer, a complete proof chain is formed to achieve traceability of the entire evaluation process. Among them, the data layer includes the hash value of the original data in the original data set and the hash value of the metadata. The process layer includes the hash value of the key node information of the model evaluation processing. The key node information can be the relevant information of the node that plays a key role in the model evaluation processing, and the key node information is, for example, the node used for model evaluation, the node for task allocation, etc. The result layer stores the hash value of the final evaluation data and verification data. Each layer is associated with the data of the previous layer through the hash value, thus forming a complete proof chain. For example, the process layer is associated with the data layer through the hash value of the key node information processed by the model evaluation, and the result layer is associated with the process layer through the hash value of the evaluation data and verification data.

[0054] After building a multi-layer hash chain, the evaluation data of the original data can be stored in the blockchain based on the multi-layer hash chain. Blockchain refers to a distributed ledger technology, which is an unalterable and decentralized data structure built and maintained by cryptographic principles. In the disclosed embodiment, the blockchain is used to store the hash value of the evaluation data and record the evaluation data to ensure the transparency and unalterability of the model evaluation process.

[0055] Exemplarily, in the process of storing the evaluation data of the original data in the blockchain based on a multi-layer hash chain, the hash value of the data layer, the hash value of the process layer and the hash value of the result layer corresponding to the evaluation data can be obtained; further, the hash value of the data layer, the hash value of the process layer and the hash value of the result layer can be unified into a target hash value through the Merkle tree, and the target hash value can be stored in the blockchain through the blockchain evidence storage protocol.

[0056] Merkle tree refers to a tree-shaped data structure, which generates a unique root hash value by hashing the data blocks represented by the evaluation data and merging them layer by layer. The Merkle tree consists of a group of leaf nodes, a group of intermediate nodes and a root node. The leaf node contains the stored data or its hash value, the intermediate node is the hash value of the content of its two child nodes, and the topmost root node is also composed of the hash value of the content of its two child nodes. The hash value of each data is calculated and saved as the leaf node of the binary tree. The nodes in the binary tree are combined in pairs to calculate the hash value as the hash value of the parent node. And so on, finally the hash value of the root node is obtained. In the disclosed embodiment, the Merkle tree can be a structural method for organizing hash values ​​of different layers such as the data layer, the process layer and the result layer. The tree nodes of the Merkle tree contain hash values, timestamps and associated information. When verifying based on the Merkle tree, data integrity proof can be achieved through the minimum verification path.

[0057] Exemplarily, the hash value of the data layer, the hash value of the process layer, and the hash value of the result layer can be combined to obtain a target hash value. For example, the hash value of the data layer and the hash value of the process layer can be combined to obtain a combined hash value, and then the combined hash value and the hash value of the result layer can be further combined to obtain the target hash value. After obtaining the target hash value, the target hash value can be stored in a block of the blockchain. In addition, the hash value of the data layer, the hash value of the process layer, and the hash value of the result layer can also be stored in the blockchain. In the disclosed embodiment, the integrity of the data is evaluated by the Merkle tree verification model, and the verification of partial data is supported.

[0058] It should be added that, in the process of storing the target hash value in the blockchain, the target hash value can be stored based on the blockchain evidence protocol in the evaluation data evidence protocol. The blockchain evidence protocol is used to limit the storage structure of the target hash value in the blockchain. The blockchain evidence protocol can include evidence identification, hash information, version control, and signature information, as shown in Table 4.

[0059] Table 4

[0060]

[0061] When storing the target hash value in the blockchain, the blockchain can verify the data format of the evaluation data corresponding to the target hash value to be stored, and pack the blocks storing the evaluation data corresponding to the target hash value into blocks to generate a certificate. The certificate refers to a digital certificate that records key information of the evaluation process, including data hash, timestamp, signature and other elements, which is used for subsequent result verification and traceability, and is an important guarantee for the credibility of the evaluation results represented by the evaluation data.

[0062] In the disclosed embodiment, a multi-layer hash chain is combined with the Merkle tree technology to construct a complete evidence storage and certification system. The multi-layer hash chain ensures that the entire evaluation process is traceable, and the Merkle tree structure optimizes the verification efficiency of large-scale data. This combination mechanism not only ensures the reliability of evidence storage, but also improves verification efficiency, improves system performance, and reduces storage costs.

[0063] After the evidence certificate is generated, the transaction information can be returned, which is used to inform the transaction to which the interaction in the blockchain belongs and in which block the transaction is packaged. Furthermore, the blockchain service provides blockchain confirmation. After the blockchain nodes verify and agree with the above information, they will provide blockchain confirmation.

[0064] In the disclosed embodiment, a layered evidence storage strategy is adopted to reduce the storage pressure on the chain. The Merkle tree structure is used to organize the storage of large-scale evaluation data, supporting efficient partial data verification. The tree node design contains hash values, timestamps and associated information, and realizes data integrity proof through the minimum verification path.

[0065] After the deposited voucher is generated, it can be sent to the verification stage. The verification service can respond to the verification request and verify the evaluation data. Exemplarily, data integrity, process consistency, and result validity verification can be performed to generate a verification report, and the verification report is used as the verification result, and the verification result is stored in the blockchain. In the disclosed embodiment, the verification mechanism based on the deposited voucher supports integrity verification and result traceability of the evaluation process. Through standardized verification processes and proof methods, a trusted consensus on evaluation data is achieved, providing the industry with a reliable evaluation infrastructure and promoting mutual recognition and sharing of evaluation results represented by evaluation data.

[0066] refer to Figure 2 The overall system architecture of the evaluation data evidence system shown in the figure can include a data layer, an evaluation layer, an evidence layer, and a verification layer. The data layer can include a data set management module and an encryption storage module, and the data layer implements the full life cycle management of the evaluation data. The data set management module is responsible for data standardization processing, including format conversion, quality verification, and metadata extraction. The version management module maintains the historical versions and change records of the original data set. The encryption storage module uses the AES-256 algorithm to encrypt data and generates data fingerprints through the SHA-256 algorithm to ensure data security and integrity.

[0067] The evaluation rule engine in the evaluation layer is responsible for the configuration and management of evaluation rules and evaluation standards, and supports the definition and weight setting of multi-dimensional evaluation indicators. The evaluation execution module is used to implement the scheduling and execution of tasks, including resource allocation, parallel processing and exception handling mechanisms. The result processing module in the evaluation execution module is used to standardize and encrypt the generated evaluation data.

[0068] The evidence layer may include an evidence generation module and a data chain module. The evidence generation module is used to build a multi-layer hash chain. The structure of the multi-layer hash chain includes a data layer, a process layer, and a result layer, where each layer contains specific proof information. The specific proof information refers to the hash value corresponding to each layer. The evidence generation module may include evidence packaging and evidence generation. The data chain module is used to process data chaining and transaction confirmation, realize the tamper-proof storage of evaluation data, and store the hash value of evaluation data in the blockchain. The system adopts a layered evidence strategy to reduce the storage pressure on the chain.

[0069] In addition, a verification layer may also be included. The verification layer may include a verification rule module and a result verification module. The verification rule module is used to implement rule management and rule execution, and the result verification module is used to perform integrity verification and consistency verification. The verification results may be further stored in the blockchain.

[0070] In the disclosed embodiment, the hierarchical protocol design standardizes the data set management, uses encrypted storage to protect data security, uses blockchain technology to store the evaluation process, and establishes a data integrity verification mechanism to ensure the credibility of the results. Through a system including a data layer, an evaluation layer, a storage layer, and a verification layer, standardized management of evaluation data, transparent processes, and verifiable results are achieved, effectively improving the credibility of model evaluation.

[0071] Figure 3 The specific flow chart of evaluation data storage is shown schematically, refer to Figure 3 As shown in the figure, it mainly includes the data preparation stage, the evaluation execution stage, the evidence storage stage and the verification stage, which specifically includes the following steps:

[0072] In step S301, the data service submits the original data in the original data set to the data preparation stage.

[0073] In step S302, the data preparation stage performs data standardization processing and original data encryption.

[0074] In step S303, AES256 encryption is performed.

[0075] In step S304, a data fingerprint is generated.

[0076] In step S305, the metadata information of the original data and the initial evidence of the original data are recorded. The initial evidence refers to the evidence information before the original data begins to be evaluated, which can be stored in the blockchain.

[0077] In step S306, the evaluation execution phase configures the evaluation rules and initializes the evaluation environment.

[0078] In step S307, the evaluation execution phase obtains the data to be evaluated from the data service, and the data to be evaluated may be original data.

[0079] In step S308, the model evaluation is performed to obtain evaluation data.

[0080] In step S309, the evaluation execution phase records the evaluation process.

[0081] In step S310, the evaluation data is encrypted to generate an evaluation log.

[0082] In step S311, the evaluation data is submitted to the evidence storage stage.

[0083] In step S312, the evidence storage stage calculates the hash value of the evaluation process and the hash value of the evaluation data.

[0084] In step S313, a multi-layer hash chain is constructed in the evidence storage stage.

[0085] In step S314, the evidence storage phase submits the evidence storage request to the blockchain service.

[0086] In step S315, the blockchain service verifies the data format, packages the blocks, and generates a certificate.

[0087] In step S316, blockchain confirmation.

[0088] In step S317, the blockchain service passes the stored evidence to the verification stage.

[0089] In step S318, a verification request is sent to the verification service.

[0090] In step S319, the verification service performs data integrity verification, process consistency verification, and result validity verification.

[0091] In step S320, a verification result is generated.

[0092] In step S321, the verification result is returned.

[0093] In step S322, the verification result is confirmed.

[0094] In step S323, the verification evidence is sent to the blockchain service to be stored in the blockchain.

[0095] In step S324, the final result is recorded.

[0096] In the disclosed embodiment, a blockchain-based evaluation data storage system is provided to solve the problems of disordered evaluation data management, opaque process, and difficult verification of results in the industry. By building a standardized data management mechanism and a credible storage system, the fairness of the evaluation process and the credibility of the evaluation results represented by the evaluation data are ensured.

[0097] Through the special large evaluation data evidence protocol (MEDAP), combined with multi-layer hash chains and Merkle trees, standardized management and full-process traceability of evaluation data are achieved. This solution not only solves the problem of secure storage of evaluation data, but also establishes a complete set of evidence verification mechanisms, providing the industry with a reliable evaluation infrastructure, which can effectively improve the standardization and credibility of large model evaluation, realize standardized management of evaluation data, reliable evidence storage of processes and reliable verification of results, and improve the reliability of model evaluation. Through unified data evidence protocols and evaluation process specifications, standardized management of evaluation data is achieved, solving the pain points of different model evaluation standards and difficult mutual recognition of results, significantly improving data processing efficiency, reducing system maintenance costs, and saving storage space. Based on blockchain technology, reliable evidence storage is provided for the entire evaluation process to ensure the security of evaluation data, transparency of processes, and verifiable results. In addition, through unified data protocols and verification mechanisms, the costs of repeated development and system docking are reduced, and the overall implementation cost of evaluation projects is reduced by about 35%. At the same time, the trusted sharing mechanism of evaluation results avoids repeated evaluations, saves a lot of resources for the industry, and reduces the operating costs of evaluation institutions.

[0098] The present disclosure also provides a blockchain-based evaluation data storage system. Figure 4 As shown, the blockchain-based evaluation data storage system 400 mainly includes the following modules:

[0099] The evaluation data generation module 401 is used to perform model evaluation processing on the original data according to the evaluation rules generated based on the evaluation data storage protocol to obtain evaluation data of the original data;

[0100] The evaluation data storage module 402 is used to construct a multi-layer hash chain, and in response to an evaluation evidence request, store the evaluation data in the blockchain through the constructed multi-layer hash chain based on the Merkle tree.

[0101] In an exemplary embodiment of the present disclosure, the evaluation data evidence protocol includes an evaluation rule protocol; the model evaluation processing is performed on the original data according to the evaluation rules generated based on the evaluation data evidence protocol to obtain the evaluation data of the original data, including: configuring the evaluation rules according to the evaluation rule protocol and initializing the evaluation environment; using the original data as the data to be evaluated, and performing model evaluation processing on the data to be evaluated in the initialized evaluation environment based on the evaluation rules to generate evaluation data corresponding to the original data.

[0102] In an exemplary embodiment of the present disclosure, the constructing of a multi-layer hash chain includes: constructing the multi-layer hash chain based on a data layer, a process layer, and a result layer; each layer in the multi-layer hash chain is associated with the previous layer through a hash value.

[0103] In an exemplary embodiment of the present disclosure, the evaluation data is stored in the blockchain through a multi-layer hash chain constructed based on a Merkle tree, including: obtaining a hash value of a data layer, a hash value of a process layer, and a hash value of a result layer corresponding to the evaluation data; unifying the hash value of the data layer, the hash value of the process layer, and the hash value of the result layer into a target hash value through a Merkle tree, and storing the target hash value in the blockchain through a blockchain evidence storage protocol.

[0104] In an exemplary embodiment of the present disclosure, the hash value of the data layer is the hash value corresponding to the original data and metadata, the hash value of the process layer is the hash value of the key node information of the model evaluation processing, and the hash value of the result layer is the hash value of the evaluation data and verification data.

[0105] In an exemplary embodiment of the present disclosure, the target hash value is stored in the blockchain through the blockchain evidence protocol, including: verifying the data format of the evaluation data corresponding to the target hash value, and packaging the evaluation data corresponding to the target hash value into blocks to generate an evidence voucher.

[0106] In an exemplary embodiment of the present disclosure, the method further includes: verifying the evaluation data to obtain a verification result, and storing the verification result in the blockchain.

[0107] It should be noted that the specific details of each module in the above-mentioned blockchain-based evaluation data notarization system have been described in detail in the corresponding blockchain-based evaluation data notarization method, so they will not be repeated here.

[0108] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0109] In addition, although the steps of the method in the present disclosure are described in a specific order in the drawings, this does not require or imply that the steps must be performed in this specific order, or that all the steps shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps, etc.

[0110] In an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above method is also provided.

[0111] Those skilled in the art will appreciate that various aspects of the present disclosure may be implemented as systems, methods or program products. Therefore, various aspects of the present disclosure may be specifically implemented in the following forms, namely: complete hardware implementation, complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software, which may be collectively referred to herein as "circuits", "modules" or "systems".

[0112] Refer to the following Figure 5 The electronic device 500 according to this embodiment of the present disclosure is described. Figure 5 The electronic device 500 shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0113] like Figure 5 As shown, the electronic device 500 is in the form of a general computing device. The components of the electronic device 500 may include, but are not limited to: the at least one processing unit 510, the at least one storage unit 520, a bus 530 connecting different system components (including the storage unit 520 and the processing unit 510), and a display unit 540.

[0114] The storage unit stores program codes, which can be executed by the processing unit 510, so that the processing unit 510 performs the steps according to various exemplary embodiments of the present disclosure described in the above “Exemplary Method” section of this specification. For example, the processing unit 510 can perform the following steps: Figure 1 Follow the steps shown in .

[0115] The storage unit 520 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 5201 and / or a cache storage unit 5202 , and may further include a read-only storage unit (ROM) 5203 .

[0116] The storage unit 520 may also include a program / utility 5204 having a set (at least one) of program modules 5205, such program modules 5205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0117] Bus 530 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0118] The electronic device 500 may also communicate with one or more external devices 600 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 500, and / or communicate with any device that enables the electronic device 500 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 550. Furthermore, the electronic device 500 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 560. As shown, the network adapter 560 communicates with other modules of the electronic device 500 via a bus 530. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 500, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0119] It should be noted that, in some embodiments of the present disclosure, a computer program product is also provided. The computer program product includes a computer program. When the computer program is executed by a processor, the above method is implemented.

[0120] In one embodiment, a computer program product may be a tangible product containing a computer program, such as a computer-readable storage medium storing a computer program. The readable storage medium may be a storage medium based on electrical, magnetic, optical, electromagnetic, infrared, and other signals, including but not limited to: random access memory (RAM), read-only memory (ROM), magnetic tape, floppy disk, flash memory (Flash), mechanical hard disk (HDD), solid-state drive (SSD), and the like. Exemplarily, a computer program product may be implemented as a non-volatile storage medium storing a computer program, such as a read-only memory, a NAND flash memory (Nand Flash), and the like. In one embodiment, a computer program product may be an intangible product containing a computer program. Exemplarily, a computer program product may be implemented as a virtual digital product, such as a digital file such as an executable file, an installation package, etc. storing a computer program.

[0121] The code of the computer program can be written in one or more programming languages. Programming languages ​​such as C language, Java, C++, etc. The program code can be executed entirely on the user computing device, or partially on the user computing device, or as a separate software package, or partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, such as a local area network (LAN), a wide area network (WAN), etc., or can be connected to an external computing device (e.g., an Internet connection provided by an operator).

[0122] The computer program may be carried or transmitted via electrical, magnetic, optical, electromagnetic, infrared, or other signals. The electronic device may convert the signal carrying the computer program into a digital signal, thereby running the computer program. When the computer program is run on an electronic device, its code is used to enable the electronic device to execute (more specifically, the processor of the electronic device may execute) the method steps of various exemplary embodiments of the present disclosure.

[0123] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the implementation of the present disclosure.

[0124] In addition, the above-mentioned figures are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the present disclosure, and are not intended to be limiting. It is easy to understand that the processes shown in the above-mentioned figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.

[0125] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0126] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing what is disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not disclosed in the present disclosure. The specification and embodiments are to be considered merely as exemplary, and the true scope and spirit of the present disclosure are indicated by the claims.

[0127] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A blockchain-based model evaluation data storage method, characterized in that: include: Perform model evaluation processing on the original data according to the evaluation rules generated based on the evaluation data notarization protocol to obtain evaluation data of the original data; Construct a multi-layer hash chain, and in response to the evaluation evidence request, store the evaluation data in the blockchain through the constructed multi-layer hash chain based on the Merkle tree.

2. The blockchain-based model evaluation data storage method according to claim 1 is characterized in that: The evaluation data storage protocol includes an evaluation rule protocol; the model evaluation processing is performed on the original data according to the evaluation rule generated based on the evaluation data storage protocol to obtain the evaluation data of the original data, including: Configure the evaluation rules according to the evaluation rule protocol and initialize the evaluation environment; The original data is used as the data to be evaluated, and based on the evaluation rules, a model evaluation process is performed on the data to be evaluated in an initialized evaluation environment to generate evaluation data corresponding to the original data.

3. The blockchain-based model evaluation data storage method according to claim 1 is characterized in that: The construction of a multi-layer hash chain includes: The multi-layer hash chain is constructed based on the data layer, the process layer and the result layer; each layer in the multi-layer hash chain is associated with the previous layer through a hash value.

4. The blockchain-based model evaluation data storage method according to claim 1 is characterized in that: The method of storing the evaluation data in the blockchain through a multi-layer hash chain based on the Merkle tree includes: Obtaining a hash value of a data layer, a hash value of a process layer, and a hash value of a result layer corresponding to the evaluation data; The hash value of the data layer, the hash value of the process layer, and the hash value of the result layer are unified into a target hash value through the Merkle tree, and the target hash value is stored in the blockchain through the blockchain evidence protocol.

5. The blockchain-based model evaluation data storage method according to claim 4 is characterized in that: The hash value of the data layer is the hash value corresponding to the original data and metadata, the hash value of the process layer is the hash value of the key node information of the model evaluation processing, and the hash value of the result layer is the hash value of the evaluation data and the verification data.

6. The blockchain-based model evaluation data storage method according to claim 1 is characterized in that: The storing the target hash value to the blockchain through the blockchain evidence storage protocol includes: Verify the data format of the evaluation data corresponding to the target hash value, and pack the evaluation data corresponding to the target hash value into blocks to generate an evidence voucher.

7. The blockchain-based model evaluation data storage method according to claim 1 is characterized in that: The method further comprises: The evaluation data is verified to obtain a verification result, and the verification result is stored in the blockchain.

8. A blockchain-based evaluation data storage system, characterized in that: include: An evaluation data generation module is used to perform model evaluation processing on the original data according to the evaluation rules generated based on the evaluation data storage protocol to obtain evaluation data of the original data; The evaluation data storage module is used to construct a multi-layer hash chain, and in response to an evaluation evidence request, store the evaluation data in the blockchain through the constructed multi-layer hash chain based on the Merkle tree.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the blockchain-based evaluation data storage method described in any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: include: processor; as well as A memory, configured to store executable instructions of the processor; Wherein, the processor is configured to execute the blockchain-based evaluation data storage method described in any one of claims 1-7 by executing the executable instructions.

Images