Block chain-based evidence fixing method and system after multi-party consensus

Abstract

The invention belongs to the technical field of evidence fixing, and particularly relates to an evidence fixing method and system after multi-party consensus based on a block chain. The block chain-based evidence fixing method after multi-party consensus comprises the following steps that S10, a notarization node receives a notarization application submitted by a participant, and the notarization application comprises first evidence, assisting party information associated with the first evidence, third-party server information and geographical location information of the assisting party and the participant; and S20, the notarization node obtains the electronic evidence related to the first evidence from the assisting party as a second evidence, obtains the electronic evidence related to the first evidence from the third-party server as an assisting evidence, and selects a verification party from the third-party server and the notarization node. According to the scheme, the problem of high working difficulty of the notarization institution in evidence fixing and verification links in a single-party evidence obtaining mode is solved, the working difficulty of the notarization institution and the judicial institution is reduced, and the dispute solving efficiency is improved.

Description

Technical Field

[0001] This solution belongs to the field of evidence fixation technology, and specifically involves a method and system for fixing evidence after multi-party consensus based on blockchain. Background Art

[0002] In today's digital age, electronic evidence plays a key role in legal proceedings, commercial disputes and intellectual property protection. When the parties communicate through third-party software, the personal devices they use (such as mobile phones and computers) and the third-party platform server will form a "double-end recording mechanism" - the personal device stores the communication data locally (such as chat records and file transfer records), and the third-party platform retains the cloud backup in accordance with the service agreement, thus forming a complete communication content database. Once a dispute arises between the two parties, these records will enter the collection and notarization process as key electronic evidence: first, the right holder or the notary agency will apply to the personal device holder or the third-party platform for data extraction in accordance with legal procedures. After technical verification and confirmation that it has not been tampered with, it will be solidified and stored by a professional notary agency to form an electronic evidence file with legal effect. When the right holder files a lawsuit with the local court (the court where the right holder is geographically located), the local court will initiate an application for the retrieval of electronic evidence files to the professional notary agency based on the case accepted. After accepting the court's retrieval application, the professional notary agency will inquire and deliver the electronic evidence files to the local court.

[0003] At present, electronic evidence is mainly obtained from electronic devices through automatic evidence collection devices, such as an electronic evidence collection device disclosed in Chinese patent CN107688754 A, which automatically obtains preservation data by recording the screen or taking screenshots of linked pages and submits it to the evidence collection server. Although it saves manpower and time costs to a certain extent, it has obvious limitations such as relying solely on device-side operations, which may lead to incomplete data, limited platform background data acquisition, and insufficient timeliness when facing data tampering or deletion. Especially when obtaining evidence unilaterally from the right holder, electronic evidence is extremely susceptible to factors such as human tampering, equipment failure or system error, which makes it difficult to guarantee the credibility of the evidence and increases the difficulty of the notary agency in the evidence fixation and verification process. In addition, such electronic evidence also requires complex technical verification. First, hash value verification and other technical means must be used to verify whether the entire process of data generation, storage, and extraction is complete and has not been tampered with. If the hash values are inconsistent, it is very likely that the data has been changed. Secondly, the source of evidence must be clarified to prove that it comes from legal, compliant, and normally functioning equipment or software, and has not been illegally manipulated. At the same time, the extraction process must also follow legal procedures, and the time, location, and method of extraction must be recorded in detail to ensure that the extraction is legal. However, in actual operations, completing these complex verifications not only requires high technical requirements, but also takes a lot of time and effort, further highlighting the limitations of current automatic evidence collection methods. Summary of the invention

[0004] The purpose of this solution is to provide a method and system for evidence fixation after multi-party consensus based on blockchain, so as to solve the problem of high work difficulty for notary institutions in the evidence fixation and verification links under the single-party evidence collection method.

[0005] To achieve the above purpose, this solution provides a method for evidence fixation after multi-party consensus based on blockchain, including the following steps S10: The notary node receives the notarization application submitted by the participating party. The notarization application includes the first evidence, the information of the assisting party and the third-party server associated with the first evidence, and the geographical location information of the assisting party and the participating party; S20: The notary node obtains the electronic evidence related to the first evidence from the assisting party as the second evidence, and obtains the electronic evidence related to the first evidence from the third-party server as the assisting evidence. The notary node selects the verification party from the third-party server and the notary node. The verification party compares the first evidence with the assisting evidence or the second evidence, and analyzes the comparison result to obtain the notarization result; S30: The notary node stores the first evidence according to the notarization result, and then obtains the court nodes corresponding to the participating party and the assisting party according to the geographical location information, and sends the first evidence to the court nodes. After receiving the first evidence, the court nodes store the first evidence.

[0006] And, a system for evidence fixation after multi-party consensus based on blockchain that uses the method for evidence fixation after multi-party consensus based on blockchain.

[0007] The principle and technical effect of this solution are as follows: First, after receiving the first evidence, this solution obtains the second evidence and the assisting evidence from the assisting party (i.e., the other party of the first evidence) and the third-party server associated with the first evidence, and conducts cross-comparison of multi-source data on the first evidence, the second evidence and the assisting evidence to form a logical closed-loop verification. Such a method can effectively identify the tampering of single-party (participating party) evidence. At the same time, this solution automatically executes the verification rules through smart contracts, thereby reducing manual intervention. The electronic evidence obtained by single-party evidence collection requires complex technical verification. This solution analyzes the notarization result based on the verification results of multi-source data, reduces the complex technical verification process, and reduces the workload and work difficulty of notary institutions.

[0008] Secondly, based on the geographical location information of the participating party and the assisting party, this solution automatically associates the court nodes in their respective jurisdictions to achieve territorial storage and judicial filing of evidence. In the traditional process, the notary institution needs to manually submit evidence to the court, which is time-consuming and error-prone. This solution realizes "verification and filing immediately" by automatically matching court nodes and synchronizing evidence, reducing intermediate links, alleviating the workload of the court and the notary institution, and improving their work efficiency. At the same time, the notary node synchronizes the verified evidence to the court node, forming a "notarization-judiciary" linkage mechanism to ensure the validity of evidence in legal procedures. As a judicial institution, the court node directly participates in evidence storage, and its endorsement enhances the authority of the evidence. In litigation, the court can directly call the blockchain evidence storage data without repeated verification, reducing the cost of presenting evidence.

[0009] In addition, in this solution, data such as the first evidence, the second evidence, and the assisting evidence are scattered and stored in different nodes (notary nodes, assisting party nodes, third-party servers, court nodes), rather than being centrally stored in a single center. Each node only stores data related to its own role (for example, the court node only stores evidence passed by the notarization result). Data access permissions are ensured through public-private key encryption. This distributed storage avoids the loss of evidence caused by the failure of a single node (for example, the downtime of the notary node does not affect the data storage of the court node). At the same time, the data redundancy is improved through the multi-copy mechanism. Moreover, if a certain assisting party node fails to provide the second evidence due to a fault, other nodes can still complete the verification through the existing evidence chain to ensure that the process is not interrupted.

[0010] In summary, this solution solves the problem of the high work difficulty of the notary institution in the evidence fixation and verification links in the single-party evidence collection method, while reducing the work difficulty of the notary institution and the judicial organ, and improving the efficiency of dispute resolution.

[0011] Furthermore, when the notary node cannot obtain the assisting evidence from the third-party server, the notary node compares the second evidence with the first evidence. If the first evidence is the same as the second evidence, the first evidence is regarded as true as the notarization result. If the second evidence is different from the first evidence, the first evidence is regarded as doubtful as the notarization result.

[0012] When the notary node cannot obtain the assisting evidence from the third-party server, this solution compares the first evidence with the second evidence, shortening the verification time and avoiding process interruption caused by the unavailability of the third-party service. At the same time, this downgrading mechanism also guarantees the basic operation ability of the system, enabling the notarization service to continue to operate under non-ideal conditions. If the first evidence is consistent with the second evidence, the evidence is determined to be true, which not only reduces the evidence collection cost but also maintains the basic credibility. However, if there are differences between the two, a doubt mark is immediately triggered, effectively preventing fraud risks and providing a clear direction for subsequent manual review.

[0013] Further, the verification party constructs a sensitivity scoring function to perform a sensitivity score on the electronic evidence, as shown in the following formula (1): (1), The verification party names and identifies the dynamic weight adjustment factor through the keyword matching function. The matching formula is as shown in the following formula (2): (2), When is satisfied, the privacy protection mode is triggered. The threshold is calculated through the following formula (3): (3), where, is the mean sensitivity of historical data, is the standard deviation, is the confidence coefficient; The verification party proves the consistency between the second evidence or the assisting evidence and the first evidence through the zero-knowledge proof technology.

[0014] Further, when the verification party proves the consistency between the second evidence or the assisting evidence and the first evidence through the zero-knowledge proof technology, the following steps are included: A10: Define the verification relationship , as shown in the following formula (4): (4), where is the first evidence, is the second evidence or the assisting evidence; A20: The verification party constructs a proof using zk-STARKs , as shown in the following formula (5): (5), where the commitment function satisfies ; A30: When verifying whether is consistent with , the verification party executes , and confirms that is consistent with only when ; where is the security parameter, is the negligible function.

[0015] Through the constructed dynamic sensitivity scoring system, this solution can not only accurately identify explicit sensitive information such as ID numbers and medical records, but also capture context-related privacy through named entity recognition and adaptive weight adjustment, improving the accuracy of privacy content recognition. At the same time, the verification protocol based on zk-STARKs not only ensures the non-interactivity of the verification process, but also ensures zero leakage of the original evidence through the cryptographic commitment function, thus meeting strict privacy requirements such as HIPAA / GDPR while reducing the evidence storage overhead. This solution realizes intelligent risk control through a dynamic threshold mechanism: when the entropy value of the historical data distribution increases, the confidence coefficient automatically increases, enabling this solution to improve the response speed to new sensitive patterns while maintaining a low false positive rate. In particular, in scenarios such as cross-institutional medical notarization, due to the adoption of quantum-secure hash commitment technology, it can not only resist future computational attacks, but also complete a large number of concurrent verifications in a short time, ultimately forming a complete closed loop of "accurate identification - efficient verification - compliant evidence storage".

[0016] Furthermore, when the notary node sends the first evidence to the court node, a consortium is established between the local notary node and each court node based on blockchain technology, and each node within the consortium follows the consensus mechanism; a dedicated sub-chain is established in the consortium chain for the first evidence. The sub-chain adopts the sharding storage technology, performs a hash operation on the first evidence to generate a unique hash value, and triggers the chain-upload operation through a smart contract, storing the hash value and the metadata of the first evidence in the sub-chain block; after each court node monitors the chain-upload operation of the sub-chain through the consensus algorithm, it automatically synchronizes and verifies the hash value, encrypts and stores the uploaded content and the corresponding hash value in the blockchain database of the local node, and at the same time updates the evidence index table of the local node to ensure the integrity verification of the stored data through the Merkle-Patricia tree structure.

[0017] When a notarization node sends the first piece of evidence to each court node, it constructs a consortium chain network based on blockchain technology and forms a consortium with each court node. Each node follows a consensus mechanism to achieve trusted collaboration, avoiding the trust risk dominated by a single node. At the same time, the permission control mechanism of the consortium chain ensures that only judicial institution nodes can join, guaranteeing the security of evidence circulation. For the first piece of evidence, a dedicated sub-chain is established in the consortium chain. The sharding storage technology is adopted to reduce the data pressure on the main chain. Through a smart contract, a hash operation is triggered to generate a unique hash value, and the hash value, as well as metadata such as evidence type, generation time, and geographical location information, are stored on the chain. This not only reduces the bandwidth consumption caused by the transmission of the complete evidence file but also realizes evidence fixation through the immutability of the hash value - if the evidence content is modified, the hash value will change completely, enabling quick identification of tampering. After each court node monitors the on-chain operation of the sub-chain through a consensus algorithm, it automatically synchronizes and verifies the hash value, encrypts and stores the on-chain content and the hash value in the local blockchain database, and updates the evidence index table simultaneously. The Merkle-Patricia tree structure is used for integrity verification to ensure that the locally stored data is consistent with that on the chain. During this process, the blockchain timestamp accurately records the time of evidence storage, meeting the requirements of judicial timeliness. Distributed storage disperses the evidence hash values among various nodes, and the failure of a single node does not affect the integrity of the evidence. The hierarchical design of the sub-chain and the main chain enhances the risk resistance of this solution. Finally, it realizes the efficient synchronous verification of cross-regional evidence, the strengthening of legal effect, and the guarantee of data security, optimizing the efficiency of the judicial process and ensuring the credibility and immutability of evidence in the judicial process through technical mechanisms.

[0018] Furthermore, when a court node retrieves electronic evidence, it first queries the electronic evidence locally as the initial evidence, then obtains the court node corresponding to the geographical location information associated with the initial evidence as the assisting court, and retrieves the electronic evidence related to the initial evidence from the assisting court as the verification evidence. The verification evidence is compared with the initial evidence. If the results are the same, the initial evidence is output as the final evidence. If the results are different, the electronic evidence related to the initial evidence is retrieved from the notarization node as the re-verification evidence. The re-verification evidence, verification evidence, and initial evidence are compared, and the final evidence is selected and output from the re-verification evidence, verification evidence, or initial evidence according to the comparison results.

[0019] Furthermore, if the re-verification evidence, verification evidence, or initial evidence are all different, then the re-verification evidence, verification evidence, or initial evidence is divided into equal-length segments, and the cross-similarity matrix of the three-party evidence is calculated for the th segment. The segment of the party with complete evidence content is determined according to the calculated similarity, and then the segments with complete evidence content are combined into a complete piece of evidence and output as the final evidence.

[0020] First, the three - level verification process of "initial evidence - verification evidence - re - verification evidence" established in this solution (court node → assisting court → notary node) not only improves the credibility of evidence through multi - party cross - verification, but also provides an authoritative data source for subsequent block - by - block repair. At the same time, the block - by - block similarity analysis mechanism introduced in this solution can not only locate byte - level differences when the three - party evidence is completely inconsistent, but also intelligently select the optimal evidence block through dynamic weights, thereby improving the evidence verification speed in complex scenarios. It is particularly worth noting that when basic inconsistencies are detected, the fault - tolerant recombination ability of this solution can automatically repair evidence, which shortens the processing cycle and reduces the storage cost in some complex scenarios (such as cross - border contract disputes, etc.).

[0021] Further, after the notary node divides the electronic evidence, it calculates the dispute coefficient of the th subsection according to the calculation result of the similarity matrix , and the calculation formula of is shown in the following formula (11): When , mark the th subsection as a disputed block, is a preset dispute threshold; allocate verification computing power to the disputed block, as shown in the following formula (12): (12), where is the set of all disputed blocks within the current window, is the benchmark computing resource; use the hybrid verification model to deeply verify the disputed blocks, and the hybrid verification model is shown in the following formula (13): (13), where the cryptographic similarity is calculated based on the Merkle - Patricia tree, and the semantic similarity is calculated using the BERT model, and the weight coefficient: ; When m consecutive blocks satisfy: immediately release the verification computing power allocated for this batch.

[0022] The quantitative evaluation model of the dispute coefficient adopted in this solution can not only accurately identify highly disputed blocks, but also achieve intelligent allocation of computing power through elastic resource allocation, improving resource utilization in block testing. At the same time, the hybrid verification model not only retains the cryptographic rigor of Merkle tree verification, but also introduces BERT semantic analysis to capture content tampering traces, thus reducing the misjudgment rate for complex evidence.

[0023] Further, the notary node compares the final evidence with the electronic evidence stored by the court node to obtain the completeness of the electronic evidence, and scores the confidence of the court node according to the completeness. The court node selects the assisting court according to the confidence score ranking.

[0024] Based on the integrity quantization model, this solution not only realizes the objective evaluation of evidence quality, but also establishes an adaptive reputation score for the court node through the exponential smoothing algorithm, improving the recognition rate of malicious nodes. At the same time, the topological optimization algorithm is used to screen the assisting court, which not only ensures the priority selection of high-reputation nodes, but also shortens the court cooperation delay and improves the efficiency of the court node comparing electronic evidence. Brief Description of the Drawings

[0025] Figure 1 It is a flowchart of the evidence fixation method after multi-party consensus based on blockchain in the embodiment of the present invention. Detailed Embodiment

[0026] The following will clearly and completely describe the concept of the present invention and the technical effects produced in combination with the embodiments, so as to fully understand the purpose, features and effects of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all embodiments. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention: As Figure 1 shown, the evidence fixation method after multi-party consensus based on blockchain includes the following steps: S10: The notary node receives the notarization application submitted by the participating party. The notarization application includes the first evidence, the information of the assisting party and the third-party server associated with the first evidence, and the geographical location information of the assisting party and the participating party; S20: The notary node obtains the electronic evidence related to the first evidence as the second evidence from the assisting party, and obtains the electronic evidence related to the first evidence as the assisting evidence from the third-party server. The notary node selects a verification party from the third-party server and the notary node. The verification party compares the first evidence with the assisting evidence or the second evidence, and analyzes the comparison result to obtain the notarization result; S30: The notary node stores the first evidence according to the notarization result, and then obtains the court nodes corresponding to the participating party and the assisting party according to the geographical location information, and sends the first evidence to the court nodes. After receiving the first evidence, the court nodes store the first evidence.

[0027] Among them, the verification party constructs a sensitivity scoring function to score the sensitivity of the electronic evidence, as shown in the following formula (1): (1), The verifier identifies and statistically calculates the dynamic weight adjustment factor through the keyword matching function. The matching formula is shown in the following formula (2): (2), When is satisfied, the privacy protection mode is triggered. The threshold is calculated through the following formula (3): (3), Among them, is the mean value of the historical data sensitivity, is the standard deviation, is the confidence coefficient; the verifier proves the consistency between the second evidence or the assisting evidence and the first evidence through the zero-knowledge proof technology.

[0028] Specifically, when the verifier proves the consistency between the second evidence or the assisting evidence and the first evidence through the zero-knowledge proof technology, the following steps are included: A10: Define the verification relationship , as shown in the following formula (4): (4), Among them is the first evidence, is the second evidence or the assisting evidence; A20: The verifier constructs a proof using zk-STARKs , as shown in the following formula (5): (5), where the commitment function satisfies ; A30: When verifying whether is consistent with , the verifier executes , and when and only when , it is confirmed that is consistent with in content; among them is the security parameter, is the negligible function.

[0029] Among them, when the notarization node sends the first piece of evidence to each court node, it constructs a consortium chain network based on blockchain technology and forms a consortium with each court node. Each node follows the consensus mechanism to achieve trusted collaboration, avoiding the trust risk dominated by a single node. At the same time, the permission control mechanism of the consortium chain ensures that only judicial institution nodes can join, guaranteeing the security of evidence transfer. A dedicated sub-chain is established in the consortium chain for the first piece of evidence. The sharding storage technology is adopted to reduce the data pressure on the main chain. Through the smart contract, a hash operation is triggered to generate a unique hash value, and the hash value and metadata such as evidence type, generation time, and geographical location information are stored on the chain. This not only reduces the bandwidth consumption caused by the transmission of the complete evidence file but also realizes evidence solidification through the immutability of the hash value. If the evidence content is modified, the hash value will change completely, enabling quick identification of tampering. After each court node monitors the sub-chain's on-chain operation through the consensus algorithm, it automatically synchronizes and verifies the hash value, encrypts and stores the on-chain content and the hash value in the local blockchain database, and updates the evidence index table simultaneously. The Merkle-Patricia tree structure is used for integrity verification to ensure that the locally stored data is consistent with that on the chain. During this process, the blockchain timestamp accurately records the time of evidence deposit, meeting the requirements of judicial timeliness. The distributed storage disperses the evidence hash values among various nodes, and the failure of a single node does not affect the integrity of the evidence. The hierarchical design of the sub-chain and the main chain enhances the risk resistance of this solution. Ultimately, it realizes the efficient synchronous verification of cross-regional evidence, the strengthening of legal effect, and the guarantee of data security, optimizing the efficiency of the judicial process and ensuring the credibility and immutability of evidence on the judicial chain through technical mechanisms.

[0030] Among them, when a court node retrieves electronic evidence, it first queries the electronic evidence locally as the initial evidence, then obtains the court node corresponding to the geographical location information associated with the initial evidence as the assisting court, retrieves the electronic evidence related to the initial evidence from the assisting court as the verification evidence, and compares the verification evidence with the initial evidence. If the results are consistent, the initial evidence is output as the final evidence. If the results are inconsistent, it retrieves the electronic evidence related to the initial evidence from the notarization node as the re-verification evidence, compares the re-verification evidence, the verification evidence, and the initial evidence, and selects the final evidence for output from the re-verification evidence, the verification evidence, or the initial evidence according to the comparison results.

[0031] Specifically, if the re-verification evidence, the verification evidence, or the initial evidence are all different, then the re-verification evidence, the verification evidence, or the initial evidence is divided into equal-length subsections, and the cross-similarity matrix of the three-party evidence is calculated for the th subsection. The subsection with complete evidence content is determined based on the calculated similarity, and then the subsections with complete evidence content are combined into a complete piece of evidence for output as the final evidence.

[0032] More specifically, when the re-verification evidence and the verification evidence When it is inconsistent with the initial evidence Generate the final evidence according to the following steps : B10: The notary node divides each piece of evidence into equal-length subsections, and the block length satisfies the following formula (6): (6), where is the minimum block threshold (default 4KB), represents the byte length of the evidence.

[0033] B20: For the th subsection ( ), calculate the cross-similarity matrix of the three-party evidence , as shown in the following formula (7): (7), where the similarity function is defined as ; is the locally sensitive hashing similarity, is the information entropy function, is the smoothing factor.

[0034] B30: The notary node selects the final version of the th subsection satisfying the following formula (8): (8), The weight is assigned according to the credibility of the evidence source, as shown in the following formula (9): (9), B40: The final evidence needs to satisfy the consistency check, as shown in the following formula (10): (10), where is the similarity threshold, is the indicator function.

[0035] More specifically, the block processing adopts a dynamic adjustment strategy. When , the sliding window mechanism is enabled: . After the evidence is reorganized, the consistency of the Merkle hash tree needs to be verified.

[0036] Specifically, after the notary node divides the electronic evidence, it calculates the Dispute coefficient of a subsection , The calculation formula of is shown in formula (11) below: (11), When , mark the th subsection as a disputed block,[[]]END]] is a preset dispute threshold, defaulting to 0.7; allocate verification computing power to the disputed block , as shown in formula (12) below: (12), where is the set of all disputed blocks within the current window,[[]]END]] is the benchmark computing resource; use a hybrid verification model to deeply verify the disputed block, and the hybrid verification model is shown in formula (13) below: (13), where the cryptographic similarity is calculated based on the Merkle-Patricia tree,[[]]END]] The calculation formula of is shown in formula (14) below: (14), The semantic similarity is calculated using the BERT model,[[]]END]] The calculation formula of is shown in formula (15) below: (15), The weight coefficient ; When m consecutive blocks satisfy: Immediately release the verification computing power allocated for this batch.[[]]END]]

[0037] The resource allocation formula ensures (total system resources).[[]]END]]

[0038] Among them, the notary node compares the final evidence with the electronic evidence stored by the court node, obtains the completeness of the electronic evidence, and gives a confidence score to the court node according to the completeness. The court node selects to assist the court according to the confidence score ranking.[[]]END]]

[0039] Specifically, the court node calculates the evidence integrity by comparing the final evidence with the evidence stored by the court node,[[]]END]] , and the calculation formula is shown in formula (16) below: (16), wherein, is the number of bytes for evidence matching, is the timestamp difference, is the maximum allowable time delay (default 24 hours).

[0040] Specifically, the score of the court node is updated by the exponential smoothing method, and its calculation formula is as shown in formula (17): (17), where The score is uploaded to the chain in real time and affects the subsequent selection of the assisting court.

[0041] Specifically, when selecting an assisting court, preference is given to:[[]] ; where is the reputation weight,

[0042] is the network delay between nodes.

[0043] This embodiment further includes a blockchain-based evidence fixation system using a blockchain-based multi-party consensus evidence fixation method.

[0044] Specifically, in implementation, a technology company and a cooperative enterprise had a dispute over the authenticity of an electronic contract. The technology company (participant) claimed that the payment terms stipulated in the contract had been tampered with, while the cooperative enterprise insisted that the contract content had never changed. To resolve this dispute, the two parties decided to adopt a blockchain-based multi-party consensus evidence fixation system, and through the coordinated operation of the notary node, the assisting party node (cooperative enterprise), the third-party server, and the court node, the trustworthy fixation and verification of the electronic contract were achieved.

[0045] The technology company (participant) submitted a notarization application to the notary node, and the application included the electronic contract (the first evidence), the information of the cooperative enterprise (the assisting party), the information of the third-party cloud storage server, and the geographical location information of both parties (e.g., the technology company is located in City A and the cooperative enterprise is located in City B).

[0046] After receiving the application, the notary node sent a request to the cooperative enterprise (the assisting party) to obtain a copy of the same electronic contract stored by it as the second evidence. At the same time, the notary node accessed the third-party cloud storage server to retrieve the historical version of the electronic contract as the assisting evidence.

[0046] The notary node selects a verification party from a third-party server or its own node to compare the first evidence, the second evidence, and the assisting evidence. The verification party first compares the first evidence with the assisting evidence and finds that there are differences in their hash values, indicating that the electronic contract may have been tampered with. Since valid assisting evidence cannot be obtained from the third-party server, the verification party then compares the first evidence with the second evidence. After comparing the content page by page, obvious differences are found in the payment terms section of the contract (the contract provided by the technology company shows that "the payment period is 30 working days", while the contract provided by the cooperative enterprise shows that "the payment period is 60 working days"). The verification party analyzes the electronic contract through a sensitivity scoring function and identifies that the payment terms involve commercially sensitive information. Since the sensitivity score exceeds the preset threshold, the system automatically triggers the privacy protection mode and uses zero-knowledge proof technology to verify the evidence consistency, ensuring that without disclosing the specific content of the contract, it can be proven that there are substantial differences between the contract versions provided by both parties.

[0047] Based on the comparison results, the notary node marks the electronic contract as "doubtful" and stores the first evidence. Subsequently, the system automatically associates the corresponding court nodes according to the geographical location information of both parties: the technology company corresponds to the node of the Intermediate People's Court of City A, and the cooperative enterprise corresponds to the node of the Basic People's Court of City B. The notary node synchronizes the doubtful electronic contract to these two court nodes for storage, realizing the territorial record-filing of the evidence.

[0048] The court node first retrieves the electronic contract from local storage as the initial evidence. Then, according to the geographical location information associated with the initial evidence, it applies to the Basic People's Court of City B (the assisting court) to obtain verification evidence. The Basic People's Court of City B returns a copy of the electronic contract it stores (which is consistent with the second evidence provided by the cooperative enterprise).

[0049] The presiding court compares the verification evidence with the initial evidence and finds that their contents are inconsistent. So it applies to the notary node to obtain re-verification evidence. Since there are differences in the initial evidence, the verification evidence, and the re-verification evidence, the notary node divides the three pieces of evidence into multiple equally long sections (such as dividing by page numbers), and calculates the cross-similarity matrix for each section. By analyzing the similarity of each section, it is found that the section where the payment terms are located has a large difference between the re-verification evidence (the evidence of the notary node) and the initial evidence (the evidence of the presiding court), and a lower similarity with the corresponding section in the verification evidence (the evidence of the assisting court). For the sections with abnormal similarity, the notary node marks them as disputed blocks and allocates additional verification computing power for in-depth verification. Through a hybrid verification model (combining Merkle-Patricia tree cryptographic verification and BERT semantic analysis), traces of tampering with the payment terms are found, confirming the possibility that the contract version provided by the cooperative enterprise is forged.

[0050] The notary nodes regularly compare the final evidence with the evidence stored in the court nodes, calculate the integrity of each court node (for example, the evidence integrity of the intermediate people's court node in City A is 98%, and the evidence integrity of the grass-roots people's court node in City B is 75%), and conduct confidence scoring on the court nodes according to the integrity. In subsequent cases, the presiding court will preferentially select the court node with a high confidence score (such as the intermediate people's court in City A) as the assisting court to ensure the accuracy and efficiency of evidence verification.

[0051] The above are only embodiments of the present invention, and common knowledge such as specific structures and characteristics known in the solution is not described in detail here. It should be noted that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can still be made, which should also be regarded as the protection scope of the present invention, and these will not affect the implementation effect of the present invention and the practicality of the patent. The protection scope required by this application should be subject to the content of its claims, and the specific implementation manners and the like recorded in the specification can be used to interpret the content of the claims.

Claims

1. A method for fixing evidence after multi - party consensus based on blockchain, characterized in that, It includes the following steps: S10: The notary node receives the notarization application submitted by the participating party. The notarization application contains the first evidence, the information of the assisting party and the third-party server associated with the first evidence, and the geographical location information of the assisting party and the participating party; S20: The notary node obtains the electronic evidence related to the first evidence from the assisting party as the second evidence, and obtains the electronic evidence related to the first evidence from the third-party server as the assisting evidence. The notary node selects a verification party from the third-party server and the notary node. The verification party compares the first evidence with the assisting evidence or the second evidence, and analyzes the comparison result to obtain the notarization result; S30: The notary node stores the first evidence according to the notarization result, and then obtains the court nodes corresponding to the participating party and the assisting party according to the geographical location information, and sends the first evidence to the court nodes. After receiving the first evidence, the court nodes store the first evidence.

2. The method for fixing evidence after multi-party consensus based on blockchain according to claim 1, wherein: When the notary node fails to obtain the assisting evidence from the third-party server, the notary node compares the second evidence with the first evidence. If the first evidence is the same as the second evidence, the first evidence is regarded as true as the notarization result. If the second evidence is different from the first evidence, the first evidence is regarded as doubtful as the notarization result.

3. The method for evidence fixation after multi-party consensus based on blockchain according to claim 2, wherein: The verifying party constructs a sensitivity scoring function to perform a sensitivity score on the electronic evidence, as shown in the following formula (1): （1）， The verification party names the entity recognition statistical dynamic weight adjustment factor through the keyword matching function, and the matching formula is shown in formula (2) below: （2）， When the following conditions are met the privacy protection mode is triggered, and the threshold is calculated by the following formula (3): （3）， wherein, is the mean historical data sensitivity, is the standard deviation, is the confidence coefficient; The verification party proves the consistency of the second evidence or the assisting evidence with the first evidence through the zero-knowledge proof technology.

4. The method for fixing evidence after multi-party consensus based on blockchain according to claim 3, characterized in that: When the verification party proves the consistency of the second evidence or the assisting evidence with the first evidence through the zero-knowledge proof technology, it includes the following steps: A10: Define the verification relationship , as shown in the following formula (4): （4）， Among them is the first evidence is the second evidence or assisting evidence; A20: The verifier uses zk-STARKs to construct a proof , as shown in formula (5) below: （5）， where the commitment function satisfies ; A30: Verification With When verifying whether they are consistent, the verifying party executes If and only if Then confirm And Have the same content; where Is a security parameter,[[]] Is a negligible function.

5. The method for fixing evidence after multi-party consensus based on blockchain according to claim 4, characterized in that: When the notary node sends the first evidence to the court nodes, it establishes an alliance between the local notary node and each court node based on the blockchain technology. Each node in the alliance follows the consensus mechanism; a dedicated sub-chain is established in the alliance chain for the first evidence. The sub-chain adopts the sharding storage technology, performs a hash operation on the first evidence to generate a unique hash value, triggers the on-chain operation through the smart contract, and stores the hash value and the metadata of the first evidence in the sub-chain block; After each court node monitors the on-chain operation of the sub-chain through the consensus algorithm, it automatically synchronizes and verifies the hash value, encrypts and stores the on-chain content and the corresponding hash value in the blockchain database of the local node, and at the same time updates the evidence index table of the local node to ensure the integrity verification of the stored data through the Merkle-Patricia tree structure.

6. The method for fixing evidence after multi-party consensus based on blockchain according to claim 5, characterized in that: When the court node retrieves the electronic evidence, it first queries the electronic evidence from the local as the initial evidence, then obtains the court node corresponding to the geographical location information associated with the initial evidence as the assisting court, obtains the electronic evidence related to the initial evidence from the assisting court as the verification evidence, and compares the verification evidence with the initial evidence. If the results are consistent, the initial evidence is output as the final evidence. If the results are inconsistent, it obtains the electronic evidence related to the initial evidence from the notary node as the re-verification evidence, compares the re-verification evidence, the verification evidence and the initial evidence, and selects the final evidence from the re-verification evidence, the verification evidence or the initial evidence according to the comparison result for output.

7. The method for evidence fixation after multi-party consensus based on blockchain according to claim 6, characterized in that: If the retest evidence, verification evidence, or initial evidence are all different, then the retest evidence, verification evidence, or initial evidence is split into equally long sections, and for the th section, calculate the cross-similarity matrix of the three-party evidence, determine the section of the party with complete evidence content based on the calculated similarity, and then combine the sections with complete evidence content into a complete piece of evidence for output as the final evidence.

8. The method for evidence fixation after multi-party consensus based on blockchain according to claim 7, wherein: After the notarization node divides the electronic evidence, calculate the controversy coefficient of the th subsection according to the calculation result of the similarity matrix , The calculation formula of is as shown in formula (11) below: （11）， When the th subsection is marked as a disputed block, where is a preset dispute threshold; allocate verification computing power to the disputed block as shown in the following formula (12): （12）， Among them, is the set of all disputed blocks within the current window, is the reference computing resource; a hybrid verification model is used to deeply verify the disputed blocks, and the hybrid verification model is shown in the following formula (13): （13）， Among them, the cryptographic similarity is calculated based on the Merkle-Patricia tree, and the semantic similarity is calculated using the BERT model. Weight coefficient: ; When m consecutive blocks meet the following conditions: Immediately release the verification computing power allocated for this batch.

9. The method for fixing evidence after multi-party consensus based on blockchain according to claim 8, wherein: The notary node compares the final evidence with the electronic evidence stored in the court node to obtain the completeness of the electronic evidence, and scores the confidence level of the court node according to the completeness. The court node selects to assist the court according to the confidence level score ranking.

10. An evidence fixation system after multi-party consensus based on blockchain, characterized in that, The evidence fixation method based on multi-party consensus on the blockchain according to any one of claims 1-9 is used.

Images