A blockchain-based settlement service data management system and method

Abstract

The present application relates to the field of business data management, and particularly relates to a settlement business data management system and method based on a block chain, the method comprising constructing a consortium chain architecture with settlement business data participants, standardizing the collection of meta events at the settlement business data participants, converting explicit rules and implicit rules into dynamic rules based on neural symbol reasoning, and binding the dynamic rules to HTLC; integrating the corresponding associated paths of each settlement business data participant to construct an event association network. The present application can quickly identify unexpected emergent events generated by multi-agent interaction by integrating multi-agent associated paths to construct an event association network, real-time monitoring of the emergence threshold with the aid of encryption algorithms, shortening the discovery and processing cycle of abnormal events, and avoiding the repeated occurrence of similar abnormalities through rule iteration, thereby reducing the operational risk of settlement business.

Description

Technical Field

[0001] This invention relates to the field of business data management, and specifically to a blockchain-based settlement business data management system and method. Background Technology

[0002] In various settlement business scenarios such as healthcare and commodity trading, such as doctors, clinics, pharmacies, and patients in the medical field, and buyers, sellers, and logistics providers in commodity trading, there is often a multi-entity and decentralized nature. Data is stored in the local systems of each participant, which leads to many industry pain points in settlement business data management. Existing technologies are difficult to meet the needs of efficient, reliable, and compliant management.

[0003] Furthermore, the data is fragmented and lacks credibility. In traditional settlement processes, prescriptions issued by doctors, records of medical services provided by clinics, drug dispensing data from pharmacies, and patient payment information are stored separately in the independent systems of each participating party, lacking a unified data sharing and evidence preservation mechanism. Inconsistent data formats across systems easily lead to inconsistencies, duplicate entries, or missing information, resulting in inefficient settlement and reconciliation. Simultaneously, data storage relies on a single entity, posing a risk of tampering and forgery. In the event of a dispute, the lack of reliable original data as a basis for tracing makes it difficult to determine responsibility.

[0004] Settlement operations involve rules including industry policy requirements, compliance standards, and transaction agreement stipulations. These rules include both explicit rules and implicit patterns based on historical data. Existing technologies mostly adopt static rule configuration methods, executing rules through fixed code or manual review. This approach cannot quickly adapt to dynamic rule changes and struggles to transform implicit rules into executable computational logic. Consequently, rule execution lacks flexibility and accuracy, easily leading to compliance verification omissions or over-verification issues. Summary of the Invention

[0005] This invention addresses the technical problems existing in the prior art by providing a blockchain-based settlement business data management system and method.

[0006] The technical solution of this invention to solve the above-mentioned technical problems is as follows: A blockchain-based settlement business data management method, comprising the following steps: A consortium blockchain architecture is built with settlement business data participants. Meta events are collected in a standardized manner at the settlement business data participants and HTLC is extended to build an event-driven underlying architecture. The meta events of the smallest unit of settlement are stored on the blockchain in real time to provide a data foundation for the future. Based on neural symbolic reasoning, explicit and implicit rules are transformed into dynamic rules, and the dynamic rules are bound to HTLC; The corresponding association paths of various settlement business data participants are integrated and constructed into an event association network. Based on the encryption algorithm, the dynamic rule adjustment of HTLC is driven by the emergent events of the association paths. The settlement business data is verified in real time to complete the authenticity verification, compliance verification and anomaly handling, and an index chain is built to realize the traceability of settlement business data.

[0007] In a preferred embodiment, the construction of the consortium blockchain architecture based on settlement business data participants includes: Entity nodes with business generation capabilities; collect and submit business action events related to settlement business, such as transaction initiation, logistics status change or service completion, and deploy local data processing and consortium blockchain architecture transmission channels; The regulatory node is responsible for verifying rule compliance and allocating consensus weights; the regulatory node has fixed permissions and a fixed consensus weight at a preset ratio to ensure the priority of regulation. Auxiliary nodes provide historical compliance data and computing power support to assist in dynamic weight calculation; The standardized data collection meta-events at the settlement business data participants include: The entity nodes, regulatory nodes, and auxiliary nodes preprocess local events, breaking them down into meta-events, the smallest data units of the settlement business. The data structure of the meta-event includes meta-event ID, business target ID, event chain ID, parent event ID, event type based on the enumeration value of the business action, timestamp, and participant signature. The meta-event ID uses a combination of UUID and node blockchain address to ensure global uniqueness and traceability. The business target ID refers to the unique code associated with the specific settlement object, such as product code, service code, etc. The event chain ID refers to the parent identifier shared by meta-events in the same settlement process, used to associate all events in the same business link. The parent event ID refers to the unique identifier of the directly preceding event, used to build event dependency relationships and used in the subsequent event association network. The timestamp uses the UTC time standard, and the participant signature is a digital signature of the time initiator and the associated party, and is encrypted based on the SM2 national cryptographic algorithm; The physical nodes, monitoring nodes, and auxiliary nodes capture business actions in real time through IoT interfaces and business system APIs, and generate local events. In summary, the on-chain verification process for metadata entering a consortium blockchain architecture built by settlement business data participants includes: Entity nodes, regulatory nodes, and auxiliary nodes perform local operations such as disassembly, preprocessing, and data cleaning on the acquired local events to form a data format that conforms to metadata standards; The meta-event hash and associated field information are calculated using the SM3 algorithm. The settlement business data is sent to the participants in the consortium blockchain architecture. Nodes confirm consensus through a dynamic weight consensus algorithm. This algorithm, which determines the consensus confirmation of meta-events on-chain and the ranking of node weights, dynamically allocates / calculates consensus weights based on node type and business needs, while retaining fixed weights for key nodes to ensure regulatory priority. Furthermore, the dynamic weight consensus algorithm used in this application combines a basic weight with dynamic bonuses. For example, the basic weight of an entity node is preset to a fixed value, such as 30 points, and then dynamically bonused, such as 0.2 for business contribution + 0.3 for historical compliance + 0.2 for computing power support. The business contribution is quantified as the proportion of the number of valid meta-events submitted by the node in the past 30 days to the total number of meta-events in the consortium blockchain. The historical compliance score is based on the on-chain stored meta-event compliance verification records, and the computing power support score is based on the consensus computing resources provided by the node to the consortium blockchain. The weight of auxiliary nodes is based on the base weight combined with dynamic additions, while the weight of regulatory nodes is a preset ratio.

[0008] After consensus is reached, the meta-event hash and identifier field information are stored on the blockchain, the sensitive data in the attribute field are encrypted and stored off-chain, and the related fields are updated in real time to the on-chain event relationship graph.

[0009] In a preferred embodiment, after obtaining the meta-event, the method further includes: collecting the meta-events of multiple identical business actions on the same settlement link on the consortium blockchain architecture to form associated meta-events and their associated meta-event set IDs; The hash of the associated meta-events is calculated using the Merkle tree algorithm to obtain the hash of the meta-event set, ensuring the integrity and immutability of the meta-event set; The dynamic time window of HTLC is preset according to the business action type. The dynamic time window will change under different rules. See the binding of HLTC and dynamic rules later. The HTLC uses a combination of meta-event set hashing and dynamic time windows to achieve contract locking.

[0010] Traditional HTLCs combine a single hash value with a static timestamp. This application extends the technology by using existing smart contract pre-set logic to automatically execute payment unlocking and achieve business settlement after the hash of the on-chain meta-event set is matched with the calculation result of the entity node.

[0011] In a preferred embodiment, the transformation of explicit and implicit rules into dynamic rules based on neural symbolic reasoning includes: Symbolic logic layer: Based on first-order predicate logic, a rule base is built, and textual descriptions and quantifiable explicit rules are transformed into symbolic logic formulas through logic programming to ensure the interpretability and auditability of the rules. First-order predicate logic refers to transforming these unstructured rules into symbolic logic formulas that can be recognized and deduced by machines through logical elements such as predicates, individual words, and quantifiers. Neural network layer; based on OCR and CNN to process image data, and BERT to process text data, it transforms explicit rules into structured symbolic variables, such as text policies, image documents, and voice recordings; The historical settlement data containing element events, rule execution results, and anomaly records are classified and associated with patterns through a deep learning network, and implicit coefficients are output to adjust the symbolic logic formula. The main approach involves fusing three types of vectors—text data, image data, and historical settlement data—using a Transformer encoder to output fused features. The model uses an MLP to output latent coefficients, and the loss function is set to mean squared error. The resulting latent coefficients are indicators that quantify the influence of historical data on the current dynamic rules, with values ​​ranging from 0 to 1.

[0012] After obtaining the symbolic variables and implicit coefficients, the dynamic rules are bound to the extended HTLC, including the following specific steps: Neural symbolic reasoning provides a basic formula framework based on symbolic logic formulas. It takes symbolic variables and implicit coefficients as inputs and generates dynamic rule formulas and rule hashes through a weighted fusion algorithm. Dynamic rule formulas refer to mathematical expressions that can be directly used in calculations, and rule hashes refer to the values ​​used to verify the hashes of the meta-event set. If the rule hashes and the values ​​of the hashes of the meta-event set are inconsistent, the HTLC locking condition is met, and the business payment is returned. After obtaining the dynamic rule formula, rule hash, and associated meta-event set ID, neural symbolic reasoning sends the dynamic rule formula, rule hash, and associated meta-event set ID to the settlement business data participants. The settlement business data participants verify the rules through symbolic logic formulas, and the regulatory node responsible for rule compliance verification and consensus weight allocation exercises veto through a fixed preset ratio of consensus weight. After the settlement business data participants are verified, HTLC sets the locking conditions with rule hash and meta-event set hash, and derives the time window parameters based on the dynamic rule formula, and presets the dynamic time window change of HTLC according to the business action type.

[0013] Furthermore, in some other specific implementations, the locked HTLC parameters, namely the rule hash, time window, and participant address, need to be stored on the blockchain for evidence. Rule changes need to be confirmed through multi-node consensus. Only relevant entity nodes and necessary regulatory nodes can ensure the security management and real-time response of the consortium blockchain in settlement business data.

[0014] In a preferred embodiment, the step of integrating the corresponding association paths of each settlement business data participant to construct an event association network includes the following specific steps: For each meta-event, the interaction strength and event impact of the settlement business data participants are quantified through parameters. The parameter quantification includes correlation strength, event propagation coefficient, and emergence threshold. The correlation strength is quantified by the ratio of the historical cooperation frequency of two settlement business data participants to the total cooperation frequency. The correlation strength is calculated in real time using on-chain data of the consortium blockchain architecture. The event propagation coefficient is based on the historical correlation strength of two settlement business data participants and is generated through training a learning model. The emergence threshold is dynamically adjusted based on the event type and an event correlation network is constructed through HTLC preset initial value and consensus update.

[0015] The event propagation coefficient refers to the weight of the impact of a certain type of event on other events. The correlation strength refers to the business dependence of two settlement business data participants, which is called Agent in traditional emergent algorithms. In this application, the Agent is uniformly the settlement business data participant. The emergence threshold refers to the risk trigger threshold of different event types.

[0016] In a preferred embodiment, the emergence threshold is dynamically adjusted based on the event type, including: The on-chain data of the consortium blockchain architecture meets the following requirements: the event types in the meta-events decomposed from local events include clearly adjusted risk levels, and / or, the historical risk event occurrence rate of the meta-events of the same business action on the consortium blockchain architecture exceeds a preset threshold, such as a certain type of medical consumables being listed as high risk due to quality issues, and / or the failure rate of emerging events in low-risk scenarios in the past 3 months is >10%; Once the on-chain data in the consortium blockchain architecture meets any condition, the current emergence threshold, the risk level of the event type, the correlation strength, and the event propagation coefficient are used as inputs. The current emergence threshold and the risk level of the event type are used as feature variables through a regression tree model. The training output is the predicted threshold [T1,T2]. The product of the correlation strength and the event propagation coefficient with the predicted threshold boundary is used as the fusion value of the emergence threshold at this time, which is the final value.

[0017] In a preferred embodiment, the dynamic rule adjustment of HTLC driven by emergent events of associated paths based on cryptographic algorithms includes: Once the emergence threshold in the event association network detected by the encryption algorithm exceeds the preset threshold, the participant with the highest weight in the event association network, i.e. the participant with the highest real-time node weight based on the dynamic weight consensus algorithm, is selected as the triggering participant. The triggering participant calls a local event to verify whether the hash values ​​of all meta-events on the associated path are consistent with the on-chain evidence stored in the consortium blockchain architecture. If they are consistent, it is a false trigger. If they are inconsistent, the triggering participant broadcasts the associated meta-event set ID and associated path to all settlement business data participants in the event-associated network. Other settlement business data participants independently recalculated based on local event data stored locally. If more than half of the settlement business data participants found that the hash value of the involved meta-events was inconsistent with the on-chain evidence stored in the consortium blockchain architecture, then there were emergent events in the consortium blockchain architecture and its event association network. This indicates that the unexpected impact generated by the interaction of multiple decentralized meta-events is unforeseeable under normal prediction. The set of related meta-events that deconstructs the hash value and the on-chain evidence storage of the consortium blockchain architecture includes: ID, event type, business action, and HTLC. The meta-event set hash is combined with the contract ID locked by the dynamic time window as an event package. The encryption algorithm generates a rule adjustment vector with event type and business action as parameters, including: the corrected rule dimension, adjustment direction, and impact weight. The triggering participant uses the SM2 national cryptographic algorithm to jointly sign the rule adjustment vector and the event packet, sends it to the consortium blockchain architecture for notarization, and binds it with the HTLC contract ID to generate an adjustment request ID. The adjustment request ID is retained to ensure the traceability of adjustment requirements.

[0018] In a preferred embodiment, after acquiring the event package, neural symbolic reasoning adds adjustment factors to the symbolic logic formula through rule adjustment vectors, including: changing the symbolic definition of the symbolic logic formula with the modified rule dimension, changing the positive and negative of the symbolic logic formula with the adjustment direction, changing the weight allocation of the symbolic logic formula with the influence weight, and changing the symbolic logic formula of neural symbolic reasoning to the modified symbolic formula framework. Therefore, neural symbolic reasoning provides a basic formula framework with new symbolic logic formulas. It takes symbolic variables and implicit coefficients as inputs, generates new dynamic rule formulas and new rule hashes through a weighted fusion algorithm, and then rebinds HTLC through the above steps.

[0019] In a preferred embodiment, the triggering participant broadcasts the modified dynamic rule formula and rule hash of the neural symbol reasoning to all settlement business data participants associated with the HTLC contract, and the other settlement business data participants execute it.

[0020] The present invention also provides a blockchain-based settlement business data management system, the system comprising: The consortium blockchain architecture building and meta-event collection module is used to construct a consortium blockchain architecture with entity nodes, regulatory nodes, and auxiliary nodes based on settlement business data participants. It deploys local data processing and transmission channels, captures business actions in real time through IoT interfaces and business system APIs to generate local events, preprocesses local events into standardized meta-events, calculates meta-event hashes and associated domain information through the SM3 algorithm, and achieves consensus through a dynamic weight consensus algorithm. It then realizes on-chain storage of meta-event hashes and identifier domain information and updates associated domains, and extends HTLC. The Neural Symbolic Reasoning and Dynamic Rule Generation Module is used to construct a rule base based on first-order predicate logic and transform symbolic logic formulas. It generates symbolic variables by processing image and text data through OCR, CNN and BERT, analyzes historical settlement data using deep learning networks to output implicit coefficients, and generates dynamic rule formulas and rule hashes through a weighted fusion algorithm using symbolic logic formulas as a framework. It sends relevant information to settlement business data participants and regulatory nodes to complete verification, and then binds the dynamic rules to the extended HTLC and derives time window parameters. The event association network construction and dynamic rule adjustment module is used to integrate the association paths of various settlement business data participants. It quantifies the interaction strength and event impact through association strength, event propagation coefficient, and emergence threshold, constructs an event association network, dynamically adjusts the emergence threshold based on a regression tree model, detects the emergence threshold exceeding the limit through an encryption algorithm, identifies the triggering participants and verifies the consistency of the meta-event hash, generates event packages and rule adjustment vectors, links with neural symbolic reasoning to update symbolic logic formulas and dynamic rules, and broadcasts the new rules to relevant participants. The real-time verification and full-chain traceability module for settlement data is used to perform authenticity verification, compliance verification, and anomaly handling on settlement business data. It marks the abnormal status of data that fails verification and generates abnormal meta-events for on-chain storage. It constructs an index chain to associate meta-event IDs, business target IDs, event chain IDs, and on-chain storage addresses to achieve full-process traceability of settlement business data.

[0021] The beneficial effects of this invention are: it integrates multi-subject association paths to construct an event association network, uses encryption algorithms to monitor the emergence threshold in real time, and can quickly identify unexpected emerging events generated by multi-subject interactions. When an anomaly is detected, it uses a dynamic weight consensus algorithm to determine the triggering participant, and uses neural symbol reasoning to generate a rule adjustment vector, thereby achieving rapid updating and synchronization of dynamic rules, shortening the discovery and processing cycle of abnormal events. It can also work with existing encryption algorithms to avoid the recurrence of similar anomalies through rule iteration, thereby reducing the operational risks of settlement business. Extended HTLC uses meta-event set hashes and dynamic time windows as locking conditions, combined with smart contracts to achieve automatic unlocking of payments. This eliminates the need for third-party intermediaries to intervene in clearing and reconciliation, simplifying the settlement process and reducing payment delays or unlocking disputes. The division of labor among entity nodes, regulatory nodes, and auxiliary nodes in the consortium blockchain architecture, coupled with a dynamic weight consensus algorithm, ensures the priority of regulatory nodes while fully mobilizing the business contributions and computing power support of each participant. This forms a clear and efficient ecosystem with well-defined responsibilities, reducing communication and trust costs in multi-party collaboration. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0025] In the description of this application, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0026] As attached Figure 1 As shown, this embodiment provides a blockchain-based settlement business data management method, including the following steps: S1. Consortium blockchain architecture is built with settlement business data participants. Meta events are collected in a standardized manner at the settlement business data participants and HTLC is extended to build an event-driven underlying architecture. The meta events of the smallest unit of settlement are stored on the blockchain in real time to provide a data foundation for the future. A consortium blockchain architecture is built around the participants in the settlement business data, including: Entity nodes with business generation capabilities; collect and submit business action events related to settlement business, such as transaction initiation, logistics status change or service completion, and deploy local data processing and consortium blockchain architecture transmission channels; The regulatory node is responsible for verifying rule compliance and allocating consensus weights; the regulatory node has fixed permissions and a fixed consensus weight at a preset ratio to ensure the priority of regulation. Auxiliary nodes provide historical compliance data and computing power support to assist in dynamic weight calculation; Standardized data collection meta-events at the settlement business data participants include: Entity nodes, regulatory nodes, and auxiliary nodes preprocess local events, breaking them down into meta-events, the smallest data units of settlement business. The data structure of a meta-event includes meta-event ID, business target ID, event chain ID, parent event ID, event type based on enumeration values ​​of business actions, timestamp, and participant signature. The meta-event ID uses a combination of UUID and node blockchain address to ensure global uniqueness and traceability. The business target ID refers to the unique code associated with the specific settlement object, such as product code, service code, etc. The event chain ID refers to the parent identifier shared by meta-events in the same settlement process, used to associate all events in the same business link. The parent event ID refers to the unique identifier of the directly preceding event, used to build event dependency relationships and used in the subsequent event association network. The timestamp uses the UTC time standard, and the participant signature is a digital signature of the time initiator and the associated party, and is encrypted based on the SM2 national cryptographic algorithm; Entity nodes, monitoring nodes, and auxiliary nodes capture business actions in real time through IoT interfaces and business system APIs, and generate local events. In summary, the on-chain verification process for metadata entering a consortium blockchain architecture built by settlement business data participants includes: Entity nodes, regulatory nodes, and auxiliary nodes perform local operations such as disassembly, preprocessing, and data cleaning on the acquired local events to form a data format that conforms to metadata standards; The meta-event hash and associated field information are calculated using the SM3 algorithm. The settlement business data is sent to the participants in the consortium blockchain architecture. Nodes confirm consensus through a dynamic weight consensus algorithm. This algorithm, which determines the consensus confirmation of meta-events on-chain and the ranking of node weights, dynamically allocates / calculates consensus weights based on node type and business needs, while retaining fixed weights for key nodes to ensure regulatory priority. Furthermore, the dynamic weight consensus algorithm used in this application combines a basic weight with dynamic bonuses. For example, the basic weight of an entity node is preset to a fixed value, such as 30 points, and combined with dynamic bonuses, such as 0.2 for business contribution + 0.3 for historical compliance + 0.2 for computing power support. The business contribution is quantified as the proportion of the number of valid meta-events submitted by the node in the past 30 days to the total number of meta-events in the consortium blockchain. The historical compliance score is based on the on-chain stored meta-event compliance verification records, and the computing power support score is based on the consensus computing resources provided by the node to the consortium blockchain. The weight of auxiliary nodes is based on the base weight combined with dynamic additions, while the weight of regulatory nodes is a preset ratio.

[0027] After consensus is reached, the meta-event hash and identifier field information are stored on the blockchain, the sensitive data in the attribute field are encrypted and stored off-chain, and the related fields are updated in real time to the on-chain event relationship graph.

[0028] After obtaining the meta-event, the method also includes: collecting the meta-events of multiple identical business actions on the same settlement link on the consortium blockchain architecture to form associated meta-events and their associated meta-event set IDs; The hash of the associated meta-events is calculated using the Merkle tree algorithm to obtain the hash of the meta-event set, ensuring the integrity and immutability of the meta-event set; The dynamic time window of HTLC is preset according to the business action type. The dynamic time window will change under different rules. See the binding of HLTC and dynamic rules later. HTLC uses a combination of meta-event set hashing and dynamic time windows to achieve contract locking.

[0029] Traditional HTLCs combine a single hash value with a static timestamp. This application extends the technology by using existing smart contract pre-set logic to automatically execute payment unlocking and achieve business settlement after the hash of the on-chain meta-event set is matched with the calculation result of the entity node.

[0030] S2. Based on neural symbolic reasoning, explicit and implicit rules are transformed into dynamic rules, and the dynamic rules are bound to HTLC; Neural symbolic reasoning transforms explicit and implicit rules into dynamic rules, including: Symbolic logic layer: Based on first-order predicate logic, a rule base is built, and textual descriptions and quantifiable explicit rules are transformed into symbolic logic formulas through logic programming to ensure the interpretability and auditability of the rules. First-order predicate logic refers to transforming these unstructured rules into symbolic logic formulas that can be recognized and deduced by machines through logical elements such as predicates, individual words, and quantifiers. Neural network layer; based on OCR and CNN to process image data, and BERT to process text data, it transforms explicit rules into structured symbolic variables, such as text policies, image documents, and voice recordings; The historical settlement data containing element events, rule execution results, and anomaly records are classified and associated with patterns through a deep learning network, and implicit coefficients are output to adjust the symbolic logic formula. The main approach involves fusing three types of vectors—text data, image data, and historical settlement data—using a Transformer encoder to output fused features. The model uses an MLP to output latent coefficients, and the loss function is set to mean squared error. The resulting latent coefficients are indicators that quantify the influence of historical data on the current dynamic rules, with values ​​ranging from 0 to 1.

[0031] After obtaining the symbolic variables and implicit coefficients, the dynamic rules are bound to the extended HTLC, including the following specific steps: Neural symbolic reasoning provides a basic formula framework based on symbolic logic formulas. It takes symbolic variables and implicit coefficients as inputs and generates dynamic rule formulas and rule hashes through a weighted fusion algorithm. Dynamic rule formulas refer to mathematical expressions that can be directly used in calculations, and rule hashes refer to the values ​​used to verify the hashes of the meta-event set. If the rule hashes and the values ​​of the hashes of the meta-event set are inconsistent, the HTLC locking condition is met, and the business payment is returned. After obtaining the dynamic rule formula, rule hash, and associated meta-event set ID, neural symbolic reasoning sends the dynamic rule formula, rule hash, and associated meta-event set ID to the settlement business data participants. The settlement business data participants verify the rules through the symbolic logic formula, and the regulatory node responsible for rule compliance verification and consensus weight allocation exercises veto through a fixed preset ratio of consensus weight. After the settlement business data participants are verified, HTLC sets the locking conditions with rule hash and meta-event set hash, and derives the time window parameters based on the dynamic rule formula, and presets the dynamic time window change of HTLC according to the business action type.

[0032] Furthermore, in some other specific implementations, the locked HTLC parameters, namely the rule hash, time window, and participant address, need to be stored on the blockchain for evidence. Rule changes need to be confirmed through multi-node consensus. Only relevant entity nodes and necessary regulatory nodes can ensure the security management and real-time response of the consortium blockchain in settlement business data.

[0033] S3. Integrate the corresponding association paths of various settlement business data participants to construct an event association network, and drive the dynamic rule adjustment of HTLC based on the emergent events of association paths using encryption algorithms. The data paths of all settlement business participants are integrated and constructed into an event association network, which includes the following specific steps: For each meta-event, the interaction strength and event impact of the settlement business data participants are quantified through parameters, including correlation strength, event propagation coefficient, and emergence threshold. The correlation strength is quantified by the ratio of the historical cooperation frequency of two settlement business data participants to the total cooperation frequency. The correlation strength is calculated in real time using on-chain data of the consortium blockchain architecture. The event propagation coefficient is based on the historical correlation strength of two settlement business data participants and is generated through training a learning model. The emergence threshold is dynamically adjusted based on the event type. An event correlation network is constructed by preset initial values ​​and consensus updates through HTLC.

[0034] The event propagation coefficient refers to the weight of the impact of a certain type of event on other events. The correlation strength refers to the business dependence of two settlement business data participants, which is called Agent in traditional emergent algorithms. In this application, the Agent is uniformly the settlement business data participant. The emergence threshold refers to the risk trigger threshold of different event types.

[0035] Emergence thresholds are dynamically adjusted based on event type, including: The on-chain data of the consortium blockchain architecture meets the following requirements: the event types in the meta-events decomposed from local events include clearly adjusted risk levels, and / or, the historical risk event occurrence rate of the meta-events of the same business action on the consortium blockchain architecture exceeds a preset threshold, such as a certain type of medical consumables being listed as high risk due to quality issues, and / or the failure rate of emerging events in low-risk scenarios in the past 3 months is >10%; Once the on-chain data in the consortium blockchain architecture meets any condition, the current emergence threshold, the risk level of the event type, the correlation strength, and the event propagation coefficient are used as inputs. The current emergence threshold and the risk level of the event type are used as feature variables through a regression tree model. The training output is the predicted threshold [T1,T2]. The product of the correlation strength and the event propagation coefficient with the predicted threshold boundary is used as the fusion value of the emergence threshold at this time, which is the final value.

[0036] Dynamic rule adjustment of HTLC based on emergent events of associated paths using cryptographic algorithms includes: Once the emergence threshold in the event association network detected by the encryption algorithm exceeds the preset threshold, the participant with the highest weight in the event association network, i.e. the participant with the highest real-time node weight based on the dynamic weight consensus algorithm, is selected as the triggering participant. The participant is triggered to call a local event to verify whether the hash values ​​of all meta-events on the associated path are consistent with the on-chain evidence stored in the consortium blockchain architecture. If they are consistent, it is a false trigger. If they are inconsistent, the participant is triggered to broadcast the associated meta-event set ID and associated path to all settlement business data participants in the event-related network. Other settlement business data participants independently recalculated based on local event data stored locally. If more than half of the settlement business data participants found that the hash value of the involved meta-events was inconsistent with the on-chain evidence stored in the consortium blockchain architecture, then there were emergent events in the consortium blockchain architecture and its event association network. This indicates that the unexpected impact generated by the interaction of multiple decentralized meta-events is unforeseeable under normal prediction. The set of related meta-events that deconstructs the hash value and the on-chain evidence storage of the consortium blockchain architecture includes: ID, event type, business action, and HTLC. The meta-event set hash is combined with the contract ID locked by the dynamic time window as an event package. The encryption algorithm generates a rule adjustment vector with event type and business action as parameters, including: the corrected rule dimension, adjustment direction, and impact weight. Examples are as follows: Event types include cost fluctuations, time delays, and compliance risks; business actions include cost increases, extended delays, and increased compliance deviations. Therefore, for the above situations, the rule adjustment vector includes the corrected rule dimensions, adjustment directions, and impact weights, which respectively refer to the unit price calculation logic, payment timeliness, compliance coefficient, or increases, decreases, and new constraints, or the priority of adjustments in each dimension, based on the severity of the event.

[0037] The triggering party uses the SM2 national cryptographic algorithm to jointly sign the rule adjustment vector and event packet, sends it to the consortium blockchain architecture for notarization, and binds it with the HTLC contract ID to generate an adjustment request ID. The adjustment request ID is retained to ensure the traceability of adjustment requirements.

[0038] After acquiring the event package, neural symbolic reasoning adds adjustment factors to the symbolic logic formula through rule adjustment vectors, including: changing the symbolic definition of the symbolic logic formula with the modified rule dimension, changing the positive and negative of the symbolic logic formula with the adjustment direction, changing the weight allocation of the symbolic logic formula with the influence weight, and changing the symbolic logic formula of neural symbolic reasoning to the modified symbolic formula framework. Therefore, neural symbolic reasoning provides a basic formula framework with new symbolic logic formulas. It takes symbolic variables and implicit coefficients as inputs, generates new dynamic rule formulas and new rule hashes through a weighted fusion algorithm, and then rebinds HTLC through the above steps.

[0039] The triggering party broadcasts the modified dynamic rule formula and rule hash of the neural symbol reasoning to all settlement business data participants associated with the HTLC contract, and other settlement business data participants execute it.

[0040] S4. Real-time verification of settlement business data is performed to complete authenticity verification, compliance verification, and anomaly handling. An index chain is then built to enable traceability of settlement business data.

[0041] This invention also provides a blockchain-based settlement business data management system, the system comprising: The consortium blockchain architecture building and meta-event collection module is used to construct a consortium blockchain architecture with entity nodes, regulatory nodes, and auxiliary nodes based on settlement business data participants. It deploys local data processing and transmission channels, captures business actions in real time through IoT interfaces and business system APIs to generate local events, preprocesses local events into standardized meta-events, calculates meta-event hashes and associated domain information through the SM3 algorithm, and achieves consensus through a dynamic weight consensus algorithm. It then realizes on-chain storage of meta-event hashes and identifier domain information and updates associated domains, and extends HTLC. The Neural Symbolic Reasoning and Dynamic Rule Generation Module is used to construct a rule base based on first-order predicate logic and transform symbolic logic formulas. It generates symbolic variables by processing image and text data through OCR, CNN and BERT, analyzes historical settlement data using deep learning networks to output implicit coefficients, and generates dynamic rule formulas and rule hashes through a weighted fusion algorithm using symbolic logic formulas as a framework. It sends relevant information to settlement business data participants and regulatory nodes to complete verification, and then binds the dynamic rules to the extended HTLC and derives time window parameters. The event association network construction and dynamic rule adjustment module is used to integrate the association paths of various settlement business data participants. It quantifies the interaction strength and event impact through association strength, event propagation coefficient, and emergence threshold, constructs an event association network, dynamically adjusts the emergence threshold based on a regression tree model, detects the emergence threshold exceeding the limit through an encryption algorithm, identifies the triggering participants and verifies the consistency of the meta-event hash, generates event packages and rule adjustment vectors, links with neural symbolic reasoning to update symbolic logic formulas and dynamic rules, and broadcasts the new rules to relevant participants. The real-time verification and full-chain traceability module for settlement data is used to perform authenticity verification, compliance verification, and anomaly handling on settlement business data. It marks the abnormal status of data that fails verification and generates abnormal meta-events for on-chain storage. It constructs an index chain to associate meta-event IDs, business target IDs, event chain IDs, and on-chain storage addresses to achieve full-process traceability of settlement business data.

[0042] This application provides specific examples as follows: The core nodes consist of doctors, pharmacies, clinics, and patients as physical nodes, along with regulatory nodes responsible for rule compliance verification and consensus weighting, such as medical insurance regulatory agencies, and auxiliary nodes providing historical compliance data and computing power support, such as regional medical data centers, to jointly construct the settlement business alliance chain architecture.

[0043] 1. All four entities deploy local data processing modules on their physical nodes, such as prescription processing and drug purchase record storage modules, and use a transmission channel with the consortium blockchain to ensure that business action events can be uploaded to the consortium blockchain in real time.

[0044] Based on their respective settlement business scenarios, the four entities capture business actions in real time and generate local events, which are then broken down into the smallest settlement data units according to the weighted event standard, as follows: The doctor's entity node includes business action events such as issuing electronic prescriptions, which include drug name, dosage, and patient ID. The core fields of the decomposed meta-event include: meta-event ID, business target ID, event chain ID, event type, timestamp, and doctor's SM2 signature. The entity node of the clinic includes business action events such as completing medical services, containing meta event ID, business target ID, event chain ID, event type, timestamp, and clinic SM2 signature.

[0045] The operations for the remaining two entity nodes are similar.

[0046] 2. The physical nodes of the four entities perform a preprocessing-hash calculation-consensus-on-chain process for meta-events, including: Local preprocessing; data cleaning of metadata events to ensure compliance with metadata standards; Hash and Encryption; Meta-event hashes and associated field information are calculated using the SM3 algorithm, and sensitive data is encrypted and stored off-chain; Consensus confirmation: The meta-event hash and identifier field information are sent to the consortium blockchain, and the entity nodes complete the consensus through the dynamic weight consensus algorithm. The four events belonging to the same event chain ID, i.e., the meta-events of a single patient's medical treatment settlement, such as doctor's prescription, clinic treatment, pharmacy dispensing, and patient payment, are set together to generate a set of related meta-events and a unique set ID. The hash of the associated meta-event set is calculated using the Merkle tree algorithm; 3. Based on the type of business action, such as drug settlement, a dynamic time window for HTLC is preset, and the dynamic time window is combined with the hash of the meta-event set as the HTLC locking condition.

[0047] Then, the explicit and implicit rules of medical settlement are transformed into dynamic rules that can be bound to HTLCs, specifically as follows: The explicit rules of medical settlement, such as prescriptions requiring a doctor's signature and matching the drug code issued by the pharmacy, and treatment costs matching the medical insurance catalog, are transformed into symbolic logic formulas through first-order predicate logic. OCR is used to process doctor's prescription images to extract signatures and drug codes. BERT is used to process clinic treatment text, transforming explicit rules into structured symbolic variables, such as prescription signature validity = 1 and drug code = A001. At the same time, a deep learning network is used to analyze historical settlement data, namely the historical meta-events, rule execution results, and abnormal records of the four elements, and output implicit coefficients, such as patient historical compliant payment rate = 0.95 and pharmacy drug matching accuracy rate = 0.98. Using symbolic logic formulas as a framework, symbolic variables and implicit coefficients are input, and dynamic rule formulas are generated through a weighted fusion algorithm. 4. Neural symbolic reasoning sends the dynamic rule formula, rule hash, and associated meta-event set ID to the four entity nodes and the supervisory node. The four entities verify the rationality of the rule through the symbolic logic formula, and the supervisory node exercises the veto right according to the fixed consensus weight. After successful verification, HTLC sets the rule hash combined with the meta-event set hash as the unlocking condition, and derives the dynamic time window parameters based on the dynamic rule formula to complete the binding of the dynamic rule with HTLC.

[0048] 5. Integrate the settlement paths of doctors, pharmacies, clinics, and patients, such as the settlement chain between patients, doctors, clinics, and pharmacies. Quantify the interaction relationships among the four through parameters to construct an event association network: Quantify the business dependence of two entity nodes, such as the correlation strength between patients and pharmacies, which is the frequency of patients purchasing medicines at that pharmacy in the past 3 months / the total frequency of patients purchasing medicines. Generated based on historical association strength through a learning model training process; The threshold is dynamically adjusted based on the event type. For example, the initial threshold for drug quality abnormality events is set to 0.7. If the failure rate of handling such events is greater than 10% in the past 3 months, the threshold is adjusted to 0.5 through a regression tree model.

[0049] 6. When the encryption algorithm detects that the emergence threshold of the event association network exceeds a preset value, such as when the hash of the patient payment meta-event is inconsistent with the on-chain evidence, and the associated pharmacy outbound meta-event has an encoding error, the HTLC rules are dynamically adjusted, including: The real-time node weights are calculated using a dynamic weight consensus algorithm, and the entity node with the highest weight is set as the triggering participant. The triggering participant verifies whether the hash values ​​of all meta-events on the associated path are consistent with the on-chain evidence. If they are inconsistent, the associated meta-event set ID is broadcast to the four entities and other nodes in combination with the associated path. The four entity nodes independently recalculate based on local data. If more than half of the nodes confirm that the meta-event hashes are inconsistent with the on-chain evidence, it is determined that there is an emerging event, such as a mismatch between prescription and drug codes leading to settlement anomalies. The encryption algorithm generates a rule adjustment vector using event type and business action as parameters; Neural symbolic reasoning is based on rule adjustment vectors. Adjustment factors are added to the symbolic logic formulas, such as incorporating drug code verification into compliance score calculation. New dynamic rule formulas and rule hashes are generated through weighted fusion algorithms, re-bound to HTLC, and broadcast by the triggering party to the four parties, which then execute the new rules.

[0050] 7. The settlement data of the four parties must pass three levels of verification, including: Verify the authenticity of the data source; verify whether the SM2 signature of the meta-event is a valid signature of the four parties. For example, the pharmacy outbound meta-event needs to verify the pharmacy's signature to ensure the authenticity of the data source. Compliance verification; check whether the settlement data conforms to the dynamic rule formula. If it does not conform, such as a prescription without a doctor's signature, trigger an exception handling. Exception handling: Mark the data that fails to be verified as abnormal and store it in the off-chain exception database. At the same time, generate an exception meta-event and store it on the chain.

[0051] An index chain is constructed to associate the meta-event ID, business target ID, and event chain ID of the four entities with the on-chain evidence storage address, enabling full-chain traceability of settlement data: for example, when querying a patient's settlement data, the entire process data can be quickly located through the patient's meta-event ID, event chain ID, and associated doctor's prescription meta-event ID / clinic treatment meta-event ID / pharmacy outbound meta-event ID.

[0052] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0053] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0054] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0055] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0056] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0057] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0058] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A blockchain-based method for managing settlement business data, characterized in that, Includes the following steps: A consortium blockchain architecture is built around the settlement business data participants, and meta-events are collected in a standardized manner at the settlement business data participants, and HTLC is extended; Based on neural symbolic reasoning, explicit and implicit rules are transformed into dynamic rules, and the dynamic rules are bound to HTLC; The system integrates the corresponding association paths of various settlement business data participants to construct an event association network. Based on an encryption algorithm, it drives the dynamic rule adjustment of HTLC using emergent events of the association paths. This includes: when the encryption algorithm detects that the emergence threshold in the event association network exceeds a preset threshold (the emergence threshold refers to the risk trigger threshold for different event types), the settlement business data participant with the highest weight in the event association network is designated as the triggering participant; the triggering participant invokes local events to verify whether the hash values ​​of all meta-events on the association path are consistent with the on-chain evidence stored in the consortium blockchain architecture. If they are consistent, it is a false trigger; if they are inconsistent, the triggering participant broadcasts the associated meta-event set ID and association path to all settlement business data participants in the event association network; other settlement business data participants use their locally stored local events... If, during independent data recalculation, more than half of the settlement business data participants have inconsistent hash values ​​of the involved meta-events with the on-chain notarization of the consortium blockchain architecture, then there are emerging events in the consortium blockchain architecture and its event-related network. The associated meta-event set with inconsistent hash values ​​and on-chain notarization of the consortium blockchain architecture is broken down into: ID, event type, business action, and HTLC contract ID locked by combining the hash of the meta-event set with a dynamic time window, forming an event packet. An encryption algorithm generates a rule adjustment vector with event type and business action as parameters, including: the corrected rule dimension, adjustment direction, and impact weight. The triggering participant jointly signs the rule adjustment vector and event packet using the SM2 national cryptographic algorithm and sends it to the consortium blockchain architecture for notarization, binding it with the HTLC contract ID to generate an adjustment request ID. After acquiring the event package, neural symbolic reasoning adds adjustment factors to the symbolic logic formula through rule adjustment vectors, including: changing the symbolic definition of the symbolic logic formula with the modified rule dimension, changing the positive and negative of the symbolic logic formula with the adjustment direction, changing the weight allocation of the symbolic logic formula with the influence weight, and changing the symbolic logic formula of neural symbolic reasoning to the modified symbolic formula framework. The settlement business data is verified in real time to complete the authenticity verification, compliance verification and anomaly handling, and an index chain is built to realize the traceability of settlement business data.

2. The blockchain-based settlement business data management method according to claim 1, characterized in that, The aforementioned consortium blockchain architecture, constructed by participants in settlement business data, includes: Entity nodes with business generation capabilities; collect and submit business action events related to settlement business, and deploy local data processing and consortium blockchain architecture transmission channels; The regulatory node is responsible for verifying rule compliance and allocating consensus weights; the regulatory node has fixed permissions and the consensus weight is fixed at a preset ratio. Auxiliary nodes that provide historical compliance data and computing power support; The standardized collection of meta-events at the settlement business data participants, the preprocessing of local events by the entity nodes, regulatory nodes and auxiliary nodes, and the decomposition of local events into meta-events, the smallest data unit of the settlement business. The data structure of the meta-event includes meta-event ID, business target ID, event chain ID, parent event ID, event type based on the enumeration value of business action, timestamp and participant signature. The entity nodes, monitoring nodes, and auxiliary nodes capture business actions in real time through IoT interfaces and business system APIs, and generate local events.

3. The blockchain-based settlement business data management method according to claim 2, characterized in that, After obtaining the meta-event, the method further includes: collecting the meta-events of multiple identical business actions on the same settlement link on the consortium blockchain architecture to form associated meta-events and their associated meta-event set IDs; The hash of the associated meta-events is calculated using the Merkle tree algorithm to obtain the hash of the meta-event set; Preset the dynamic time window of HTLC based on the type of business action; HTLC uses a combination of meta-event set hashing and dynamic time windows to achieve contract locking.

4. The blockchain-based settlement business data management method according to claim 3, characterized in that, The transformation of explicit and implicit rules into dynamic rules based on neural symbolic reasoning includes: A rule base is built based on first-order predicate logic, and textual descriptions and quantifiable explicit rules are transformed into symbolic logic formulas through logic programming. Image data is processed using OCR and CNN, and text data is processed using BERT, transforming explicit rules into structured symbolic variables; The historical settlement data containing element events, rule execution results, and anomaly records are classified and associated with patterns through a deep learning network, and implicit coefficients are output to adjust the symbolic logic formula. After obtaining the symbolic variables and latent coefficients, neural symbolic reasoning provides a basic formula framework with symbolic logic formulas. With symbolic variables and latent coefficients as input, it generates dynamic rule formulas and rule hashes through a weighted fusion algorithm. After obtaining the dynamic rule formula, rule hash, and associated meta-event set ID, neural symbolic reasoning sends the dynamic rule formula, rule hash, and associated meta-event set ID to the settlement business data participants. The settlement business data participants verify the rules through symbolic logic formulas, and the regulatory node responsible for rule compliance verification and consensus weight allocation exercises veto through a fixed preset ratio of consensus weight. After the settlement business data participants are verified, HTLC sets the locking conditions with rule hash and meta-event set hash, and derives the time window parameters based on the dynamic rule formula. It then presets the dynamic time window change of HTLC according to the business action type, thus completing the binding of dynamic rules to HTLC.

5. The blockchain-based settlement business data management method according to claim 1, characterized in that, The process of integrating the corresponding association paths of each settlement business data participant to construct an event association network includes the following specific steps: For each meta-event, the interaction strength and event impact of the settlement business data participants are quantified by parameters. The parameter quantification includes association strength, event propagation coefficient, and emergence threshold. The association strength is quantified by the ratio of the historical cooperation frequency of two settlement business data participants to the total cooperation frequency. The event propagation coefficient is based on the historical association strength of two settlement business data participants and is generated by training a learning model. The emergence threshold is dynamically adjusted based on the event type and an event association network is constructed by preset initial value and consensus update through HTLC.

6. The blockchain-based settlement business data management method according to claim 5, characterized in that, The emergence threshold is dynamically adjusted based on event type, including: The on-chain data of the consortium blockchain architecture meets the following requirements: the event types in the meta-events decomposed from local events include clearly adjusted risk levels, and / or the historical risk event occurrence rate of the meta-events of the same business action on the consortium blockchain architecture exceeds a preset threshold. Once the on-chain data in the consortium blockchain architecture meets any condition, the current emergence threshold, the risk level of the event type, the correlation strength, and the event propagation coefficient are used as inputs. The current emergence threshold and the risk level of the event type are used as feature variables through a regression tree model. The training output is the prediction threshold, and the value obtained by multiplying the correlation strength and the event propagation coefficient by the boundary of the prediction threshold is used as the current emergence threshold value.

7. The blockchain-based settlement business data management method according to claim 6, characterized in that, The triggering participant broadcasts the modified dynamic rule formula and rule hash of the neural symbol reasoning to all settlement business data participants associated with the HTLC contract, and other settlement business data participants execute it.

8. A blockchain-based settlement business data management system, applied to the blockchain-based settlement business data management method described in any one of claims 1-7, characterized in that, The system includes: Consortium blockchain architecture setup and meta-event collection module; used to build consortium blockchain architecture, deploy local data processing and transmission channels, preprocess local events into standardized meta-events, and extend HTLC; The Neural Symbolic Reasoning and Dynamic Rule Generation Module is used to construct symbolic logic formulas, analyze historical settlement data using deep learning networks to output implicit coefficients, generate dynamic rule formulas and rule hashes through a weighted fusion algorithm using symbolic logic formulas as a framework, bind dynamic rules to the extended HTLC and derive time window parameters. The Event Association Network Construction and Dynamic Rule Adjustment Module is used to integrate the association paths of various settlement business data participants, construct the event association network, generate event packages and rule adjustment vectors, and link neural symbolic reasoning to update symbolic logic formulas and dynamic rules. The real-time verification and full-chain traceability module for settlement data is used to perform authenticity verification, compliance verification, and anomaly handling on settlement business data, build an index chain, and realize full-process traceability of settlement business data.

Images