Data processing method and device, electronic equipment and storage medium

Abstract

This invention discloses a data processing method, apparatus, electronic device, and storage medium. The method includes: in response to a financing request from a financing applicant, acquiring the applicant's transaction performance data, enterprise registration data, and enterprise tax credit data via an online interface; performing data fusion processing on the transaction performance data, enterprise registration data, and enterprise tax credit data to obtain multi-dimensional fused data; determining the financing applicant's financing access result based on the multi-dimensional fused data; in response to a successful financing access result, acquiring supplementary unstructured financing data uploaded by the financing applicant and performing cross-validation processing with the multi-dimensional fused data; generating a final financing result based on the cross-validation result and displaying the final financing result to the financing applicant. This invention achieves multi-source data connectivity through multi-dimensional data fusion, solving the problem of data silos; data cross-validation significantly reduces the data error rate compared to traditional manual verification of only paper materials.

Description

Technical Field

[0001] This invention relates to the field of big data technology, and in particular to a data processing method, apparatus, electronic device, and storage medium. Background Technology

[0002] During the procurement process, financing applicants need to obtain credit support based on procurement transaction contracts in order to ensure contract performance and their own business operations.

[0003] Currently, financing-related business processing for applicants mainly relies on traditional offline processes and manual operation. This model has many shortcomings: the data collection stage requires applicants to submit paper materials offline, and service providers then manually enter the data into the system. Furthermore, data from multiple platforms, such as enterprise information registration, tax administration, and procurement transactions, are isolated and not effectively connected, resulting in low data acquisition efficiency and difficulty in ensuring data integrity. In terms of document processing, unstructured materials such as contracts and financial statements submitted by applicants require manual parsing of key information, which is not only time-consuming and labor-intensive but also prone to data errors due to human error, thus affecting the efficiency of financing approval, the accuracy of judgments, and the effectiveness of risk management. Summary of the Invention

[0004] This invention provides a data processing method, apparatus, electronic device, storage medium, and computer program product.

[0005] According to one aspect of the present invention, a data processing method is provided, the method comprising: In response to a financing applicant's financing request for a target financing product, the system obtains the applicant's transaction performance data, corporate registration data, and corporate tax credit data through an online interface. The transaction performance data, corporate registration data, and corporate tax credit data of the financing applicant are integrated to obtain multi-dimensional integrated data of the financing applicant; Based on the multi-dimensional fusion data of the financing applicant, the financing access result of the financing applicant is determined; When the financing access result is approved, the supplementary financing data uploaded by the financing applicant is obtained and cross-validated with the multi-dimensional fusion data. The final financing result for the financing applicant is generated based on the cross-validation results, and then displayed to the financing applicant.

[0006] According to another aspect of the present invention, a data processing apparatus is provided, comprising: The data acquisition module is used to respond to the financing applicant's financing request for the target financing product by acquiring the financing applicant's transaction performance data, enterprise registration data, and enterprise tax credit data through an online interface; The data fusion module is used to perform data fusion processing on the transaction performance data, enterprise registration data and enterprise tax credit data of the financing applicant to obtain multi-dimensional fused data of the financing applicant; The financing access judgment module is used to determine the financing access result of the financing applicant based on the multi-dimensional fusion data of the financing applicant; The data cross-validation module is used to obtain the supplementary financing data uploaded by the financing applicant when the financing access result is approved, and to perform cross-validation processing with the multi-dimensional fused data. The financing determination module is used to generate the final financing result for the financing applicant based on the cross-validation results, and then display the final financing result to the financing applicant.

[0007] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and A memory that is communicatively connected to at least one processor; wherein, The memory stores a computer program that can be executed by at least one processor, such that the at least one processor is able to perform the data processing method of the embodiments of the present invention.

[0008] According to another aspect of the present invention, a computer-readable storage medium is provided, which stores computer instructions for causing a processor to execute and implement the data processing method of the embodiments of the present invention.

[0009] According to another aspect of the present invention, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the steps in the above-described method.

[0010] The technical solution of this invention directly obtains the transaction performance data, enterprise registration data, and enterprise tax credit data of the financing applicant through an online interface, replacing offline submission and improving the timeliness of data acquisition; it integrates the three types of scattered data into multi-dimensional fused data, realizing data connectivity and solving the problem of data isolation; for the unstructured supplementary financing data provided by the financing applicant, key fields are extracted and cross-validated with the multi-dimensional fused data to double-verify the authenticity of the data, which significantly reduces the data error rate compared to the traditional manual verification of only the surface information of paper materials.

[0011] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart illustrating the data processing method provided in an embodiment of the present invention; Figure 2 This is a flowchart illustrating another data processing method provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of the data processing device provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device that implements the data processing method of the present invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0015] The data processing method involved in this invention can be applied to a financing service platform. To facilitate understanding of the technical solution of this invention, the technical architecture of the financing service platform involved in this invention will first be explained. The financing service platform is based on a layered architecture + external system integration + internal support foundation, and is divided into four functional layers: data interface layer, data aggregation layer, data model layer, and data persistence layer. It also integrates with external platforms such as online procurement and transaction platforms, enterprise information registration platforms, and tax collection and management platforms, and links with internal bank credit management platforms, online banking, and other supporting systems to realize the digital processing of data in financing scenarios.

[0016] Specifically, the data interface layer serves as the data interaction entry point between the platform and external systems, providing unified external access capabilities. Its core functions include: 1. Providing a unified interface to support online requests from external platforms such as the online procurement and transaction platform, enterprise information registration platform, and tax administration platform, receiving transaction performance data, enterprise registration data, enterprise tax credit data, and authorization verification data from financing applicants; 2. Ensuring data security through encryption algorithms and digital signatures for secure dedicated line transmission, while also supporting load balancing and circuit breaker mechanisms to ensure interface stability; 3. Triggering business processes: When a financing applicant initiates a financing request from the online procurement and transaction platform, the platform pushes transaction performance data and authorization verification data to the data interface layer via a dedicated line, initiating subsequent processes such as financing access determination and credit limit calculation.

[0017] The data aggregation layer serves as the cleaning and standardization center for multi-source data, responsible for transforming scattered data into usable datasets. Its specific functions include: 1. Data processing: primarily verifying, deduplicating, and supplementing transaction performance data, enterprise registration data, and enterprise tax credit data, integrating them into a standardized dataset to lay the foundation for subsequent data fusion processing; 2. Unstructured data parsing: mainly relying on the optical character recognition technology and multi-layer perceptron model of the large model platform to extract key transaction elements from the supplementary unstructured financing data uploaded by financing applicants, and cross-validating them with multi-dimensional fused data to ensure data consistency.

[0018] The data model layer serves as the core execution layer for business logic and intelligent decision-making, automating credit assessment, financing approval, and risk control. Core functions include: 1. Automatically determining the financing applicant's eligibility and estimated credit limit using a pre-set credit assessment form and calculation formula, synchronously updating the information to the bank's credit management platform, and transmitting the results back to the online procurement transaction platform for display; 2. Online approval: After the financing applicant confirms the estimated credit limit on the online procurement transaction platform, the system pushes an approval notification to the account manager via instant messaging. After online approval, the system notifies the financing applicant to log in to online banking to withdraw funds; 3. Post-loan management: Relying on the risk prediction model of the big data processing platform, the system regularly collects the financing applicant's corporate registration data, transaction performance data, corporate tax credit data, and repayment information, dynamically updating the credit whitelist and predicting post-loan repayment risks. Simultaneously, it generates post-loan monitoring reports and risk control anomaly notifications, pushing them to the account manager via the communication platform.

[0019] The data persistence layer serves as the platform's data storage and lifecycle management center, providing stable data support for business operations. Its core functions include: 1. Data storage: Primarily responsible for synchronizing and updating business data such as financing applicant information, transaction records, credit assessment results, and repayment information. It supports frequency-based backup and data cleanup to ensure high data availability; 2. Data anonymization and reuse: This involves anonymizing database data and storing it in a data lake, providing a data foundation for subsequent customer profiling and post-loan management optimization.

[0020] The internal support framework includes a designated cloud, primary and backup databases, a data lake, a big data processing platform, a large model, and an interface portal. The designated cloud provides computing and storage resources to support the deployment and operation of each functional layer; the primary and backup databases store business data, ensuring high availability; the data lake stores anonymized historical data for subsequent analysis; the big data processing platform provides computing power support for massive data processing; the large model provides intelligent parsing capabilities such as character recognition technology and multi-layer perceptron models; and the interface portal supports unified interface management for the data interface layer.

[0021] The internal support platform includes an internal credit management platform, online banking, and an internal instant messaging platform. The internal credit management platform is used to manage credit limits and maintain whitelists. Online banking is used to support customer withdrawals and repayments. The internal instant messaging platform is used to push approval notifications and post-loan information.

[0022] Based on the above, the data processing method applied to the financing service platform can be seen in the following embodiment. Embodiment 1 Figure 1 This is a flowchart of a data processing method provided in an embodiment of the present invention. This embodiment can be applied to scenarios where a financing applicant seeks financing. The method can be executed by a data processing device, which can be implemented in hardware and / or software and can be configured in an electronic device.

[0023] like Figure 1 As shown, the data processing methods applied to the government procurement loan financing platform include: S101. In response to the financing request from the financing applicant for the target financing product, obtain the financing applicant's transaction performance data, corporate registration data, and corporate tax credit data through the online interface.

[0024] In this context, "financing applicant" refers to the enterprise entity participating in the procurement transaction and initiating the financing request; "target financing product" refers to a credit support product designed for the financing applicant, including contractual financing based on supply chain contracts and transactional financing based on historical procurement flows, which is the specific target of the financing applicant's financing request; this product is embedded in the online procurement transaction platform, which serves as the front-end entry point for the financing applicant to initiate the financing request, possessing core functions such as target financing product display, financing request submission, and data push. "Online interface" refers to the standardized data interaction channel between the platform executing the data processing method of this invention and external systems (including the online procurement transaction platform, enterprise information registration platform, and tax administration platform); "transaction performance data" refers to data containing the financing applicant's procurement platform registration information, procurement contract information, and historical transaction flow data (such as transaction-related identifiers, contract amounts, performance periods, etc.), originating from the online procurement transaction platform. "Enterprise entity registration data" refers to data recording the basic information of the financing applicant enterprise, such as enterprise name, registered address, legal representative information, business scope, and years of establishment; "enterprise tax credit data" refers to data reflecting the financing applicant's tax payment status, such as tax amount, tax credit rating, and whether there are any tax arrears records. In some embodiments, in response to a financing request from a financing applicant for a target financing product, the transaction performance data, enterprise registration data, and enterprise tax credit data of the financing applicant are obtained through an online interface, including S1011-S1012: S1011. In response to a financing request for a target financing product initiated by a financing applicant on a pre-set online procurement and transaction platform, the system receives the transaction performance data and authorization verification data of the financing applicant pushed by the online procurement and transaction platform through an online interface.

[0025] Specifically, after logging into the online procurement and transaction platform, the financing applicant selects either a contract-based or transaction-based financing product from the embedded target financing product section, fills in basic application information such as the financing amount and term, and clicks the "Submit Financing Request" button, triggering the online data push process of the procurement and transaction platform. The online procurement and transaction platform transmits the financing applicant's transaction performance data (including the applicant's procurement platform registration information, structured data related to completed procurement contracts, and recent historical procurement transaction details) and authorization verification data via a pre-set dedicated online interface (connected to the financing service platform's data interface layer). The authorization verification data refers to the electronic data corresponding to the data query authorization document signed by the financing applicant, containing information such as the authorization mark, validity period, and scope of authorization allowing the service provider to access third-party data such as the applicant's corporate registration and tax credit information. This serves as a prerequisite compliance basis for subsequent steps in S1012.

[0026] During data transmission, data packets are first encrypted using an encryption algorithm before being transmitted through a dedicated network channel between the financing service platform and the online procurement and transaction platform to prevent data leakage. After receiving the transmitted data, the data interface layer of the financing service platform first verifies the data packet integrity using its built-in mechanism (e.g., checking for missing data fields and whether the data packet size matches a preset value). If the data is correct, it is forwarded to the data aggregation layer for further processing. If the verification fails, the interface automatically returns a data retransmission request to the online procurement and transaction platform to ensure the reliability of data transmission.

[0027] S1012. Based on the authorized verification data of the financing applicant, obtain the enterprise registration data of the financing applicant from the enterprise information registration platform through the online interface, and obtain the enterprise tax credit data of the financing applicant from the tax administration platform.

[0028] Among them, transaction performance data, enterprise registration data, and enterprise tax credit data are all encrypted using encryption algorithms and transmitted via dedicated network lines. Dedicated network lines refer to the dedicated communication lines built between the financing service platform and the online procurement and transaction platform, the enterprise information registration platform, and the tax collection and management platform. These are not public internet channels and are the physical foundation for ensuring the security of cross-platform sensitive data transmission.

[0029] In practice, the data aggregation layer of the financing service platform first performs compliance verification on the authorization verification data received in step S1011. This includes verifying whether the authorizing entity is consistent with the registered name of the financing applicant's enterprise, whether the authorization is valid, and whether the scope of authorization includes querying enterprise registration data and enterprise tax credit data. If the verification fails (e.g., the authorization has expired or the scope of authorization is insufficient), a supplementary authorization notification is automatically generated and pushed to the online procurement and transaction platform through the online interface. The platform then prompts the financing applicant to re-authorize the authorization.

[0030] After authorization verification is passed, the data interface layer of the financing service platform calls the dedicated online interfaces that connect with the enterprise information registration platform and the tax administration platform respectively. It sends a data query request (carrying the financing applicant's unified social credit code and authorization certificate) to the enterprise information registration platform and a data query request (carrying the financing applicant's tax identification number and authorization certificate) to the tax administration platform. After the enterprise information registration platform and the tax administration platform verify the validity of the authorization certificate, they respectively transmit the financing applicant's enterprise registration data and enterprise tax credit data to the financing service platform through a dedicated network line. Before transmission, the data is encrypted using a preset encryption algorithm to ensure data transmission security.

[0031] After receiving encrypted enterprise registration data and enterprise tax credit data, the data interface layer of the financing service platform first decrypts the data using the corresponding decryption algorithm (matching the encryption algorithm) and then forwards the data to the data aggregation layer. The data aggregation layer unifies the field formats of the enterprise registration data and enterprise tax credit data according to the standardized dataset specifications (such as unifying the establishment year into years and the tax amount into yuan), laying the foundation for the data fusion processing in the subsequent S102 step.

[0032] In other embodiments, for high-concurrency scenarios, the financing service platform can deploy a multi-node online interface cluster to support a load balancing mechanism. When an interface node fails, the system automatically switches to a backup interface node to ensure the continuous acquisition of transaction performance data, enterprise registration data, and enterprise tax credit data. This avoids data collection interruption due to interface failure, further improving the stability and efficiency of data collection and adapting to the needs of large-scale financing business.

[0033] Understandably, step S101 replaces traditional offline material submission and manual data entry, reducing the manpower and time costs of data collection and solving the technical problems of low efficiency and error-proneness in offline operations; through network dedicated line encryption and digital signature, it avoids leakage and tampering during data transmission, ensuring data security; and it acquires three types of core data at once, laying the foundation for subsequent integrated processing and breaking down data silos between different platforms.

[0034] S102. Perform data fusion processing on the transaction performance data, enterprise registration data and enterprise tax credit data of the financing applicant to obtain multi-dimensional fused data of the financing applicant.

[0035] The fusion processing refers to the cleaning, verification, and integration of three types of heterogeneous data collected, including transaction performance data, enterprise registration data, and enterprise tax credit data, to eliminate data redundancy and format differences, forming a dataset with a unified structure. Multidimensional fused data refers to a standardized data set that includes dimensions of financing applicant procurement transactions (e.g., contract amount, transaction details), enterprise qualifications (legal representative, years of registration), and taxation (tax amount, tax credit rating). Data from each dimension is linked through a unique primary key, supporting subsequent financing access determination and credit limit calculation.

[0036] In practice, the transaction performance data, enterprise registration data, and enterprise tax credit data of the financing applicant are integrated to obtain multi-dimensional integrated data, including steps S1021-S1025: S1021. Standardize the data format of the transaction performance data, enterprise registration data, and enterprise tax credit data of the financing applicant.

[0037] Data format standardization refers to the process of adjusting three types of data from different sources and with different formats according to a pre-set unified data format specification, so that the data types (such as text, numerical, and date), field naming rules, data length, etc. are consistent.

[0038] In practice, a unified format specification is first established. For example, a data format standard for financing scenarios is predefined, specifying that date fields should uniformly adopt the YYYY-MM-DD format (e.g., "2023.10.05" and "2023 / 10 / 05" in enterprise registration data should be uniformly adjusted to "2023-10-05"); amount fields should uniformly retain two decimal places and be in yuan (e.g., "500 million yuan" in transaction performance data should be converted to "500,000,000.00 yuan"); text fields should uniformly adopt a preset encoding, and field naming should follow the rules of business type and field meaning (e.g., the name of the legal representative in enterprise registration data should be uniformly named "Entity_Legal Representative Name").

[0039] Furthermore, format conversion tools are used to configure conversion rules based on the format characteristics of data from different sources. For example, for corporate tax credit data in Excel format and corporate registration data in XML format, the tool's built-in format parsing and conversion functions automatically convert them into structured data that conforms to unified standards. Finally, data type validation and correction are performed. The converted data is validated for type. If the registered capital field in the corporate registration data is found to be mistakenly stored as text, the numerical part is extracted using regular expressions and converted to a numeric type. If the performance period field in the transaction performance data is a text description (such as "six months"), it is converted to the numeric type "6" and the term unit field is marked "month".

[0040] The benefits of this step are as follows: it solves the problem of heterogeneous formats in multi-source data, unifying the three types of data in terms of type, naming, and encoding, providing a foundation for subsequent data cleaning and fusion, and avoiding data processing anomalies caused by format differences; it reduces the difficulty of data parsing, allowing subsequent technical steps such as optical character recognition extraction and cross-validation to directly call standardized data, thereby improving data processing efficiency; and it ensures data consistency, providing standardized input for credit limit calculation and risk model calculation, and reducing calculation errors caused by inconsistent formats.

[0041] S1022. Through preset data cleaning rules, invalid data filtering is performed on the transaction performance data, corporate registration data, and corporate tax credit data of the financing applicant.

[0042] The pre-defined data cleaning rules refer to a set of rules pre-established based on financing business needs and data quality requirements to identify and remove invalid data. These rules cover data integrity rules, reasonableness rules, and business relevance rules, and can be dynamically adjusted according to actual business scenarios. Invalid data refers to data that cannot provide valid information for financing access determination or quota calculation, or data that does not conform to business logic or data specifications. This includes data missing key fields, data with values ​​exceeding reasonable ranges, and data unrelated to financing business.

[0043] In practice, multi-dimensional data cleaning rules are established. For example, from the perspective of data integrity, procurement contract-related data must include three key fields: contract number, contract amount, and performance period; data lacking any of these fields is considered invalid. From the perspective of data rationality, corporate income tax annual declaration amounts in corporate tax credit data that are negative and whose absolute value exceeds a preset percentage of the company's annual revenue are considered invalid. From the perspective of business relevance, transaction performance data with a transaction date earlier than the establishment date of the financing applicant is considered invalid. Furthermore, the specified data cleaning rules are embedded in the data aggregation layer of the financing service platform, and invalid data in these three categories is filtered based on these rules.

[0044] For example, in the transaction performance data of a certain financing applicant, a "purchase contract data in November 2023" only contains "contract number: CG20231XXXX", but lacks the fields of "contract amount" and "performance period". According to the preset cleaning rules, this data is judged as invalid data and removed from the transaction performance dataset.

[0045] The benefits of this step are: eliminating invalid data such as missing key fields and data that exceeds reasonable limits, reducing the interference of redundant information on the data processing flow, and improving data processing speed; improving data quality, preventing invalid data from entering the admission judgment and credit limit calculation stages, and reducing the risk of credit decision-making due to data distortion.

[0046] S1023. Remove duplicate data from the transaction performance data, enterprise registration data, and enterprise tax credit data of the financing applicant through data duplication verification.

[0047] Data duplication verification refers to the process of identifying duplicate data with identical content or the same core fields by comparing records in a dataset. Duplicate data removal refers to the process of deleting redundant and duplicate records according to preset retention rules after identifying duplicate data, retaining only one valid record.

[0048] In specific implementation, the core verification fields for each data type are determined as follows: For transaction performance data, the contract number is used as the core verification field (the same contract number corresponds to a unique contract; if multiple records with the same contract number exist, they are considered duplicates); for enterprise registration data, the unified social credit code is used as the core verification field (the unified social credit code uniquely identifies an enterprise; enterprise registration data with the same code are considered duplicates); for enterprise tax credit data, the taxpayer identification number and the tax declaration period are used as the core verification fields (the tax data of the same enterprise in the same declaration period should be unique; if both are the same, they are considered duplicates).

[0049] A multi-stage verification method is adopted: The first stage performs intra-dataset duplication verification, using the deduplication function of the database's data processing tools to deduplicatize the transaction performance data table, enterprise registration data table, and enterprise tax credit data table (e.g., deleting records with duplicate contract numbers in the transaction performance data table); The second stage performs cross-dataset duplication verification, comparing the core fields between different datasets. For example, if the unified social credit code in the enterprise registration data is the same as the taxpayer identification number in the enterprise tax credit data, and the enterprise names in the two data records are also the same, then it is determined to be a cross-dataset duplication, and the updated data from the source is retained (e.g., retaining the latest updated enterprise tax credit data and deleting redundant information in the duplicate enterprise registration data).

[0050] Set rules for retaining duplicate data. When duplicate data is identified, prioritize retaining the record with the most recent data generation time; if the data generation times are the same, retain the record with higher source credibility.

[0051] The benefits of this step are: eliminating duplicate data caused by repeated pushes from multiple platforms or system failures, avoiding data redundancy that occupies storage resources and reducing system operating costs; preventing duplicate data from affecting the calculation results and ensuring the accuracy of credit decisions; simplifying data association logic so that duplicate matching issues do not need to be handled during subsequent data fusion, thereby improving data fusion efficiency.

[0052] S1024. Through data completion technology, repair missing or abnormal fields in the transaction performance data, enterprise registration data, and enterprise tax credit data of the financing applicant.

[0053] Data completion technology refers to the technical methods used to supplement or correct missing or abnormal field values ​​in a dataset to reasonable values. These methods include rule-based completion, data-related completion, and statistical method-based completion.

[0054] In practice, rule-based completion and repair are implemented: for fields with clear business rules, they are processed according to preset rules. For example, if the registered capital field in the enterprise registration data is missing, and the enterprise type is a limited liability company, a default value can be set based on the industry average registered capital; if the tax credit rating field in the enterprise tax credit data is abnormal, it is corrected to the lowest normal rating by default, and a data abnormality correction mark is marked.

[0055] Data completion and repair based on related datasets: This involves using the relationships between different datasets to supplement missing fields or correct abnormal fields. For example, if the legal representative field for financing applications is missing in transaction performance data, it can be added to the transaction performance data by linking the enterprise registration data to the unique primary key (Unified Social Credit Code). Similarly, if the enterprise name field in the enterprise tax credit data is inconsistent with the enterprise name in the enterprise registration data, the enterprise tax credit data will be corrected based on the enterprise name in the enterprise registration data to ensure consistency.

[0056] Statistical method-based completion and repair: For missing fields without clear rules and unrelated data, statistical methods (such as mean, median, and mode) are used to complete them.

[0057] The benefits of this step are as follows: supplementing missing fields, correcting abnormal data, improving data integrity, ensuring that multi-dimensional integrated data can fully reflect the true situation of the financing applicant, and providing sufficient basis for pre-loan approval; avoiding process interruptions due to missing or abnormal data, ensuring the continuity of data processing, and improving business processing efficiency; and ensuring data traceability by marking repair indicators, reducing compliance risks caused by data repair, and providing a reference for subsequent data quality optimization.

[0058] S1025. Using a preset unique primary key, the processed transaction performance data, corporate registration data, and corporate tax credit data of the financing applicant are integrated to obtain multi-dimensional integrated data.

[0059] Among them, the unique primary key is the unique identifier of the financing applicant (such as the unified social credit code of the financing applicant).

[0060] In practice, a unique primary key is determined and uniformly labeled. The unified social credit code of the financing applicant is set as the unique primary key. In the three types of data that have been processed in the previous step, a unique primary key field is added and the corresponding unified social credit code is filled in to ensure that each data contains a unique primary key field.

[0061] The integration is achieved using a relational database table structure. For example, five core data tables are created in the database (financing applicant information table, purchase order table, purchase history table, tax information table, and annual profit table), each with a unique primary key as a foreign key. Multiple table joins are implemented using SQL statements. For instance, the financing applicant information table and the tax information table are joined internally using a unique primary key, allowing simultaneous access to the financing applicant's corporate information and tax information. Simultaneously, a data fusion module is developed in the data aggregation layer of the financing service platform to automatically write the processed data from various sources into the corresponding data tables, maintaining the relationships between the tables using unique primary keys.

[0062] The table includes: Financing Applicant Information Table (unique primary key, company name, registered address, registration date, registered capital, legal representative's name, and data date); Purchase Order Table (order number, unique primary key, order effective date, total order amount, and data date); Purchase History Transaction Table (serial number, unique primary key, transaction date, order quantity, total order amount, and data date); Tax Information Table (serial number, unique primary key, tax payment date, year of tax payment, tax amount, tax credit rating, and data date); Annual Profit Statement (serial number, unique primary key, financial statement year, total revenue, total profit, and data date).

[0063] After data fusion, the fusion results are verified to ensure that each table can be accurately linked through a unique primary key, and that the linked data has no logical conflicts. If a linking failure or data conflict is found, an exception handling mechanism is triggered, the abnormal data is temporarily stored in a temporary table, and technical personnel are notified to investigate the cause.

[0064] The benefits of this step are: breaking down data silos between different platforms, integrating the three types of scattered data into a unified dataset, realizing multi-dimensional data correlation, and providing data support for a comprehensive assessment of the qualifications of financing applicants; forming a standardized data table structure and linking it with a unique primary key, which facilitates the system to quickly query and call data, and improves the response speed of credit limit calculation and approval.

[0065] S103. Based on the multi-dimensional fusion data of the financing applicant, determine the financing access result of the financing applicant.

[0066] Among them, the financing access result refers to the determination of whether the financing applicant meets the basic conditions for the application of the target financing product based on multi-dimensional integrated data. The result is divided into approval and disapproval, which is the key basis for triggering subsequent processes.

[0067] In some embodiments, based on the supplier's multi-dimensional fusion data, the system automatically determines whether the supplier meets the basic application conditions for government procurement loan financing through preset business rules or scoring models, outputs standardized access judgment results, and realizes the automation and standardization of access review.

[0068] Taking the scoring model as an example, based on the supplier's multi-dimensional fusion data, the supplier's financing access results are determined, including S1031-S1033: S1031. Obtain a pre-constructed credit assessment table; wherein the credit assessment table includes at least one assessment dimension, each assessment dimension corresponds to at least one scoring factor, each scoring factor corresponds to a factor weight and a factor range, each factor range includes at least two factor values, and different factor values ​​correspond to different credit scores.

[0069] The assessment dimensions are the areas for evaluating the applicant's qualifications as defined in the credit assessment table. These typically include basic business operations, operational stability, and creditworthiness, covering core assessment dimensions related to the applicant's repayment ability and performance risk. Scoring factors are the specific assessment indicators broken down under each assessment dimension, such as registered capital and years of registration under the basic business operations dimension, and tax credit rating and historical performance records under the creditworthiness dimension. Factor weights refer to the proportion of importance of each scoring factor within its respective dimension (the total weight is 100%), with factors having higher risk impact having higher weights (e.g., tax credit rating weight). (Possibly higher than the registered address area); Factor range refers to the division of the value range or classification of scoring factors, such as the range of the registration period factor being [1,2) years, [2,3) years, ≥3 years, and the range of the tax credit rating factor being A level, B level; Factor value refers to the specific content within the factor range (categorized as options for types, and range for numerical types), such as the A level and B level of the tax credit rating factor range, and the [100,500) million yuan and ≥5 million yuan of the registered capital factor range; Credit score refers to the quantitative score corresponding to each factor value, such as 20 points for a tax credit rating of A level and 15 points for a tax credit rating of B level.

[0070] S1032. Based on the multi-dimensional fusion data and credit assessment form of the financing applicant, determine the credit assessment result of the financing applicant according to the preset credit assessment calculation formula.

[0071] The preset credit assessment calculation formula refers to the mathematical formula set in advance for calculating the total score. The core logic is to first calculate the score of each factor (factor weight × corresponding credit score), then accumulate all factor scores to obtain the dimension score, and finally accumulate the scores of each dimension to obtain the total credit assessment score.

[0072] In practice, the actual values ​​of each scoring factor are extracted from the multi-dimensional integrated data of the financing applicant. For example, the registered capital of 5 million yuan and the registration period of 3 years are extracted from the financing applicant information table; the tax credit rating of A is extracted from the tax information table; and the transaction amount of 6 million yuan in the past year, 12 orders in the past year, and historical performance rate of 98% are extracted from the purchase history transaction record table. The credit score corresponding to each factor value is matched with the credit assessment table obtained by S1031. For example, the registered capital of 5 million yuan corresponds to 10 points, the registration period of 3 years corresponds to 20 points, the tax credit rating of A corresponds to 20 points, the transaction amount of 6 million yuan in the past year corresponds to 40 points, the 12 orders in the past year corresponds to 25 points, and the historical performance rate of 98% corresponds to 20 points.

[0073] Calculate the scores for each dimension using the following formulas: Basic Business Score = (10% × 10) + (20% × 20) = 1 + 4 = 5 points; Business Stability Score = (15% × 40) + (25% × 25) = 6 + 6.25 = 12.25 points; Credit Status Score = (10% × 20) + (20% × 20) = 2 + 4 = 6 points; Calculate the total credit assessment score: 5 + 12.25 + 6 = 23.25 points (In this example, the total score is accumulated based on the dimension scores. In practice, it can be adjusted to a percentage system according to the business, such as accumulating the scores of each dimension after converting them according to the weight ratio); Generate a credit assessment result table, which includes a unique primary key, the values ​​of each factor, the scores of each factor, the scores of each dimension, and the total credit assessment score.

[0074] S1033. Based on the credit assessment results, determine the financing access outcome for the financing applicant.

[0075] In practice, based on the risk control objectives of the financing business, an access judgment threshold is set. For example, when using a percentage system, an access is granted if the total credit assessment score is ≥60; if the total credit assessment score is ≤40 and <60, manual review is required; and if the total credit assessment score is <40, access is denied. The financing access result can be obtained based on the calculated credit assessment results and the judgment threshold.

[0076] The benefits of this step are as follows: it avoids bias in access determination caused by relying solely on procurement contracts based on multi-dimensional data, thereby improving the rationality of access determination; it replaces manual due diligence, thereby improving the efficiency of financing access determination; and the access results are automatically recorded to the bank's credit management platform, which facilitates subsequent traceability and access in the approval process, thus achieving the continuity of business processes.

[0077] S104. In response to the financing access result being approved, obtain the supplementary financing data uploaded by the financing applicant and perform cross-validation processing with the multi-dimensional fusion data.

[0078] Among them, supplementary financing data refers to the application materials data uploaded by the financing applicant that are not in a fixed format, such as scanned copies of purchase contracts, photos of business licenses, financial statements and other unstructured data that cannot be directly stored in database fields; cross-validation processing refers to comparing the key information after parsing the unstructured supplementary financing data with the existing multidimensional fused data to verify the authenticity and completeness of the data.

[0079] In some embodiments, in response to a successful financing approval, supplementary unstructured financing data uploaded by the financing applicant is obtained and cross-validated with multidimensional fused data, including S1041-S1042: S1041. Obtain the supplementary unstructured financing data uploaded by the financing applicant, and extract key fields from the supplementary unstructured financing data using optical character recognition technology and large model recognition technology.

[0080] Specifically, optical character recognition (OCR) technology is used to scan and parse unstructured financing supplementary data to extract text content. The extracted text content is then filtered to remove invalid text (such as blank characters and irrelevant notes), and the remaining text content is converted into a dataset in a preset format. This dataset, along with preset prompts, is input into a large model for processing, outputting candidate fields. The preset prompts could be something like, "You are a financial document analysis expert. Please extract the total contract amount field from the following procurement contract. This field is a number between 0 and 99,999,999,999, with two decimal places. Note that this field may contain units such as 'ten thousand yuan' or 'million yuan,' which need to be uniformly converted to units (yuan)." The candidate fields undergo consistency and validity checks (such as verifying whether the contract amount is a reasonable value and matches the procurement scenario), and the candidate fields that pass the checks are used as key fields.

[0081] S1042. The extracted key fields are compared with the corresponding fields in the multidimensional fused data to achieve cross-validation.

[0082] For example, key fields such as the purchase contract amount and performance period extracted from the supplementary data of unstructured financing are compared with the corresponding fields in the purchase order table of the multidimensional fusion data. If the comparison is consistent, the cross-validation is deemed to have passed; if there is a discrepancy in the comparison (such as the difference between the extracted contract amount and the amount in the multidimensional fusion data exceeding a preset threshold), the cross-validation is deemed to have failed, triggering the exception handling process and notifying the financing applicant to provide supplementary explanations or re-upload materials.

[0083] The benefits of this step are: it replaces traditional manual analysis of unstructured materials, reduces human error, and improves data analysis efficiency; through cross-validation, it prevents applicants from submitting false materials (such as falsifying contract amounts), reduces financing risks, and solves the problem of high risks associated with manual verification.

[0084] S105. Generate the final financing result for the financing applicant based on the cross-validation results, and present the final financing result to the financing applicant.

[0085] If the cross-validation results are consistent, a final financing result is generated, including the applicant's final credit limit, loan interest rate, and loan term. This final result is then sent to the online procurement and transaction platform via an online interface for display, allowing the applicant to promptly access the financing information. The applicant can view the final financing result on the online platform and adjust the application information within the modifiable scope. After verification and confirmation, the applicant fills in their bank account information.

[0086] For financing applicants who have entered their account information, the financing service platform automatically generates a loan contract based on the loan contract template and the financing application information, and notifies the applicant to log in to their corporate online banking to complete the loan contract signing and financing authorization. After the applicant completes the contract signing and authorization, the financing application record status is updated to "withdrawable" on the bank's internal credit management platform, and the applicant is notified of the withdrawal via corporate online banking or offline notification. For financing applicants who have already withdrawn funds, their withdrawal information (withdrawal amount, withdrawal time, used amount, and remaining loan amount) is synchronized daily, and for financing applicants making their first withdrawal, the financing application record status is updated to "withdrawn" on the bank's internal credit management platform.

[0087] In this embodiment, transaction performance data, enterprise registration data, and enterprise tax credit data are directly obtained through an online interface, replacing offline submission and improving the timeliness of data acquisition. The three types of scattered data are integrated into multi-dimensional fused data, realizing data connectivity across multiple platforms and solving the problem of data isolation. For the unstructured supplementary financing data provided by the financing applicant, key fields are extracted and cross-validated with the multi-dimensional fused data to double-verify the authenticity of the data. Compared with the traditional manual verification of only the surface information of paper materials, this significantly reduces the data error rate.

[0088] Example 2 Figure 2 This invention provides a flowchart of a data processing method. Based on Embodiment 1, this embodiment adds functions for estimated credit limit calculation, final credit limit adjustment, and post-loan repayment risk prediction, further improving the entire process of the data processing method in financing scenarios. See [link to previous document]. Figure 2 The method includes the following steps: S201. In response to the financing applicant's financing request for the target financing product, obtain the financing applicant's transaction performance data, corporate registration data, and corporate tax credit data through the online interface.

[0089] The specific implementation method of this step is completely the same as S101 in Embodiment 1, and will not be repeated here.

[0090] S202. Data fusion processing is performed on the transaction performance data, enterprise registration data, and enterprise tax credit data of the financing applicant to obtain multi-dimensional fused data of the financing applicant.

[0091] The specific implementation method of this step is completely the same as S102 in Embodiment 1, and will not be repeated here.

[0092] S203. Based on the multi-dimensional fusion data of the financing applicant, determine the financing access result and estimated credit limit of the financing applicant.

[0093] The specific implementation of determining the financing access result based on the multi-dimensional fusion data of the financing applicant is completely consistent with S1031-S1033 in Example 1; determining the estimated credit limit based on the multi-dimensional fusion data of the financing applicant includes the following two sub-steps: Sub-step 1: Determine the value of the credit assessment coefficient based on the credit assessment score range in which the credit assessment result falls; the value of the credit assessment coefficient will vary depending on the credit assessment score range.

[0094] Specifically, based on historical data of financing transactions and risk control objectives, multiple continuous credit assessment score ranges are pre-defined, and corresponding credit assessment coefficient values ​​are set for each range. The credit assessment coefficient (usually represented by M) is a coefficient between 0 and 1. The higher the credit assessment score (the better the applicant's qualifications), the larger the credit assessment coefficient (a higher estimated credit limit can be obtained). The value needs to be verified through a historical credit limit calculation model (ensuring that the coefficient is negatively correlated with the bad debt rate). The mapping relationship between credit assessment score ranges and credit assessment coefficients is stored in a structured table. Subsequently, only the applicant's current credit assessment score needs to be matched with the mapping relationship to determine the value of the credit assessment coefficient.

[0095] Sub-step 2: Based on the multi-dimensional integrated data of the financing applicant and the value of the credit assessment coefficient, determine the estimated credit limit according to the preset credit limit calculation formula.

[0096] Specifically, the preset credit limit calculation formula can be set in conjunction with the purchase transaction flow, for example: Estimated credit limit = MIN(2×A×M, 0.8×B×M, 0.6×C×M); where A is the online purchase transaction amount of the financing applicant in the past 6 months, B is the online purchase transaction amount in the past year, C is the average annual transaction amount of the online purchase transaction in the past 2 years, and M is the credit assessment coefficient.

[0097] S204. In response to the financing access result being approved, obtain the supplementary financing data uploaded by the financing applicant and perform cross-validation processing with the multi-dimensional fusion data.

[0098] The specific implementation method of this step is completely the same as S104 in Embodiment 1, and will not be repeated here.

[0099] S205. Adjust the estimated credit limit based on the cross-validation results to determine the final credit limit, and present the final credit limit as the final financing result to the financing applicant.

[0100] In practice, if cross-validation passes, the estimated credit limit can be directly used as the final credit limit; if cross-validation finds discrepancies (such as small discrepancies between the extracted key fields and the multidimensional fused data, but not complete consistency), the estimated credit limit is adjusted according to the degree of discrepancy to obtain the final credit limit; if the cross-validation discrepancy is large, it is judged as a validation failure, and the financing applicant needs to be notified to supplement materials and re-validate, and the final financing result is not generated for the time being.

[0101] Furthermore, the final credit limit, along with information such as loan interest rate and loan term, will be displayed on the online procurement and transaction platform as the final financing result, so that the financing applicant can be aware of and confirm it.

[0102] S206. After the financing is completed, regularly collect the enterprise registration data, transaction performance data, enterprise tax credit data and repayment information of the financing applicant; and predict the post-loan repayment risk of the financing applicant based on the enterprise registration data, transaction performance data, enterprise tax credit data and repayment information through a risk prediction model.

[0103] In practice, after the financing is completed, the financing service platform will continuously collect enterprise registration data (such as changes in legal representative, changes in business scope, etc.), transaction performance data (such as the latest purchase orders, transaction flow, etc.), enterprise tax credit data (such as the latest tax declaration information, changes in tax credit rating, etc.), and repayment information (such as repayment amount, repayment date, whether it is overdue, etc.) from the financing applicant during the financing period.

[0104] Based on the risk prediction model, various types of data acquired regularly are used to predict risks. Through pre-set risk control verification rules (such as changes in legal representatives or actual controllers, reduced activity in purchase orders, reduced tax credit ratings, and overdue repayments), the system automatically triggers credit limit adjustments or invalidation of financing applications for applicants with potential risks. The system also notifies the corresponding account manager in real time to intervene and follow up, thereby reducing the risk of bad debts after the loan is issued.

[0105] In this embodiment, not only is the efficiency of financing data processing improved, but also the bad debt rate is effectively reduced through multi-source data verification, credit limit calculation, and post-loan dynamic early warning. The operating costs of both financing applicants and service providers are significantly reduced, while providing technical support for the large-scale and standardized development of financing business, and helping to solve the problems of difficult and slow financing for small and medium-sized enterprises.

[0106] Example 3 Figure 3 This is a schematic diagram of a data processing apparatus provided in an embodiment of the present invention. This apparatus can execute any data processing method of the present invention. For example... Figure 3 As shown, the data processing device includes: The data acquisition module 301 is used to respond to the financing request of the financing applicant for the target financing product by acquiring the financing applicant's transaction performance data, enterprise registration data and enterprise tax credit data through the online interface; The data fusion module 302 is used to perform data fusion processing on the transaction performance data, enterprise registration data and enterprise tax credit data of the financing applicant to obtain multi-dimensional fused data of the financing applicant; The financing access judgment module 303 is used to determine the financing access result of the financing applicant based on the multi-dimensional fusion data of the financing applicant; The data cross-validation module 304 is used to obtain the supplementary financing data uploaded by the financing applicant when the financing access result is approved, and to perform cross-validation processing with the multi-dimensional fused data. The financing determination module 305 is used to generate the final financing result for the financing applicant based on the cross-validation results and display the final financing result to the financing applicant.

[0107] In some embodiments, the data cross-validation module 304 includes: The identification unit is used to obtain the unstructured supplementary financing data uploaded by the financing applicant, and extract key fields from the unstructured supplementary financing data through optical character recognition technology and large model recognition technology. The comparison unit compares the extracted key fields with the corresponding fields in the multidimensional fused data to achieve cross-validation.

[0108] In some embodiments, the identification unit is specifically used for: Using optical character recognition technology, unstructured financing supplementary data is scanned and analyzed to extract its text content; The extracted text content is filtered to remove invalid text content, and the remaining text content is converted into a dataset in a preset format. Input the dataset in the preset format and the preset prompt words into the large model for processing, and output candidate fields; Perform field consistency and field validity checks on candidate fields, and use the candidate fields that pass the checks as key fields.

[0109] In some embodiments, the data fusion module 302 is specifically used for: Standardize the data format of the transaction performance data, enterprise registration data, and enterprise tax credit data of the financing applicants; By using preset data cleaning rules, invalid data is filtered out from the transaction performance data, corporate registration data, and corporate tax credit data of financing applicants. By performing data duplication checks, duplicate data is removed from the transaction performance data, corporate registration data, and corporate tax credit data of financing applicants. Data completion technology is used to repair missing or abnormal fields in the transaction performance data, corporate registration data, and corporate tax credit data of financing applicants. By using a pre-defined unique primary key, the processed transaction performance data, corporate registration data, and corporate tax credit data of the financing applicant are integrated to obtain multi-dimensional integrated data; among them, the unique primary key is the unique identifier of the financing applicant.

[0110] In some embodiments, the data acquisition module 301 is specifically used for: In response to a financing request for a target financing product initiated by a financing applicant on a pre-defined online procurement and transaction platform, the system receives transaction performance data and authorization verification data pushed by the financing applicant from the online procurement and transaction platform via an online interface. Based on the authorized verification data of the financing applicant, the enterprise registration data of the financing applicant is obtained from the enterprise information registration platform through an online interface, and the enterprise tax credit data of the financing applicant is obtained from the tax administration platform. Among them, transaction performance data, enterprise registration data, and enterprise tax credit data are all encrypted using encryption algorithms and transmitted via dedicated network lines.

[0111] In some embodiments, the financing access determination module 303 is specifically used for: Obtain a pre-built credit assessment table; wherein the credit assessment table includes at least one assessment dimension, each assessment dimension corresponds to at least one scoring factor, each scoring factor corresponds to a factor weight and a factor range, each factor range includes at least two factor values, and different factor values ​​correspond to different credit scores; Based on the multi-dimensional fusion data and credit assessment form of the financing applicant, the credit assessment result of the financing applicant is determined according to the preset credit assessment calculation formula; Based on the credit assessment results, the financing access outcome for the financing applicant is determined.

[0112] In some embodiments, the data processing apparatus further includes a credit limit determination module, configured to: The value of the credit assessment coefficient is determined based on the credit assessment score range in which the credit assessment result falls; the value of the credit assessment coefficient is different for different credit assessment score ranges. Based on the multi-dimensional integrated data of the financing applicant and the value of the credit assessment coefficient, the estimated credit limit is determined according to the preset credit limit calculation formula. Accordingly, the financing confirmation module 305 is specifically used for: The estimated credit limit is adjusted based on the cross-validation results to determine the final credit limit, which is then used as the financing result for the applicant.

[0113] In some embodiments, the data processing apparatus further includes a risk prediction module, for: After financing is completed, the company's corporate registration data, transaction performance data, corporate tax credit data, and repayment information are collected regularly from the financing applicant. The risk prediction model is used to predict the post-loan repayment risk of financing applicants based on their corporate registration data, transaction performance data, corporate tax credit data, and repayment information.

[0114] The data processing apparatus provided in the embodiments of the present invention can execute the data processing method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the method.

[0115] According to embodiments of the present invention, the present invention also provides an electronic device, a readable storage medium, and a computer program product.

[0116] Example 4 Figure 4 A schematic diagram of the structure of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the invention described and / or claimed herein.

[0117] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory 12 or a random access memory 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the read-only memory 12 or loaded from storage unit 18 into the random access memory 13. The random access memory 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, read-only memory 12, and random access memory 13 are interconnected via a bus 14. An input / output interface 15 is also connected to the bus 14.

[0118] Multiple components in electronic device 10 are connected to input / output interface 15, including: input unit 16; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disks, optical disks, etc.; and communication unit 19, such as network interface cards, modems, wireless transceivers, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0119] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, central processing units, graphics processing units, various special-purpose artificial intelligence computing chips, various processors running machine learning model algorithms, digital signal processors, and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as performing data processing methods.

[0120] In some embodiments, the data processing method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via read-only memory 12 and / or communication unit 19. When the computer program is loaded into random access memory 13 and executed by processor 11, one or more steps of the data processing method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the data processing method by any other suitable means (e.g., by means of firmware).

[0121] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays, application-specific integrated circuits (ASICs), application-specific standard products (ASICs), systems-on-a-chip (SoCs), complex programmable logic devices, computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0122] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0123] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, optical fibers, portable compact disk read-only memory, optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0124] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device or liquid crystal display for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0125] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet. The computing system can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having client-server relationships with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, a host product within the cloud computing service system, addressing the shortcomings of traditional physical hosts and virtual private servers, such as high management difficulty and weak business scalability.

[0126] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0127] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A data processing method, characterized in that, include: In response to a financing applicant's financing request for a target financing product, the system obtains the applicant's transaction performance data, corporate registration data, and corporate tax credit data through an online interface. The transaction performance data, corporate registration data, and corporate tax credit data of the financing applicant are fused to obtain multi-dimensional fused data of the financing applicant. Based on the multi-dimensional fusion data of the financing applicant, the financing access result of the financing applicant is determined; When the financing access result is approved, the supplementary financing data uploaded by the financing applicant is obtained and cross-validated with the multi-dimensional fusion data. The final financing result for the financing applicant is generated based on the cross-validation results, and the final financing result is displayed to the financing applicant.

2. The method according to claim 1, characterized in that, The step of obtaining the supplementary financing data uploaded by the financing applicant and performing cross-validation processing with the multi-dimensional fused data includes: Obtain the supplementary unstructured financing data uploaded by the financing applicant, and extract key fields from the supplementary unstructured financing data using optical character recognition technology and large model recognition technology; The extracted key fields are compared with the corresponding fields in the multidimensional fused data to achieve cross-validation.

3. The method according to claim 2, characterized in that, The extraction of key fields from the unstructured financing supplementary data using optical character recognition (OCR) and large model recognition technologies includes: The unstructured financing supplementary data is scanned and parsed using the optical character recognition technology to extract the text content. The extracted text content is filtered to remove invalid text content, and the remaining text content is converted into a dataset in a preset format. The dataset in the preset format and the preset prompt words are input into the large model for processing, and candidate fields are output. Perform field consistency and field validity checks on the candidate fields, and use the candidate fields that pass the checks as key fields.

4. The method according to claim 1, characterized in that, The process of fusing the transaction performance data, enterprise registration data, and enterprise tax credit data of the financing applicant yields multi-dimensional fused data of the financing applicant, including: The transaction performance data, corporate registration data, and corporate tax credit data of the financing applicants are processed in a standardized format. By using preset data cleaning rules, invalid data is filtered out from the transaction performance data, corporate registration data, and corporate tax credit data of the financing applicant. By performing data duplication checks, duplicate data is removed from the transaction performance data, corporate registration data, and corporate tax credit data of the financing applicant. Data completion technology is used to repair missing or abnormal fields in the transaction performance data, corporate registration data, and corporate tax credit data of the financing applicant. By using a preset unique primary key, the processed transaction performance data, corporate registration data, and corporate tax credit data of the financing applicant are integrated to obtain multi-dimensional integrated data; wherein, the unique primary key is the unique identifier of the financing applicant.

5. The method according to claim 1, characterized in that, In response to a financing request from a financing applicant for a target financing product, the system obtains the applicant's transaction performance data, business registration data, and tax credit data via an online interface, including: In response to a financing request for a target financing product initiated by a financing applicant on a pre-defined online procurement and transaction platform, the system receives transaction performance data and authorization verification data pushed by the financing applicant from the online procurement and transaction platform via an online interface. Based on the authorized verification data of the financing applicant, the enterprise entity registration data of the financing applicant is obtained from the enterprise information registration platform through the online interface, and the enterprise tax credit data of the financing applicant is obtained from the tax collection and management platform. The transaction performance data, the enterprise registration data, and the enterprise tax credit data are all encrypted using encryption algorithms and transmitted via dedicated network lines.

6. The method according to claim 1, characterized in that, The determination of the financing applicant's financing access result based on the multi-dimensional fusion data of the financing applicant includes: Obtain a pre-constructed credit assessment table; wherein the credit assessment table includes at least one assessment dimension, each assessment dimension corresponds to at least one scoring factor, each scoring factor corresponds to a factor weight and a factor range, each factor range includes at least two factor values, and different factor values ​​correspond to different credit scores; Based on the multidimensional fusion data of the financing applicant and the credit assessment form, the credit assessment result of the financing applicant is determined according to the preset credit assessment calculation formula. Based on the credit assessment results, the financing access outcome for the financing applicant is determined.

7. The method according to claim 6, characterized in that, The method further includes: The value of the credit assessment coefficient is determined based on the credit assessment score range in which the credit assessment result falls; the value of the credit assessment coefficient is different for different credit assessment score ranges. Based on the multidimensional fusion data of the financing applicant and the value of the credit assessment coefficient, the estimated credit limit is determined according to the preset credit limit calculation formula. Accordingly, generating the final financing result for the financing applicant based on the cross-validation results includes: The estimated credit limit is adjusted based on the cross-validation results to determine the final credit limit, which is then used as the financing result for the financing applicant.

8. The method according to claim 1, characterized in that, The method further includes: After the financing is completed, the enterprise registration data, transaction performance data, enterprise tax credit data and repayment information of the financing applicant will be collected periodically. By using a risk prediction model, the post-loan repayment risk of the financing applicant is predicted based on the enterprise registration data, transaction performance data, enterprise tax credit data, and repayment information of the financing applicant, so as to achieve dynamic monitoring and early warning of post-loan repayment risk.

9. A data processing apparatus, characterized in that, include: The data acquisition module is used to respond to the financing request of the financing applicant for the target financing product by acquiring the transaction performance data, enterprise registration data and enterprise tax credit data of the financing applicant through the online interface; The data fusion module is used to perform data fusion processing on the transaction performance data, enterprise registration data and enterprise tax credit data of the financing applicant to obtain multi-dimensional fused data of the financing applicant; The financing access judgment module is used to determine the financing access result of the financing applicant based on the multi-dimensional fusion data of the financing applicant; The data cross-validation module is used to obtain the supplementary financing data uploaded by the financing applicant when the financing access result is approved, and to perform cross-validation processing with the multi-dimensional fused data. The financing determination module is used to generate the final financing result for the financing applicant based on the cross-validation results, and to display the final financing result to the financing applicant.

10. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the method of any one of claims 1-8.

12. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-8.

Images