Data processing method and device based on block chain

Abstract

The invention discloses a data processing method and device based on a block chain, and relates to the technical field of computers. A specific embodiment of the method comprises the steps that a data verification request for a second object transmitted by a first object is transmitted to a block chain, and the block chain obtains a transaction according to an identifier of the second object; an intelligent contract configured for the first object in advance is called, the related data meet the triggering condition in the intelligent contract, execution logic processing is used, and an execution result is obtained; extracting specific information from the execution result for sentiment analysis, identifying a transaction intention, determining risk information by using a risk analysis strategy corresponding to the transaction intention, and generating an analysis result; and converting the analysis result into a feature vector, and transmitting the feature vector to an automatic approval node, so that the approval node processes the feature vector by using an approval strategy to obtain an approval result, and returns the approval result to the first object. According to the embodiment, the intelligent contract technology and the decision engine are combined, and automatic decision making in the supply chain is achieved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a data processing method and device based on blockchain. Background Art

[0002] The operation and management information of the company's carriers, suppliers, service providers, real estate developers, etc. is the core basis for judging whether the corresponding services can be provided. In this field, the approval process of the above core basis involves a large amount of manual operation, paper documents and the dissemination of electronic documents, resulting in low efficiency, prone to errors and high costs. In order to improve efficiency and accuracy, companies have promoted some digital means in the online approval process, but this process can only work at the manual operation level of the internal system, and there are problems with low approval efficiency and high error rate. Summary of the invention

[0003] In view of this, the embodiments of the present invention provide a data processing method and device based on blockchain, which can at least solve the problem that the existing technology relies on manual approval, resulting in low approval efficiency and high error rate.

[0004] To achieve the above-mentioned purpose, according to one aspect of an embodiment of the present invention, a data processing method based on blockchain is provided, including: receiving a data verification request for a second object transmitted by a first object, transmitting the data verification request to the blockchain, so that the blockchain obtains a transaction according to the identifier of the second object; wherein the transaction includes relevant data of the second object; calling a smart contract pre-configured for the first object, and in response to the relevant data satisfying a trigger condition in the smart contract, using the execution logic in the smart contract to process the relevant data to obtain an execution result; extracting specific information from the execution result, performing sentiment analysis on the specific information to identify transaction intent, using a risk analysis strategy corresponding to the transaction intent to determine the risk information of the specific information, and then generating an analysis result; converting the analysis result into a feature vector, transmitting the feature vector to an automated approval node, so that the automated approval node uses the approval strategy to process the feature vector, obtains the approval result and returns it to the first object.

[0005] Optionally, the method further includes: acquiring the relevant data of the second object within a preset time period according to the identifier of the second object; packaging the relevant data into a transaction, and broadcasting the transaction and the identifier of the second object to the blockchain.

[0006] Optionally, the transaction is obtained by decrypting the signed transaction, and the broadcasting of the transaction and the identifier of the second object to the blockchain includes: obtaining a private key pre-registered by the second object, using the private key to sign the transaction to obtain the signed transaction, and broadcasting the signed transaction and the identifier to each block of the blockchain; the obtaining of the transaction according to the identifier of the second object includes: retrieving the corresponding block according to the identifier, obtaining the signed transaction from the retrieved block, obtaining the public key corresponding to the identifier or obtaining the general authorization key, and using the public key or the authorization key to decrypt the signed transaction to obtain the transaction.

[0007] Optionally, broadcasting the signed transaction and the identifier to each block of the blockchain includes: broadcasting the signed transaction and the identifier to each block of the blockchain so that each block obtains a public key corresponding to the identifier and uses the public key to verify the signature; in response to one or more blocks passing the signature verification, creating a new block in the blockchain, packaging the signed transaction and the identifier into the new block, adding the new block to the blockchain, and deleting the signed transaction and the identifier in all original blocks.

[0008] Optionally, the transaction also includes a hash value, which is obtained by encapsulating the relevant data into a data object in a structured format, encrypting the data object, and hashing the encrypted data object; calling the smart contract pre-configured for the first object includes: encapsulating the relevant data into a new data object in a structured format, encrypting the new data object, and hashing the encrypted new data object to obtain a new hash value; in response to the new hash value being consistent with the hash value in the transaction, calling the smart contract pre-configured for the first object.

[0009] Optionally, obtaining the relevant data of the second object within a preset time period based on the identifier of the second object includes: obtaining original relevant data of the second object within a preset time period based on the identifier of the second object, and processing the original relevant data using a preset data processing strategy to obtain the relevant data.

[0010] Optionally, the original relevant data includes real-time operation data and operation result data; after obtaining the original relevant data of the second object within a preset time period, the method further includes: processing the real-time operation data and operation result data through a risk identification model to obtain risk data.

[0011] Optionally, the method also includes: receiving business logic data transmitted by the first object in one or more ways, converting the business logic data into a feature space vector; converting the feature space vector into information in a computer-readable form, and generating smart contract rules based on the information in the computer-readable form; using a programming language program preset in the blockchain to generate a smart contract code corresponding to the smart contract rules; compiling the smart contract code into bytecode for deployment to the blockchain, and assigning a storage address to the smart contract, and storing the smart contract in the storage address.

[0012] Optionally, converting the business logic data into a feature space vector includes: mapping business logic data of different modalities into the same feature space to obtain a feature space vector for each business logic data; wherein the feature space includes one or more feature dimensions, and the feature space vector is composed of feature values ​​under different feature dimensions; calculating the similarity between the feature values ​​of business logic data of different modalities under the same feature dimension, comparing the similarity with a preset similarity threshold, and obtaining a comparison result; using a fusion strategy corresponding to the comparison result, fusing the feature values ​​of business logic data of different modalities under the same feature dimension, and then integrating to obtain the feature space vectors of business logic data of all modalities.

[0013] Optionally, the use of a fusion strategy corresponding to the comparison result is to fuse the feature values ​​of different modal business logic data under the same feature dimension, including: in response to the comparison result being greater than or equal to, selecting any one of the feature values ​​of similar different modal business logic data under the same feature dimension; in response to the comparison result being less than, sorting and integrating the feature values ​​of dissimilar different modal business logic data under the same feature dimension.

[0014] Optionally, extracting specific information from the execution result includes: extracting specific information from the execution result using a pre-trained large language model; wherein, an original large language model corresponding to the first object is determined, and the original large language model is trained using the feature space vector to adjust parameters of the original large language model to obtain the pre-trained large language model.

[0015] To achieve the above-mentioned purpose, according to another aspect of an embodiment of the present invention, a data processing device based on blockchain is provided, including: a transaction acquisition module, which is used to receive a data verification request for a second object transmitted by a first object, and transmit the data verification request to the blockchain, so that the blockchain obtains the transaction according to the identifier of the second object; wherein the transaction includes relevant data of the second object; a smart contract module, which is used to call a smart contract pre-configured for the first object, and in response to the relevant data satisfying the trigger condition in the smart contract, use the execution logic in the smart contract to process the relevant data to obtain an execution result; an analysis module, which is used to extract specific information from the execution result, perform sentiment analysis on the specific information to identify the transaction intention, use the risk analysis strategy corresponding to the transaction intention, determine the risk information of the specific information, and then generate an analysis result; an approval module, which is used to convert the analysis result into a feature vector, and transmit the feature vector to an automated approval node, so that the automated approval node uses the approval strategy to process the feature vector, obtain the approval result and return it to the first object.

[0016] Optionally, the device also includes a data chain module, which is used to: obtain the relevant data of the second object within a preset time period according to the identifier of the second object; package the relevant data into a transaction, and broadcast the transaction and the identifier of the second object to the blockchain.

[0017] Optionally, the transaction is obtained by decrypting the signed transaction, and the data on-chain module is used to: obtain a private key pre-registered by the second object, use the private key to sign the transaction to obtain the signed transaction, and broadcast the signed transaction and the identifier to each block of the blockchain; the obtaining of the transaction according to the identifier of the second object includes: retrieving the corresponding block according to the identifier, obtaining the signed transaction from the retrieved block, obtaining the public key corresponding to the identifier or obtaining the general authorization key, and using the public key or the authorization key to decrypt the signed transaction to obtain the transaction.

[0018] Optionally, the data on-chain module is used to: broadcast the signed transaction and the identifier to each block of the blockchain, so that each block obtains the public key corresponding to the identifier and uses the public key to verify the signature; in response to one or more blocks passing the signature verification, create a new block in the blockchain, package the signed transaction and the identifier into the new block, add the new block to the blockchain, and delete the signed transaction and the identifier in all the original blocks.

[0019] Optionally, the transaction also includes a hash value, which is obtained by encapsulating the relevant data into a data object in a structured format, encrypting the data object, and hashing the encrypted data object; the smart contract module is used to: encapsulate the relevant data into a new data object in a structured format, encrypt the new data object, and hash the encrypted new data object to obtain a new hash value; in response to the new hash value being consistent with the hash value in the transaction, call the smart contract pre-configured for the first object. Optionally, the data chain module is used to: obtain the original relevant data of the second object within a preset time period according to the identifier of the second object, and process the original relevant data using a preset data processing strategy to obtain the relevant data.

[0020] Optionally, the original relevant data includes real-time operation data and operation result data; the data uploading module is also used to: process the real-time operation data and operation result data through a risk identification model to obtain risk data.

[0021] Optionally, the device also includes a smart contract generation module, which is used to: receive business logic data transmitted by the first object through one or more methods, and convert the business logic data into a feature space vector; convert the feature space vector into information in a computer-readable form, and generate smart contract rules based on the information in the computer-readable form; use a programming language program preset in the blockchain to generate a smart contract code corresponding to the smart contract rules; compile the smart contract code into bytecode for deployment to the blockchain, and assign a storage address to the smart contract and store the smart contract in the storage address.

[0022] Optionally, the smart contract generation module is used to: map business logic data of different modalities into the same feature space to obtain a feature space vector for each business logic data; wherein the feature space includes one or more feature dimensions, and the feature space vector is composed of feature values ​​under different feature dimensions; calculate the similarity between the feature values ​​of business logic data of different modalities under the same feature dimension, compare the similarity with a preset similarity threshold, and obtain a comparison result; use a fusion strategy corresponding to the comparison result to fuse the feature values ​​of business logic data of different modalities under the same feature dimension, and then integrate to obtain the feature space vectors of business logic data of all modalities.

[0023] Optionally, the smart contract generation module is used to: in response to the comparison result being greater than or equal to, select any one of the feature values ​​of similar different modal business logic data under the same feature dimension; in response to the comparison result being less than, sort and integrate the feature values ​​of dissimilar different modal business logic data under the same feature dimension.

[0024] Optionally, the analysis module is used to: extract specific information from the execution results using a pre-trained large language model; wherein, an original large language model corresponding to the first object is determined, and the original large language model is trained using the feature space vector to adjust the parameters of the original large language model to obtain the pre-trained large language model.

[0025] To achieve the above objectives, according to another aspect of an embodiment of the present invention, a blockchain-based data processing electronic device is provided.

[0026] The electronic device of an embodiment of the present invention includes: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement any of the above-mentioned blockchain-based data processing methods.

[0027] To achieve the above-mentioned purpose, according to another aspect of an embodiment of the present invention, a computer-readable medium is provided, on which a computer program is stored, and when the program is executed by a processor, any of the above-mentioned blockchain-based data processing methods is implemented.

[0028] To achieve the above-mentioned purpose, according to another aspect of an embodiment of the present invention, a computing program product is provided. A computing program product of an embodiment of the present invention includes a computer program, and when the program is executed by a processor, the blockchain-based data processing method provided by an embodiment of the present invention is implemented.

[0029] According to the solution provided by the present invention, one embodiment of the above invention has the following advantages or beneficial effects: Combining smart contract technology and decision-making engine, it aims to realize automated decision-making in the supply chain, solving the problem of low approval efficiency and high error rate caused by existing reliance on manual approval. Blockchain technology is used to ensure the security and immutability of data, which significantly enhances the security and transparency of data. Smart contracts automatically verify the validity of data, reduce manual intervention and improve efficiency. Through sentiment analysis of specific information, transaction intentions are identified to achieve accurate risk assessment. The analysis results are converted into feature vectors to accelerate the automated approval process and improve overall efficiency. The entire process is based on a decentralized blockchain network, which enhances trust among participants and provides an efficient, secure and reliable operating platform for the supply chain.

[0030] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The accompanying drawings are used to better understand the present invention and do not constitute an improper limitation of the present invention.

[0032] Figure 1This is a schematic diagram of the main process of a blockchain-based data processing method according to an embodiment of the present invention;

[0033] Figure 2 is a flowchart of an optional blockchain-based data processing method according to an embodiment of the present invention;

[0034] Figure 3 (a) is a schematic diagram of data import;

[0035] Figure 3 (b) is a schematic diagram of the data uploading process;

[0036] Figure 3 (c) is a schematic diagram of the data verification process;

[0037] Figure 4 is a flowchart of another optional blockchain-based data processing method according to an embodiment of the present invention;

[0038] Figure 5 This is a schematic diagram of the smart contract generation process;

[0039] Figure 6 is a flowchart of another optional blockchain-based data processing method according to an embodiment of the present invention;

[0040] Figure 7 is a flowchart of a data processing method based on blockchain according to an embodiment of the present invention;

[0041] Figure 8 This is a schematic diagram of main modules of a blockchain-based data processing device according to an embodiment of the present invention;

[0042] Fig. 9 is an exemplary system architecture diagram to which embodiments of the present invention may be applied;

[0043] Fig.10 It is a schematic diagram of the structure of a computer system of a mobile device or a server suitable for implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0044] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0045] It should be noted that the collection, collection, updating, analysis, processing, use, transmission, storage and other aspects of user personal information involved in the technical solution of this disclosure are in compliance with the provisions of relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. Necessary measures are taken for user personal information to prevent illegal access to user personal information data and maintain the security of user personal information, network security and national security.

[0046] The logic of this solution is applicable to all scenarios based on trust or contracts, such as the introduction, signing and withdrawal of carriers and suppliers in the process of investment promotion, which will involve the identification and approval of various types of data. They can all adopt the current process described in this solution, except that the parameters introduced are different, such as the introduction of basic data and data preprocessing process, but the overall idea is the same. This solution can solve the following problems existing in the existing technology:

[0047] 1. Data island problem. The existing data information comes from the internal system. In order to ensure data security, it is not connected with the upstream and downstream systems (i.e. upstream demanders and downstream service providers), which makes information sharing difficult and the approval process slow.

[0048] 2. Insufficient security. Paper documents and electronic data are easily tampered with and lost. The trust of downstream service providers is greatly reduced, making it more difficult to facilitate transactions.

[0049] 3. Manual operations are cumbersome, involving online, offline, internal and external approvals, which increases the risk of errors.

[0050] See also Figure 1 , which shows a main flow chart of a data processing method based on blockchain provided by an embodiment of the present invention, comprising the following steps:

[0051] S101: receiving a data verification request for a second object transmitted by a first object, and transmitting the data verification request to a blockchain, so that the blockchain obtains a transaction according to an identifier of the second object; wherein the transaction includes relevant data of the second object;

[0052] S102: calling a smart contract pre-configured for the first object, and in response to the relevant data satisfying a trigger condition in the smart contract, using the execution logic in the smart contract to process the relevant data to obtain an execution result;

[0053] S103: extracting specific information from the execution result, performing sentiment analysis on the specific information to identify the transaction intention, using a risk analysis strategy corresponding to the transaction intention, determining risk information of the specific information, and then generating an analysis result;

[0054] S104: Convert the analysis result into a feature vector, and transmit the feature vector to the automated approval node, so that the automated approval node processes the feature vector using the approval strategy, obtains the approval result, and returns it to the first object.

[0055] In the above implementation, for step S101, this solution is applied to the automated approval system of the supply chain, and the system is divided into the front end and the back end, and the back end is the blockchain. Blockchain: A distributed ledger technology that can ensure the immutability and transparency of data. The automated approval system is located between the upstream system and the downstream system. The upstream system is mainly used to serve the upstream demand side, and the downstream system is mainly used to serve the downstream service side. For example, the upstream is a financing customer and the downstream is a service customer. It is determined whether the information of the upstream financing customer meets the requirements of the downstream service customer. If it meets the requirements, the upstream financing customer can sign a contract with the downstream service customer.

[0056] This is just an example. This solution can be applied to scenarios where it is necessary to determine whether the upstream and downstream are matched. The first object proposed can be a downstream service provider or an upstream demand provider, and the second object is also not limited. However, the first object and the second object corresponding to the data verification request are different parties. This solution takes the first object as the downstream service provider X and the second object as the upstream demand provider Y as an example for explanation.

[0057] When the downstream service provider X needs to verify data, such as multimodal processing, decision engines, or financial institutions approving financing, a data verification request can be sent to the blockchain network. Since there are usually multiple upstream demand parties, the request needs to specify a specific upstream demand party, such as the aforementioned upstream demand party Y. The automated approval system receives the data verification request for the upstream demand party Y transmitted by the downstream service provider X, and retrieves the transaction corresponding to the identifier (such as ID, user name, identity information, etc.) of the upstream demand party Y from the block of the blockchain network according to the request. The transaction here refers to the relevant data of the upstream demand party Y over a period of time, such as transaction data, operation results, analysis results, and auxiliary data in the business link.

[0058] For step S102, after obtaining the relevant data in the transaction, call the smart contract pre-configured for the downstream service provider. The smart contract configuration process is described in the following section. Figure 4 As shown in the description. Smart contracts include trigger conditions and execution logic to determine whether the relevant data of the upstream demander meets the trigger condition. For example, if the trigger condition is "when the goods arrive at the destination and are signed for, payment will be automatically executed", assuming that the current relevant data includes the delivery status of the goods as successfully signed for, it means that the trigger condition is met and the execution logic of the smart contract is automatically executed. For example, when the logistics information in the relevant data shows that the goods have been delivered, the contract automatically triggers the payment function and transfers the payment to the merchant.

[0059] For step S103, the execution result of the smart contract will be recorded on the blockchain, and relevant parties such as suppliers, financial institutions, buyers, etc., such as downstream service provider X and upstream demander Y, can be notified through events or logs. It can also include staff of the automated approval system. Only the result record with a successful execution result is passed to the LLM (Large Language Model).

[0060] LLM is used to extract key information (i.e. specific information) from the execution results, such as key terms, urgency of the transaction, potential risk factors, etc., which can be extracted by keywords. Using LLM for deep learning and feature extraction can achieve a comprehensive understanding and analysis of the data. This solution takes into account the different characteristics of different downstream service providers. In order to avoid the use of a general LLM that causes inaccurate information extraction about the downstream service provider or extracts some unnecessary information, it is preferred to set up an LLM for each downstream service provider.

[0061] For the processing of feature space vectors, see the following Figure 4~Figure 6 Here, the pre-trained LLM is fine-tuned on the feature space vector representation after fusion by the downstream service provider X, in order to make the model better adaptable to specific application scenarios, such as contract review and risk assessment in supply chain finance. Fine-tuning helps the model learn features and patterns related to specific tasks.

[0062] The steps of the fine-tuning process are as follows, taking the upstream demand side as supply chain finance as an example:

[0063] 1. Data preparation: Collect data related to supply chain finance, including but not limited to contract texts, transaction records, market analysis reports, etc.

[0064] 2. Feature fusion: Integrate data of different modalities (such as text, voice, and image) into a unified feature space through feature extraction and fusion technology to obtain a feature space vector.

[0065] 3. Pre-trained model selection: Select a suitable pre-trained large language model that has been trained on a large amount of data and has good language understanding capabilities.

[0066] 4. Fine-tuning training: Retraining the pre-trained model using the feature space vectors. This process usually involves adjusting the parameters of the model so that it can capture specific information and context related to supply chain finance.

[0067] 5. Model optimization: During the fine-tuning process, optimization algorithms such as gradient descent are used to minimize the loss function of the model on a specific task and improve the prediction accuracy of the model.

[0068] 6. Evaluation and Iteration: Evaluate the model after fine-tuning to check its performance on the supply chain task. Depending on the evaluation results, further adjustments and iterations of the fine-tuning process may be required.

[0069] Then, sentiment analysis is performed on key information, such as using natural language processing technology to analyze the emotional tendency in the text, judging the overall tendency based on the sentiment polarity (positive or negative) and intensity of the words, and analyzing key information to identify transaction intentions. Machine learning models can be used to improve recognition accuracy. Use risk analysis methods corresponding to transaction intentions to determine the risk level of specific information, such as assessing the risk level of a transaction or contract.

[0070] For example, the smart contract involves a large loan transaction. The key information extracted includes: loan amount: 500,000, repayment period: 24 months, interest rate: 6%, collateral: real estate, borrower credit score: 720, and borrower comments. Sentiment analysis and risk assessment of these key information show that the sentiment tendency is positive, and the transaction intention is to obtain a loan to buy a new house. The risk assessment results are: Credit score: high (720), indicating that the borrower has a good repayment record. Collateral value: The value of the property exceeds the loan amount, reducing the risk of default. Interest rate: 6% is at the market average and will not add additional burden. Repayment period: 24 months, the time is moderate, and the borrower has enough buffer period to arrange repayment. Taking all the above factors into consideration, the risk level of this transaction is low. Through sentiment analysis of key information and transaction intention identification, combined with specific transaction details, the risk level of the transaction can be effectively assessed.

[0071] For step S104, the analysis results are converted into feature vectors that can be used by the decision engine, including sentiment analysis results, intent recognition, key information extraction, and risk assessment. The rule engine is designed to handle structured decisions, such as automatic approval based on contract terms and risk assessment. The feature vector is transmitted to the automated approval node, so that the automated approval node processes the feature vector through the rule engine to obtain the approval result, such as the final decision, whether the upstream demand party Y meets the requirements of the downstream service party X, credit score, transaction approval status, risk warning, etc. The approval results and result records are recorded on the blockchain to increase transparency and traceability. The approval results and result records are fed back to the downstream service party X, financial institutions, or other participants in the supply chain.

[0072] The method provided in the above embodiment combines smart contract technology and decision engine to achieve automated decision making in the supply chain. Its core advantage is to use blockchain technology to verify and record key data, ensure the immutability and integrity of data, and thus build a transparent and reliable operating environment.

[0073] First, blockchain technology ensures the security and immutability of the data of the second object during transmission, and its distributed characteristics improve the security of data storage, thereby significantly enhancing the security and transparency of data. Secondly, the use of smart contracts to automatically verify the validity of data reduces the need for manual intervention and improves efficiency. At the same time, smart contracts can quickly respond to data changes and achieve instant processing, further improving the level of automation.

[0074] In addition, by identifying transaction intentions through sentiment analysis of specific information, potential risks can be assessed more accurately, and corresponding risk analysis strategies can be adopted for different transaction intentions, which improves the pertinence and accuracy of risk assessment. The analysis results are converted into feature vectors to facilitate processing by the automated approval system. The application of automated approval nodes greatly speeds up decision-making, shortens the approval cycle, and improves the efficiency of the overall process.

[0075] See also Figure 2 , shows a schematic flow chart of an optional blockchain-based data processing method according to an embodiment of the present invention, comprising the following steps:

[0076] S201: acquiring the relevant data of the second object within a preset time period according to the identifier of the second object;

[0077] S202: Packing the relevant data into a transaction, and broadcasting the transaction and the identifier of the second object to the blockchain;

[0078] S203: receiving a data verification request for a second object transmitted by the first object, and transmitting the data verification request to the blockchain, so that the blockchain obtains a transaction according to the identifier of the second object; wherein the transaction includes relevant data of the second object;

[0079] S204: calling a smart contract pre-configured for the first object, and in response to the relevant data satisfying a trigger condition in the smart contract, using the execution logic in the smart contract to process the relevant data to obtain an execution result;

[0080] S205: extracting specific information from the execution result, performing sentiment analysis on the specific information to identify the transaction intention, using a risk analysis strategy corresponding to the transaction intention, determining risk information of the specific information, and then generating an analysis result;

[0081] S206: Convert the analysis result into a feature vector, and transmit the feature vector to the automated approval node, so that the approval node processes the feature vector using the approval strategy, obtains the approval result, and returns it to the first object.

[0082] In the above implementation, for steps S204 to S206, see Figure 1 The description shown is not repeated here.

[0083] For step S201, in the supply chain's automated approval system, the data input node is the foundation of the entire system, responsible for collecting and integrating data from all links of the supply chain, specifically, the relevant data of each upstream demand party, such as the relevant data of upstream demand party Y in the past three days, or the relevant data in the past month. The data input interface is shown in Figure 3 (a), taking database import as an example. Here, taking supply chain finance as an example, the three-level division of supply chain data and the core data types contained in each level are shown.

[0084] 1. The first level corresponds to the business perspective. Business represents results. Through the process of reverse query of results, it is called business result data, such as 1) Order information: including order number, product details, quantity, price, delivery date, etc. 2) Invoice data: invoice number, invoice date, amount, payment terms, etc. 3) Payment records: payer, payee, payment amount, payment date, payment method, etc. 4) Contract documents: contract terms, price terms, delivery terms, breach of contract liability, etc.

[0085]

[0086] 2. The second level corresponds to the operational perspective. Based on the results, you can check the data of the operational process, which is called real-time operational data. Taking the logistics scenario as an example, real-time operational data includes logistics and warehousing data. For example: 1) Cargo tracking information: transportation tools, cargo status, estimated arrival time, actual arrival time, etc. 2) Inventory level: inventory volume of each warehouse, inventory turnover rate, safety stock level, etc. 3) Warehouse management records: warehousing records, outbound records, inventory counts, damage and scrap records, etc. 4) Transportation contract: transportation service terms, fees, responsibilities, insurance, etc.

[0087]

[0088] 3. The third layer is based on the first and second layers, through the risk identification model, combined with the conclusions of data analysis, for advanced analysis and auxiliary data. For example, based on real-time operation data and business results data, risk data is generated. Such as 1) Market analysis report: industry trends, market demand forecasts, competitor analysis, etc. 2) Credit score data: corporate credit score based on historical transactions and payment behaviors. 3) Compliance inspection results: anti-money laundering inspections, sanctions list screening, tax compliance, etc. 4) Risk assessment report: assessment results based on various risk models, such as credit risk, market risk, operational risk, etc.

[0089]

[0090] After obtaining the above data, the original relevant data can be preprocessed in advance. Different levels have different data processing methods, so the data processing methods for different levels are different. Taking the first-level contract document as an example, text processing: segment the contract text, remove stop words, extract stems or restore the word form. Feature extraction: extract key terms in the contract, such as delivery period, breach of contract liability, etc. Taking the second level of inventory level as an example, data verification: check whether the inventory level is within a reasonable range to avoid negative inventory or unreasonably high inventory. Data specification: The inventory level may need to be normalized for analysis. Taking the third-level credit score data as an example, data cleaning: check the distribution of credit scores and remove abnormal or erroneous scores. Data normalization: convert credit scores into a unified scoring standard.

[0091] For step S202, after obtaining the relevant data pre-processed by the upstream demand party Y, the data verification stage is entered. The data verification stage is implemented based on blockchain technology, mainly through two core capabilities: data chaining and data verification. Here we only explain data chaining. For data verification, see Figure 1 Description shown.

[0092] In the first embodiment, the data on-chain stage includes the steps of creating transactions and packaging transactions into blocks. 1) Create a transaction, package the aforementioned related data into a transaction, and send it to the blockchain network. Packaging into transactions is to ensure the security and integrity of the data, and the data is finally packaged and encrypted before verification and recording on the blockchain. In this scheme, it is preferred to package the relevant data about the upstream demand party Y obtained this time into a transaction. Assuming that the original relevant data of 10 upstream demand parties is received on the same day, 10 transactions need to be generated after processing. 2) Package the transaction into a block, and the blockchain can package the transaction and the identifier into any block.

[0093] In Example 2, the data chaining stage includes steps such as creating transactions, packaging transactions into new blocks, and adding to the blockchain. 1) Create a transaction, package the aforementioned related data into a transaction (Transaction), and send it to the blockchain network. 2) Package the transaction into a new block, the blockchain packages the transaction and the identifier into a new block, and calculates the hash value of the new block based on the transaction and the identifier of the upstream demander Y. 3) Add to the blockchain, add the new block to the blockchain, and form an unalterable chain structure.

[0094] In the third embodiment, the data on-chain stage includes the steps of creating transactions, signing transactions, broadcasting transactions, and block verification.

[0095] 1) Create a transaction and package the aforementioned related data into a transaction.

[0096] 2) Sign the transaction, use the private key of the upstream demander Y to sign the transaction to ensure the legitimacy of the initiator's identity. The private key here is pre-registered by the upstream demander Y in the automatic audit system, and can be maintained in advance to the front end of the automatic audit system. If the private key corresponding to the upstream demander Y's identity is not found, the private key registration interface is displayed to automatically generate a private key for the upstream demander Y based on the identity information entered and save it to the front end.

[0097] 3) Broadcast transactions, broadcasting signed transactions to each block in the blockchain network;

[0098] 4) Block verification: After receiving the signed transaction, each block in the blockchain network uses the public key of the upstream demander Y to verify the validity of the transaction signature. Here, the public key of the upstream demander Y also corresponds to its identity and is pre-stored in the front end. Private keys and public keys are usually generated in pairs, so when the upstream demander Y registers the private key, a private key and a paired public key will be automatically generated for it. The design of the blockchain network allows nodes to independently obtain and use public keys when verifying transactions, without pre-storing all possible public keys. This approach maintains the flexibility and scalability of the system.

[0099] In the fourth embodiment, the data on-chain stage includes steps such as data encapsulation, data encryption, hash value generation, transaction creation, transaction signing, transaction broadcasting, and block verification.

[0100] 1) Data encapsulation, which is used to encapsulate the above-mentioned related data into a structured format (JSON object), including contract documents, inventory data, credit data, etc.

[0101] 2) Data encryption, which is used to encrypt the encapsulated data objects to ensure the security of the data during transmission. Symmetric encryption algorithms such as AES (Advanced Encryption Standard) are preferably used.

[0102] 3) Generate a hash value for the encrypted data object, preferably using the SHA-256 (SecureHash Algorithm 256-bit) algorithm to ensure the integrity and uniqueness of the data.

[0103] 4) Create a transaction, package the aforementioned related data and hash value into a transaction, and send it to the blockchain network.

[0104] 5) Sign the transaction using the private key of the upstream demander Y to ensure the legitimacy of the initiator of the transaction.

[0105] 6) Broadcast transactions, broadcast signed transactions to each block in the blockchain network;

[0106] 7) Block verification: After receiving the signed transaction, each block in the blockchain network uses the public key of the upstream demander Y to verify the validity of the transaction signature.

[0107] In the fifth embodiment, the data on-chain stage includes steps such as data encapsulation, data encryption, transaction creation, hash value generation, transaction signature, transaction broadcasting, block verification, transaction packaging to a new block, and adding to the blockchain, as shown in FIG3 (b).

[0108] 1) Data encapsulation, which is used to encapsulate the above-mentioned related data into a structured format (JSON object), including contract documents, inventory data, credit data, etc.

[0109] 2) Data encryption, which is used to encrypt the encapsulated data objects to ensure the security of the data during transmission. It is preferred to use symmetric encryption algorithms such as AES.

[0110] 3) Generate a hash value for the encrypted data object, preferably using the SHA-256 algorithm to ensure the integrity and uniqueness of the data.

[0111] 4) Create a transaction and package the aforementioned related data and hash value into a transaction.

[0112] 5) Sign the transaction using the private key of the upstream demander Y to ensure the legitimacy of the initiator of the transaction.

[0113] 6) Broadcast transactions, broadcast signed transactions to each block in the blockchain network.

[0114] 7) Block verification: After receiving the signed transaction, each block in the blockchain network uses the public key of the upstream demander Y to verify the validity of the transaction signature.

[0115] 8) Package the transaction into a block. If the signature verification is successful, the transaction is considered valid. The blockchain packages the verified signed transaction and identifier into a new block, and calculates the hash value of the new block based on the signed transaction and the identifier of the upstream demander Y.

[0116] 9) Add to the blockchain, add the new block to the blockchain to form an unalterable chain structure.

[0117] For step S203, after obtaining the transaction or signed transaction from the block of the blockchain, verification is required. Specifically, the signed transaction is verified. For details, see Figure 3 (c). The public key can be determined according to the identification of the upstream demand party Y, or the signed transaction can be decrypted using the universal authorization key to obtain the original transaction, that is, the pre-processed related data and hash value. Then enter the hash value comparison stage, recalculate the hash value of the decrypted related data, and compare it with the hash value stored in the transaction. If the hash value matches, it proves that the data has not been tampered with after being uploaded to the chain, and the verification is passed. Only when the verification is passed, the smart contract of the downstream service party X is called. However, if the hash value does not match, it is necessary to trigger the exception handling mechanism to notify the relevant parties (such as the downstream service party), suspend the transaction, etc.

[0118] The method provided in the above embodiment packages the relevant data of the second object into a transaction form and stores it on the chain, so as to ensure the security and non-tamperability of the data through blockchain technology.

[0119] See also Figure 4 , shows a schematic flow chart of another optional blockchain-based data processing method according to an embodiment of the present invention, comprising the following steps:

[0120] S401: receiving business logic data transmitted by a first object in one or more ways, and converting the business logic data into a feature space vector;

[0121] S402: Convert the feature space vector into information in a computer-readable form, and generate smart contract rules according to the information in the computer-readable form;

[0122] S403: Generate a smart contract code corresponding to the smart contract rules using a programming language program preset in the blockchain;

[0123] S404: Compile the smart contract code into bytecode to be deployed into the blockchain, assign a storage address to the smart contract, and store the smart contract into the storage address.

[0124] In the above implementation, usually the smart contract is set for the party that needs to use it. For example, if the downstream service provider X extracts the request, the smart contract of the downstream service provider X is used. If the upstream demander Y makes a demand, the smart contract of the upstream demander Y is used. Here, how to generate a smart contract is described, taking the generation of the smart contract of the downstream service provider X as an example.

[0125] For step S401, the downstream service provider Y transmits the business logic data to the automated approval system in one or more ways. The ways here are mainly used to limit the form of business logic data, such as text data: contracts, reports, communication records, etc., voice data: customer service calls, etc., image data: product pictures, signatures, contract scans, etc. The downstream service provider Y can explain its business logic in the form of text, voice, pictures, etc., and the business logic data in different ways may overlap or not overlap, such as text represents event 1, voice represents event 2, and image represents event 3. This solution first needs to use a model such as CLIP (Contrastive Language-Image Pre-training) to convert these business logic data into feature space vectors.

[0126] Multimodal processing: A technology that can process and analyze multiple types of data simultaneously. The CLIP model can map data of different modalities (text, speech, image) to a common feature space. For example, CNN (Convolutional Neural Network) is used for feature extraction of image data, and ASR (Automatic Speech Recognition) technology is used for transcription of speech data. These different modal data are then converted into a unified feature representation so that the model can be jointly analyzed. For details, see the following Figure 6 Description shown.

[0127] For steps S402~S404, according to the feature space vector, it is converted into computer-recognizable business logic, and the terms of the smart contract are designed, including payment conditions: defining when the payment behavior is triggered and the specific amount of payment, etc. Delivery deadline: clarify the time requirements for delivery of goods. Liability for breach of contract: stipulate the liability and penalties for breach of contract terms. Then write the smart contract code in the programming language supported by the blockchain platform (Solidity is used for Ethereum), clearly define the triggering conditions and execution logic. Deploy the written smart contract to the blockchain network. When deployed, the smart contract will be compiled into bytecode and associated with a specific address on the blockchain. The overall process can be seen in Figure 5 shown.

[0128] In the blockchain, an address is a unique identifier for a smart contract and is used to identify and access it. It is similar to an email address and allows users to send transactions to a smart contract to invoke its functions. The address is generated when the smart contract is deployed and is used when they need to interact with the smart contract. For example, when a user wants to trigger a function in a smart contract or execute a rule in the contract, they send a transaction to this address. In addition, there are no pre-set addresses on the blockchain. Each smart contract is compiled into bytecode when it is deployed and automatically assigned a unique address. This address is randomly generated to ensure that each smart contract has a unique identifier. Different smart contracts are associated with different addresses.

[0129] In addition, when designing smart contracts, it is necessary to consider possible future upgrades and maintenance requirements to ensure the sustainability and flexibility of the contract. Before the deployment of smart contracts, a thorough security audit is required to check for potential security vulnerabilities and logical errors, and to ensure that the logic and execution of smart contracts comply with relevant laws, regulations and industry standards. Suppose there is a lending agreement in a decentralized financial application. The agreement has undergone a comprehensive security audit before deployment, including but not limited to: Code review: Check whether the code has known security vulnerabilities, such as reentrancy attacks, overflow errors, etc. Logic verification: Ensure that the business logic is correct, such as interest rate calculation, collateral processing, etc. Compliance check: Confirm that all operations comply with local financial regulatory requirements. Stress testing: Simulate system performance under extreme conditions to ensure that smart contracts can run stably.

[0130] The method provided in the above embodiment converts business logic data into feature space vectors, which can adapt to different business scenario requirements. The smart contract rules are automatically generated through computer-readable information, which simplifies the smart contract development process. The smart contract code is generated using the programming language program preset by the blockchain, ensuring the standardization and compatibility of the code. The smart contract code is compiled into bytecode and deployed to the blockchain, which improves the deployment efficiency. At the same time, a storage address is assigned to the smart contract for easy management and tracking.

[0131] See also Figure 6 , shows a schematic flow chart of another optional blockchain-based data processing method according to an embodiment of the present invention, comprising the following steps:

[0132] S601: Mapping business logic data of different modes into the same feature space to obtain a feature space vector of each business logic data; wherein the feature space includes one or more feature dimensions, and the feature space vector is composed of feature values ​​under different feature dimensions;

[0133] S602: Calculate the similarity between the feature values ​​of different modal business logic data under the same feature dimension, and compare the similarity with a preset similarity threshold to obtain a comparison result;

[0134] S603: Using a fusion strategy corresponding to the comparison result, the feature values ​​of different modal business logic data under the same feature dimension are merged, and then the feature space vectors of all modal business logic data are integrated.

[0135] In the above implementation, for step S601, after obtaining the business logic data of different modalities, different processing methods can be used for preprocessing first, such as text: word segmentation, standardization, and removal of stop words. Voice: transcribe into text using ASR technology. Image: extract visual features using models such as CNN. Or text: use NLP (Natural Language Processing) technology to extract key information, such as payment terms, delivery date, etc. Image: If the contract contains a scanned copy, use OCR technology to identify the text content therein, and then combine NLP technology to extract key information. Voice: For customer service call recordings, first use ASR technology to transcribe them into text, and then extract key information through NLP technology.

[0136] Modality-specific key information: such as key terms in text, key discussion points in speech, product quality status in images, etc. Use CLIP or other multimodal learning frameworks to merge business logic data of different modalities into a unified feature space. Unified features do not mean that the features of all modalities are exactly the same, but map data of different modalities to a common feature space, so that data from different sources can be represented by the same feature dimension. Assuming that there are dimensions 1, 2, 3, ..., n (n ≥ 1) in the feature space, it is necessary to extract the feature values ​​corresponding to the dimensions from the business logic data of each modality, and then generate feature space vectors in the order of dimension 1 to dimension n. For example, if the downstream service provider X has business logic in text form and business logic in voice form for price increase, it is necessary to calculate the feature space vector of the text and the feature space vector of the voice.

[0137] For step S602, after calculating the feature space vector of each modal business logic data, enter the comparative learning stage, and enhance the model's understanding of the data by comparing the similarities and differences between the feature space vectors of different modalities. Contrastive learning refers to finding the similarities and differences between them by comparing the feature values ​​between different modal data. This is not a comparison of different features under the same modality, but a comparison of the same features across modalities. This is achieved by calculating the distance or similarity between the feature values ​​of different modalities, for example, using metrics such as cosine similarity or Euclidean distance.

[0138] For step S603, the comparison result is nothing more than similarity or similarity. It is determined to be similar when the similarity is greater than or equal to the preset similarity threshold, otherwise it is dissimilar and there are differences. Here, the fusion feature vector of the model can be used to represent the features fused after the conversion of different modal data. Identifying similarities and differences can help the model better understand the intrinsic connections and structures of the data. For example, in multimodal learning, similarities can be used to strengthen the model's learning of consistency information, while differences can be used to distinguish information specific to different modalities. By understanding the similarities and differences between different data modalities, the model can more accurately capture the key information in the data, thereby improving the ability to classify, identify and predict the data.

[0139] In the case of similarity, select one of the feature values ​​of similar business logic data of different modalities under the same feature dimension; in the case of dissimilarity, sort and integrate the feature values ​​of dissimilar business logic data of different modalities under the same feature dimension. Assuming that there are feature values ​​A1, A2 and A3 under dimension 1, and it is calculated that all three are similar, select one of them. Assuming A1 is selected, A2 is used as the final feature value of dimension 1. Or if A1 and A2 are similar, select one of them. Assuming A1 is selected, calculate the similarity between A1 and A3. If it is less than the threshold, A1+A3 is used as the final feature value of dimension 1.

[0140] The method provided in the above embodiment can integrate and process business logic data from different sources and types, including text, voice, and images, etc., and use multimodal processing technology to map business logic data of different modalities into the same feature space for easy comparison and fusion. The appropriate fusion strategy is selected according to the similarity between the feature values, thereby improving the flexibility and accuracy of data processing. This operation significantly improves the insight into complex scenarios in the supply chain and provides richer and more accurate data support for the subsequent generation of smart contracts.

[0141] See also Figure 7 , shows a flowchart of a data processing method based on blockchain according to an embodiment of the present invention, including the following steps:

[0142] 1. The second object’s data is uploaded to the blockchain

[0143] Data input: Collect data from the second object, including order information, transaction records, logistics information, etc. Data preprocessing: Before the data enters the big model for analysis, preprocessing steps such as cleaning, standardization, and formatting are performed to obtain relevant data.

[0144] The relevant data is packaged into a transaction, and the transaction and the identifier of the second object are broadcast to the blockchain. Here, the relevant data can also be signed and encrypted to generate a hash value, and then packaged into a signed transaction and broadcast to the blockchain. The blockchain can store transactions (or signed transactions) in an existing block, or create a new block and store them in the new block.

[0145] 2. The first object is the smart contract generation phase, which is used to deploy smart contracts on the blockchain

[0146] Receive business logic data transmitted by the first object in one or more ways, including multiple data types such as text, voice, and image. Convert the business logic data into a feature space vector through the CLTP model; convert the feature space vector into information in a computer-readable form, and generate smart contract rules according to the information in the computer-readable form; use a programming language program preset in the blockchain to generate smart contract code corresponding to the smart contract rules; compile the smart contract code into bytecode to be deployed in the blockchain, and assign a storage address to the smart contract, and store the smart contract in the storage address.

[0147] 3. The first object requests the data verification stage of the second object to ensure the authenticity and integrity of the data, which is one of the core advantages of blockchain technology.

[0148] Receive a data verification request for a second object transmitted by the first object, transmit the data verification request to the blockchain, so that the blockchain obtains the transaction according to the identifier of the second object; call a smart contract pre-configured for the first object, and the relevant data meets the triggering condition in the smart contract, and use the execution logic in the smart contract to process the relevant data to obtain an execution result;

[0149] Extract specific information from the execution results, conduct sentiment analysis on the specific information to identify transaction intentions, use risk analysis strategies corresponding to transaction intentions, determine the risk information of specific information, and then generate analysis results; for example, through LLM, extract valuable information from the above unstructured data and provide decision-making judgments on financing conditions.

[0150] Based on the judgment results of LLM, the decision engine (a system component that automatically completes decision-making based on data analysis results) automatically makes approval decisions. Combined with the execution results of smart contracts, more accurate and automated decision suggestions are provided. Finally, the approval results are recorded on the blockchain to ensure the transparency and traceability of the results. The approval results are notified to the relevant parties and their feedback is collected. The relevant parties can execute: the financing service provider makes loans, the financing demander collects payments, etc.

[0151] This solution can reduce costs and increase revenue at the same time. The specific effects are as follows

[0152] 1. Savings on manual review and operating costs. By introducing automated approval processes and large-scale language model analysis, the reliance on manual review is reduced, and the costs associated with human resources are reduced.

[0153] 1) Economic benefits: Automation reduces approval time, reduces error rates, and improves efficiency. This leads to direct labor cost savings, including employee wages, training costs, office and management costs. Due to reduced manual intervention, the cost of rework and error correction caused by human errors is also reduced.

[0154] 2) Cost savings: For example, originally 10 employees were required to conduct audit work, but after automation, only 2 employees are needed to supervise and handle abnormal situations, thus achieving significant cost savings. Based on the current labor cost of around 2,000 per day, 4.17 million can be saved throughout the year.

[0155] 2. Expand downstream institutional partners through blockchain technology. Blockchain technology provides a transparent, secure, and tamper-proof data sharing platform, which increases the trust of downstream institutions and encourages them to establish cooperative relationships with upstream demand parties.

[0156] 1) Economic benefits: Taking the financial industry as an example, as the number of cooperative financial institutions increases, as a financial service provider, it can improve its bargaining power through economies of scale. This not only helps to attract more sources of funds, but also helps to obtain a more favorable profit-sharing ratio in cooperation with financial institutions.

[0157] 2) Cost savings: Through blockchain technology, upstream demand parties can prove the reliability and efficiency of their services to more financial institutions, thereby adding new financial institution partners on the basis of existing financial institutions. The profit sharing brought by each new partner, plus the additional benefits brought by the improvement of bargaining power.

[0158] See also Figure 8, showing a schematic diagram of the main modules of a blockchain-based data processing device 800 provided by an embodiment of the present invention, including: a transaction acquisition module 801, used to receive a data verification request for a second object transmitted by a first object, and transmit the data verification request to the blockchain, so that the blockchain obtains the transaction according to the identifier of the second object; wherein the transaction includes relevant data of the second object; a smart contract module 802, used to call a smart contract pre-configured for the first object, and in response to the relevant data satisfying the trigger condition in the smart contract, use the execution logic in the smart contract to process the relevant data to obtain an execution result; an analysis module 803, used to extract specific information from the execution result, perform sentiment analysis on the specific information to identify the transaction intention, use the risk analysis strategy corresponding to the transaction intention, determine the risk information of the specific information, and then generate an analysis result; an approval module 804, used to convert the analysis result into a feature vector, and transmit the feature vector to an automated approval node, so that the automated approval node uses the approval strategy to process the feature vector, obtain the approval result and return it to the first object.

[0159] The implementation device of the present invention also includes a data chain module, which is used to: obtain the relevant data of the second object within a preset time period according to the identifier of the second object; package the relevant data into a transaction, and broadcast the transaction and the identifier of the second object to the blockchain.

[0160] In the implementation device of the present invention, the transaction is obtained by decrypting the signed transaction, and the data on-chain module is used to: obtain a private key pre-registered by the second object, use the private key to sign the transaction to obtain the signed transaction, and broadcast the signed transaction and the identifier to each block of the blockchain; the acquisition of the transaction according to the identifier of the second object includes: retrieving the corresponding block according to the identifier, obtaining the signed transaction from the retrieved block, obtaining the public key corresponding to the identifier or obtaining the general authorization key, and using the public key or the authorization key to decrypt the signed transaction to obtain the transaction.

[0161] In the implementation device of the present invention, the data chain module is used to: broadcast the signed transaction and the identifier to each block of the blockchain, so that each block obtains the public key corresponding to the identifier and uses the public key to verify the signature; in response to one or more blocks passing the signature verification, create a new block in the blockchain, package the signed transaction and the identifier into the new block, add the new block to the blockchain, and delete the signed transaction and the identifier in all the original blocks.

[0162] In the implementation device of the present invention, the transaction also includes a hash value, which is obtained by encapsulating the relevant data into a data object in a structured format, encrypting the data object, and hashing the encrypted data object; the smart contract module 802 is used to: encapsulate the relevant data into a new data object in a structured format, encrypt the new data object, and hash the encrypted new data object to obtain a new hash value; in response to the new hash value being consistent with the hash value in the transaction, call the smart contract pre-configured for the first object.

[0163] In the implementation device of the present invention, the data uplink module is used to: obtain the original relevant data of the second object within a preset time period according to the identifier of the second object, and process the original relevant data using a preset data processing strategy to obtain the relevant data.

[0164] In the implementation device of the present invention, the original relevant data includes real-time operation data and operation result data; the data chain module is also used to: process the real-time operation data and operation result data through a risk identification model to obtain risk data.

[0165] The implementation device of the present invention also includes a smart contract generation module, which is used to: receive business logic data transmitted by the first object through one or more methods, and convert the business logic data into a feature space vector; convert the feature space vector into information in a computer-readable form, and generate smart contract rules based on the information in the computer-readable form; use a programming language program preset in the blockchain to generate a smart contract code corresponding to the smart contract rules; compile the smart contract code into bytecode for deployment to the blockchain, and assign a storage address to the smart contract, and store the smart contract in the storage address.

[0166] In the implementation device of the present invention, the smart contract generation module is used to: map business logic data of different modalities into the same feature space to obtain a feature space vector for each business logic data; wherein the feature space includes one or more feature dimensions, and the feature space vector is composed of feature values ​​under different feature dimensions; calculate the similarity between the feature values ​​of business logic data of different modalities under the same feature dimension, compare the similarity with a preset similarity threshold, and obtain a comparison result; use a fusion strategy corresponding to the comparison result to fuse the feature values ​​of business logic data of different modalities under the same feature dimension, and then integrate to obtain the feature space vectors of business logic data of all modalities.

[0167] In the implementation device of the present invention, the smart contract generation module is used to: in response to the comparison result being greater than or equal to, select any one from the feature values ​​of similar different modal business logic data under the same feature dimension; in response to the comparison result being less than, sort and integrate the feature values ​​of dissimilar different modal business logic data under the same feature dimension.

[0168] In the implementation device of the present invention, the analysis module 803 is used to: extract specific information from the execution result using a pre-trained large language model; wherein, the original large language model corresponding to the first object is determined, and the original large language model is trained using the feature space vector to adjust the parameters of the original large language model to obtain the pre-trained large language model.

[0169] In addition, the specific implementation content of the device described in the embodiment of the present invention has been described in detail in the method described above, so the repeated content will not be described again here.

[0170] Fig. 9 An exemplary system architecture 900 to which embodiments of the present invention may be applied is shown, including terminal devices 901 , 902 , 903 , a network 904 and a server 905 (only an example).

[0171] Terminal devices 901, 902, 903 can be various electronic devices with display screens and supporting web browsing, and various communication client applications are installed. Users can use terminal devices 901, 902, 903 to interact with server 905 through network 904 to receive or send messages, etc.

[0172] The network 904 is used to provide a medium for communication links between the terminal devices 901, 902, 903 and the server 905. The network 904 may include various connection types, such as wired, wireless communication links or optical fiber cables.

[0173] The server 905 may be a server that provides various services, such as a backend management server that provides support for shopping websites browsed by users using terminal devices 501, 502, and 503 (for example only). The backend management server may analyze and process the received data such as product information query requests, and feed back the processing results (such as target push information, product information - for example only) to the terminal device. It should be noted that the method provided in the embodiment of the present invention is generally executed by the server 905, and accordingly, the device is generally set in the server 905.

[0174] It should be understood that Fig. 9 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to the implementation requirements.

[0175] Reference below Fig.10 , which shows a schematic diagram of the structure of a computer system 1000 of a terminal device suitable for implementing an embodiment of the present invention. Fig.10 The terminal device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0176] like Fig.10 As shown, the computer system 1000 includes a central processing unit (CPU) 1001, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage part 1008 into a random access memory (RAM) 1003. In the RAM 1003, various programs and data required for the operation of the system 1000 are also stored. The CPU 1001, the ROM 1002, and the RAM 1003 are connected to each other through a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0177] The following components are connected to the I / O interface 1005: an input section 1006 including a keyboard, a mouse, etc.; an output section 1007 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 1008 including a hard disk, etc.; and a communication section 1009 including a network interface card such as a LAN card, a modem, etc. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to the I / O interface 1005 as needed. A removable medium 1011, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 1010 as needed, so that a computer program read therefrom is installed into the storage section 1008 as needed.

[0178] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 1009, and / or installed from the removable medium 1011. When the computer program is executed by the central processing unit (CPU) 1001, the above-mentioned functions defined in the system of the present invention are executed.

[0179] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0180] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0181] The modules involved in the embodiments of the present invention may be implemented in software or hardware. The modules described may also be set in a processor. For example, they may be described as follows: a processor includes a transaction acquisition module, a smart contract module, an analysis module, and an approval module. The names of these modules do not, in some cases, constitute limitations on the modules themselves. For example, the approval module may also be described as an "automatic approval module."

[0182] As another aspect, the present invention further provides a computer-readable medium, which may be included in the device described in the above embodiment; or may exist independently without being assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by a device, the device executes any of the above-described blockchain-based data processing methods.

[0183] The computer program product of the present invention includes a computer program, which, when executed by a processor, implements the blockchain-based data processing method in an embodiment of the present invention.

[0184] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A data processing method based on blockchain, characterized in that: include: Receiving a data verification request for a second object transmitted by a first object, and transmitting the data verification request to a blockchain, so that the blockchain obtains a transaction according to an identifier of the second object; wherein the transaction includes relevant data of the second object; Calling a smart contract pre-configured for the first object, and in response to the relevant data satisfying a trigger condition in the smart contract, using the execution logic in the smart contract to process the relevant data to obtain an execution result; Extract specific information from the execution results, perform sentiment analysis on the specific information to identify transaction intentions, use risk analysis strategies corresponding to transaction intentions to determine risk information of the specific information, and then generate analysis results; The analysis result is converted into a feature vector, and the feature vector is transmitted to the automated approval node, so that the automated approval node processes the feature vector using an approval strategy, obtains an approval result, and returns it to the first object.

2. The method according to claim 1, characterized in that The method further comprises: Acquire the relevant data of the second object within a preset time period according to the identifier of the second object; The relevant data is packaged into a transaction, and the transaction and an identifier of the second object are broadcast to the blockchain.

3. The method according to claim 2, characterized in that The transaction is obtained by decrypting the signed transaction, and the transaction and the identifier of the second object are broadcast to the blockchain, including: Obtain the private key pre-registered by the second object, use the private key to sign the transaction, obtain the signed transaction, and broadcast the signed transaction and the identifier to each block of the blockchain; The acquiring the transaction according to the identifier of the second object includes: Retrieve the corresponding block according to the identifier, obtain the signed transaction from the retrieved block, obtain the public key corresponding to the identifier or obtain the universal authorization key, use the public key or the authorization key to decrypt the signed transaction, and obtain the transaction.

4. The method according to claim 3, characterized in that The broadcasting of the signed transaction and the identification to each block of the blockchain includes: Broadcast the signed transaction and the ID to each block of the blockchain so that each block obtains the public key corresponding to the ID and uses the public key to verify the signature; In response to one or more blocks passing the signature verification, a new block is created in the blockchain, the signed transaction and the identifier are packaged into the new block, the new block is added to the blockchain, and the signed transaction and the identifier in all the original blocks are deleted.

5. The method according to any one of claims 1 to 4, characterized in that The transaction also includes a hash value, which is obtained by encapsulating the relevant data into a data object in a structured format, encrypting the data object, and hashing the encrypted data object; The calling of the smart contract pre-configured for the first object includes: Encapsulating the relevant data into a new data object in a structured format, encrypting the new data object, and performing hash processing on the encrypted new data object to obtain a new hash value; In response to the new hash value being consistent with the hash value in the transaction, a smart contract pre-configured for the first object is called.

6. The method according to claim 2, characterized in that The acquiring, according to the identifier of the second object, the relevant data of the second object within a preset time period includes: According to the identifier of the second object, original relevant data of the second object within a preset time period is acquired, and the original relevant data is processed using a preset data processing strategy to obtain the relevant data.

7. The method according to claim 6, characterized in that The original relevant data includes real-time operation data and business result data; After obtaining the original relevant data of the second object within the preset time period, the method further includes: processing the real-time operation data and the operation result data through a risk identification model to obtain risk data.

8. The method according to claim 1, characterized in that: The method further comprises: Receiving business logic data transmitted by the first object in one or more ways, and converting the business logic data into a feature space vector; Converting the feature space vector into information in a computer-readable form, and generating smart contract rules based on the information in the computer-readable form; Generate smart contract code corresponding to the smart contract rules using a programming language program preset in the blockchain; Compile the smart contract code into bytecode for deployment to the blockchain, assign a storage address to the smart contract, and store the smart contract in the storage address.

9. The method according to claim 8, characterized in that The converting the business logic data into a feature space vector comprises: Mapping business logic data of different modes into the same feature space to obtain a feature space vector of each business logic data; wherein the feature space includes one or more feature dimensions, and the feature space vector is composed of feature values ​​under different feature dimensions; Calculate the similarity between the feature values ​​of business logic data of different modalities under the same feature dimension, compare the similarity with the preset similarity threshold, and obtain the comparison result; Using the fusion strategy corresponding to the comparison results, the feature values ​​of different modal business logic data under the same feature dimension are fused, and then the feature space vectors of all modal business logic data are integrated.

10. The method according to claim 9, characterized in that The method of using a fusion strategy corresponding to the comparison result to fuse the feature values ​​of different modal business logic data under the same feature dimension includes: In response to the comparison result being greater than or equal to, selecting one of the feature values ​​of similar business logic data of different modalities under the same feature dimension; In response to the comparison result being less than, the feature values ​​of the dissimilar business logic data of different modalities under the same feature dimension are sorted and integrated.

11. The method according to claim 8, characterized in that The extracting specific information from the execution result includes: The specific information is extracted from the execution result using a pre-trained large language model; wherein, an original large language model corresponding to the first object is determined, and the original large language model is trained using the feature space vector to adjust the parameters of the original large language model to obtain the pre-trained large language model.

12. A data processing device based on blockchain, characterized in that: include: A transaction acquisition module, configured to receive a data verification request for a second object transmitted by a first object, and transmit the data verification request to a blockchain, so that the blockchain acquires a transaction according to an identifier of the second object; wherein the transaction includes relevant data of the second object; A smart contract module, configured to call a smart contract pre-configured for the first object, and in response to the relevant data satisfying a trigger condition in the smart contract, use the execution logic in the smart contract to process the relevant data to obtain an execution result; An analysis module is used to extract specific information from the execution results, perform sentiment analysis on the specific information to identify transaction intentions, use risk analysis strategies corresponding to the transaction intentions, determine risk information of the specific information, and then generate analysis results; The approval module is used to convert the analysis result into a feature vector, and transmit the feature vector to the automated approval node, so that the automated approval node processes the feature vector using the approval strategy, obtains the approval result and returns it to the first object.

13. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 11.

14. A computer readable medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 11 is implemented.

15. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 11 is implemented.

Images