Intelligent contract whole-process automatic management method and system

Abstract

The application provides a kind of intelligent contract whole process automation management method and system, the method is by obtaining business negotiation content, and input to pre-trained legal big model generates initial contract clause, realizes the conversion from negotiation minutes to structured contract text;By calling risk knowledge graph, real-time compliance pre-trial and risk labeling are carried out on the initial contract clause, and potential legal risks are effectively identified;Through the comparison conflict detection based on role permission, support multi-party efficient, controllable contract revision;The revised contract version and signing process are credibly notarized by using blockchain technology, and security authentication and signing are completed by combining with biological characteristic electronic seal;Finally, by interfacing business system to obtain performance data, and with the aid of smart contract, the performance conditions are automatically verified.The method is automated and intelligent from contract generation to risk preliminary examination, solves the problem of low efficiency of contract generation under the premise of ensuring the quality of clauses.

Description

Technical Field

[0001] This invention relates to the field of legal technology and artificial intelligence, and in particular to a method and system for automated management of the entire intelligent contract process. Background Technology

[0002] In today's increasingly complex and specialized business environment, intricate commercial contracts in fields such as construction engineering and equipment procurement serve as the core carriers of rights and obligations, and their management quality directly impacts transaction security and corporate operational efficiency. With the deep integration of legal technology and artificial intelligence, contract management is transforming from traditional manual methods to intelligent and automated processes. The industry is placing higher demands on efficient collaboration, risk warning, reliable evidence storage, and performance tracking throughout the entire contract lifecycle, urgently requiring intelligent solutions that can cover the entire process of drafting, revising, signing, and fulfilling contracts to adapt to the customized needs and compliance requirements of complex business scenarios.

[0003] Existing contract management technologies and models rely on pre-set template libraries for contract drafting. The drafting process requires manual adjustments to the content based on the templates, and also relies on legal and business personnel to manually review clauses, manage versions, and track performance.

[0004] Traditional contract drafting relies entirely on fixed template libraries, making it impossible to dynamically adjust the content of the terms based on specific business negotiation results, industry characteristics, and customized needs. This results in a high average drafting time in the industry, which is not only inefficient but also prone to potential disputes due to the mismatch between template terms and actual business scenarios. Summary of the Invention

[0005] This invention provides a smart contract full-process automated management method and system to solve the problem of low efficiency in generating contract terms in the prior art.

[0006] On one hand, the present invention provides a method for automated management of the entire process of smart contracts, including:

[0007] Obtain the content of business negotiations; the content of business negotiations includes text information, voice information, and image information;

[0008] Input the business negotiation content into a pre-trained legal model, and output the initial contract terms text;

[0009] The risk knowledge graph is invoked to perform compliance pre-review and risk labeling on the initial contract terms text, and the contract terms text is output.

[0010] Based on role-based access control, a difference comparison algorithm is used to detect clause conflicts in the contract text, and a collaboratively revised version of the contract is output.

[0011] The collaboratively revised contract version and signing process information are stored in the blockchain network, and multi-party authentication and signing are completed through electronic signatures bound to biometric features, outputting the signed contract version and evidence storage information;

[0012] Obtain the contract terms performance data corresponding to the signed contract version, verify whether the performance conditions are met through a smart contract, and generate performance exception information when the performance conditions are not met.

[0013] Optional, constructing a large legal model includes;

[0014] Collect training data, which includes contract templates, judicial judgments, legal texts and corresponding clause annotations;

[0015] An initial large language model is constructed based on the Transformer architecture, and pre-trained using the training corpus to enable the initial large language model to learn general legal language representations, thereby obtaining a large legal model.

[0016] The legal big data model is fine-tuned in a supervised manner using contract clause generation and risk labeling tasks.

[0017] Optionally, a risk knowledge graph can be invoked to perform compliance pre-review and risk labeling on the initial contract terms text, outputting the contract terms text, including:

[0018] Input the initial contract terms text into the risk knowledge graph to identify the legal entities, contractual obligations, and key conditions in the initial contract terms text;

[0019] The legal entity, the contractual obligation, and the key conditions are associated and matched with the judicial precedent nodes, legal provision nodes, and industry risk model nodes stored in the risk knowledge graph to obtain the association matching results;

[0020] Based on the association matching results, risk scores are calculated and risk labels are generated for the clauses in the initial contract text according to the preset risk assessment rules;

[0021] Output the contract terms text with the aforementioned risk score and risk label.

[0022] Optionally, constructing a risk knowledge graph includes:

[0023] Judicial case data and regulatory policy texts were collected as source data for constructing the graph;

[0024] Using a sequence labeling model based on BERT-BiLSTM-CRF, legal entities, points of contention, and judgment results are extracted from the source data as nodes of a knowledge graph;

[0025] Based on the semantic connections and legal logic between the legal entities and the points of contention, the relationships between nodes are constructed to form a risk knowledge graph;

[0026] Based on the judgment results and penalty information in the regulatory policies, the risk weights of each node and its relationships in the risk knowledge graph are dynamically calculated and updated.

[0027] Optionally, based on role-based access control, a difference comparison algorithm is used to detect clause conflicts in the contract text, and a collaboratively revised contract version is output, including:

[0028] Configure editable contract terms for different user roles involved in the revision;

[0029] Receive revision suggestions from users of various roles within their respective authority for the contract terms text, and generate multiple revised versions;

[0030] Based on the Levenshtein-CRF hybrid difference comparison algorithm, text difference analysis and conflict detection are performed on the multiple revision versions to identify contradictory clause revisions.

[0031] Based on the revised terms, a solution is matched and generated from a pre-built solution library;

[0032] Integrate the conflict-free clause revisions and the proposed solutions to output a collaboratively revised contract version.

[0033] Optionally, the collaboratively revised contract version and signing process information are stored in a blockchain network, and multi-party authentication and signing are completed through electronic signatures bound to biometrics, outputting the signed contract version and evidence information, including:

[0034] The key content of the collaboratively revised contract version is generated into a digital fingerprint, which, together with the timestamp and version identifier, constitutes evidence storage information;

[0035] The evidence information is uploaded to a blockchain network based on a consortium blockchain architecture for trusted evidence storage, and a blockchain transaction receipt is obtained.

[0036] When a signatory initiates a signing, the signatory's biometric information is collected via a mobile terminal. The biometric information includes at least one of iris features and voiceprint features.

[0037] The biometric information is compared and authenticated with the pre-stored signer identity information. After successful authentication, an electronic seal conforming to national cryptographic standards is invoked to sign the collaboratively revised contract version. The signing action, signing time, and blockchain transaction receipt are associated and bound together, and the signed contract version and complete evidence storage information are output.

[0038] Optionally, obtain the contract terms performance data corresponding to the signed contract version, verify whether the performance conditions are met through a smart contract, and generate performance exception information when the performance conditions are not met, including:

[0039] Analyze the signed contract version and extract the performance conditions, time nodes and delivery standards stipulated in the signed contract version;

[0040] Obtain performance data related to the performance conditions, the performance data including at least one of payment status, logistics information, and acceptance report;

[0041] The performance data is input into a smart contract deployed on the blockchain to determine whether the performance data meets the performance conditions; the smart contract has built-in condition-action verification logic corresponding to the contract terms;

[0042] When the performance data does not meet the performance conditions, the performance exception information is generated through the smart contract; the performance exception information includes at least the exception type, the reference to the default clause, and the default amount or liability definition calculated according to the contract.

[0043] On the other hand, the present invention also provides an intelligent contract end-to-end automated management system, comprising:

[0044] The content acquisition module is configured to acquire business negotiation content; the business negotiation content includes text information, voice information, and image information.

[0045] The contract generation module is configured to input the business negotiation content into a pre-trained legal big data model and output the initial contract terms text;

[0046] The risk analysis module is configured to call a risk knowledge graph to perform compliance pre-review and risk labeling on the initial contract terms text, and output the contract terms text.

[0047] The collaborative revision module is configured to detect clause conflicts in the contract text based on role-based access control and a difference comparison algorithm, and output a collaboratively revised version of the contract.

[0048] The evidence storage and signing module is configured to store the collaboratively revised contract version and signing process information in the blockchain network, and complete multi-party authentication and signing through electronic signatures bound to biometric features, and output the signed contract version and evidence storage information.

[0049] The performance monitoring module is configured to obtain the performance data of the contract terms corresponding to the signed contract version, verify whether the performance conditions are met through smart contracts, and generate performance exception information when the performance conditions are not met.

[0050] On the other hand, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the smart contract full-process automated management method as described above.

[0051] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the smart contract full-process automated management method as described above.

[0052] On the other hand, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the smart contract full-process automated management method as described above.

[0053] This invention provides a fully automated management method and system for intelligent contracts. The method acquires business negotiation content and inputs it into a pre-trained legal model to generate initial contract terms, achieving intelligent conversion from negotiation minutes to structured contract text. By invoking a risk knowledge graph, it performs real-time compliance pre-review and risk labeling of the initial contract terms, effectively identifying and alerting potential legal risks. Through role-based collaborative revision and discrepancy comparison and conflict detection, it supports efficient and controllable contract revision by multiple parties. Blockchain technology is used for trusted storage of the revised contract version and signing process, combined with biometric electronic signatures for secure authentication and signing. Finally, by connecting to business systems to obtain performance data, and using smart contracts to automatically verify performance conditions, the method automates and intelligently manages the contract generation and initial risk review process, improving overall efficiency in the contract generation and preliminary drafting stages while ensuring the quality of the terms. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0055] Figure 1 This is a flowchart illustrating the fully automated management method for smart contracts provided in this embodiment of the invention.

[0056] Figure 2 This is a schematic diagram of the structure of the intelligent contract full-process automated management system provided in an embodiment of the present invention;

[0057] Figure 3 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0059] Figure 1 This is a flowchart illustrating the fully automated management method for smart contracts provided in this embodiment of the invention.

[0060] like Figure 1 As shown in the figure, the fully automated management method for smart contracts provided in this embodiment of the invention mainly includes the following steps:

[0061] 101. Obtain the content of business negotiations.

[0062] Business negotiation content includes textual, audio, and visual information. For example, textual information can be obtained from negotiation records, emails, etc.; audio information can be obtained by recording the negotiation process; and visual information may come from pictures, charts, and other materials from the negotiation site. Obtaining business negotiation content allows for a comprehensive and accurate capture of the key points and details of the negotiation.

[0063] 102. Input the content of the business negotiation into the pre-trained legal model and output the initial contract terms text.

[0064] The legal big data model is based on the Transformer architecture. The Transformer architecture is a deep learning model architecture that uses a self-attention mechanism, which can effectively capture long-distance dependencies in text sequences.

[0065] The construction of a large legal model includes:

[0066] Collect training data, which includes contract templates, judicial judgments, legal texts, and corresponding clause annotations.

[0067] The training corpus needs to cover both the legal professional field and the target business scenario, specifically including various templates, judicial judgments, currently effective legal texts, and authoritative annotations of the corresponding clauses. The corpus must undergo deduplication, compliance screening, and format standardization to ensure its professionalism, accuracy, and scenario adaptability, providing training data for the model to learn legal expertise and business scenario rules.

[0068] An initial large language model is constructed based on the Transformer architecture, and pre-trained using training corpus to enable the initial large language model to learn general legal language representations, thus obtaining a large legal model.

[0069] Among them, an initial large language model is built based on the Transformer architecture. The Transformer architecture has good context association capture ability and natural language understanding ability, and can adapt to the long sequence and strong logic characteristics of legal texts.

[0070] The pre-training phase uses a diverse set of organized training corpora as input. Through unsupervised learning, the model learns the semantic representation, grammatical structure, and usage of professional terms in general legal language, masters the logical framework and expression norms of different legal texts, understands the correspondence between legal clauses and factual scenarios, and gradually forms a general understanding of legal professional knowledge. Ultimately, a large legal model with basic legal language processing capabilities is obtained.

[0071] The legal big data model is fine-tuned in a supervised manner using contract clause generation and risk labeling tasks.

[0072] During the fine-tuning process, the input training samples must include specific business scenario requirements descriptions and corresponding adapted contract clause output samples, contract clause texts and corresponding 28 types of legal risk labels labeled samples.

[0073] By continuously inputting labeled training data, the parameters of the legal big data model are iteratively optimized, enabling the model to learn the logic of generating compliance clauses from business needs, while mastering the identification standards and labeling rules of risk clauses. This improves the adaptability of the legal big data model in generating clauses and the accuracy of risk identification in complex commercial contract scenarios.

[0074] 103. Use the risk knowledge graph to conduct a compliance pre-review and risk labeling of the initial contract terms text, and output the contract terms text.

[0075] Among them, the risk knowledge graph is a knowledge representation and reasoning model based on graph databases. It materializes various types of knowledge in the legal field, constructing a complex semantic network in the form of nodes and edges. Specifically, constructing the risk knowledge graph includes:

[0076] Judicial case data and regulatory policy texts were collected as source data for constructing the graph.

[0077] The source data collected must cover core scenarios of legal practice and regulatory requirements, specifically including over 5 million publicly available judicial cases related to contract disputes from courts at all levels nationwide. This data includes basic case information, factual findings, reasoning, and judgments, as well as texts related to civil and commercial law, economic law, and industry regulatory policies. During the collection process, the data must be deduplicated and anonymized, and a data quality verification mechanism must be established to ensure the authenticity, completeness, and timeliness of the source data.

[0078] Using a sequence labeling model based on BERT-BiLSTM-CRF, legal entities, points of contention, and judgment results are extracted from the source data as nodes of a knowledge graph.

[0079] A sequence labeling model based on BERT-BiLSTM-CRF is used to extract elements from the collected source data. The sequence labeling model has semantic understanding and sequence labeling capabilities, and can effectively adapt to the strong logic and professional terminology of legal texts.

[0080] The extracted knowledge nodes specifically include three core elements: first, legal entities, such as contract parties, subject matter, performance period, liability for breach of contract, as well as regulatory bodies and those subject to penalties; second, points of contention, such as ambiguous terms, breach of contract, and compliance defects; and third, judgment results, such as the determination of victory / defeat, the proportion of liability, and the amount of compensation. The extracted results will be directly used as nodes in the risk knowledge graph.

[0081] Based on the semantic connections and legal logic between legal entities and the focus of the dispute, the relationships between nodes are constructed to form a risk knowledge graph.

[0082] Specifically, the relationships include, but are not limited to, the triggering relationship between legal entities and the points of contention, the correspondence between the points of contention and the judgment, and the compliance constraint relationship between legal entities and regulatory policies. By clarifying the logical connections between nodes, the risk knowledge graph can clearly present the path of contract risk and its legal consequences.

[0083] Based on the judgment results and penalty information in regulatory policies, the risk weights of each node and its relationships in the risk knowledge graph are dynamically calculated and updated.

[0084] Specifically, based on the records of lost cases in court judgments and the penalty information in regulatory policies, a dynamic risk weight calculation algorithm is used to calculate and update the risk weights of each node and its relationships in the risk knowledge graph in real time. For example, Risk_score = case loss rate × 0.6 + frequency of regulatory penalties × 0.4. When new judicial cases are added or regulatory policies are adjusted, relevant information is automatically retrieved, and the risk scores of the corresponding nodes and their relationships are recalculated, enabling the risk knowledge graph to dynamically reflect changes in the risk level of different contract terms and business scenarios.

[0085] After the risk knowledge graph is constructed, it is used to perform compliance pre-review and risk labeling on the initial contract terms text, and output the contract terms text, including:

[0086] Input the initial contract terms into the risk knowledge graph to identify the legal entities, contractual obligations, and key conditions in the initial contract terms.

[0087] The initial contract terms are input into the risk knowledge graph system, and natural language processing technology is used to perform in-depth analysis of the text to identify the legal entities, contractual obligations and key conditions in the initial contract terms, ensuring that the extracted elements fully cover the core dimensions of contract compliance review and risk assessment.

[0088] The legal entities, contractual obligations, and key conditions are matched with judicial precedent nodes, legal provisions nodes, and industry risk model nodes stored in the risk knowledge graph to obtain the matching results.

[0089] Specifically, the identified legal entities, contractual obligations, and key conditions are intelligently matched with judicial precedent nodes, authoritative legal provisions nodes, and industry risk model nodes stored in the risk knowledge graph. Matching legal entities with judicial precedent nodes identifies historical disputes involving similar entities; matching contractual obligations with legal provisions nodes verifies whether the clauses comply with current civil and commercial laws and industry regulatory policies; and matching key conditions with industry risk model nodes identifies known high-risk performance scenarios, ultimately forming a multi-dimensional, full-scenario matching result.

[0090] Based on the correlation matching results, risk scores are calculated and risk labels are generated for the clauses in the initial contract text according to the pre-set risk assessment rules.

[0091] Specifically, based on the matching results, each clause in the initial contract text is scored for risk according to pre-set risk assessment rules. Based on the scoring results and the matched risk types, corresponding risk labels are automatically generated, such as clause ambiguity risk, compliance defect risk, and high probability of breach of contract risk, achieving dual labeling of risk level and risk type.

[0092] Output the contract terms text with risk scores and risk labels.

[0093] This involves integrating the original text of the terms, corresponding risk scores, and risk labels to generate a contract text with complete risk information. The contract text presents the risk level, risk label, and the basis for the risk association for each clause.

[0094] 104. Based on role-based access control, use a difference comparison algorithm to detect clause conflicts in the contract text and output a collaboratively revised version of the contract.

[0095] Specifically, based on role-based access control, a discrepancy comparison algorithm is used to detect clause conflicts in the contract text, and a collaboratively revised contract version is output, including:

[0096] Configure editable contract terms for different user roles involved in the revision;

[0097] Receive revision suggestions from users of various roles within their respective permissions for the contract terms text, and generate multiple revised versions;

[0098] Based on the Levenshtein-CRF hybrid difference comparison algorithm, text difference analysis and conflict detection were performed on multiple revisions to identify contradictory clause revisions.

[0099] Based on the revised terms, a solution is matched and generated from a pre-built solution library;

[0100] Integrate conflict-free clause revisions and solutions to output a collaboratively revised contract version.

[0101] Specifically, based on the user roles of the parties involved in the contract revision, the scope of editable contract terms for each role is configured. Among them, the legal role can only edit compliance terms, liability for breach of contract, and dispute resolution; the business role can edit payment settlement, delivery cycle, and performance guarantee; and the technical role is limited to quality standards, technical parameters, and acceptance procedures. By isolating permissions, disorderly revisions across roles are avoided, ensuring the professionalism and relevance of the revised terms.

[0102] It receives revision suggestions submitted by users in each role within their authorized scope, records the specific content, revision time, and operator identity of each revision in real time, and generates independent revision versions for revision operations of different roles, ensuring that the revision trajectory of each version is traceable and verifiable, and providing a complete data foundation for subsequent difference analysis and conflict detection.

[0103] By using the Levenshtein algorithm to calculate text edit distance, the system quickly locates the additions, deletions, and modifications to the revised content. Leveraging the sequence labeling capabilities of the CRF model, it deeply analyzes the semantic logic of the clauses, identifying contradictions and conflicts in the revisions made by different roles. Examples include conflicts between the business role's extended delivery cycle and the technical role's key node acceptance requirements, and logical contradictions between the legal role's compliance clauses and the business role's payment settlement clauses.

[0104] For each identified type of clause conflict, the optimal solution is matched from a pre-built solution library. This library is built upon judicial cases, industry practices, and regulatory policies, and includes various common conflict scenarios. Based on the elements of the conflicting clauses, it automatically retrieves the corresponding standardized solutions, while also indicating the legal basis and practical feasibility of each solution.

[0105] Finally, the system automatically integrates the conflict-free revisions from all parties, embeds the matched conflict resolutions into the corresponding clauses, and generates a unified, collaboratively revised contract version. The version clearly marks the revision history and conflict resolution basis for each clause, ensuring the revision process is transparent and traceable, and providing a logically consistent and compliant standardized text for subsequent contract signing and performance.

[0106] 105. Store the collaboratively revised contract version and signing process information in the blockchain network, and complete multi-party authentication and signing through electronic signatures bound to biometrics, outputting the signed contract version and evidence information.

[0107] Specifically, the collaboratively revised contract version and signing process information are stored in a blockchain network, and multi-party authentication and signing are completed through electronic signatures bound to biometrics. The signed contract version and evidence information are then output, including:

[0108] The key content of the collaboratively revised contract version will be generated into a digital fingerprint, which, together with the timestamp and version identifier, will constitute evidence information.

[0109] This process involves extracting clauses and complete text from the collaboratively revised contract version, generating a unique digital fingerprint using a hash algorithm to ensure the contract content is tamper-proof. The digital fingerprint is then linked and integrated with a timestamp and a version identifier that uniquely identifies the contract's iteration status, together forming complete evidence information encompassing contract details, time dimensions, and version traceability.

[0110] The evidence is uploaded to a blockchain network based on a consortium blockchain architecture for trusted evidence storage, and a blockchain transaction receipt is obtained.

[0111] Specifically, the completed evidence storage information is uploaded to a consortium blockchain network based on the Hyperledger Fabric architecture for cross-chain evidence storage. This consortium blockchain network boasts high security and efficient processing capabilities, with a block generation time of less than 2 seconds. The network employs a dual verification mechanism of timestamp solidification and digital fingerprinting to encrypt and distribute the evidence storage information, ensuring data immutability and full traceability. Upon completion of the evidence storage, the blockchain network automatically returns a blockchain transaction receipt containing key information such as the evidence storage address and block height, serving as proof of the evidence's validity.

[0112] When a signatory initiates a signing, the signatory's biometric information is collected via a mobile terminal. The biometric information includes at least one of iris features and voiceprint features.

[0113] When a signatory initiates a signing request, the signatory's biometric information is collected via a mobile signing app, supporting the collection of at least one or more combinations of iris and voiceprint features. During the collection process, liveness detection technology is used to prevent the risk of proxy signing via photos or audio recordings, ensuring the authenticity and uniqueness of the biometric information and providing a reliable basis for identity authentication.

[0114] The biometric information is compared and authenticated with the pre-stored signer identity information. After successful authentication, an electronic seal conforming to national cryptographic standards is used to sign the collaboratively revised contract version. The signing action, signing time, and blockchain transaction receipt are linked and bound together, and the signed contract version and complete evidence storage information are output.

[0115] Specifically, the collected biometric information is compared and authenticated with pre-stored signer identity information. Upon successful authentication, an electronic signature conforming to the national cryptographic standard SM2 / SM9 is invoked to sign the collaboratively revised contract version. Simultaneously, the signing details, signing time, and signatory identity information are automatically linked to previously obtained blockchain transaction receipts, forming a complete signing chain record. The final output is a signed contract version with an electronic signature, along with a complete evidence package containing storage information, transaction receipts, and signing records, meeting subsequent compliance review and dispute evidence requirements.

[0116] 106. Obtain the contract terms performance data corresponding to the signed contract version, verify whether the performance conditions are met through smart contracts, and generate performance exception information when the performance conditions are not met.

[0117] Specifically, it retrieves the contract terms performance data corresponding to the signed contract version, verifies whether the performance conditions are met through smart contracts, and generates performance exception information when the performance conditions are not met, including:

[0118] Analyze the signed contract version and extract the performance conditions, time nodes, and delivery standards stipulated in the signed contract version.

[0119] Among them, parsing the signed contract version can use natural language processing technology to perform structured parsing of the signed contract version, extract the core performance elements clearly stipulated in the contract, including performance conditions such as payment triggering conditions, delivery obligations, and quality compliance requirements, time parameters of key nodes, and delivery standards such as subject specifications, technical parameters, and acceptance procedures.

[0120] During the parsing process, the original contract terms stored on the blockchain are linked simultaneously to ensure that the extracted performance elements are completely consistent with the contract stipulations.

[0121] Obtain performance data related to the performance conditions.

[0122] The performance data includes at least one of the following: payment status, logistics information, and acceptance reports. By capturing performance data directly related to the performance conditions, the data is encrypted and transmitted to a blockchain network for verification, ensuring that the data source is trustworthy and the content has not been tampered with.

[0123] The data is input into a smart contract deployed on the blockchain to determine whether the data meets the performance conditions; the smart contract has built-in condition-action verification logic corresponding to the contract terms.

[0124] Specifically, the verified performance data is input into the smart contract. The smart contract contains built-in condition-action verification logic that corresponds one-to-one with the contract terms. For example, milestone acceptance corresponds to the fulfillment of payment conditions, and the delivery of the subject matter meeting technical parameters corresponds to the completion of delivery obligations. Based on the immutable contract terms and performance data on the blockchain, the smart contract executes the verification logic to determine whether the performance data meets the performance conditions stipulated in the contract and whether the delivery standards are met within the specified time. The entire verification process requires no manual intervention, ensuring the objectivity and accuracy of the results.

[0125] When the performance data does not meet the performance conditions, a performance exception information is generated through a smart contract; the performance exception information includes at least the exception type, the reference to the default clause, and the default amount or liability definition calculated according to the contract.

[0126] Specifically, when a smart contract determines that the performance data does not meet the performance conditions, it triggers an anomaly generation mechanism, outputting standardized performance anomaly information. This information includes at least the anomaly type, reference to default clauses, and definition of liability for breach of contract. Once generated, the anomaly information is simultaneously pushed to the performance monitoring dashboard for visual alerts and stored on the blockchain for future traceability.

[0127] Based on the same inventive concept, this invention also protects an automated management system for the entire process of smart contracts. The automated management system for the entire process of smart contracts provided by this invention will be described below. The automated management system for the entire process of smart contracts described below can be referred to in correspondence with the automated management method for the entire process of smart contracts described above.

[0128] like Figure 2 As shown, this embodiment of the invention also provides a smart contract end-to-end automated management system, including:

[0129] The content acquisition module 210 is configured to acquire business negotiation content; the business negotiation content includes text information, voice information, and image information.

[0130] The contract generation module 220 is configured to input business negotiation content into a pre-trained legal model and output the initial contract terms text.

[0131] Risk analysis module 230 is configured to call the risk knowledge graph to perform compliance pre-review and risk labeling on the initial contract terms text, and output the contract terms text;

[0132] The collaborative revision module 240 is configured to detect clause conflicts in the contract text based on role-based access control and a difference comparison algorithm, and output a collaboratively revised version of the contract.

[0133] The evidence storage and signing module 250 is configured to store the collaboratively revised contract version and signing process information in the blockchain network, and complete multi-party authentication and signing through electronic signatures bound to biometric features, and output the signed contract version and evidence storage information.

[0134] The performance monitoring module 260 is configured to obtain performance data of the contract terms corresponding to the signed contract version, verify whether the performance conditions are met through smart contracts, and generate performance exception information when the performance conditions are not met.

[0135] Figure 3 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention.

[0136] like Figure 3 As shown, the electronic device may include a processor 310, a communications interface 320, a memory 330, and a communication bus 340. The processor 310, communications interface 320, and memory 330 communicate with each other via the communication bus 340. The processor 310 can call logical instructions from the memory 330 to execute a fully automated management method for the smart contract process.

[0137] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0138] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the smart contract full-process automated management method provided by the above methods.

[0139] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the smart contract full-process automated management method provided by the above methods.

[0140] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0141] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0142] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for fully automated management of the entire intelligent contract process, characterized in that, include: Obtain the details of business negotiations; The business negotiation content includes text information, voice information, and image information; Input the business negotiation content into a pre-trained legal model, and output the initial contract terms text; The risk knowledge graph is invoked to perform compliance pre-review and risk labeling on the initial contract terms text, and the contract terms text is output. Based on role-based access control, a difference comparison algorithm is used to detect clause conflicts in the contract text, and a collaboratively revised version of the contract is output. The collaboratively revised contract version and signing process information are stored in the blockchain network, and multi-party authentication and signing are completed through electronic signatures bound to biometric features, outputting the signed contract version and evidence storage information; Obtain the contract terms performance data corresponding to the signed contract version, verify whether the performance conditions are met through a smart contract, and generate performance exception information when the performance conditions are not met.

2. The intelligent contract full-process automated management method according to claim 1, characterized in that, The construction of a large legal model includes: Collect training data, which includes contract templates, judicial judgments, legal texts and corresponding clause annotations; An initial large language model is constructed based on the Transformer architecture, and pre-trained using the training corpus to enable the initial large language model to learn general legal language representations, thereby obtaining a large legal model. The legal big data model is fine-tuned in a supervised manner using contract clause generation and risk labeling tasks.

3. The intelligent contract full-process automated management method according to claim 1, characterized in that, The risk knowledge graph is invoked to perform compliance pre-review and risk labeling on the initial contract terms text, outputting the contract terms text, including: Input the initial contract terms text into the risk knowledge graph to identify the legal entities, contractual obligations, and key conditions in the initial contract terms text; The legal entity, the contractual obligation, and the key conditions are associated and matched with the judicial precedent nodes, legal provision nodes, and industry risk model nodes stored in the risk knowledge graph to obtain the association matching results; Based on the association matching results, risk scores are calculated and risk labels are generated for the clauses in the initial contract text according to the preset risk assessment rules; Output the contract terms text with the aforementioned risk score and risk label.

4. The intelligent contract full-process automated management method according to claim 1, characterized in that, Building a risk knowledge graph includes: Judicial case data and regulatory policy texts were collected as source data for constructing the graph; Using a sequence labeling model based on BERT-BiLSTM-CRF, legal entities, points of contention, and judgment results are extracted from the source data as nodes of a knowledge graph; Based on the semantic connections and legal logic between the legal entities and the points of contention, the relationships between nodes are constructed to form a risk knowledge graph; Based on the judgment results and penalty information in the regulatory policies, the risk weights of each node and its relationships in the risk knowledge graph are dynamically calculated and updated.

5. The intelligent contract full-process automated management method according to claim 1, characterized in that, Based on role-based access control, a difference comparison algorithm is used to detect clause conflicts in the contract text, and a collaboratively revised contract version is output, including: Configure editable contract terms for different user roles involved in the revision; Receive revision suggestions from users of various roles within their respective authority scopes for the contract terms text, and generate multiple revised versions; Based on the Levenshtein-CRF hybrid difference comparison algorithm, text difference analysis and conflict detection were performed on the multiple revision versions to identify contradictory clause revisions. Based on the revised terms, a solution is matched and generated from a pre-built solution library; Integrate the conflict-free clause revisions and the proposed solutions to output a collaboratively revised contract version.

6. The intelligent contract full-process automated management method according to claim 1, characterized in that, The collaboratively revised contract version and signing process information are stored in a blockchain network, and multi-party authentication and signing are completed through electronic signatures bound to biometrics. The signed contract version and evidence information are then output, including: The key content of the collaboratively revised contract version is generated into a digital fingerprint, which, together with the timestamp and version identifier, constitutes evidence storage information; The evidence information is uploaded to a blockchain network based on a consortium blockchain architecture for trusted evidence storage, and a blockchain transaction receipt is obtained. When a signatory initiates a signing, the signatory's biometric information is collected via a mobile terminal. The biometric information includes at least one of iris features and voiceprint features. The biometric information is compared and authenticated with the pre-stored signer identity information. After successful authentication, an electronic seal conforming to national cryptographic standards is invoked to sign the collaboratively revised contract version. The signing action, signing time, and blockchain transaction receipt are associated and bound together, and the signed contract version and complete evidence storage information are output.

7. The intelligent contract full-process automated management method according to claim 1, characterized in that, Obtain the contract terms performance data corresponding to the signed contract version, verify whether the performance conditions are met through a smart contract, and generate performance exception information when the performance conditions are not met, including: Analyze the signed contract version and extract the performance conditions, time nodes and delivery standards stipulated in the signed contract version; Obtain performance data related to the performance conditions, the performance data including at least one of payment status, logistics information, and acceptance report; The performance data is input into a smart contract deployed on the blockchain to determine whether the performance data meets the performance conditions; the smart contract has built-in condition-action verification logic corresponding to the contract terms; When the performance data does not meet the performance conditions, the performance exception information is generated through the smart contract; the performance exception information includes at least the exception type, the reference to the default clause, and the default amount or liability definition calculated according to the contract.

8. A fully automated intelligent contract management system, characterized in that, include: The content acquisition module is configured to acquire business negotiation content; The business negotiation content includes text information, voice information, and image information; The contract generation module is configured to input the business negotiation content into a pre-trained legal big data model and output the initial contract terms text; The risk analysis module is configured to call a risk knowledge graph to perform compliance pre-review and risk labeling on the initial contract terms text, and output the contract terms text. The collaborative revision module is configured to detect clause conflicts in the contract text based on role-based access control and a difference comparison algorithm, and output a collaboratively revised version of the contract. The evidence storage and signing module is configured to store the collaboratively revised contract version and signing process information in the blockchain network, and complete multi-party authentication and signing through electronic signatures bound to biometric features, and output the signed contract version and evidence storage information. The performance monitoring module is configured to obtain the performance data of the contract terms corresponding to the signed contract version, verify whether the performance conditions are met through smart contracts, and generate performance exception information when the performance conditions are not met.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the intelligent contract full-process automated management method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the intelligent contract full-process automated management method as described in any one of claims 1 to 7.

Images