A cross-industry data sharing method and system supporting dynamic policy negotiation

Abstract

The application discloses a cross-industry data sharing method and system supporting dynamic policy negotiation, relates to the technical field of data sharing, and comprises the following steps: guiding a data provider and a data user to register respective industry attributes and submit data sharing policy elements and constraint conditions; loading industry semantic mapping rules and compliance rules according to the industry attributes; continuously collecting dynamic context information during data sharing; based on the data sharing policy elements, the constraint conditions, the dynamic context information, the industry semantic mapping rules and the compliance rules, executing a negotiation algorithm, and generating at least one data sharing policy negotiation scheme; performing expected effect evaluation and compliance verification to obtain evaluation results and verification results; and the data provider and the data user reach a consensus to generate a verifiable policy contract and execute a data sharing operation. The application solves the problem that the sharing policy cannot be dynamically adjusted in the prior art, data sharing rules are ambiguous, and the communication effect of the two sharing parties is poor.

Description

Technical Field

[0001] This invention relates to the field of data sharing technology, and specifically to a cross-industry data sharing method and system that supports dynamic strategy negotiation. Background Technology

[0002] There are significant differences between different industries in terms of data definition, business terminology, security standards, and the laws and regulations they follow, resulting in numerous unresolved practical problems when sharing data.

[0003] Currently, traditional data sharing technologies, based on natural language-based bilateral negotiations or static contract templates, struggle to accurately obtain control parameters and dynamically adjust sharing strategies. This results in ambiguous data sharing rules, poor communication between the data sharing parties, and incomplete verification processes for negotiated solutions.

[0004] Therefore, a new data sharing method and system is needed that can understand cross-industry semantics and generate negotiation solutions to promote the safe, compliant, and efficient cross-industry circulation of data elements. Summary of the Invention

[0005] This application provides a cross-industry data sharing method and system that supports dynamic strategy negotiation, addressing the problems in existing technologies such as difficulty in accurately obtaining technical control parameters, inability to dynamically adjust sharing strategies, resulting in ambiguous data sharing rules and poor communication between data sharing parties.

[0006] In view of the above problems, this application provides a cross-industry data sharing method and system that supports dynamic strategy negotiation.

[0007] Firstly, this application provides a cross-industry data sharing method that supports dynamic strategy negotiation, the method comprising:

[0008] The system guides data providers and data users to register their respective industry attributes with the trusted negotiation engine and submit data sharing strategy elements and constraints.

[0009] The trusted negotiation engine loads the corresponding industry semantic mapping rules and compliance rules based on the industry attributes;

[0010] During the data sharing process, dynamic context information is continuously collected, including data processing environment security certification, business scenario identification, and external regulatory events.

[0011] The trusted negotiation engine, based on the data sharing strategy elements, constraints, dynamic context information, and the loaded industry semantic mapping rules and compliance rules, executes a negotiation algorithm to generate at least one data sharing strategy negotiation scheme.

[0012] The expected effects and compliance verification of the data sharing strategy negotiation scheme were evaluated, and the evaluation results and verification results were obtained.

[0013] Based on the evaluation and verification results, the data provider and the data user reach a consensus, generate a verifiable strategy contract, and execute data sharing operations according to the verifiable strategy contract.

[0014] Secondly, the present invention provides a cross-industry data sharing system that supports dynamic strategy negotiation, comprising:

[0015] The attribute registration module guides data providers and data users to register their respective industry attributes with the trusted negotiation engine and submit data sharing strategy elements and constraints.

[0016] The rule acquisition module is used by the trusted negotiation engine to load the corresponding industry semantic mapping rules and compliance rules according to the industry attributes.

[0017] The information collection module is used to continuously collect dynamic context information during the data sharing process. The dynamic context information includes data processing environment security certification, business scenario identification, and external regulatory events.

[0018] The negotiation scheme generation module is used by the trusted negotiation engine to execute a negotiation algorithm based on the data sharing strategy elements, constraints, dynamic context information, and the loaded industry semantic mapping rules and compliance rules to generate at least one data sharing strategy negotiation scheme.

[0019] The results evaluation module is used to evaluate the expected effects and verify the compliance of the data sharing strategy negotiation scheme, and obtain the evaluation results and verification results.

[0020] The data sharing module is used to facilitate consensus between the data provider and the data user based on the evaluation and verification results, generate a verifiable strategy contract, and execute data sharing operations according to the verifiable strategy contract.

[0021] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0022] First, this application addresses the issue of ambiguous intent expression in traditional data sharing by registering industry attributes and submitting strategy elements and constraints through a trusted negotiation engine. This provides precise input for subsequent processing and improves negotiation efficiency. Second, the trusted negotiation engine generates semantic mapping rules and compliance rules, enabling the system to be compatible with different industry attributes and overcoming the problem of unclear rules in cross-industry collaboration, ensuring the professionalism and legality of the negotiation context. Third, it collects dynamic contextual information including security certifications, business scenarios, and regulatory events, responding in real time to changes in security status, adjustments to business intent, and external regulatory dynamics, achieving continuous adaptability of the negotiation strategy and improving environmental adaptability.

[0023] Simultaneously, the algorithm generates multiple negotiation solutions and uses intelligent search to find the optimal solution within the strategy space, improving the efficiency and quality of solution generation. Furthermore, it conducts expected effect assessments and compliance verifications, performing business judgments and legal reviews to reduce the risk of misjudgment. Finally, it facilitates consensus between both parties, generating a verifiable strategy contract to ensure the arbitrability of shared terms and disputes, ultimately resolving key technical challenges in cross-industry data sharing, including difficulties in negotiation, compliance, dynamic adaptation, and execution. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating a cross-industry data sharing method that supports dynamic strategy negotiation, as described in this application.

[0025] Figure 2 This is a schematic diagram of the structure of a cross-industry data sharing system that supports dynamic strategy negotiation, as described in this application.

[0026] In the attached diagram, the components represented by each number are as follows:

[0027] Attribute registration module 11, rule acquisition module 12, information collection module 13, negotiation scheme generation module 14, result evaluation module 15, and data sharing module 16. Detailed Implementation

[0028] This application provides a cross-industry data sharing method that supports dynamic strategy negotiation, specifically addressing the problems in existing technologies such as difficulty in accurately obtaining control parameters, inability to dynamically adjust sharing strategies, resulting in ambiguous data sharing rules and poor communication between data sharing parties.

[0029] The present invention will now be described in detail with reference to the accompanying drawings.

[0030] Example 1, as Figure 1 As shown, this application provides a cross-industry data sharing method that supports dynamic strategy negotiation, the method comprising:

[0031] S10: Guide data providers and data users to register their respective industry attributes with the trusted negotiation engine and submit data sharing strategy elements and constraints.

[0032] In this embodiment, the data provider is an entity that owns data assets and is willing to share them with other parties under specific conditions; the data user is an entity that wants to obtain and use other people's data due to business or R&D needs; the trusted negotiation engine is a software system or service trusted by all parties; and the industry attribute is the identifier of the specific industry field to which the data provider or data user belongs.

[0033] Specifically, before data sharing, identity authentication is first performed through a trusted negotiation engine. This involves registering industry attributes and invoking corresponding semantic mapping rules and compliance rules. By submitting their respective data sharing strategy elements and constraints to the trusted negotiation engine, negotiation is initiated within a rule-based framework. This ensures compliance with different industry standards and avoids data sharing due to mismatched industry attributes, which could lead to data security issues.

[0034] Step S10 in the method provided in this application embodiment includes:

[0035] The data provider registers the first industry attribute with the trusted negotiation engine and submits the data provider's data sharing strategy elements and data provider constraints. The data provider's data sharing strategy elements include at least one of the following: data usage purpose scope, data usage timeliness, and data usage geographical restrictions.

[0036] The data user registers a second industry attribute with the trusted negotiation engine and submits the data user's data sharing strategy elements and data user constraints. The data user's data sharing strategy elements include at least one of the following: data processing technology solution, data processing environment certification, and data output format requirements.

[0037] The constraints imposed on the data provider and the constraints imposed on the data user both include their respective non-negotiable bottom-line policy clauses.

[0038] In this embodiment, firstly, a first industry attribute, such as a medical attribute, is obtained from the data provider. Then, the data provider registers the first industry attribute, which is recorded in the trusted negotiation engine. Simultaneously, the data provider submits data sharing strategy elements and data provider constraints. The sharing strategy elements specify the restrictions in data sharing, including at least one of the following: the scope of data use purpose, the timeliness of data use, and geographical restrictions on data use. These can usually be negotiated and are used to balance data value and risk control to ensure data security in data sharing. The data provider constraints are mandatory requirements that the data provider must meet in data sharing. They are based on the first industry attribute and the relevant laws, regulations, or mandatory requirements and standards of the industry.

[0039] Defining the true purpose of data sharing by defining the scope of data usage objectives, such as commercial promotion, prevents data from being used in unintended processes that pose ethical or commercial risks. Controlling the duration of data use helps manage data lifecycle risks and prevents uncontrollable risks arising from indefinite data retention. Controlling the spatial flow of data through geographical restrictions, such as limiting processing to within the country and prohibiting transmission to regions without data protection, ensures compliance with data sharing regulations.

[0040] By defining the primary industry attributes, data sharing strategy elements, and data provider constraints, a solid foundation is laid for subsequent automated semantic alignment, compliance verification, and solution generation, ensuring the core interests and compliance requirements of the data sharing provider are met.

[0041] Secondly, the data user registers a second industry attribute with the trusted negotiation engine and submits the data user's data sharing strategy elements and data user constraints. The data user's data sharing strategy elements include at least one of the following: data processing technology solution, data processing environment certification, and data output format requirements. The data processing technology solution is the specific technology used by the data user to process the shared data; the data processing environment certification is a certificate or report confirming that the data user meets security standards or commitments; the data output format requirements are the result format that the data user expects to obtain from this data sharing; and the data user constraints are the conditions that the data user must meet.

[0042] Specifically, the data user determines the corresponding industry attribute. Then, in the trusted negotiation engine, the data user generates data sharing strategy elements and data user constraints based on the second industry attribute, thus obtaining the second industry attribute and the legal and mandatory requirements that exist in the data use process.

[0043] The key elements of a data sharing strategy for data users include at least one of the following: data processing technology solutions, proof of the data processing environment, and requirements for the data output format. The data processing technology solutions, by defining the technical implementation methods used in the data sharing process, ensure the suitability of data use. The data processing environment demonstrates the security of the data usage environment and establishes technical trust. The data output format requirements specify the format of the results obtained through data sharing, defining the supporting conditions for data sharing and avoiding limitations imposed by data format asymmetry. Similarly, the data users are also bound by constraints, including technical requirements for data use, such as a data interface latency of less than 200ms, or business compliance requirements, such as the provider guaranteeing the legality of the data source and bearing joint liability for any data infringement.

[0044] If the requirements of the data provider are included but those of the data user are not, the trusted negotiation engine may only generate an extremely conservative solution that is of no practical value to the user. By determining the data sharing strategy elements and constraints of the data user, the data provider and the data sharing party can generate corresponding data sharing strategy elements and constraints, which helps to negotiate the feasibility of the solution through the data sharing strategy.

[0045] The data provider's and data user's constraints both include their respective non-negotiable bottom-line clauses. These non-negotiable bottom-line clauses are core principles that the participating parties will not compromise on, abandon, or modify under any circumstances during negotiations. If violated, the transaction will lose its foundation or bring unbearable risks.

[0046] During negotiations, it is crucial to accurately distinguish between negotiable strategic elements and non-negotiable constraints. Non-negotiable strategic bottom-line clauses serve as hard constraints or filters, and all candidate solutions must fully satisfy these bottom-line clauses. Any solution that violates any party's bottom-line clause should be directly excluded. For example, if the provider's bottom line is that data must not leave the country, any technical solution proposed by the user involving cross-border data transfer will be judged as conflicting by the rules during the initial semantic alignment stage and will therefore not enter the subsequent optimization search space.

[0047] In this embodiment of the application, by specifying the categories of strategic elements and non-negotiable constraints that data providers and users need to submit, vague and unstructured business intentions are transformed into structured and dimensional inputs that can be accurately processed by machines, and the hard boundaries of security and compliance in the negotiation are determined, laying a reliable data foundation for automated negotiation.

[0048] S20: The trusted negotiation engine loads the corresponding industry semantic mapping rules and compliance rules according to the industry attributes;

[0049] In this application embodiment, industry compliance rules are normative requirements such as laws, regulations, administrative rules, national standards, and industry standards that must be followed by a specific industry.

[0050] Specifically, after both parties submit their industry attributes, the trusted negotiation engine understands and processes these attributes. Internally, the engine connects to industry semantic mapping rules and compliance rules. After identifying the data provider's industry attributes, it loads the applicable semantic mapping rules and compliance rules from the rule base. This ensures that the industry meets the mandatory requirements at each stage of the data lifecycle, guaranteeing that any solutions developed through subsequent negotiations remain within the basic compliance framework of each industry and avoid legal risks.

[0051] Step S20 in the method provided in this application embodiment includes:

[0052] Based on the data provider's primary industry attribute, load the corresponding primary industry semantic mapping rules and primary industry compliance rules from the pre-built rule base;

[0053] Based on the second industry attribute of the data user, load the corresponding second industry semantic mapping rules and second industry compliance rules from the rule base;

[0054] When the first industry attribute is different from the second industry attribute, a bridging mapping rule for policy semantic conversion between the first industry and the second industry is loaded from the rule base;

[0055] The bridging mapping rule is used to implement at least one of the following transformations:

[0056] The business compliance requirements defined in the first industry semantic mapping rules are converted into equivalent technical control parameters defined in the second industry semantic mapping rules;

[0057] The data processing constraints defined in the second industry compliance rules are converted into equivalent constraint expressions that meet the requirements of the first industry compliance rules;

[0058] Transform the domain-specific business terms in the data provider's data sharing strategy elements into the corresponding cross-domain technical implementation terms in the data user's data sharing strategy elements.

[0059] In this embodiment of the application, firstly, based on the first industry attribute of the data provider, the corresponding first industry semantic mapping rule and first industry compliance rule are loaded from the pre-built rule base. The first industry semantic mapping rule is a conversion rule that matches the industry to which the data provider belongs and transforms the business concepts and strategic intentions within the industry into executable technical parameters. The first industry compliance rule is a law, regulation or standard that matches the industry to which the data provider belongs and has mandatory binding force on industry data processing activities.

[0060] After the trusted negotiation engine receives the first industry attribute registered by the data provider, it uses this first industry attribute as the query keyword to load the corresponding first industry semantic mapping rules and first industry compliance rules from the rule base. The first industry semantic mapping rules convert industry semantics into rules for specific technical control parameters; the first industry compliance rules include mandatory laws and regulations and are structured and encoded in formats such as IF-THEN. For example, IF Data Category == Personal Health Information THEN Must Obtain Separate Consent.

[0061] Secondly, based on the second industry attribute of the data user, the corresponding second industry semantic mapping rules and second industry compliance rules are loaded from the rule base. The second industry semantic mapping rules are a set of rules that match the industry to which the data user belongs, transform the technical solutions and business processes within the industry into standardized parameters, and understand the definition and conversion rules of the industry-specific terms. The second industry compliance rules are a structured set of mandatory rules that match the industry to which the data user belongs and regulate the industry data acquisition, use and output activities.

[0062] After obtaining the semantic mapping rules and compliance rules for the first industry, the trusted negotiation engine uses the second industry attribute registered by the data user as a keyword to retrieve and load the semantic mapping rules and compliance rules for the second industry from the pre-set rule base.

[0063] Furthermore, when the attributes of the first industry are different from those of the second industry, a bridging mapping rule for policy semantic conversion between the first and second industries is loaded from the rule base.

[0064] When data providers and users come from different industries, simply loading two sets of rules is insufficient for analyzing industry terminology from both sectors. Furthermore, if the data quality requirements of the first and second industries differ, the resulting industry standards and rules may conflict. Therefore, bridging mapping rules are needed to perform policy semantic transformation, enabling mutual understanding between the rules of the first and second industries. The existence of bridging mapping rules is contingent upon the following condition: attributes of the first industry ≠ attributes of the second industry. If the first and second industries belong to the same sector, then identity mapping rules may not be necessary.

[0065] The bridging mapping rules are used to implement at least one of the following transformations:

[0066] Business compliance requirements should be translated into equivalent technical control parameters, which are directly applicable industry technical system execution parameters. For example, compliance requirements in the healthcare industry ensure data anonymization and prevent re-identification, but it's not appropriate to directly require the financial system to prevent re-identification. Bridging rules need to be converted into parameters that are understandable and configurable by the financial technology system. For instance, before federated learning feature exchange, differential privacy mechanisms should be applied, and the noise scale must be ≤1.0.

[0067] Using industry-specific data processing constraints, we can interpret them as equivalent constraints that can be recognized within the industry's compliance framework. For example, the compliance constraints in the financial industry require external data to meet anti-money laundering screening requirements, while the healthcare industry may not have a direct corresponding concept. The bridging rules need to convert these constraints into expressions that comply with the ethics and compliance of healthcare data sharing.

[0068] The bridging rule directly converts specific business terms within the data sharing strategy elements from domain-specific business terminology to technical implementation terminology. For example, the scope of data usage purpose in the data provider's strategy element, such as "comparative clinical efficacy study," might not be easily understood by the data user. This bridging rule maps it to the technical implementation terminology used by the data user, making it easier for them to comprehend.

[0069] In step S20 of the method provided in this application embodiment, the rule base construction process includes:

[0070] We collect normative texts and historical contract data from different industries through standards organization publication platforms, public announcements from industry regulatory agencies, and de-identified historical data sharing contracts.

[0071] The collected normative texts and historical contract data are structured to construct an industry rule knowledge graph that includes industry semantic mapping rules and industry compliance rules.

[0072] Based on the historical contract data, cross-industry bridging mapping rules are generated through machine learning training;

[0073] Consistency and validity verifications are performed on the rules in the industry rule knowledge graph and the cross-industry bridging mapping rules, and the verified rules are classified and stored to form a rule base.

[0074] In this application embodiment, the standards organization publishing platform is an information publishing channel established at the international, national, or industry level; the public announcements of industry regulatory agencies are documents with universal binding force; the anonymized historical data sharing contracts are contract texts generated in past data sharing cooperation; the normative texts are texts derived from standards and regulatory provisions; and the historical contract data are texts that have been practically implemented based on practical experience.

[0075] First, we collected normative texts and historical contract data from different industries from standards organization publication platforms, public announcements from industry regulatory agencies, and de-identified historical data sharing contracts.

[0076] Relying solely on standard regulations may result in rules that are too general and lack operability; relying solely on historical contracts may lead to rules that lack authority and are limited by past cases. Combining normative texts with historical contract data ensures that the generated rules are both regulatory compliant and practically feasible. Furthermore, collecting normative texts from multiple levels ensures the breadth of the rule base, enabling it to support a wide range of cross-industry scenarios.

[0077] Secondly, the collected normative texts and historical contract data are structured to construct an industry rule knowledge graph that includes industry semantic mapping rules and industry compliance rules. The structured processing is the process of using natural language processing (NLP) technology to convert unstructured legal provisions and other natural language texts into discrete data units and relationships that can be recognized and associated by computers. The industry rule knowledge graph uses graph structures to model and store industry rule knowledge.

[0078] Specifically, the natural language processing model can perform organizational processing including entity identification, relation extraction, and attribute labeling. Entity identification involves using the natural language processing model to find key entities such as personal information, sensitive information, and explicit consent from normative texts and historical contract data. Relationships are then extracted from the identified entities to identify the logical relationships between them.

[0079] After entity identification and relation extraction are completed, entities are labeled according to their attribute features. For example, technical attributes are labeled for "anonymized" entities. After structuring all entities, nodes, attributes, and relationships are created based on the results of the structuring process, following the knowledge graph paradigm. These relationships are then persistently stored to construct an industry rule knowledge graph. In this process, identified entities are used as nodes in the industry rule knowledge graph, entity relationships are used as edges (i.e., relationships between nodes), and attribute labels are used as node attributes. Finally, nodes are determined, and edges are connected to them. Then, the attributes of the nodes are determined to construct the industry rule knowledge graph.

[0080] Furthermore, based on historical contract data, cross-industry bridging mapping rules are generated through machine learning training.

[0081] The process of using graph neural networks for machine learning to train and generate cross-industry bridging mapping rules is as follows: using de-identified historical data sharing contracts as training data, the model's goal is to predict or generate feasible and equivalent technical constraints or terms in the target industry for a given business requirement and the context of the target industry.

[0082] For example, using anonymized historical data sharing contracts as training data for machine learning analysis, the model can learn bridging mapping rules. The embedded industry nodes gradually absorb information about related business concepts, technical parameters, and regulations, extracting positive samples from historical contracts. Simultaneously, negative examples are generated through negative sampling. The model's learning objective is to maximize the similarity between the source and target nodes in the embedding space within the positive samples.

[0083] After multiple iterations, the position of each node in the high-dimensional vector space encodes the structural, semantic, and contextual information of the entire historical contract graph, resulting in bridging mapping rules, such as: IF (First industry attribute == Healthcare AND Second industry attribute == Finance & Insurance AND Data usage purpose includes risk assessment) THEN Recommended data processing technology solution = {Federated learning, secure multi-party computation}; Recommended data output format requirements = {Model parameters, statistical reports}.

[0084] Finally, consistency and validity verification are performed on the rules in the industry rule knowledge graph and the cross-industry bridging mapping rules. The verified rules are then categorized and stored to form a rule base. Consistency verification verifies whether there are logical contradictions or conflicts between different rules; validity verification assesses whether the rules are correct, useful, and in line with the current legal and technological environment.

[0085] Consistency verification determines whether the rules in the industry rule knowledge graph and the cross-industry bridging mapping rules are consistent; validity verification checks whether the rules can be effectively executed. If both consistency and validity verifications pass, the metadata is organized and indexed according to the rules and stored in the database or knowledge base.

[0086] In this embodiment, industry semantic mapping rules and industry compliance rules for data providers and data users are obtained by corresponding industry attributes. Simultaneously, bridging mapping rules for policy semantic transformation are acquired, enabling the system to adapt to any industry attribute. Subsequently, normative texts and historical contract data are collected from platforms such as standards and regulatory notices to ensure the compliance and authority of the rules. A knowledge graph is constructed to identify the relationships between entities. Based on machine learning from historical contracts, cross-industry bridging rules are generated, reducing reliance on expert experience. Finally, consistency verification and categorized storage ensure the inherent consistency and application reliability of the rule base, making the knowledge base source upon which the trusted negotiation engine relies reliable and capable of self-optimization based on practical cases.

[0087] S30: During the data sharing process, dynamic context information is continuously collected, wherein the dynamic context information includes data processing environment security certification, business scenario identification and external regulatory events;

[0088] In this embodiment, dynamic context information refers to environmental factors and state information that may change during the sharing process but cannot be fully determined at the beginning of the negotiation; security proof is evidence that proves the computing environment of the data user meets specific security standards or technical requirements; and business scenario identifier is the specific application scenario identified by the data user when initiating a data access or processing request.

[0089] Specifically, a remote verification mechanism is used to collect reports and verify whether the environment corresponding to the trusted negotiation engine is operating normally and has not been tampered with during the sharing process. Dynamic context information is collected to obtain data processing environment security certificates, business scenario identifiers, and external regulatory events. This dynamic context information is continuously collected and fed back to the trusted negotiation engine to assess the security and compliance of the original protocol, or to trigger renegotiation when necessary.

[0090] Step S30 in the method provided in this application embodiment includes:

[0091] By deploying a monitoring agent in the data processing environment of the data user, security certificates of the data processing environment are continuously collected, wherein the security certificates of the data processing environment include Trusted Execution Environment (TEE) remote certification reports or secure container runtime credentials.

[0092] By parsing the business request messages from the data user, the current business scenario identifier can be extracted;

[0093] By subscribing to regulatory agencies' publishing interfaces, external regulatory events can be obtained in real time;

[0094] The data processing environment security certificate, business scenario identifier, and external regulatory event are all accompanied by timestamp information.

[0095] In this embodiment, the regulatory agency release interface is an application programming interface provided by industry regulatory agencies, government departments, or standards organizations to obtain information such as official announcements and policy updates; external regulatory events are event information initiated by authoritative institutions outside the data sharing participants; the Trusted Execution Environment (TEE) remote certification report is a report that can be generated internally by the environment and signed by a hardware chip; and the secure container runtime certificate is a document that proves the integrity of the container, runtime isolation, and compliance with security policies.

[0096] Specifically, firstly, before or simultaneously with data sharing, a monitoring agent is deployed at the data user's location. Within the server or container cluster processing the shared data, it retrieves the latest remote verification report or operational credentials and performs verification using the verification service or corresponding verification signature. If the verification is invalid or does not match the baseline value, the trusted negotiation engine can determine that the environment may have been compromised, thereby triggering an alarm or suspending data sharing.

[0097] Secondly, at the data access entry point, the system intercepts or listens to every business request message sent by the data user. By parsing the header, parameters, or payload of this message, the system can extract predefined business scenario identifiers. Here, the business request message is a structured data unit sent by the data user to its data processing system or to the data provider's gateway when specifically using shared data; the business scenario identifier is a classification label or code used to clearly distinguish the specific use scenario of the data.

[0098] Secondly, the trusted negotiation engine pre-registers and subscribes to the interfaces published by relevant regulatory agencies. Regulatory dynamic event information is pushed to the trusted negotiation engine in real time through the interface, and then parsed and classified.

[0099] Among them, security certificates obtained from monitoring agents, scenario identifiers extracted from request messages, and events pushed from regulatory interfaces are accompanied by timestamp information. The timestamps can be used to determine data leakage events or compliance violations, and can prevent attackers from intercepting old, valid certificates for replay.

[0100] In this embodiment, by deploying monitoring, the data processing environment is transformed into a data processing environment security certificate, thereby achieving dynamic verification of the data processing environment; secondly, scenario identifiers are extracted by parsing business request messages; finally, external events are obtained by subscribing to regulatory interfaces, enabling the system to proactively detect and quickly respond to changes in the compliance environment.

[0101] S40: The trusted negotiation engine, based on the data sharing strategy elements, constraints, dynamic context information, and the loaded industry semantic mapping rules and compliance rules, executes a negotiation algorithm to generate at least one data sharing strategy negotiation scheme.

[0102] In this embodiment, the negotiation algorithm seeks a dynamic solution that maximizes the satisfaction of all parties' strategic elements while meeting all hard constraints; the data sharing strategy negotiation scheme is an executable data sharing protocol scheme generated by the negotiation algorithm.

[0103] Specifically, the first step is to align the needs of both parties at the semantic level, integrate all input information: data sharing strategy elements, constraints, dynamic context information, and loaded industry semantic mapping rules and compliance rules, execute negotiation algorithms, search and optimize in a multi-dimensional solution space composed of various parameters, and find one or more specific sets of strategy parameters as a data sharing strategy negotiation scheme.

[0104] Step S40 in the method provided in this application embodiment includes:

[0105] Input the data sharing strategy elements of the data provider, the data sharing strategy elements of the data user, the constraints of the data provider, and the constraints of the data user into the negotiation algorithm;

[0106] The negotiation algorithm, based on the loaded first industry semantic mapping rule, second industry semantic mapping rule, and bridging mapping rule, performs semantic alignment and standardization transformation on all input strategy elements and constraints to generate standard strategy elements and standard constraints.

[0107] The negotiation algorithm, based on the standard strategy elements and standard constraints, and combined with the currently collected dynamic context information, generates multiple candidate data sharing strategy negotiation schemes through optimized search.

[0108] In this embodiment of the application, firstly, the data sharing strategy elements of the data provider, the data sharing strategy elements of the data user, the constraints of the data provider, and the constraints of the data user, that is, all the static negotiation basic information collected and temporarily stored in the early stage, are integrated in a format and submitted to the negotiation algorithm for processing.

[0109] Secondly, the negotiation algorithm identifies and eliminates ambiguities caused by differences in terminology, conceptual systems, or expression habits between information from different sources through semantic alignment, based on the first industry semantic mapping rules, the second industry semantic mapping rules, and the bridging mapping rules. Then, it re-expresses and encodes the aligned semantic information through standardization transformation. After rule-driven transformation, standard strategy elements and standard constraints are obtained.

[0110] Ultimately, standard strategy elements are used as objectives, and standard constraints are conditions that must be absolutely satisfied. Then, the current dynamic context information is used to adjust the optimization direction in real time, assessing the degree of satisfaction with the standard strategy elements and whether the standard constraints are violated. This context-dependent optimization search is performed throughout the entire parameter combination space, exploring a large number of possible solutions, and finally selecting candidate solutions that perform well in different dimensions or have different focuses.

[0111] In step S40 of the method provided in this application embodiment, the negotiation algorithm, based on the standard strategy elements and standard constraints, and combined with the currently collected dynamic context information, generates multiple data sharing strategy negotiation schemes through optimized search, including:

[0112] Based on data from historical successful negotiation cases, multiple initial strategy parameter sets are generated using control strategy parameter sets to form an initial search space;

[0113] For each set of strategy parameters in the initial search space, a comprehensive quality score is calculated, wherein the comprehensive quality score is determined based on the degree to which the set of strategy parameters satisfies the standard strategy elements, the degree to which it conforms to the standard constraints, and the degree to which it matches the current dynamic context information.

[0114] Based on the comprehensive quality score, the set of strategy parameters with high scores is selected as the high-quality set, and a new set of strategy parameters is generated and introduced based on the high-quality set according to the preset update ratio, forming an updated search space.

[0115] The process of iteratively calculating the overall quality score, selecting the high-quality set, and updating the search space continues until a preset convergence condition is reached. The convergence condition includes reaching a preset maximum number of iterations or the change in the highest overall quality score in a consecutive preset number of iterations being lower than a preset threshold.

[0116] When the convergence condition is met, the set of one or more policy parameters with the highest comprehensive quality score constitutes a candidate data sharing policy negotiation scheme.

[0117] In this embodiment, firstly, multiple historical successful negotiation cases are obtained from a database storing such cases. Then, a set of control strategy parameters is extracted from these cases, and multiple initial strategy parameter sets are generated through adjustments, forming the initial search space for the first round of searching.

[0118] For example, the system's successful negotiation case library contains parameter sets of successful cases. Parameters are extracted from each case to generate an initial strategy parameter set, which constitutes the initial search space: set 1, set 2, set 3, etc., a total of N newly generated strategy parameter sets, which serve as the initial search space for the first round of algorithm iteration.

[0119] Secondly, for each set of strategy parameters in the initial search space, the set of strategy parameters is compared with standard strategy elements and standard constraints. Based on the degree of matching of the current dynamic context information, a comprehensive quality score is calculated. In the comparison with standard strategy elements, for example, if the standard elements require high data utility and the current scheme sets a high update frequency, the score is high; if low processing latency is required and complex homomorphic encryption leads to high latency, the score is low.

[0120] By comparing the solution with standard constraints, it is checked whether the solution violates any standard constraints. For example, if a standard constraint requires ≥5, but the current data processing flow may cause the effective k-value to drop to 4, then the compliance level is non-compliance. By comparing the solution with the current dynamic context information, the solution is analyzed in conjunction with the real-time dynamic context. Finally, a weighted fusion is used to calculate the comprehensive quality score. The weights of the three factors can be determined according to their influence on the comprehensive quality score, and technical personnel can adjust them according to the actual situation. The sum of the weights of the three factors is 1.

[0121] For example, suppose the initial set of strategy parameters in the search space has a satisfaction level of 0.8 with standard strategy elements and a compliance level of 1 with standard constraints. Its matching degree with the current dynamic is low at 0.7. w1, w2, and w3 are 0.3, 0.4, and 0.3 respectively. The overall quality score is: Overall Score = 0.3 × 0.7 + 0.4 × 1 + 0.3 × 0.8 = 0.85, and the overall quality score is 85 points.

[0122] Next, based on the comprehensive quality score, all solutions are ranked, and a certain percentage of the ranked solutions are selected as the high-quality set for this round. Then, according to the preset update ratio, the update range of the search space is determined, and the solution with the lowest update ratio is eliminated. To maintain the stability of the search space size, new solutions with the same update ratio are generated based on the high-quality set through crossover and mutation: First, two high-quality solutions are randomly selected, and some of their parameters are swapped to form two new solutions; then, a high-quality solution is randomly selected, and one or more of its parameters are randomly and slightly adjusted. The newly generated solution, together with the high-quality solutions that were not eliminated, forms the updated search space for the next iteration.

[0123] For example, suppose the initial search space has 100 sets of strategy parameters. Select the top 30 sets with the highest overall scores as the high-quality sets, with a preset update ratio of 70%, discard the 70 sets with the lowest scores, and then regenerate 70 new sets.

[0124] Subsequently, following the steps described above, the updated search space is used to recalculate the overall quality score, filter the best set, and update the space again, repeating this process until the preset maximum number of iterations is reached (e.g., a maximum of 200 iterations); or the change in the highest overall quality score within a preset number of iterations is less than a preset threshold (e.g., the change in the highest score is less than 1 point across 10 consecutive iterations). If the change is less than the preset threshold, convergence is achieved, and iteration can stop. During this process, the highest overall quality score among all solutions in each generation is continuously tracked, representing the score of the current optimal solution. The convergence condition is the termination criterion for determining when the optimization search algorithm stops iterating and finds the best solution; the preset maximum number of iterations is a safe number of iterations to prevent the algorithm from looping indefinitely, after which iteration will be forcibly stopped.

[0125] For example, suppose that from generation N to generation N+9, the highest scores for these 10 consecutive generations are: [92.1, 92.3, 92.0, 92.2, 92.1, 92.1, 92.2, 92.0, 92.1, 92.1], where the fluctuations are very small, and the maximum change is much less than 1 point. In this case, the search is considered to have converged, and the iteration is stopped.

[0126] Finally, after the convergence condition is met and iteration stops, one or more sets of policy parameters with the highest overall quality score are selected from the search space updated last time. For example, the solution with the highest absolute score can be selected, or the top three solutions in terms of score can be selected. The one or more sets of policy parameters with the highest overall quality score represent the optimal or suboptimal solution found after multiple rounds of evolution, and these constitute the candidate data-sharing policy negotiation scheme.

[0127] In this embodiment, according to the negotiation algorithm, standard strategy elements and standard constraints are generated through semantic alignment and standardization transformation to eliminate semantic ambiguity; then, an initial solution is generated based on historical successful cases to ensure the high quality of the search starting point and accelerate the convergence process; by designing a comprehensive quality scoring function, standard strategy elements, standard constraints, and dynamic context are considered in a unified manner to ensure the comprehensive optimality of the solution; through iteration, directional exploration in the solution space is carried out while effective updates are performed to avoid getting trapped in local optima, so that the system can select the optimal or suboptimal solution in the strategy combination and obtain candidate data sharing strategy negotiation schemes.

[0128] S50: Evaluate the expected effects and verify the compliance of the data sharing strategy negotiation scheme, and obtain the evaluation results and verification results;

[0129] In this application embodiment, the expected effect assessment is the process of predicting and quantifying the business value, utility, or benefits that a certain negotiation plan may bring to the data user; the compliance verification is the process of comparing and checking the specific terms and control measures in the negotiation plan with the applicable industry compliance rules to confirm whether there are any conflicts or risks of violation.

[0130] Specifically, through a pre-trained value prediction model, based on the effect data of historical data sharing cases, the predicted utility value is output as the evaluation result, and the generated expected effect evaluation is verified for compliance: the trusted negotiation engine compares each control point in the data sharing strategy negotiation scheme with the first industry compliance rules and the second industry compliance rules to verify the compliance of the scheme.

[0131] Step S50 in the method provided in this application embodiment includes:

[0132] The data sharing strategy negotiation scheme is evaluated using a pre-trained value prediction model. The expected data utility value that can be generated after the data sharing strategy negotiation scheme is executed is predicted, and the evaluation result is obtained.

[0133] The set of control strategy parameters in the data sharing strategy negotiation scheme is compared one by one with the clauses in the loaded first industry compliance rules and second industry compliance rules to identify potential compliance conflict points, and compliance modification suggestions are generated based on bridging mapping rules to obtain verification results.

[0134] The evaluation results and the verification results are formatted into a machine-readable evaluation report.

[0135] In this embodiment of the application, firstly, a pre-trained value prediction model is used to evaluate the data sharing strategy negotiation scheme, predict the expected data utility value that can be generated after the data sharing strategy negotiation scheme is executed, and obtain the evaluation result. The pre-trained value prediction model is a prediction function trained by machine learning algorithm in the offline stage using a large amount of historical data sharing project execution effect data as training samples.

[0136] Specifically, the strategy parameters of the data sharing strategy negotiation scheme are used as input, and the data utility values ​​of similar historical schemes are used as labels. A neural network model is used for supervised training. By initializing the model parameters, forward propagation is used to calculate the error between the predicted result and the actual result using the loss function. Backpropagation is used to reduce the error, and finally the value prediction model is trained.

[0137] For example, a value prediction model can be constructed using a fully connected neural network. The structure includes an input layer, hidden layers, and an output layer. The input layer directly receives the input data. The hidden layers perform complex nonlinear transformations on the data to extract deep features. The output layer generates the final prediction result.

[0138] Subsequently, the strategy parameters of the data sharing negotiation scheme are used as input, and the utility values ​​of historical similar schemes are used as labels. The training, validation, and test sets are then divided in a 7:2:1 ratio. A learning rate of 0.001 is set, and the Adam optimizer is used. Through forward propagation, a non-linear transformation of the input data is performed using an activation function. The gradient of the error is calculated through backpropagation, and the weights and biases are updated using gradient descent to reduce the error. After training, the model's performance on the reserved validation set is evaluated, ultimately yielding the value prediction model.

[0139] Secondly, the control parameter set is traversed, and the control strategy parameter set in the data sharing strategy negotiation scheme is compared one by one with the clauses in the loaded first industry compliance rules and second industry compliance rules. Some provisions in the scheme may have potential compliance conflicts with the currently effective mandatory industry rules. Based on the bridging mapping rules, the clause modification or supplement schemes for the identified conflict points are generated. Finally, all comparison conclusions, conflict points and modification suggestions are summarized to form the verification results.

[0140] Finally, the evaluation and verification results are formatted and integrated into a well-structured, machine-readable evaluation report according to the output format, ensuring that the negotiated results can be securely transformed into actual data sharing operations.

[0141] In this embodiment, a pre-trained value prediction model is used to evaluate the expected results. Based on historical data, the business results of the predicted scheme are quantified, and the merits of the scheme are transformed into data comparisons to obtain evaluation results. Subsequently, the scheme is compared with control rules to form verification results, thereby achieving automated compliance review and reducing identification risks. Finally, the evaluation and verification results are formatted into an evaluation report to ensure the efficiency and accuracy of information flow.

[0142] S60: Based on the evaluation and verification results, the data provider and the data user reach a consensus, generate a verifiable strategy contract, and execute data sharing operations according to the verifiable strategy contract.

[0143] In this embodiment of the application, the verifiable strategy contract is an electronic contract that can be automatically executed and verified by forming a consensus-reached data sharing strategy scheme.

[0144] Specifically, the evaluation results and verification results are sent to the data provider and the data user respectively through the trusted negotiation engine. If they accept the same solution, it means that the two parties have reached a consensus. Subsequently, the trusted negotiation engine can programmable and verifiable strategy contracts.

[0145] Step S60 in the method provided in this application embodiment includes:

[0146] The assessment report will be sent to both the data provider and the data user.

[0147] Based on the assessment report, the data provider and the data user shall conduct at least one round of confirmation or modification feedback through the trusted negotiation engine;

[0148] When both the data provider and the data user confirm acceptance of the modified version of the negotiated plan based on the same data sharing strategy, it is determined that a consensus has been reached.

[0149] The trusted negotiation engine compiles the final data sharing strategy negotiation scheme that has reached a consensus into a verifiable strategy contract. The verifiable strategy contract is deployed on the blockchain in the form of a smart contract, and the contract address and content hash of the verifiable strategy contract are returned to the data provider and the data user.

[0150] In this embodiment, the verifiable strategy contract is a smart contract containing all the logic of the consensus scheme; the contract address is a unique identifier of the smart contract on the blockchain; and the content hash is a fixed-length string obtained by performing cryptographic hashing on the final bytecode or source code of the smart contract.

[0151] Specifically, firstly, by calling the communication interface reserved during registration, the evaluation report is sent to both the data provider and the data user to avoid unfairness caused by information transmission delays or deviations.

[0152] For example, the assessment report is sent to the built-in message center of the Health Cloud Hospital Data Governance Platform through a secure channel, triggering an internal notification within the platform; it is also sent to the user's federated learning operations management platform through a secure channel.

[0153] Secondly, the data provider and the data user review the evaluation report. After the review, they confirm each candidate solution through a dedicated negotiation interface provided by the trusted negotiation engine. They confirm the acceptance of the solution or propose adjustments to certain specific terms, and conduct at least one round of confirmation or modification feedback until an agreement is reached.

[0154] Furthermore, after one or more rounds of modification and feedback, until after a certain modification, when both the data provider and the data user confirm acceptance of the modified version based on the same data sharing strategy negotiation plan, it is determined that a consensus has been reached and the data sharing strategy negotiation plan has been recognized by both parties.

[0155] Finally, the contract code is deployed to the pre-agreed blockchain. Upon successful deployment, the blockchain network assigns a unique contract address. Simultaneously, the content hash of the contract code is calculated and returned to both the data provider and the data user. Upon receiving this, both parties can independently query the contract via the address on the blockchain explorer and ensure consistency by comparing the hash values. Ultimately, data sharing no longer relies on continuous scheduling by a central engine; it is automatically executed and arbitrated by verifiable policy contracts deployed on the blockchain.

[0156] In this embodiment, the evaluation reports are sent and structured feedback is collected to establish an online negotiation channel, thereby improving communication efficiency. Subsequently, through objective analysis, a confirmed data sharing strategy negotiation scheme is obtained under the consensus of mutual acceptance. The data sharing strategy negotiation scheme is then compiled into a blockchain smart contract and deployed on the blockchain to realize the execution transformation of the negotiation result, making the sharing strategy an immutable and automatically executed code. At the same time, the content hash is calculated to provide integrity proof for the contract address.

[0157] The embodiments of this application, through the above specific implementation methods, achieve the following technical effects:

[0158] In this embodiment, by first specifying the categories of strategic elements and non-negotiable constraints that data providers and users need to submit, the vague and unstructured business intentions are transformed into structured and dimensional inputs that can be accurately processed by machines, and the hard boundaries of security and compliance in the negotiation are determined, thus laying a reliable data foundation for automated negotiation.

[0159] Secondly, by matching industry attributes with the industry semantic mapping rules and industry compliance rules of data providers and data users, and simultaneously acquiring bridging mapping rules for policy semantic transformation, the system can adapt to any industry attribute. Subsequently, by collecting normative texts and historical contract data from platforms such as standards and regulatory notices, the compliance and authority of the rules are ensured. A knowledge graph is constructed to identify the relationships between entities. Based on machine learning from historical contracts, cross-industry bridging rules are generated, reducing reliance on expert experience. Finally, consistency verification and categorized storage ensure the inherent consistency and application reliability of the rule base, making the knowledge base source upon which the trusted negotiation engine relies reliable and capable of self-optimization based on practical cases.

[0160] Secondly, by deploying monitoring, the data processing environment is transformed into a data processing environment security certificate, enabling dynamic verification of the data processing environment. Thirdly, by parsing business request messages, scenario identifiers are extracted. Finally, by subscribing to regulatory interfaces to obtain external events, the system acquires the ability to proactively detect and quickly respond to changes in the compliance environment.

[0161] Furthermore, based on the negotiation algorithm, standard policy elements and standard constraints are generated through semantic alignment and standardization transformation to eliminate semantic ambiguity. Then, initial solutions are generated based on historical successful cases to ensure the high quality of the search starting point and accelerate the convergence process. By designing a comprehensive quality scoring function, standard policy elements, standard constraints, and dynamic context are considered in a unified manner to ensure the comprehensive optimality of the solution. Through iteration, directional exploration in the solution space is carried out while effective updates are performed to avoid getting trapped in local optima. This enables the system to select the optimal or suboptimal solution from the policy combination and obtain candidate data sharing policy negotiation solutions.

[0162] Meanwhile, by using a pre-trained value prediction model to evaluate the expected results, the business outcomes of the proposed solution are quantified based on historical data, and the merits and demerits of the solution are transformed into data comparisons to obtain evaluation results. Subsequently, the solution is compared with control rules to form verification results, thereby achieving automated compliance review and reducing identification risks. Finally, the evaluation and verification results are formatted into an evaluation report to ensure the efficiency and accuracy of information flow.

[0163] Finally, the evaluation reports were sent out separately and structured feedback was collected. An online negotiation channel was established, which improved communication efficiency. Subsequently, through objective analysis, a confirmed data sharing strategy negotiation plan was obtained under the consensus of mutual acceptance. The data sharing strategy negotiation plan was then compiled into a blockchain smart contract and deployed on the blockchain to realize the execution transformation of the negotiation results. This made the sharing strategy an immutable and automatically executed code. At the same time, the content hash was calculated to provide integrity proof for the contract address.

[0164] Example 2, as Figure 2 As shown, based on the same inventive concept as the cross-industry data sharing method supporting dynamic strategy negotiation provided in Embodiment 1, this embodiment of the invention also provides a cross-industry data sharing system supporting dynamic strategy negotiation, comprising:

[0165] The attribute registration module 11 is used to guide data providers and data users to register their respective industry attributes with the trusted negotiation engine and submit data sharing strategy elements and constraints.

[0166] The rule acquisition module 12 is used by the trusted negotiation engine to load the corresponding industry semantic mapping rules and compliance rules according to the industry attributes.

[0167] The information collection module 13 is used to continuously collect dynamic context information during the data sharing process, wherein the dynamic context information includes data processing environment security certification, business scenario identification and external regulatory events;

[0168] The negotiation scheme generation module 14 is used by the trusted negotiation engine to execute a negotiation algorithm based on the data sharing strategy elements, constraints, dynamic context information, and the loaded industry semantic mapping rules and compliance rules to generate at least one data sharing strategy negotiation scheme.

[0169] The result evaluation module 15 is used to evaluate the expected effect and verify the compliance of the data sharing strategy negotiation scheme, and obtain the evaluation results and verification results.

[0170] Data sharing module 16 is used to facilitate consensus between the data provider and the data user based on the evaluation and verification results, generate a verifiable strategy contract, and execute data sharing operations according to the verifiable strategy contract.

[0171] In one embodiment, the attribute registration module 11 is used for:

[0172] The data provider registers the first industry attribute with the trusted negotiation engine and submits the data provider's data sharing strategy elements and data provider constraints. The data provider's data sharing strategy elements include at least one of the following: data usage purpose scope, data usage timeliness, and data usage geographical restrictions.

[0173] The data user registers a second industry attribute with the trusted negotiation engine and submits the data user's data sharing strategy elements and data user constraints. The data user's data sharing strategy elements include at least one of the following: data processing technology solution, data processing environment certification, and data output format requirements.

[0174] The constraints imposed on the data provider and the constraints imposed on the data user both include their respective non-negotiable bottom-line policy clauses.

[0175] In one embodiment, the rule acquisition module 12 is used for:

[0176] Based on the data provider's primary industry attribute, load the corresponding primary industry semantic mapping rules and primary industry compliance rules from the pre-built rule base;

[0177] Based on the second industry attribute of the data user, load the corresponding second industry semantic mapping rules and second industry compliance rules from the rule base;

[0178] When the first industry attribute is different from the second industry attribute, a bridging mapping rule for policy semantic conversion between the first industry and the second industry is loaded from the rule base;

[0179] The bridging mapping rule is used to implement at least one of the following transformations:

[0180] The business compliance requirements defined in the first industry semantic mapping rules are converted into equivalent technical control parameters defined in the second industry semantic mapping rules;

[0181] The data processing constraints defined in the second industry compliance rules are converted into equivalent constraint expressions that meet the requirements of the first industry compliance rules;

[0182] Transform the domain-specific business terms in the data provider's data sharing strategy elements into the corresponding cross-domain technical implementation terms in the data user's data sharing strategy elements.

[0183] The process of building the rule base includes:

[0184] We collect normative texts and historical contract data from different industries through standards organization publication platforms, public announcements from industry regulatory agencies, and de-identified historical data sharing contracts.

[0185] The collected normative texts and historical contract data are structured to construct an industry rule knowledge graph that includes industry semantic mapping rules and industry compliance rules.

[0186] Based on the historical contract data, cross-industry bridging mapping rules are generated through machine learning training;

[0187] Consistency and validity verifications are performed on the rules in the industry rule knowledge graph and the cross-industry bridging mapping rules, and the verified rules are classified and stored to form a rule base.

[0188] In one embodiment, the information acquisition module 13 is used for:

[0189] By deploying a monitoring agent in the data processing environment of the data user, security certificates of the data processing environment are continuously collected, wherein the security certificates of the data processing environment include Trusted Execution Environment (TEE) remote certification reports or secure container runtime credentials.

[0190] By parsing the business request messages from the data user, the current business scenario identifier can be extracted;

[0191] By subscribing to regulatory agencies' publishing interfaces, external regulatory events can be obtained in real time;

[0192] The data processing environment security certificate, business scenario identifier, and external regulatory event are all accompanied by timestamp information.

[0193] In one embodiment, the negotiation scheme generation module 14 is used for:

[0194] Input the data sharing strategy elements of the data provider, the data sharing strategy elements of the data user, the constraints of the data provider, and the constraints of the data user into the negotiation algorithm;

[0195] The negotiation algorithm, based on the loaded first industry semantic mapping rule, second industry semantic mapping rule, and bridging mapping rule, performs semantic alignment and standardization transformation on all input strategy elements and constraints to generate standard strategy elements and standard constraints.

[0196] The negotiation algorithm, based on the standard strategy elements and standard constraints, and combined with the currently collected dynamic context information, generates multiple candidate data sharing strategy negotiation schemes through optimized search.

[0197] The negotiation algorithm, based on the standard strategy elements and standard constraints, and combined with the currently collected dynamic context information, generates multiple data sharing strategy negotiation schemes through optimized search, including:

[0198] Based on data from historical successful negotiation cases, multiple initial strategy parameter sets are generated using control strategy parameter sets to form an initial search space;

[0199] For each set of strategy parameters in the initial search space, a comprehensive quality score is calculated, wherein the comprehensive quality score is determined based on the degree to which the set of strategy parameters satisfies the standard strategy elements, the degree to which it conforms to the standard constraints, and the degree to which it matches the current dynamic context information.

[0200] Based on the comprehensive quality score, the set of strategy parameters with high scores is selected as the high-quality set, and a new set of strategy parameters is generated and introduced based on the high-quality set according to the preset update ratio, forming an updated search space.

[0201] The process of iteratively calculating the overall quality score, selecting the high-quality set, and updating the search space continues until a preset convergence condition is reached. The convergence condition includes reaching a preset maximum number of iterations or the change in the highest overall quality score in a consecutive preset number of iterations being lower than a preset threshold.

[0202] When the convergence condition is met, the set of one or more policy parameters with the highest comprehensive quality score constitutes a candidate data sharing policy negotiation scheme.

[0203] In one embodiment, the result evaluation module 15 is used for:

[0204] The data sharing strategy negotiation scheme is evaluated using a pre-trained value prediction model. The expected data utility value that can be generated after the data sharing strategy negotiation scheme is executed is predicted, and the evaluation result is obtained.

[0205] The set of control strategy parameters in the data sharing strategy negotiation scheme is compared one by one with the clauses in the loaded first industry compliance rules and second industry compliance rules to identify potential compliance conflict points, and compliance modification suggestions are generated based on bridging mapping rules to obtain verification results.

[0206] The evaluation results and the verification results are formatted into a machine-readable evaluation report.

[0207] In one embodiment, the data sharing module 16 is used for:

[0208] The assessment report will be sent to both the data provider and the data user.

[0209] Based on the assessment report, the data provider and the data user shall conduct at least one round of confirmation or modification feedback through the trusted negotiation engine;

[0210] When both the data provider and the data user confirm acceptance of the modified version of the negotiated plan based on the same data sharing strategy, it is determined that a consensus has been reached.

[0211] The trusted negotiation engine compiles the final data sharing strategy negotiation scheme that has reached a consensus into a verifiable strategy contract. The verifiable strategy contract is deployed on the blockchain in the form of a smart contract, and the contract address and content hash of the verifiable strategy contract are returned to the data provider and the data user.

[0212] Compared to existing technologies, the first step is to obtain the categories of strategic elements and non-negotiable constraints that the data provider and user must submit through the attribute registration module 11. This transforms vague and unstructured business intentions into structured and dimensional inputs that can be accurately processed by machines, and determines the hard boundaries of security and compliance in the negotiation, thus laying a reliable data foundation for automated negotiation.

[0213] Secondly, through the rule acquisition module 12, industry semantic mapping rules and industry compliance rules for data providers and data users are obtained according to industry attributes. At the same time, bridging mapping rules for policy semantic transformation are also obtained, enabling the system to adapt to any industry attribute. Subsequently, normative texts and historical contract data are collected from platforms such as standards and regulatory notices to ensure the compliance and authority of the rules. By constructing a knowledge graph, the relationships between the identified entities in the knowledge graph are established. Based on machine learning of historical contracts, cross-industry bridging rules are generated to reduce reliance on expert experience. Finally, consistency verification and classified storage ensure the internal consistency and application reliability of the rule base, making the knowledge base source on which the trusted negotiation engine relies reliable and able to self-optimize based on practical cases.

[0214] Secondly, through the information collection module 13, deployment monitoring is carried out to transform the data processing environment into a data processing environment security certificate, thereby realizing the dynamic verification of the data processing environment. Thirdly, scenario identifiers are extracted by parsing business request messages. Finally, external events are obtained by subscribing to regulatory interfaces, enabling the system to have the ability to proactively detect and quickly respond to changes in the compliance environment.

[0215] In addition, through the negotiation scheme generation module 14, standard strategy elements and standard constraints are generated by semantic alignment and standardization transformation according to the negotiation algorithm to eliminate semantic ambiguity. Then, an initial solution is generated based on historical successful cases to ensure the high quality of the search starting point and accelerate the convergence process. By designing a comprehensive quality scoring function, standard strategy elements, standard constraints and dynamic context are considered in a unified manner to ensure the comprehensive optimality of the scheme. Through iteration, directional exploration in the solution space is carried out while effective updates are performed to avoid getting trapped in local optima. This enables the system to select the optimal or suboptimal solution in the strategy combination and obtain candidate data sharing strategy negotiation schemes.

[0216] Meanwhile, through the results evaluation module 15, the expected effect is evaluated using a pre-trained value prediction model. Based on historical data, the business results of the predicted solution are quantified, and the merits and demerits of the solution are transformed into data comparisons to obtain evaluation results. Subsequently, the solution is compared with the control rules to form verification results, realize automated compliance review, and reduce identification risks. Finally, the evaluation and verification results are formatted into an evaluation report to ensure the efficiency and accuracy of information flow.

[0217] Finally, through the data sharing module 16, the evaluation reports were sent and structured feedback was collected, establishing an online negotiation channel and improving communication efficiency. Subsequently, through objective analysis, a confirmed data sharing strategy negotiation plan was obtained under the consensus of mutual acceptance. The data sharing strategy negotiation plan was then compiled into a blockchain smart contract and deployed on the blockchain to realize the execution transformation of the negotiation results, making the sharing strategy an immutable and automatically executed code. At the same time, the content hash was calculated to provide integrity proof for the contract address.

Claims

1. A cross-industry data sharing method supporting dynamic strategy negotiation, characterized in that, The method includes: The system guides data providers and data users to register their respective industry attributes with the trusted negotiation engine and submit data sharing strategy elements and constraints. The trusted negotiation engine loads the corresponding industry semantic mapping rules and compliance rules based on the industry attributes; During the data sharing process, dynamic context information is continuously collected, including data processing environment security certification, business scenario identification, and external regulatory events. The trusted negotiation engine, based on the data sharing strategy elements, constraints, dynamic context information, and the loaded industry semantic mapping rules and compliance rules, executes a negotiation algorithm to generate at least one data sharing strategy negotiation scheme. The expected effects and compliance verification of the data sharing strategy negotiation scheme were evaluated, and the evaluation results and verification results were obtained. Based on the evaluation and verification results, the data provider and the data user reach a consensus, generate a verifiable strategy contract, and execute data sharing operations according to the verifiable strategy contract. The trusted negotiation engine loads corresponding industry semantic mapping rules and compliance rules based on the industry attributes, including: Based on the data provider's primary industry attribute, load the corresponding primary industry semantic mapping rules and primary industry compliance rules from the pre-built rule base; Based on the second industry attribute of the data user, load the corresponding second industry semantic mapping rules and second industry compliance rules from the rule base; When the first industry attribute is different from the second industry attribute, a bridging mapping rule for policy semantic conversion between the first industry and the second industry is loaded from the rule base; The bridging mapping rule is used to implement at least one of the following transformations: The business compliance requirements defined in the first industry semantic mapping rules are converted into equivalent technical control parameters defined in the second industry semantic mapping rules; The data processing constraints defined in the second industry compliance rules are converted into equivalent constraint expressions that meet the requirements of the first industry compliance rules; Transform the domain business terms in the data provider's data sharing strategy elements into the corresponding cross-domain technical implementation terms in the data user's data sharing strategy elements; The trusted negotiation engine, based on the data sharing strategy elements, constraints, dynamic context information, and loaded industry semantic mapping rules and compliance rules, executes a negotiation algorithm to generate at least one data sharing strategy negotiation scheme, including: Input the data sharing strategy elements of the data provider, the data sharing strategy elements of the data user, the constraints of the data provider, and the constraints of the data user into the negotiation algorithm; The negotiation algorithm, based on the loaded first industry semantic mapping rule, second industry semantic mapping rule, and bridging mapping rule, performs semantic alignment and standardization transformation on all input strategy elements and constraints to generate standard strategy elements and standard constraints. The negotiation algorithm, based on the standard strategy elements and standard constraints, and combined with the currently collected dynamic context information, generates multiple candidate data sharing strategy negotiation schemes through optimized search. The negotiation algorithm, based on the standard strategy elements and standard constraints, and combined with the currently collected dynamic context information, generates multiple data sharing strategy negotiation schemes through optimized search, including: Based on data from historical successful negotiation cases, multiple initial strategy parameter sets are generated using control strategy parameter sets to form an initial search space; For each set of strategy parameters in the initial search space, a comprehensive quality score is calculated, wherein the comprehensive quality score is determined based on the degree to which the set of strategy parameters satisfies the standard strategy elements, the degree to which it conforms to the standard constraints, and the degree to which it matches the current dynamic context information. Based on the comprehensive quality score, the set of strategy parameters with high scores is selected as the high-quality set, and a new set of strategy parameters is generated and introduced based on the high-quality set according to the preset update ratio, forming an updated search space. The process of iteratively calculating the overall quality score, selecting the high-quality set, and updating the search space continues until a preset convergence condition is reached. The convergence condition includes reaching a preset maximum number of iterations or the change in the highest overall quality score in a consecutive preset number of iterations being lower than a preset threshold. When the convergence condition is met, the set of one or more policy parameters with the highest comprehensive quality score constitutes a candidate data sharing policy negotiation scheme.

2. The cross-industry data sharing method supporting dynamic strategy negotiation according to claim 1, characterized in that, The guidelines instruct data providers and data users to register their respective industry attributes with the trusted negotiation engine and submit data sharing strategy elements and constraints, including: The data provider registers the first industry attribute with the trusted negotiation engine and submits the data provider's data sharing strategy elements and data provider constraints. The data provider's data sharing strategy elements include at least one of the following: data usage purpose scope, data usage timeliness, and data usage geographical restrictions. The data user registers a second industry attribute with the trusted negotiation engine and submits the data user's data sharing strategy elements and data user constraints. The data user's data sharing strategy elements include at least one of the following: data processing technology solution, data processing environment certification, and data output format requirements. The constraints imposed on the data provider and the constraints imposed on the data user both include their respective non-negotiable bottom-line policy clauses.

3. The cross-industry data sharing method supporting dynamic strategy negotiation according to claim 1, characterized in that, The process of building a rule base includes: We collect normative texts and historical contract data from different industries through standards organization publication platforms, public announcements from industry regulatory agencies, and de-identified historical data sharing contracts. The collected normative texts and historical contract data are structured to construct an industry rule knowledge graph that includes industry semantic mapping rules and industry compliance rules. Based on the historical contract data, cross-industry bridging mapping rules are generated through machine learning training; Consistency and validity verifications are performed on the rules in the industry rule knowledge graph and the cross-industry bridging mapping rules, and the verified rules are classified and stored to form a rule base.

4. The cross-industry data sharing method supporting dynamic strategy negotiation according to claim 1, characterized in that, During the data sharing process, dynamic context information is continuously collected. This dynamic context information includes data processing environment security certification, business scenario identifiers, and external regulatory events, including: By deploying a monitoring agent in the data processing environment of the data user, security certificates of the data processing environment are continuously collected, wherein the security certificates of the data processing environment include Trusted Execution Environment (TEE) remote certification reports or secure container runtime credentials. By parsing the business request messages from the data user, the current business scenario identifier can be extracted; By subscribing to regulatory agencies' publishing interfaces, external regulatory events can be obtained in real time; The data processing environment security certificate, business scenario identifier, and external regulatory event are all accompanied by timestamp information.

5. The cross-industry data sharing method supporting dynamic strategy negotiation according to claim 1, characterized in that, The expected effects and compliance verification of the data sharing strategy negotiation scheme were conducted to obtain the evaluation results and verification results, including: The data sharing strategy negotiation scheme is evaluated using a pre-trained value prediction model. The expected data utility value that can be generated after the data sharing strategy negotiation scheme is executed is predicted, and the evaluation result is obtained. The set of control strategy parameters in the data sharing strategy negotiation scheme is compared one by one with the clauses in the loaded first industry compliance rules and second industry compliance rules to identify potential compliance conflict points, and compliance modification suggestions are generated based on bridging mapping rules to obtain verification results. The evaluation results and the verification results are formatted into a machine-readable evaluation report.

6. The cross-industry data sharing method supporting dynamic strategy negotiation according to claim 5, characterized in that, Based on the evaluation and verification results, the data provider and the data user reach a consensus to generate a verifiable strategy contract, including: The assessment report will be sent to both the data provider and the data user. Based on the assessment report, the data provider and the data user shall conduct at least one round of confirmation or modification feedback through the trusted negotiation engine; When both the data provider and the data user confirm acceptance of the modified version of the negotiated plan based on the same data sharing strategy, it is determined that a consensus has been reached. The trusted negotiation engine compiles the final data sharing strategy negotiation scheme that has reached a consensus into a verifiable strategy contract. The verifiable strategy contract is deployed on the blockchain in the form of a smart contract, and the contract address and content hash of the verifiable strategy contract are returned to the data provider and the data user.

7. A cross-industry data sharing system supporting dynamic strategy negotiation, characterized in that, For implementing the cross-industry data sharing method supporting dynamic strategy negotiation as described in any one of claims 1-6, the system comprises: The attribute registration module guides data providers and data users to register their respective industry attributes with the trusted negotiation engine and submit data sharing strategy elements and constraints. The rule acquisition module is used by the trusted negotiation engine to load the corresponding industry semantic mapping rules and compliance rules according to the industry attributes. The information collection module is used to continuously collect dynamic context information during the data sharing process. The dynamic context information includes data processing environment security certification, business scenario identification, and external regulatory events. The negotiation scheme generation module is used by the trusted negotiation engine to execute a negotiation algorithm based on the data sharing strategy elements, constraints, dynamic context information, and the loaded industry semantic mapping rules and compliance rules to generate at least one data sharing strategy negotiation scheme. The results evaluation module is used to evaluate the expected effects and verify the compliance of the data sharing strategy negotiation scheme, and obtain the evaluation results and verification results. The data sharing module is used to facilitate consensus between the data provider and the data user based on the evaluation and verification results, generate a verifiable strategy contract, and execute data sharing operations according to the verifiable strategy contract.

Images