A blockchain+big data decision method and system for a digital economic ecosystem and a cloud service center

Abstract

The invention of a blockchain + big data decision-making method and system for a digital economic ecosystem and a cloud service center is a trusted decision-making system and method based on the deep integration of blockchain and big data. The invention involves a complete business model and computer technology and artificial intelligence field. The method comprises: obtaining raw data from multiple source heterogeneous data sources, processing and generating standardized data packages, storing the hash value and evidence information to the blockchain to form data evidence credentials; based on the credentials, cross-domain association and feature fusion of multi-source data are performed to generate a fusion feature dataset; the dataset is input into a pre-trained interpretable artificial intelligence model to generate a trusted output with decision results and decision basis explanation; finally, the output is packaged as a standardized cloud service interface. Through the deep collaboration of blockchain and AI, the invention constructs a complete technical closed loop from "trusted data" to "transparent decision" to "convenient service"; effectively solves the problems of shallow integration of the two, non-transparent decision-making process and difficult service landing in the prior art, and significantly improves the credibility and efficiency of data-driven decision-making in the digital economic ecosystem.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a system that integrates blockchain technology, big data analysis and artificial intelligence, specifically a trusted decision-making system and method based on the deep integration of blockchain and big data. Background Technology

[0002] With the deepening development of the digital economy, big data decision-making technology has gradually evolved from single-data statistical analysis to multi-source data fusion decision-making since 2010, and has been widely applied in ecosystem scenarios such as retail, finance, and supply chain. Blockchain technology, with its characteristics of "immutability" and "traceability," became a core technology for solving data trust issues after 2015, expanding from the field of digital currency to areas such as data ownership confirmation and cross-border collaboration. In recent years, big data and blockchain technologies have shown a trend of integration. The current common integration model is "blockchain notarization + big data analysis," that is, first storing the hash value of the data on the blockchain to ensure its immutability, and then using big data technology to analyze the original data. However, this combination is superficial and has the following three obvious drawbacks: Lack of deep collaboration: Blockchain only acts as a "notary" after the fact, without being deeply integrated into the process of data fusion and decision generation. It cannot guarantee the authenticity of the original data used for data analysis, or the transparency of the decision-making process. Lack of decision-making closed loop: After the decision results are generated, the business feedback data on the actual effect cannot be linked again with blockchain notarization and model optimization, thus failing to form a self-optimizing and trustworthy closed loop. Service-oriented implementation faces challenges: Existing solutions struggle to encapsulate the capabilities of "trusted data" and "intelligent decision-making" into standardized, convenient cloud services, failing to meet the digital economy ecosystem's demands for "trusted data - efficient decision-making - convenient services." The integrated needs of the digital economy are urgent. Therefore, there is a pressing need in this field for a complete technological system that enables deep technological collaboration, forms a closed-loop decision-making process, and can be easily implemented as a cloud service. In particular, the advancements in AI data quantification and cleaning technologies in the past two years have provided new solutions for the in-depth development of the digital economy. Summary of the Invention

[0003] 1. The technical problem to be solved by the present invention This invention aims to address the problem of superficial integration of big data and blockchain in existing technologies, and provides a complete closed-loop system and method that can realize data credibility, intelligent decision-making and service implementation. Combined with the quantitative analysis of AI and the cleaning of invalid data, it can meet the integrated needs of the digital economy for data credibility, efficient decision-making and convenient services. 2. Technical solution of the present invention To solve the above-mentioned technical problems, the present invention adopts the following technical solution: (1) In a first aspect, the present invention provides a trusted decision-making method based on the deep integration of blockchain and big data, comprising the following steps: Data trustworthiness and evidence preservation steps: Obtain raw data from multiple heterogeneous data sources, perform feature extraction and standardization processing on the raw data to generate standardized data packets; calculate the hash value of the standardized data packets, and store the hash value together with the data source information and timestamp in the blockchain network to generate data evidence preservation certificates; Multi-source data fusion steps: Based on the data storage certificate, cross-domain association and feature fusion are performed on standardized data packets from different data sources to generate a fused feature dataset; Intelligent decision generation step: Input the fused feature dataset into a pre-trained interpretable artificial intelligence model to generate decision results and corresponding explanations of decision basis; Decision service step: Encapsulate the decision results and explanations of decision basis into a standardized cloud service interface for output. (2) In a second aspect, the present invention provides a trusted decision-making system based on the deep integration of blockchain and big data to implement the above method, comprising: The Trusted Data Base Module is used to obtain raw data from multiple heterogeneous data sources, process it to generate standardized data packets, calculate its hash value and store it in the blockchain to generate data storage certificates. The intelligent decision engine module is used to perform cross-domain correlation and feature fusion on standardized data packets from different data sources based on the data storage certificate, generate a fused feature dataset, and input it into a pre-trained interpretable artificial intelligence model to generate decision results and corresponding explanations of decision basis. The cloud-edge-device collaborative service module is used to interpret and encapsulate the decision results and decision basis into standardized cloud service interfaces for output. (3) In a third aspect, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the above method. (4) In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above method. 3. Compared with the prior art, the present invention has the following significant advantages: (1) Deep technological synergy has been achieved: Blockchain is not only a tool for evidence storage, but also a trusted cornerstone in data fusion and decision-making processes. By using data evidence storage certificates as a prerequisite for data fusion and AI decision-making, the credibility of the input end of the entire decision-making chain is ensured, and the organic integration of blockchain with big data / AI is realized. (2) A trustworthy decision-making closed loop has been constructed: By introducing an interpretable AI model, the generated decision results are accompanied by explanations of the decision-making basis, making the decision-making process transparent and auditable. At the same time, business feedback data can be used for model retraining and optimization, and this optimization process is recorded on the blockchain, thus forming a complete and trustworthy closed-loop system of "data trustworthiness -> decision transparency -> feedback optimization -> record on the blockchain". (3) Facilitated the implementation of convenient services: By encapsulating core decision-making capabilities into standardized cloud service interfaces, the technical threshold for use has been lowered, enabling users in the retail, finance, and supply chain sectors to call trusted decision-making services in a loosely coupled and highly efficient manner, truly realizing the implementation of "Decision-making as a Service". (4) Enhanced data value and security: Under the premise of ensuring data privacy and security (for example, by combining with technologies such as federated learning), the value integration and mining of multi-source data has been realized, making data a production factor that can be identified, priced, and circulated with confidence. Attached Figure Description 1) System overall business structure topology diagram: Figure 1 2) Overall system business structure: Figure 2 3) System User Flow: Figure 3 4) Overall workflow: Figure 4 5) Data Flow Visualization S201: Figure 5 6) Feature calculation model S202: Figure 6 7) Intelligent Solution Model S203: Figure 7 8) Service and Deployment Release Model: S204: Figure 8 9) Detailed process for intelligent decision generation: Figure 9 10) Data Inference Model S301: Figure 10 11) Key Feature Acquisition Model S303: Figure 11 12) Credibility Measurement Model S304: Figure 12 13) Report generation model S305: Figure 13 14) Quality Assurance and Anomaly Handling Mechanism: Figure 14 15) Overall hardware environment topology of the system platform: Figure 15 16) Component connection topology: Figure 16 17) Data flow and control flow topology: Figure 17 18) Fault-tolerant and redundant topologies: Figure 18 19) Software Architecture and Hardware Mapping: Figure 19 20) Typical Deployment Scenarios: Figure 20 21) The seven functions and processing models of the present invention: Figure 21 Detailed Implementation The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention. (1) Example 1: Figures 1, 2 and 3 together include the overall system business organization flowchart, the overall system business structure diagram and the user flowchart. Referring to Figure 1, the system architecture of this invention mainly includes four core modules: a user data encryption module 100, a trusted data base module 200, an intelligent decision engine module 300, and a cloud-edge-device collaborative service module 400. These modules can be deployed in a distributed computing environment consisting of servers, network devices, and user terminals. Referring to Figure 4, the reliable decision-making method of the present invention includes the following steps: Meanwhile, the overall workflow in Figure 3 can be referenced. S201: Data Trustworthiness and Evidence Preservation. Figure 5 The trusted data foundation module 200 acquires raw data from multiple heterogeneous data sources (such as the user data encryption module 100 end: enterprise ERP database, IoT sensors, third-party data platforms, etc.). The data acquisition unit 201 cleans, deduplicates, and converts the format of the raw data. Subsequently, the data preprocessing unit 202 performs feature extraction and standardization on the cleaned data to generate a standardized data packet with a unified format. The blockchain notarization unit 203 calculates the hash value of the data packet and writes it, along with the data source ID, timestamp, and other information, into the blockchain network 500 through a smart contract, generating a unique and tamper-proof data notarization certificate. This certificate serves as the "pass" for the data to enter subsequent processes. S202: Multi-source data fusion. Figure 6 The feature fusion unit 301 of the intelligent decision engine module 300 is triggered. It acquires standardized data packets from different data sources but with the same evidence. Using algorithms such as graph neural networks (GNNs) or attention mechanisms, it performs cross-domain correlation analysis on these data packets to mine deep feature associations, ultimately generating a comprehensive, high-quality fused feature dataset. S203: Intelligent decision generation. Figure 7 The fused feature dataset generated in S202 is input into the interpretable AI analysis unit 202. This unit consists of a pre-trained interpretable model (e.g., a model based on SHAP or LIME). The model not only outputs the decision result (e.g., "high risk" or "low risk" in a financial risk control scenario), but also simultaneously outputs the key feature factors that led to the decision and their weights, forming an explanation of the decision basis. S204: Decision-making as a Service. Figure 8 The service encapsulation unit 401 of the cloud-edge-device collaborative service module 400 receives the decision results and interpretation information, and encapsulates them into an API interface conforming to the RESTful specification. The resource scheduling unit 402 dynamically allocates computing resources based on the current system load to ensure the API's response performance. Ultimately, users or third-party applications can conveniently obtain reliable decision services by calling this API. The above process ensures that the following key control points for critical system processes are quality checkpoints: Data storage and verification: Blockchain-based evidence storage must be successful before proceeding to the next stage. Fusion Quality Assessment: The fused datasets must pass a quality assessment. Decision confidence check: Low-confidence decisions trigger uncertainty handling Service health check: Continuous monitoring to ensure service availability And including exception handling mechanisms: Data acquisition failed → Retry mechanism + alternative data source Failed evidence storage → Retry a maximum of 3 times, log the process. If the fusion quality is substandard, provide feedback and adjust the fusion parameters. Low decision confidence level → Triggers manual review or data supplementation Service failure → Automatic capacity expansion or failover The following performance optimization strategies also exist: Asynchronous processing: Data storage and feature fusion can be performed in parallel. Caching mechanism: Frequently used data is cached to reduce redundant calculations. Incremental update: only processes changed data, not the entire dataset. Edge computing: Some decisions are made at the edge nodes. (2) Example 2: Detailed process of intelligent decision generation based on Figure 9 This embodiment provides a detailed explanation of step S203 (intelligent decision generation), as shown in Figure 9: S301: Model Inference. The fused feature dataset is input into the interpretable AI model. Figure 10 S302: Result Prediction. The model outputs preliminary prediction results (such as classification labels or regression values). S303: Based on retrospection. The interpreter inside the model analyzes the contribution of input features to the prediction results and identifies the top N key features with the greatest impact. Figure 11 S304: Quantitative Confidence Measurement. A quantitative decision confidence score is calculated by combining the model's own confidence level, the consistency of key features, and the freshness of the data. Figure 12 S305: Report Generation. Integrate the prediction results, key feature list, contribution weights, and confidence scores to generate a human-readable decision explanation report. Figure 13 Quality Assurance and Anomaly Handling Mechanism Figure 14 Quality checkpoints: Model input validation: Ensure that the data format and distribution meet the model requirements. Interpretive consistency check: Results from different interpretation methods should be consistent. Credibility threshold setting: Dynamically adjust the credibility threshold according to the application scenario. Report integrity verification: Ensure the report contains all necessary information. Exception handling strategy Performance optimization measures: Caching mechanism: Cache the model interpretation results to avoid duplicate calculations. Parallel processing: The interpretation computation of multiple samples can be performed in parallel. Incremental update: Only recalculates the interpretation of the changed parts. Approximate Algorithm: For scenarios with high real-time requirements, an approximate interpretation algorithm is used. This set of detailed flowcharts fully demonstrates the internal working mechanism of intelligent decision generation. Every step from model inference to final report generation is represented in detail and includes anomaly handling and quality control mechanisms to ensure the interpretability and reliability of the decision-making process. (3) Example 3: Topology diagram of the overall hardware environment of the platform in conjunction with Figure 15 Figure 15 illustrates the topology of a platform environment (such as a server or high-performance computer) for implementing the method of the present invention, including: At least two application algorithm processors 401 (CPU / GPU); Two data storage devices 402 are used to store and back up data hash values ​​and various types of valuable data; Communication module 403 is used for communication with other devices or networks; Optionally, input / output interface 404. The memory 402 stores a computer program, which, when executed by the processor 401, implements the steps of the method described in embodiments 1 and 2 above. Based on the description in Figure 16 above, we designed a more detailed topology diagram: Component connection topology diagram 16 Data flow and control flow topology diagram 17 Fault-tolerant and redundant topology diagram 18 Software architecture and hardware mapping diagram 19 And the following key configuration instructions 1. Processor configuration scheme: Processor Array 401: - Type: Heterogeneous Computing Architecture - CPU: Multi-core Xeon / EPYC, clock speed ≥ 2.5GHz, cores ≥ 32 - GPU: NVIDIA Tesla / AMD Instinct, with ≥32GB of video memory - Interconnection: NVLink / InfiniBand, bandwidth ≥200Gb / s - Cache: L3 cache ≥ 50MB 2.2 Memory Configuration Scheme Data storage device 402: Primary Storage: - Type: All-flash array - Capacity: ≥100TB available space - IOPS: ≥500K random read / write - Protocol: NVMe over Fabrics Backup Storage: - Type: Hybrid Storage Array - Capacity: ≥500TB - Features: Data deduplication, compression, snapshots Synchronization mechanism: - Mode: Synchronous replication + asynchronous replication - RPO: <5 minutes - RTO: <15 minutes 3. Network Configuration Scheme: - Communication Module 403: - Core Network: - Switch: 100GbE Spine-Leaf architecture - Delay: <10 microseconds - Bandwidth: ≥40Gb / s per node Storage Network: - Protocol: iSCSI / NVMe-oF - Private network: Separate traffic Management Network: - Out-of-band Management Interface - IPMI / iDRAC support 4 System Redundancy Design - N+1 processor redundancy: The failure of any processor does not affect system operation. - Active-active storage: Real-time dual data writing, automatic failover - Multipath networks: Link aggregation and failover - Hot-swappable components: Supports online replacement of faulty parts. - Dual power supply redundancy: Independent power supply for A and B circuits Figures 4-5 illustrate typical deployment scenarios. These topology diagrams demonstrate the complete hardware architecture of the platform environment for the method of this invention, including computing, storage, networking, and I / O subsystems, as well as their interconnections and data flows. The system design considers high performance, high availability, and scalability, enabling efficient execution of the method steps described in Examples 1 and 2. Figure 20 illustrates the complete hardware architecture of the platform environment for the method of this invention, including computing, storage, networking, and I / O subsystems, as well as their interconnections and data flows. The system design considers high performance, high availability, and scalability, enabling efficient execution of the method steps described in Examples 1 and 2. The entire system is executed by seven main modules. Figure 21 ). 6. Innovation points that need to be explained The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims. Based on the above, we have made a more comprehensive plan for the system we are about to implement, and the innovation of the system includes the following aspects: This invention achieves deep integration of blockchain and big data through a "four-layer collaborative architecture + five-step closed-loop process," with the following core innovations: (1) Innovation Point 1: A four-layer collaborative architecture of "raw data preservation - blockchain rights confirmation - big data standardized decision-making - cloud services". Original Data Encryption Protection Layer: In the past digital economy and blockchain structures, once a transaction or confirmation of rights was completed, the complete ownership of the data belonged to the buyer. However, problems can arise during the transaction process. Therefore, by adding an encryption protection layer for the original data, we have solved the problem of the integrity and immutability of the original data, thereby addressing the integrity and accuracy of the original data even if the cloud data has been tampered with. Blockchain ownership confirmation layer: It pioneered the "dynamic ownership mapping algorithm" to record the ownership changes of data throughout the entire process through smart contracts, generate tamper-proof certificates, and solve the problems of data trust and ownership. Big data standardization decision-making layer: Develop a "cross-chain data call interface" to synchronize trusted blockchain data; combine a "dynamic weight model" to adjust the analysis weights according to the importance of data and improve the accuracy of decision-making. Cloud service layer: Design "multi-entity customized interface" to match the needs of different entities in the ecosystem, embed blockchain verification components, and realize "service call is verification". (2) Innovation Point 2: Five-step closed-loop process of "confirmation of rights - collection - analysis - decision-making - service" Confirmation of rights: The smart contract completes the registration of data ownership; Encryption: Measures to protect the integrity of data and information; Data collection: Capturing trusted blockchain data across interconnect interfaces; AI Quantitative Analysis: Blockchain computing power assists in parallel quantitative calculation of big data, balancing efficiency and security; Decision-making: Generate traceable decision reports with timestamps; Services: Customized API push results, log synchronization on the blockchain, forming a closed loop across the entire chain. (3) Innovation Point 3: Three-dimensional integrated design of "method-system-service" The integrated design of "raw data preservation + blockchain (digital economy application) + big data decision-making method", "four-layer architecture hardware system" and "cloud service center" covers all dimensions of "encryption-algorithm-hardware-service" and fills the gaps in the single protection of existing technologies. 7. Description of the inventive concept This invention uses the Xiaoshiren company's platform as a carrier, and through the collaborative four-layer architecture of "user original data value construction layer, blockchain rights confirmation layer, big data decision-making layer, and cloud service layer", it realizes a closed loop of "data value generation - data value rights confirmation - decision-making - service", which is adapted to the business development needs of the Xiaoshiren ecosystem. The functions of each structure are as follows: (1) User Original Data Value Construction Layer: Establish the initial value recognition standard for the digital economy of Xiaoshiren Company's platform data and the identity recognition of data in this blockchain. The recognition and accuracy confirmation of identity are completed when each digital product is manufactured. Data Source and Value Confirmation: When each user registers as a platform user, they receive a unique digital data value identifier on the platform. The system automatically generates a data value tag closely associated with the user based on the platform's digital value protocol rules. This tag will be used in any future activities the user engages in on the platform, such as creating new digital assets. These digital assets will be automatically assigned a corresponding value based on their identifier, thus maximizing the user's ability to objectively reference each transaction. This also includes the redefinition of the digital value of products purchased by customers on the platform. This value will adhere to the natural market fluctuation thresholds on the trading platform, and the platform will regularly publish these standards in advance for the convenience of all parties. Customizable Value and Value Downgrade: Each user can redefine the value of their digital assets based on their understanding of and access to similar data on the platform. However, this value only becomes effective after a formal transaction is completed and ownership is confirmed, and the value is then assigned to the new owner of the digital asset. In addition, the platform will also upgrade or downgrade the value of the digital assets owned by the user based on their own digital asset value level mechanism. This allows for maximum management and control over value operations between customers, thereby preventing fraudulent activities in transactions. The User Digital Economy Value Assessment System automatically assigns an initial value assessment to each user upon joining the platform. Furthermore, it re-evaluates the user's digital economy value based on each transaction and the increase or decrease in transaction data. This system delegates the evaluation of users and merchants to the blockchain, avoiding the centralized evaluation system of traditional business platforms. This feature truly makes the system a fairer platform, allowing each user to cherish their data economic value on this platform. (2) Blockchain Rights Confirmation Layer: Strengthening the Trustworthy Foundation of Xiaoshiren Company Platform Data Core structure and adaptation 1. Xiaoshiren Exclusive Ownership Module: Built-in smart contract templates for platform ecosystem entities (enterprises, users, supply chains) to automatically define data ownership for business development, consumption, etc. 2. Hybrid consensus nodes: Composed of Xiaoshiren's operators, core enterprises, and regulatory agencies, the nodes combine platform credit scores to weighted verify ownership, ensuring fairness. 3. Certificate - Account Binding: Ownership certificates (with hash values) are directly bound to the user / enterprise's Xiaoshiren account, which can be queried on the platform and used for business endorsement. Function Solving the problems of platform data tampering and ambiguous ownership will enable enterprises to dare to share data and be willing to use data for decision-making, laying the foundation for improving the conversion rate. (3) Big Data Decision-Making Layer: Providing Xiaoshiren Company's platform with precise decision-making capabilities Core structure and adaptation 1. Cross-link interface: Directly connects to the Xiaoshiren blockchain layer and data platform, automatically captures trusted data and converts its format (e.g., JSON to CSV). 2. Contextualized weighting model: Built-in core platform contextual rules (Business development scenario: regional data 35% + customer group data 30%; Recommendation scenario: consumption data 40% + behavioral data 30%). 3. Computing power collaboration: Utilize idle nodes of Xiaoshiren (such as enterprise servers) to process complex calculations in parallel, reducing costs and improving efficiency. Function Output time-stamped decision results (such as business expansion area suggestions and user recommendation lists) to solve the problems of insufficient credibility and accuracy of traditional decision-making. (4) Cloud service layer: Realize the platform decision-making scenario implementation of Xiaoshiren Company Core structure and adaptation 1. Customized Interface: Enterprise side connects to merchant backend (check reports, confirm rights), user side embeds into APP (check ownership, raise objections), call requires platform digital signature; 2. Scheduling and monitoring: Service nodes are allocated according to requests, resources are adjusted according to the enterprise's business level, and logs are synchronized to the platform operation backend and blockchain; 3. Extended Adaptability: Supports the addition of new entities (such as local service providers) and new scenarios (such as community group buying decisions), enabling low-cost integration. Function Transform decision-making into business tools and user services, solve problems such as poor platform service adaptation and low ecosystem compatibility, and support ecosystem expansion. (5) Methodology and Flow: Xiaoshiren Company's platform full-chain closed loop Confirmation of rights: Users / enterprises transmit data, generate vouchers, and bind them to Xiaoshiren accounts; Data collection: The decision-making level collects trusted data from the blockchain and the platform; Analysis: Calculate based on scene weight and utilize idle computing power; Decision: Generate a report and synchronize it to your platform account; Service: Results are pushed to the cloud, and logs are stored as evidence, forming a closed loop.

[0004] Advantages or beneficial effects of the present invention 1. Addressing data trust issues and boosting business transactions: The blockchain ownership layer generates immutable ownership certificates for Xiaoshiren's platform data, clarifying data ownership, eliminating concerns about data sharing, making decision-making results more likely to be adopted, and directly improving business transaction rates. 2. Breaking down technological barriers and achieving deep collaboration: The four-layer architecture enables blockchain and big data to communicate with each other within Xiaoshiren's platform—blockchain provides trusted data, and big data enables efficient analysis and decision-making, forming a "data-decision" closed loop and avoiding the problem of functional fragmentation in existing technologies. 3. Comprehensive coverage to reduce vulnerability risks: Covering all dimensions of "method-system-service", it adapts to the entire process requirements of Xiaoshiren Company's platform from data processing to service implementation, with no protection blind spots and a more stable technical system. 4. Adapt to the ecosystem and improve compatibility: The cloud-based customized interface is adapted to multiple entities such as platform enterprises and users of Xiaoshiren Company. The cost of new entities joining is low, which helps the ecosystem expand and meets the goal of "multi-entity participation". 5. Balancing efficiency and security: By utilizing the idle computing power of Xiaoshiren Company's platform to assist in analysis, we can ensure data security and immutability while improving decision-making speed, thus resolving the contradiction of "difficulty in balancing efficiency and security" in existing technologies.

Claims

1. A trusted decision-making method based on the deep integration of blockchain and big data, characterized in that, Includes the following steps: Data trustworthiness processing steps: Obtain raw data from multiple heterogeneous data sources, perform feature extraction and standardization processing on the raw data, generate standardized data packets, and perform two-way notarization on the user end and blockchain nodes; Blockchain evidence storage steps: Calculate the hash value of the standardized data packet, store the hash value together with the data source information and timestamp in the blockchain network, and generate data evidence storage certificate and transaction certificate; Multi-source data fusion steps: Based on the data storage certificate, cross-domain association and feature fusion are performed on standardized data packets from different data sources to generate a fused feature dataset, and user rights are encrypted. Intelligent decision generation steps: Input the fused feature dataset into a pre-trained interpretable artificial intelligence model to generate decision results and corresponding credibility assessments; Decision-making closed-loop optimization steps: Compare the actual business feedback data with the decision results, dynamically adjust the parameters of the interpretable artificial intelligence model based on the comparison results, and store the adjustment process and results in the blockchain; Service-oriented encapsulation steps: Encapsulate the decision results and corresponding credibility assessments into standardized cloud service interfaces. It provides services to the outside world through a microservice architecture.

2. A trusted decision-making system based on the deep integration of blockchain and big data, implementing the method as described in claim 1, characterized in that, include: Trusted data base module, including: (1) Data acquisition unit, used to acquire raw data from multiple heterogeneous data sources in real time; (2) Data preprocessing unit, compares and identifies the raw data with the existing data in the blockchain, and performs cleaning, noise reduction and data standardization processing; (3) Blockchain notarization unit, used to generate data hash values ​​and write them into the blockchain; The intelligent decision engine module includes: (1) Feature fusion unit, used for cross-domain association and feature engineering processing of multi-source data; (2) An interpretable AI analysis unit that generates decision results and explanatory information based on the fused feature data; (3) Model optimization unit, which dynamically adjusts AI model parameters based on business feedback; The cloud-edge-device collaboration service module includes: (1) Service encapsulation unit, which encapsulates decision-making capabilities into standard API services; (2) Resource scheduling unit, which dynamically allocates computing resources to meet service quality requirements; (3) Access control unit, which uses blockchain digital identity for access control.

3. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method as described in claim 1.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method as described in claim 1.

5. The method as described in claim 1, characterized in that, The data trustworthiness processing steps specifically include: Raw data is obtained through a trusted acquisition agent deployed at the data source; The raw data is subjected to a quality assessment, and a data quality report is generated; When the data quality score falls below a preset threshold, a data re-collection process is triggered. Convert qualified raw data into a unified standardized format.

6. The method as described in claim 1, characterized in that, The blockchain evidence storage step also includes: The legitimacy of data sources is automatically verified through smart contracts; Generate a unique digital identity for each data packet; Store the data hash value together with the data usage permission information; Establish a data traceability chain to record the complete history of data flow.

7. The method as described in claim 1, characterized in that, The multi-source data fusion step includes: Based on federated learning technology, cross-domain feature fusion is performed while protecting data privacy; Use attention mechanisms to dynamically assess the feature importance of different data sources; Graph neural networks are used to uncover deep relationships between multiple data sources. Generate an importance weight distribution map of the fused features.

8. The method as described in claim 1, characterized in that, The interpretable artificial intelligence model in the intelligent decision generation step includes: Model prediction result generation unit; The decision-making basis traceability unit is used to identify the key features that have the greatest impact on the final decision. The credibility quantification unit comprehensively evaluates the credibility of decisions based on feature consistency, data freshness, and model confidence. The visualization and explanation unit generates decision explanation reports that are understandable to humans.

9. The method as described in claim 1, characterized in that, The decision-making closed-loop optimization steps specifically include: Establish a decision-making effectiveness evaluation index system; Decision quality scores are calculated based on business feedback data. When the quality score falls below the optimization threshold, the model retraining process is automatically triggered. The performance comparison data of the model before and after optimization is stored in the blockchain.

10. The method as described in claim 1, characterized in that, The service-oriented encapsulation step also includes: Dynamically configure service level agreements based on user roles and business scenarios; Automatically scale service instances up or down based on real-time load; Provides service composition and orchestration capabilities to support complex business scenarios; Provides service usage monitoring and billing functions.

11. The system as claimed in claim 2, characterized in that, The trusted data base module also includes a data exchange submodule, used for: Implement data ownership registration and authorized access based on smart contracts; Secure data flow that is "usable but not visible" through privacy computing technology; Record data using logs and generate audit reports; Based on the reasonable allocation of value in contributing data records using tokens.

12. The system as claimed in claim 2, characterized in that, The cloud-edge-device collaborative service module also includes an edge computing submodule, used for: Deploy lightweight decision-making models on edge devices; Enable collaborative reasoning between cloud-based and edge models; The distribution of computing tasks is dynamically adjusted based on network conditions; Ensure basic decision-making capabilities in scenarios where the internet is down.

Images