Shared economic credit evaluation and evidence storage method based on block chain

Abstract

The invention particularly relates to a shared economic credit evaluation and evidence storage method based on a block chain. According to the shared economic credit evaluation and evidence storage method based on the block chain, during first registration and login, initialization processing is performed on a user identity, service behavior data is acquired in real time, preprocessing is performed, feature engineering processing is performed, and credit factors are extracted and normalized; constructing a multi-dimensional credit model, and realizing dynamic credit score and threshold adaptive calibration; generating a multi-dimensional abstract, and storing an evidence on a chain; the target sharing economy platform obtains the credit voucher of the user in a cross-platform manner and performs verification and synchronization; multiple security and compliance strategies are provided, and privacy protection, compliance supervision and black and white list feedback closed-loop management are realized. According to the shared economy credit assessment and evidence storage method based on the block chain, credit credibility and traceability are improved, migration and verification of user credit among multiple shared economy platforms are realized, user privacy and compliance are guaranteed, and multi-platform joint management and risk early warning are realized.

Description

Technical Field

[0001] This invention relates to the fields of blockchain and credit assessment technology, and in particular to a blockchain-based method for credit assessment and evidence storage in the sharing economy. Background Technology

[0002] With the rapid development of the sharing economy, services such as ride-sharing, accommodation sharing, office sharing, and labor sharing have become increasingly popular, greatly improving the efficiency of social resource utilization. In these services, the creditworthiness of platform users has become a key factor determining service quality and transaction security. Therefore, various sharing platforms have generally introduced user credit scoring mechanisms to assess user behavior and provide services such as recommendations and credit guarantees accordingly.

[0003] However, existing sharing economy platforms generally adopt centralized credit assessment systems, which have the following technical bottlenecks and shortcomings:

[0004] First, user credit data is independently controlled by each platform, and the data cannot be shared, forming "information silos." Second, credit scoring lacks a transparent evaluation mechanism, making it susceptible to platform manipulation or malicious interference, affecting fairness. Third, user credit cannot be transferred between different platforms; users must rebuild their credit each time they enter a new platform, impacting user experience and trust building. In addition, traditional credit systems suffer from insufficient credibility in data storage and traceability, making it difficult to meet the needs of supervision and dispute arbitration.

[0005] Blockchain technology, with its decentralized, tamper-proof, and traceable characteristics, has demonstrated strong data ownership verification and credit protection capabilities in various fields such as finance, logistics, and government affairs. Therefore, there is an urgent need for a blockchain-based method for user credit assessment and notarization in the sharing economy, breaking down existing platform barriers, enabling cross-platform verification and sharing of credit, and improving the overall credit credibility and operational efficiency of the sharing economy system.

[0006] Based on the above, this invention proposes a blockchain-based method for credit assessment and evidence storage in the sharing economy. Summary of the Invention

[0007] To overcome the shortcomings of existing technologies, this invention provides a simple and efficient blockchain-based method for credit assessment and evidence storage in the sharing economy.

[0008] This invention is achieved through the following technical solution:

[0009] A blockchain-based method for credit assessment and notarization in the sharing economy includes the following steps:

[0010] Step S1: When a user registers and logs in for the first time, the user's identity is initialized, and service behavior data of the user is collected in real time during each complete service lifecycle on the sharing economy platform.

[0011] In step S1, when a user completes KYC (Know Your Customer) real-name authentication for the first time on a sharing economy platform (such as shared travel, shared accommodation, shared charging, etc.), the identity service module will cast a unique decentralized identity identifier DID for the user in the consortium blockchain identity contract, and solidify the user's public key, KYC hash, device fingerprint and signature timestamp on the consortium blockchain, and establish a mapping relationship with the local account.

[0012] At each key node of the user's subsequent complete service lifecycle, including order placement, fulfillment, and evaluation, the sharing economy platform gateway automatically collects multi-dimensional raw data through event streams, including order number, timestamp, GPS trajectory / device log and billing details, as well as mutual evaluation content and star rating of both trading partners, while recording risk signals, including cancellation, timeout, customer complaints and penalties;

[0013] All collected data is pushed to the behavioral data lake via Kafka / Pulsar streams, with BlockTime appended to the message header to ensure on-chain and off-chain time consistency.

[0014] If data collection fails, it will enter the retry and supplement queue to ensure data integrity.

[0015] Step S2: Preprocess the user service behavior data used, perform feature engineering on the original fields according to the service type, extract and normalize five major categories of credit factors, including performance factor, evaluation factor, risk factor, historical consistency factor and activity factor; all features are bound to trusted timestamps and written to the consortium blockchain front buffer or mainstream public blockchain front buffer on the sharing economy platform side, waiting for scoring to be called;

[0016] To ensure cross-platform comparability, in step S2, a unified cleaning rule is used to remove dirty data from user service behavior data, including duplicate records, attack scores, and system errors.

[0017] The performance factors include on-time rate, completion rate, and equipment loss rate;

[0018] The evaluation factors include service attitude score, overall star rating, and text sentiment score;

[0019] The risk factors include the number of complaints, arbitration results, and records of liquidated damages;

[0020] The historical consistency factor is the volatility compared to the mean / variance of the behavior over the past 90 days;

[0021] The activity factor includes the number of orders in the past 30 days, total mileage, or cumulative rental duration.

[0022] Step S3: Adopt a hierarchical hybrid architecture, construct a multi-dimensional credit model through a two-level engine, and realize dynamic credit scoring and threshold adaptive calibration.

[0023] For user-defined low-complexity scenarios, execute weighted linear or scorecard models to quickly output basic scores;

[0024] For user-defined high-complexity or high-risk scenarios, the XGBoost model (structured data modeling), GNN model (graph neural network to capture associated fraud networks) and / or Transformer model (time series feature extraction and long-term dependency analysis) are invoked to integrate historical sequence features and real-time risk signals, and output dynamic scores and confidence intervals.

[0025] In step S3, the scoring range of the multi-dimensional credit model is uniformly mapped to 0–100 using a broken line. The sharing economy platform side calculates the mean and standard deviation of the most recent N (default 10000) scores in real time.

[0026] If the mean drift exceeds the threshold range, the three thresholds of excellent, warning and risk will be automatically recalibrated according to the block height and will take effect after being submitted through multi-signature of the governance contract;

[0027] The final output includes a comprehensive score, risk level, version number of the multi-dimensional credit model combination, and timestamp (the final output structure is {score,riskLevel,modelVersion,ts}), along with a score tracking ID.

[0028] Step S4: Generate a multi-dimensional digest through hash operation, including a behavior layer digest and a user layer digest, and write it into the consortium blockchain sidechain or the mainstream public blockchain sidechain for storage through a smart contract.

[0029] To ensure data verifiability while avoiding privacy leaks, step S4 employs a two-layer digest and verifiable encryption strategy.

[0030] The order ID, important factor vector, score value and timestamp are hashed using SHA-256 to obtain Hash_act, which is used as the behavior layer summary;

[0031] Merkle tree aggregation is performed on the summaries of the most recent M behaviors, and then hashed to obtain Hash_user, which serves as the user-level summary. M is determined by the user based on the application scenario.

[0032] Use homomorphic encryption or ElGamal to encrypt user-defined key scoring fields while preserving additivity;

[0033] The smart contract CreditRecordContract receives the DID, Hash_act, Hash_user, EncScore, and PlatformSig fields, and writes them to the consortium blockchain / mainstream public blockchain sidechain after node consensus. Once the on-chain record is confirmed, it cannot be tampered with and can be traced and audited later.

[0034] Step S5: When a user jumps to another sharing economy platform, the target sharing economy platform obtains the user's credit credentials across platforms and verifies and synchronizes them.

[0035] In step S5, whenever a new on-chain certificate is confirmed, the shared economy platform ledger service automatically refreshes the user's on-chain credit index and, when necessary, mints or updates a non-transferable CreditSoulbound Token (C-SBT) for the user, encoding the latest encrypted score and risk level into the token metadata TokenMetadata.

[0036] When a user is redirected to another sharing economy platform, the target sharing economy platform completes three verification steps by calling open APIs or on-chain query interfaces:

[0037] First, the signature of the Credit Soul token holder is questioned to confirm whether the identity is consistent;

[0038] Then, the zero-knowledge proof ZKP circuit is invoked to verify whether the encryption score is not lower than 80 (EncScore≥80) or whether there is a risk-free label, without decrypting the original score;

[0039] Finally, a random check is performed on the on-chain Hash_user field, and the hash is recalculated locally to prevent tampering.

[0040] Step S6: To comply with GDPR, cybersecurity laws, and local data export regulations, multiple security and compliance strategies are provided to achieve closed-loop management of privacy protection, compliance supervision, and blacklist / whitelist feedback; specific strategies are as follows:

[0041] The principle of minimum disclosure: When a sharing economy platform queries data from other parties, it only obtains the minimum set that meets its business objectives;

[0042] Supports user revocation of authorization: Users can revoke third-party read permissions at any time on the wallet app;

[0043] Blacklist and whitelist synchronization: If a sharing economy platform determines that a user has engaged in malicious order-brushing or serious breach of contract, it supports adding the user to the blacklist through on-chain event broadcasting combined with multi-signature consensus of regulatory nodes, and synchronizing it with other sharing economy platforms in real time.

[0044] Regulatory read-only nodes: Compliance regulatory departments hold independent on-chain read-only nodes, which support thorough auditing of evidence storage data and algorithm logs, and support dispute arbitration and evidence collection;

[0045] Audit traceability: Contract calls, model inference, and threshold update operations are all recorded in WORM storage and retained for no less than 7 years for subsequent evidence collection.

[0046] A blockchain-based credit assessment and evidence storage system for the sharing economy, used to implement the above method, includes an identity service module, a data collection module, a feature extraction module, a dynamic scoring module, a blockchain evidence storage module, a credit certificate sharing module, and a security supervision module.

[0047] The identity service module is responsible for forging a unique decentralized identity identifier (DID) for users in the consortium blockchain identity contract after users complete KYC (Know Your Customer) real-name authentication for the first time on the sharing economy platform. It also solidifies the user's public key, KYC hash, device fingerprint, and signature timestamp on the consortium blockchain and establishes a mapping relationship with the local account.

[0048] The data acquisition module is responsible for collecting user service behavior data in real time during each complete service lifecycle on the sharing economy platform, and pushing all collected data to the behavior data lake via Kafka / Pulsar stream, with BlockTime appended to the message header to ensure consistency between on-chain and off-chain time sequences.

[0049] The feature extraction module is responsible for preprocessing the user service behavior data, performing feature engineering on the original fields according to the service type, and extracting and normalizing five major categories of credit factors, including performance factors, evaluation factors, risk factors, historical consistency factors, and activity factors.

[0050] The dynamic scoring module is responsible for using a hierarchical hybrid architecture to build a multi-dimensional credit model through a two-level engine, thereby achieving dynamic credit scoring and adaptive threshold calibration.

[0051] For user-defined low-complexity scenarios, execute weighted linear or scorecard models to quickly output basic scores;

[0052] For user-defined high-complexity or high-risk scenarios, the XGBoost model (structured data modeling), GNN model (graph neural network to capture associated fraud networks) and / or Transformer model (time series feature extraction and long-term dependency analysis) are invoked to integrate historical sequence features and real-time risk signals, and output dynamic scores and confidence intervals.

[0053] The blockchain evidence storage module is responsible for generating multi-dimensional summaries through hash operations, including behavioral layer summaries and user layer summaries, and storing them in the consortium blockchain sidechain or mainstream public blockchain sidechain through smart contracts.

[0054] The Credit Certificate Sharing Module is responsible for minting or updating Credit Soul Binding Tokens for users after new on-chain notarization and confirmation, encoding the latest encrypted score and risk level into the token metadata; and when users jump to other sharing economy platforms, it helps the target sharing economy platform obtain the user's Credit Soul Binding Tokens across platforms and verify and synchronize them.

[0055] The security supervision module is responsible for providing multiple security and compliance strategies to achieve closed-loop management of privacy protection, compliance supervision, and blacklist / whitelist feedback.

[0056] A blockchain-based shared economy credit assessment and evidence storage device includes a memory and a processor; the memory is used to store a computer program, and the processor is used to execute the computer program to implement the above-described method steps.

[0057] A readable storage medium storing a computer program that, when executed by a processor, implements the above-described method steps.

[0058] The beneficial effects of this invention are: the blockchain-based sharing economy credit assessment and notarization method utilizes blockchain to realize the on-chain notarization and timestamp confirmation of user credit data, ensuring that the information is tamper-proof and traceable throughout the process, thereby improving credit credibility and traceability;

[0059] Based on decentralized identity (DID) and credit token (SBT), it enables the migration and verification of user credit across multiple sharing economy platforms, supports cross-platform credit interoperability and privacy compliance, and combines encrypted scoring and zero-knowledge proof to ensure user privacy and compliance.

[0060] Meanwhile, the introduction of on-chain blacklist and whitelist mechanisms and dynamic scoring models has effectively improved the identification rate of malicious behavior, enhanced risk control capabilities and platform joint prevention and control, and achieved multi-platform joint governance and risk warning. Attached Figure Description

[0061] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0062] Figure 1 This is a schematic diagram of the blockchain-based credit assessment and evidence storage method for the sharing economy according to the present invention. Detailed Implementation

[0063] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions in the embodiments of this invention will be clearly and completely described below in conjunction with the embodiments of this invention. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0064] This blockchain-based method for credit assessment and notarization in the sharing economy includes the following steps:

[0065] Step S1: When a user registers and logs in for the first time, the user's identity is initialized, and service behavior data of the user is collected in real time during each complete service lifecycle on the sharing economy platform.

[0066] In step S1, when a user completes KYC (Know Your Customer) real-name authentication for the first time on a sharing economy platform (such as shared travel, shared accommodation, shared charging, etc.), the identity service module will cast a unique decentralized identity identifier DID for the user in the consortium blockchain identity contract, and solidify the user's public key, KYC hash, device fingerprint and signature timestamp on the consortium blockchain, and establish a mapping relationship with the local account.

[0067] At each key node of the user's subsequent complete service lifecycle, including order placement, fulfillment, and evaluation, the sharing economy platform gateway automatically collects multi-dimensional raw data through event streams, including order number, timestamp, GPS trajectory / device log and billing details, as well as mutual evaluation content and star rating of both trading partners, while recording risk signals, including cancellation, timeout, customer complaints and penalties;

[0068] All collected data is pushed to the behavioral data lake via Kafka / Pulsar streams, with BlockTime appended to the message header to ensure on-chain and off-chain time consistency.

[0069] If data collection fails, it will enter the retry and supplement queue to ensure data integrity.

[0070] Step S2: Preprocess the user service behavior data used, perform feature engineering on the original fields according to the service type, extract and normalize five major categories of credit factors, including performance factor, evaluation factor, risk factor, historical consistency factor and activity factor; all features are bound to trusted timestamps and written to the consortium blockchain front buffer or mainstream public blockchain front buffer on the sharing economy platform side, waiting for scoring to be called;

[0071] To ensure cross-platform comparability, in step S2, a unified cleaning rule is used to remove dirty data from user service behavior data, including duplicate records, attack scores, and system errors.

[0072] The performance factors include on-time rate, completion rate, and equipment loss rate;

[0073] The evaluation factors include service attitude score, overall star rating, and text sentiment score;

[0074] The risk factors include the number of complaints, arbitration results, and records of liquidated damages;

[0075] The historical consistency factor is the volatility compared to the mean / variance of the behavior over the past 90 days;

[0076] The activity factor includes the number of orders in the past 30 days, total mileage, or cumulative rental duration.

[0077] Step S3: Adopt a hierarchical hybrid architecture, construct a multi-dimensional credit model through a two-level engine, and realize dynamic credit scoring and threshold adaptive calibration.

[0078] For user-defined low-complexity scenarios, execute weighted linear or scorecard models to quickly output basic scores;

[0079] For user-defined high-complexity or high-risk scenarios, the XGBoost model (structured data modeling), GNN model (graph neural network to capture associated fraud networks) and / or Transformer model (time series feature extraction and long-term dependency analysis) are invoked to integrate historical sequence features and real-time risk signals, and output dynamic scores and confidence intervals.

[0080] In step S3, the scoring range of the multi-dimensional credit model is uniformly mapped to 0–100 using a broken line. The sharing economy platform side calculates the mean and standard deviation of the most recent N (default 10000) scores in real time.

[0081] If the mean drift exceeds the threshold range, the three thresholds of excellent, warning and risk will be automatically recalibrated according to the block height and will take effect after being submitted through multi-signature of the governance contract;

[0082] The final output includes a comprehensive score, risk level, version number of the multi-dimensional credit model combination, and timestamp (the final output structure is {score,riskLevel,modelVersion,ts}), along with a score tracking ID.

[0083] Step S4: Generate a multi-dimensional digest through hash operation, including a behavior layer digest and a user layer digest, and write it into the consortium blockchain sidechain or the mainstream public blockchain sidechain for storage through a smart contract.

[0084] To ensure data verifiability while avoiding privacy leaks, step S4 employs a two-layer digest and verifiable encryption strategy.

[0085] The order ID, important factor vector, score value and timestamp are hashed using SHA-256 to obtain Hash_act, which is used as the behavior layer summary;

[0086] Merkle tree aggregation is performed on the summaries of the most recent M behaviors, and then hashed to obtain Hash_user, which serves as the user-level summary. M is determined by the user based on the application scenario.

[0087] Use homomorphic encryption or ElGamal to encrypt user-defined key scoring fields while preserving additivity;

[0088] The smart contract CreditRecordContract receives the DID, Hash_act, Hash_user, EncScore, and PlatformSig fields, and writes them to the consortium blockchain / mainstream public blockchain sidechain after node consensus. Once the on-chain record is confirmed, it cannot be tampered with and can be traced and audited later.

[0089] Step S5: When a user jumps to another sharing economy platform, the target sharing economy platform obtains the user's credit credentials across platforms and verifies and synchronizes them.

[0090] In step S5, whenever a new on-chain certificate is confirmed, the shared economy platform ledger service automatically refreshes the user's on-chain credit index and, when necessary, mints or updates a non-transferable CreditSoulbound Token (C-SBT) for the user, encoding the latest encrypted score and risk level into the token metadata TokenMetadata.

[0091] When a user is redirected to another sharing economy platform, the target sharing economy platform completes three verification steps by calling open APIs or on-chain query interfaces:

[0092] First, the signature of the Credit Soul token holder is questioned to confirm whether the identity is consistent;

[0093] Then, the zero-knowledge proof ZKP circuit is invoked to verify whether the encryption score is not lower than 80 (EncScore≥80) or whether there is a risk-free label, without decrypting the original score;

[0094] Finally, a random check is performed on the on-chain Hash_user field, and the hash is recalculated locally to prevent tampering.

[0095] This process enables business innovations such as deposit-free unlocking, expedited check-in, and priority order dispatch, significantly improving the user experience.

[0096] Step S6: To comply with GDPR, cybersecurity laws, and local data export regulations, multiple security and compliance strategies are provided to achieve closed-loop management of privacy protection, compliance supervision, and blacklist / whitelist feedback; specific strategies are as follows:

[0097] The principle of minimum disclosure: When a sharing economy platform queries data from other parties, it only obtains the minimum set that meets its business objectives;

[0098] Supports user revocation of authorization: Users can revoke third-party read permissions at any time on the wallet app;

[0099] Blacklist and whitelist synchronization: If a sharing economy platform determines that a user has engaged in malicious order-brushing or serious breach of contract, it supports adding the user to the blacklist through on-chain event broadcasting combined with multi-signature consensus of regulatory nodes, and synchronizing it with other sharing economy platforms in real time.

[0100] Regulatory read-only nodes: Compliance regulatory departments hold independent on-chain read-only nodes, which support thorough auditing of evidence storage data and algorithm logs, and support dispute arbitration and evidence collection;

[0101] Audit traceability: Contract calls, model inference, and threshold update operations are all recorded in WORM storage and retained for no less than 7 years for subsequent evidence collection.

[0102] In summary, the six-step closed loop from data generation, factor processing, dynamic scoring, on-chain evidence storage, credential verification to compliance supervision has been fully established, significantly improving the credibility, transparency, and cross-platform usability of the sharing economy credit system, while ensuring data security and compliance, and providing the industry with a replicable and scalable decentralized credit infrastructure.

[0103] The blockchain-based sharing economy credit assessment and evidence storage system is used to implement the above methods, including an identity service module, a data collection module, a feature extraction module, a dynamic scoring module, a blockchain evidence storage module, a credit certificate sharing module, and a security supervision module.

[0104] The identity service module is responsible for forging a unique decentralized identity identifier (DID) for users in the consortium blockchain identity contract after users complete KYC (Know Your Customer) real-name authentication for the first time on the sharing economy platform. It also solidifies the user's public key, KYC hash, device fingerprint, and signature timestamp on the consortium blockchain and establishes a mapping relationship with the local account.

[0105] The data acquisition module is responsible for collecting user service behavior data in real time during each complete service lifecycle on the sharing economy platform, and pushing all collected data to the behavior data lake via Kafka / Pulsar stream, with BlockTime appended to the message header to ensure consistency between on-chain and off-chain time sequences.

[0106] The feature extraction module is responsible for preprocessing the user service behavior data, performing feature engineering on the original fields according to the service type, and extracting and normalizing five major categories of credit factors, including performance factors, evaluation factors, risk factors, historical consistency factors, and activity factors.

[0107] The dynamic scoring module is responsible for using a hierarchical hybrid architecture to build a multi-dimensional credit model through a two-level engine, thereby achieving dynamic credit scoring and adaptive threshold calibration.

[0108] For user-defined low-complexity scenarios, execute weighted linear or scorecard models to quickly output basic scores;

[0109] For user-defined high-complexity or high-risk scenarios, the XGBoost model (structured data modeling), GNN model (graph neural network to capture associated fraud networks) and / or Transformer model (time series feature extraction and long-term dependency analysis) are invoked to integrate historical sequence features and real-time risk signals, and output dynamic scores and confidence intervals.

[0110] The blockchain evidence storage module is responsible for generating multi-dimensional summaries through hash operations, including behavioral layer summaries and user layer summaries, and storing them in the consortium blockchain sidechain or mainstream public blockchain sidechain through smart contracts.

[0111] The Credit Certificate Sharing Module is responsible for minting or updating Credit Soul Binding Tokens for users after new on-chain notarization and confirmation, encoding the latest encrypted score and risk level into the token metadata; and when users jump to other sharing economy platforms, it helps the target sharing economy platform obtain the user's Credit Soul Binding Tokens across platforms and verify and synchronize them.

[0112] The security supervision module is responsible for providing multiple security and compliance strategies to achieve closed-loop management of privacy protection, compliance supervision, and blacklist / whitelist feedback.

[0113] The blockchain-based sharing economy credit assessment and evidence storage device includes a memory and a processor; the memory is used to store a computer program, and the processor is used to execute the computer program to implement the above-described method steps.

[0114] The readable storage medium stores a computer program that, when executed by a processor, implements the above-described method steps.

[0115] This blockchain-based credit assessment and evidence storage method for the sharing economy enhances the credibility and universality of user credit in the sharing economy scenario through a decentralized mechanism. It effectively solves problems such as centralization, data silos, and migration difficulties in the current credit assessment process and is applicable to multiple fields such as shared mobility, shared accommodation, and shared labor.

[0116] The embodiments described above are merely one specific implementation of the present invention. Ordinary changes and substitutions made by those skilled in the art within the scope of the technical solution of the present invention should be included within the protection scope of the present invention.

Claims

1. A blockchain-based method for credit assessment and notarization in the sharing economy, characterized in that: Includes the following steps: Step S1: When a user registers and logs in for the first time, the user's identity is initialized, and service behavior data of the user is collected in real time during each complete service lifecycle on the sharing economy platform. Step S2: Preprocess the user service behavior data, perform feature engineering on the original fields according to the service type, and extract and normalize five major categories of credit factors, including performance factor, evaluation factor, risk factor, historical consistency factor and activity factor. All features are bound to a trusted timestamp and written to the consortium blockchain front buffer or mainstream public blockchain front buffer on the sharing economy platform side, waiting for the scoring call; Step S3: Adopt a hierarchical hybrid architecture, construct a multi-dimensional credit model through a two-level engine, and realize dynamic credit scoring and threshold adaptive calibration. For user-defined low-complexity scenarios, execute a weighted linear or scorecard model and output a base score; For user-defined high-complexity or high-risk scenarios, the XGBoost model, GNN model, and / or Transformer model are invoked to integrate historical sequence features with real-time risk signals and output dynamic scores and confidence intervals. Step S4: Generate a multi-dimensional digest through hash operation, including a behavior layer digest and a user layer digest, and write it into the consortium blockchain sidechain or the mainstream public blockchain sidechain for storage through a smart contract. Step S5: When a user jumps to another sharing economy platform, the target sharing economy platform obtains the user's credit credentials across platforms and verifies and synchronizes them. Step S6: Provide multiple security and compliance strategies to achieve closed-loop management of privacy protection, compliance supervision, and blacklist / whitelist feedback; specific strategies are as follows: The principle of minimum disclosure: When a sharing economy platform queries data from other parties, it only obtains the minimum set that meets its business objectives; Supports user revocation of authorization: Users can revoke third-party read permissions at any time on the wallet app; Blacklist and whitelist synchronization: If a sharing economy platform determines that a user has engaged in malicious order-brushing or serious breach of contract, it supports adding the user to the blacklist through on-chain event broadcasting combined with multi-signature consensus of regulatory nodes, and synchronizing it with other sharing economy platforms in real time. Regulatory read-only nodes: Compliance regulatory departments hold independent on-chain read-only nodes, which support thorough auditing of evidence storage data and algorithm logs, and support dispute arbitration and evidence collection; Audit traceability: Contract calls, model inference, and threshold update operations are all recorded in WORM storage and retained for no less than 7 years.

2. The blockchain-based sharing economy credit assessment and notarization method according to claim 1, characterized in that: In step S1, when a user completes KYC real-name authentication on the sharing economy platform for the first time, the identity service module will cast a unique decentralized identity identifier (DID) for the user in the consortium blockchain identity contract, solidify the user's public key, KYC hash, device fingerprint and signature timestamp on the consortium blockchain, and establish a mapping relationship with the local account. At each key node of a user's complete service lifecycle, including order placement, fulfillment, and evaluation, the sharing economy platform gateway automatically collects multi-dimensional raw data through event streams, including order number, timestamp, GPS trajectory / device logs and billing details, as well as mutual evaluation content and star ratings of both trading partners, while recording risk signals, including cancellation, timeout, customer complaints and penalties; All collected data is pushed to the behavioral data lake via Kafka / Pulsar streams, with BlockTime appended to the message header to ensure on-chain and off-chain time consistency. If data collection fails, it will enter the retry and supplement queue to ensure data integrity.

3. The blockchain-based sharing economy credit assessment and notarization method according to claim 1, characterized in that: To ensure cross-platform comparability, in step S2, a unified cleaning rule is used to remove dirty data from user service behavior data, including duplicate records, attack scores, and system errors. The performance factors include on-time rate, completion rate, and equipment loss rate; The evaluation factors include service attitude score, overall star rating, and text sentiment score; The risk factors include the number of complaints, arbitration results, and records of liquidated damages; The historical consistency factor is the volatility compared to the mean / variance of the behavior over the past 90 days; The activity factor includes the number of orders in the past 30 days, total mileage, or cumulative rental duration.

4. The blockchain-based sharing economy credit assessment and notarization method according to claim 1, characterized in that: In step S3, the scoring range of the multi-dimensional credit model is uniformly mapped to 0–100 using a broken line. The sharing economy platform side calculates the mean and standard deviation of the most recent N scores in real time, where N is determined by the user based on the application scenario. If the mean drift exceeds the threshold range, the three thresholds of excellent, warning and risk will be automatically recalibrated according to the block height and will take effect after being submitted through multi-signature of the governance contract; The final output includes a comprehensive score, risk level, version number of the multi-dimensional credit model combination, and timestamp, along with a score link ID trace.

5. The blockchain-based shared economy credit assessment and notarization method according to claim 1, characterized in that: In step S4, a two-layer digest and verifiable encryption strategy is employed. The order ID, important factor vector, score value and timestamp are hashed using SHA-256 to obtain Hash_act, which is used as the behavior layer summary; Merkle tree aggregation is performed on the summaries of the most recent M behaviors, and then hashed to obtain Hash_user, which serves as the user-level summary. M is determined by the user based on the application scenario. Encrypt user-defined key scoring fields using homomorphic encryption or ElGamal while preserving summability; The smart contract CreditRecordContract receives the DID, Hash_act, Hash_user, EncScore, and PlatformSig fields, and writes them to the consortium blockchain or a mainstream public blockchain sidechain after node consensus.

6. The blockchain-based sharing economy credit assessment and notarization method according to claim 1, characterized in that: In step S5, whenever a new on-chain notarization is confirmed, the shared economy platform ledger service automatically refreshes the user's on-chain credit index and mints or updates the credit soul binding token for the user, encoding the latest encrypted score and risk level into the token metadata. When a user is redirected to another sharing economy platform, the target sharing economy platform completes three verification steps by calling open APIs or on-chain query interfaces: First, the signature of the Credit Soul token holder is questioned to confirm whether the identity is consistent; Then, the zero-knowledge proof ZKP circuit is invoked to verify whether the encryption score is not lower than 80 or whether there is a risk-free label, without decrypting the original score; Finally, a random check is performed on the on-chain Hash_user field, and the hash is recalculated locally to prevent tampering.

7. A blockchain-based credit assessment and evidence storage system for the sharing economy, characterized in that: The method for implementing any one of claims 1 to 6 includes an identity service module, a data acquisition module, a feature extraction module, a dynamic scoring module, a blockchain evidence storage module, a credit certificate sharing module, and a security supervision module. The identity service module is responsible for forging a unique decentralized identity identifier (DID) for users in the consortium blockchain identity contract after the user completes KYC real-name authentication for the first time on the sharing economy platform. It also solidifies the user's public key, KYC hash, device fingerprint, and signature timestamp on the consortium blockchain and establishes a mapping relationship with the local account. The data acquisition module is responsible for collecting user service behavior data in real time during each complete service lifecycle on the sharing economy platform, and pushing all collected data to the behavior data lake via Kafka / Pulsar stream, with BlockTime appended to the message header to ensure consistency between on-chain and off-chain time sequences. The feature extraction module is responsible for preprocessing the user service behavior data, performing feature engineering on the original fields according to the service type, and extracting and normalizing five major categories of credit factors, including performance factors, evaluation factors, risk factors, historical consistency factors, and activity factors. The dynamic scoring module is responsible for using a hierarchical hybrid architecture to build a multi-dimensional credit model through a two-level engine, thereby achieving dynamic credit scoring and adaptive threshold calibration. For user-defined low-complexity scenarios, execute a weighted linear or scorecard model and output a base score; For user-defined high-complexity or high-risk scenarios, the XGBoost model, GNN model, and / or Transformer model are invoked to integrate historical sequence features with real-time risk signals and output dynamic scores and confidence intervals. The blockchain evidence storage module is responsible for generating multi-dimensional summaries through hash operations, including behavioral layer summaries and user layer summaries, and storing them in the consortium blockchain sidechain or mainstream public blockchain sidechain through smart contracts. The Credit Certificate Sharing Module is responsible for minting or updating Credit Soul Binding Tokens for users after new on-chain notarization and confirmation, encoding the latest encrypted score and risk level into the token metadata; and when users jump to other sharing economy platforms, it helps the target sharing economy platform obtain the user's Credit Soul Binding Tokens across platforms and verify and synchronize them. The security supervision module is responsible for providing multiple security and compliance strategies to achieve closed-loop management of privacy protection, compliance supervision, and blacklist / whitelist feedback.

8. A blockchain-based sharing economy credit assessment and evidence storage device, characterized in that: It includes a memory and a processor; the memory is used to store a computer program, and the processor is used to execute the computer program to implement the method according to any one of claims 1 to 6.

9. A readable storage medium, characterized in that: The readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1 to 6.

Images