A cross-platform agricultural product supplier credit evaluation and risk control method

Abstract

The application relates to the field of computer data processing and discloses a cross-platform agricultural product supplier credit evaluation and risk control method. A server obtains cross-platform performance data from an agricultural product transaction platform, a logistics platform, a storage platform, a detection platform and a settlement platform, generates a supplier master identifier, and generates a batch performance unit according to agricultural product batch attributes; performance events in the same batch performance unit are generated into a performance event chain according to time, and quantity consistency parameters, time effectiveness consistency parameters and quality consistency parameters are determined based on the performance event chain; then, platform data reliability is generated according to comparison results of transaction platform data and logistics, storage, detection and settlement data, and cross-platform selective performance risk is identified under the same supplier, the same agricultural product category, the same production area and the same harvesting period; the server generates a supplier credit level according to the above, and outputs risk control instructions to an order processing module, a random inspection task generation module and a settlement review module.

Description

Technical Field

[0001] This invention relates to the field of computer data processing, and more specifically to a method for cross-platform agricultural product supplier credit assessment and risk control. Background Technology

[0002] Agricultural product supply is characterized by dispersed production areas, distinct harvesting cycles, and high requirements for timely distribution. Suppliers typically connect to multiple trading platforms simultaneously and cooperate with logistics, warehousing, testing, and settlement platforms to fulfill their supply obligations. A supplier's credit status directly impacts platform order allocation, sampling arrangements, settlement audits, and supplier access management.

[0003] Current credit assessments of agricultural product suppliers are mostly based on historical orders, complaint records, fulfillment rates, test results, or basic enterprise information on a single trading platform. Some solutions also incorporate logistics data, quality testing data, or supply chain traceability data to help determine the supplier's fulfillment capabilities and the quality risks of agricultural products.

[0004] However, existing methods typically use the supplier's overall records or records from a single platform as the basis for evaluation, making it difficult to categorize order, warehousing, logistics, testing, receipt, after-sales, and settlement data from different platforms into the same agricultural product batch for correlation verification. When suppliers selectively disclose performance data, transfer abnormal batches, or conceal quality disputes across different platforms, platforms struggle to promptly identify genuine performance risks and implement targeted controls on orders at different performance stages. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a cross-platform agricultural product supplier credit assessment and risk control method to solve the technical problems existing in the prior art.

[0006] The above-mentioned technical objective of the present invention is achieved through the following technical solution: A method for cross-platform agricultural product supplier credit assessment and risk control includes the following steps: S1: Obtain cross-platform fulfillment data from multiple agricultural product trading platforms, logistics platforms, warehousing platforms, testing platforms, and settlement platforms, and perform field standardization and timestamp unification processing on the cross-platform fulfillment data; S2: Generate a supplier master identifier based on the supplier identity data in the cross-platform fulfillment data, so that supplier accounts corresponding to the same agricultural product supplier on different platforms are merged into the same supplier master identifier; S3: Generate batch fulfillment units based on the supplier's main identifier and agricultural product batch attributes, and group data records that meet the batch association conditions into the same batch fulfillment unit; S4: Generate a performance event chain based on the occurrence time of performance events in the same batch of performance units; S5: Generate quantity consistency parameters, timeliness consistency parameters, and quality consistency parameters based on the performance event chain; S6: Compare the data provided by the agricultural product trading platform with the data provided by the logistics platform, warehousing platform, testing platform and settlement platform, and generate the platform data credibility based on the comparison results; S7: Under the same supplier main identifier, the same agricultural product category, the same production area and the same harvest cycle, based on the quantity consistency parameters, timeliness consistency parameters, quality consistency parameters and platform data credibility corresponding to different agricultural product trading platforms, identify cross-platform selective performance risks and generate cross-platform selective performance risk markers; S8: Generate the supplier credit rating corresponding to the supplier master identifier based on the quantity consistency parameter, timeliness consistency parameter, quality consistency parameter, platform data credibility, and cross-platform selective performance risk marker; S9: Based on the supplier's credit rating, cross-platform selective performance risk markers, and the performance node of the corresponding order in the performance event chain, output risk control instructions to the order processing module, sampling task generation module, and settlement review module of the corresponding agricultural product trading platform; S10: After the fulfillment of the contract by the subsequent batch of fulfillment units, the corresponding actual fulfillment results will be fed back to the batch fulfillment units, and the platform data credibility and supplier credit rating will be updated based on the feedback results.

[0007] Preferably, the cross-platform fulfillment data in S1 includes supplier identity data, order data, agricultural product batch data, harvesting data, warehousing inbound and outbound data, logistics status data, testing result data, receipt data, after-sales data, and settlement data of agricultural product suppliers; The field standardization and timestamp unification processing includes: The supplier account field in different platforms is unified as platform supplier account, the agricultural product name field is unified as agricultural product category code, the place of origin field is unified as production area code, and the order number, logistics number, testing certificate number, storage batch number and settlement flow number are written into the corresponding unique identifier field respectively. Order confirmation time, harvesting time, outbound time, testing time, shipping time, receipt time, after-sales occurrence time, and settlement time are all converted into event times under the same time base. Data records with the same unique identifier field and the same event time within the same platform are merged, while retaining the data source identifier of the data record.

[0008] Preferably, the supplier identity data in S2 includes two or more of the following: unified social credit code, business entity name, payment account, shipping address, place of origin registration information, and contact person information; The generated supplier master identifier includes: When supplier identity data on different platforms have the same unified social credit code, the corresponding supplier accounts will be merged into the same supplier master identifier. When supplier identity data on different platforms lacks a unified social credit code, the business entity name, payment account, shipping address, and place of origin registration information are compared. When the business entity name is the same, and two of the following information are consistent: the receiving account, the shipping address, and the place of origin registration information, the corresponding supplier accounts will be merged into the same supplier master identifier. When there are conflicting fields between the business entity name, payment account, shipping address and place of origin registration information, different supplier master identifiers will be generated for the corresponding supplier accounts, and the conflicting fields will be written into the supplier identity conflict record.

[0009] Preferably, the agricultural product batch attributes in S3 include agricultural product category, production area, harvest time, packaging specifications, shipment quantity, order time, logistics tracking number, testing certificate number, and storage batch number; The generation of batch fulfillment units includes: First, the supplier's main identifier, agricultural product category, and production area are used as batch aggregation fields to initially aggregate order data, warehouse inbound and outbound data, logistics status data, test result data, receipt data, after-sales data, and settlement data; Then, data records with harvest time differences not exceeding the first time threshold, delivery time differences not exceeding the second time threshold, and identical test certificate numbers or storage batch numbers are grouped into the same batch fulfillment unit. When both the testing certificate number and the storage batch number are missing, data records with consistent agricultural product category, production area, packaging specifications, logistics tracking number, and shipment quantity, and with a harvest time difference not exceeding the first time threshold, are classified into the same batch fulfillment unit. The first and second time thresholds are pre-configured in the category timeliness parameter table according to the agricultural product category; The harvesting cycle is determined as follows: Read the start date and end date of the harvest cycle from the harvest cycle parameter table according to the agricultural product category and production area; Batch fulfillment units whose harvesting time falls between the start date and end date of the same harvesting cycle are grouped into the same harvesting cycle; When there is no corresponding record for agricultural product category and production area in the harvest cycle parameter table, the earliest harvest time under the same supplier main identifier, the same agricultural product category and the same production area shall be used as the cycle start time, and the fourth time threshold after the cycle start time shall be used as the cycle end time. The fourth time threshold is pre-configured in the harvest cycle parameter table according to the agricultural product category.

[0010] Preferably, generating the performance event chain in S4 includes: The order confirmation event, inventory preparation event, outbound event, inspection event, shipment event, in-transit status event, receipt event, after-sales event, and settlement event in the same batch of fulfillment units are identified as fulfillment events; For each performance event, an event record is generated, which includes the event type, event object identifier, event time, data source identifier, and event status value; Event records are arranged from earliest to latest according to the event time; When there are duplicate event records with the same event type and the same event status value under the same event object identifier, the event record with the latest event time is retained; When there are event records with the same event type but different event status values ​​under the same event object identifier, the priority data source is determined according to the order of detection platform, warehousing platform, logistics platform, settlement platform, and agricultural product trading platform, and the event records that are not adopted are written into the event conflict record.

[0011] Preferably, the quantity consistency parameter, timeliness consistency parameter, and quality consistency parameter generated in S5 include: Quantity consistency parameters are generated based on the correspondence between order quantity, outbound quantity, transportation quantity, test sample quantity, and signed quantity in the fulfillment event chain; Calculate the first quantity difference between the order quantity and the outbound quantity, the second quantity difference between the order quantity and the transportation quantity, the third quantity difference between the order quantity and the signed quantity, and the fourth quantity difference between the outbound quantity and the number of test samples. The maximum value among the first, second, third, and fourth quantity difference values ​​is taken as the batch quantity difference value, and the quantity consistency parameter is determined based on the proportion of the batch quantity difference value to the order quantity. When the order quantity is zero or the order quantity field is missing, the batch fulfillment unit is marked as having an unavailable quantity parameter, and the calculation of supplier credit rating is stopped using the batch fulfillment unit. Generate timeliness consistency parameters based on the correspondence between the promised delivery time, actual outbound time, logistics pickup time, signing time, harvesting time, testing time, and testing validity period; The time from harvesting to delivery is determined based on the harvesting time and the actual delivery time; the time from delivery to pickup is determined based on the actual delivery time and the logistics pickup time; the transportation time is determined based on the logistics pickup time and the signing time; and the valid expiration time of testing is determined based on the testing time and the testing validity period. When the receipt time is later than the valid deadline for testing, a test validity conflict flag is generated; When the transportation time exceeds the shelf life threshold corresponding to the agricultural product category, a transportation timeliness conflict marker is generated. Based on the conflict markers of the test validity period, the conflict markers of the transportation timeliness, and the time difference between the promised delivery time and the actual delivery time, the timeliness consistency parameters are determined. Quality consistency parameters are generated based on the correspondence between test results, after-sales reasons, return reasons, and settlement deduction reasons. Write the test item code and test conclusion from the test result data into the test result field; write the after-sales reason code, return reason code and complaint object from the after-sales data into the after-sales result field; write the deduction reason code, deduction amount and review status from the settlement data into the settlement result field. When the test result is unqualified, and the after-sales reason code, return reason code, or deduction reason code points to the same quality reason of agricultural product category, a quality result consistency mark is generated; When the test result is qualified, but the after-sales reason code, return reason code, or deduction reason code points to the same quality reason of agricultural product category, a quality result conflict mark is generated. The quality consistency parameters are determined based on the quality result consistency markers and quality result conflict markers. The shelf life threshold is pre-configured in the category timeliness parameter table according to the agricultural product category.

[0012] Preferably, the credibility of the platform data generated in S6 includes: The order data, after-sales data, and settlement status data provided by the agricultural product trading platform are compared with the logistics status data provided by the logistics platform, the warehousing inbound and outbound data provided by the warehousing platform, the testing result data provided by the testing platform, and the settlement data provided by the settlement platform. Using agricultural product trading platforms, supplier main identifiers, and agricultural product categories as the objects of credibility calculation, the number of times data is missing and the number of times data conflicts occur in consecutively completed batch fulfillment units for these objects of credibility calculation. The number of missing data entries includes the number of times logistics status data is missing, the number of times test result data is missing, the number of times after-sales data is not synchronized, and the number of times the reason for settlement deduction is not reported. The number of data conflicts includes the number of times the order quantity is inconsistent with the outbound quantity, the number of times the signing time is inconsistent with the logistics signing time, and the number of times the test conclusion and after-sales result fields are inconsistent. The credibility of the platform data includes a first credibility level, a second credibility level, a third credibility level, and a fourth credibility level. The first credibility level is higher than the second credibility level, the second credibility level is higher than the third credibility level, and the third credibility level is higher than the fourth credibility level. When the same agricultural product trading platform has complete order data but missing logistics status data, missing test result data, unsynchronized after-sales data, or no feedback on the reason for settlement deduction in multiple batches of fulfillment units, and the number of missing data reaches the first threshold, the credibility of the platform data of the corresponding agricultural product trading platform under the corresponding supplier main identifier and the corresponding agricultural product category will be downgraded by one level. When the number of data conflicts reaches the second threshold, the credibility of the platform data of the corresponding agricultural product trading platform under the corresponding supplier main identifier and the corresponding agricultural product category will be downgraded by one level; When no data loss or data conflict occurs in the third consecutive batch of fulfillment units that meet the threshold number of consecutive completions, the corresponding platform data credibility will be upgraded by one level. The first number threshold, the second number threshold, and the third number threshold are pre-configured in the credibility update parameter table according to the agricultural product category.

[0013] Preferably, identifying cross-platform selective performance risks in S7 includes: Under the same supplier main identifier, the same agricultural product category, the same production area and the same harvest cycle, the batch fulfillment units corresponding to different agricultural product trading platforms are divided into the first platform batch group and the second platform batch group; When any one of the following appears in the first platform batch group: inspection failure mark, delayed receipt mark, return mark, after-sales complaint mark, or settlement deduction mark, and any one of the following appears in the second platform batch group: missing inspection certificate mark, batch change mark, order quantity decrease mark, or early settlement mark, a cross-platform selective performance risk mark is generated. The decrease in order quantity is marked as the number of orders in the second platform batch group during the current harvest cycle being lower than the product of the average number of orders in the first three completed harvest cycles of the same agricultural product category and the same production area of ​​the supplier main identifier and the first proportion threshold. The advance settlement marker is defined as a settlement time earlier than the receipt time, or a settlement time earlier than the time the test result data is written. The first proportional threshold and the preset consecutive number threshold are pre-configured in the risk identification parameter table according to the agricultural product category.

[0014] Preferably, generating supplier credit ratings and outputting risk control instructions in S8 and S9 includes: A batch risk record is generated for each batch of performance units. The batch risk record includes quantity consistency parameters, timeliness consistency parameters, quality consistency parameters, platform data credibility, and cross-platform selective performance risk markers. The risk records of the fifth most recent batch under the same supplier main identifier and the same agricultural product category were selected as the samples for credit rating calculation. When there are cross-platform selective performance risk markers in the credit rating calculation sample, or when there are quality result conflict markers and the corresponding platform data credibility is at the third or fourth credibility level, the supplier credit rating will be determined as a restricted level. When there are no cross-platform selective performance risk markers in the credit rating calculation sample, and the quantity consistency parameter, timeliness consistency parameter, and quality consistency parameter are all within the corresponding qualified range, and the data credibility of the corresponding platform is at the first or second credibility level, the supplier's credit rating will be determined as normal. In all other cases, the supplier's credit rating will be set to the observation level; The fulfillment node of the corresponding order is determined based on the latest event record in the fulfillment event chain. When the latest event record is an order confirmation event or a stock preparation event, the corresponding order is determined to be an unshipped node; when the latest event record is a shipment event or a transit status event, the corresponding order is determined to be a transit node; when the latest event record is a receipt event and there is no settlement event, the corresponding order is determined to be a signed but unsettled node. When the corresponding order is in the unshipped stage and there is a cross-platform selective fulfillment risk mark, output a new order limit instruction to the order processing module and a supplementary inspection certificate instruction to the sampling task generation module. When the corresponding order is in transit and there are conflict marks for the validity period of the inspection or the transportation timeliness, the sampling task generation module is output with instructions to increase the sampling ratio upon arrival and to supplement the cold chain transportation certificate, and the sampling ratio field of the corresponding order is updated to a target sampling ratio that is higher than the basic sampling ratio of the agricultural product category. When the corresponding order is in the "signed but not settled" stage and there is a quality result conflict mark or a settlement deduction mark, output a "temporary settlement" instruction and a "settlement review" instruction to the settlement review module, and update the settlement status field of the corresponding order to "pending review" status; When the number of times the same supplier's main identifier appears with the cross-platform selective fulfillment risk mark under the same agricultural product category reaches the preset consecutive number threshold, an instruction to suspend the supply of that agricultural product category will be output to the order processing module. The fifth threshold and qualified range are pre-configured in the credit rating parameter table according to the agricultural product category.

[0015] Preferably, the update of platform data credibility and supplier credit rating in S10 includes: After subsequent batches of fulfillment units complete signing, inspection, after-sales service, and settlement, read the actual signed quantity, actual signing time, inspection conclusion, after-sales reason code, return reason code, deduction reason code, and settlement status of that subsequent batch of fulfillment units; Write the actual quantity received, actual receipt time, test results, after-sales reason code, return reason code, deduction reason code, and settlement status into the corresponding batch risk record; The platform data credibility corresponding to the subsequent batch fulfillment unit is recalculated based on the updated batch risk record. The supplier's credit rating is reassessed based on the recalculated platform data credibility and the fifth most recent batch risk record under the supplier's main identifier.

[0016] In summary, the present invention has the following main beneficial effects: This application unifies the performance data from multiple agricultural product trading platforms, logistics platforms, warehousing platforms, testing platforms, and settlement platforms under a single supplier master identifier. Furthermore, it generates batch performance units based on agricultural product category, production area, harvest time, delivery time, tracking number, testing certificate number, and warehousing batch number. This achieves the effect of unifying order, warehousing, logistics, testing, receipt, after-sales, and settlement data, which were originally scattered across different platforms, into a single batch processing object. As a result, the server no longer relies solely on a single platform's supplier account or historical transaction records for credit assessment. Instead, it can identify the supplier's true performance status across different platforms based on specific agricultural product batches, reducing credit judgment biases caused by fragmented accounts across multiple platforms, mixed batch data, and missing testing certificates.

[0017] This application generates a fulfillment event chain within a batch fulfillment unit and determines quantity consistency parameters, timeliness consistency parameters, and quality consistency parameters based on the fulfillment event chain. This achieves consistency verification of the same batch across multiple fulfillment nodes, including order quantity, outbound quantity, transported quantity, inspection coverage quantity, signed quantity, inspection validity period, transportation duration, after-sales reasons, and settlement deduction reasons. This allows for the detection of mismatches between order data and warehousing, logistics, inspection, after-sales, and settlement data. In particular, it can identify inconsistencies between inspection results and subsequent after-sales, return, or deduction reasons, providing a clear data source, calculation basis, and fulfillment chain support for the generation of supplier credit ratings.

[0018] This application reverse-calibrates the data credibility of agricultural product trading platforms by utilizing data provided by logistics, warehousing, testing, and settlement platforms. It identifies cross-platform selective fulfillment risks under the same supplier main identifier, agricultural product category, production area, and harvesting cycle, thereby identifying and controlling risks such as suppliers selectively disclosing fulfillment data, transferring abnormal batches, concealing quality anomalies, or engaging in differentiated fulfillment across different platforms. Simultaneously, by outputting risk control instructions to the order processing module, sampling task generation module, and settlement review module based on the order's stage (unshipped, in transit, or signed but not settled), it achieves tiered processing of orders at different fulfillment stages, transforming risk control from simple risk alerts into executable platform field updates and process controls. Attached Figure Description

[0019] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0021] Example 1 refer to Figure 1 A method for cross-platform agricultural product supplier credit assessment and risk control includes the following steps: S1: Obtain cross-platform fulfillment data from multiple agricultural product trading platforms, logistics platforms, warehousing platforms, testing platforms, and settlement platforms, and perform field standardization and timestamp unification processing on the cross-platform fulfillment data; S2: Generate a supplier master identifier based on the supplier identity data in the cross-platform fulfillment data, so that supplier accounts corresponding to the same agricultural product supplier on different platforms are merged into the same supplier master identifier; S3: Generate batch fulfillment units based on the supplier's main identifier and agricultural product batch attributes, and group data records that meet the batch association conditions into the same batch fulfillment unit; S4: Generate a performance event chain based on the occurrence time of performance events in the same batch of performance units; S5: Generate quantity consistency parameters, timeliness consistency parameters, and quality consistency parameters based on the performance event chain; S6: Compare the data provided by the agricultural product trading platform with the data provided by the logistics platform, warehousing platform, testing platform and settlement platform, and generate the platform data credibility based on the comparison results; S7: Under the same supplier main identifier, the same agricultural product category, the same production area and the same harvest cycle, based on the quantity consistency parameters, timeliness consistency parameters, quality consistency parameters and platform data credibility corresponding to different agricultural product trading platforms, identify cross-platform selective performance risks and generate cross-platform selective performance risk markers; S8: Generate the supplier credit rating corresponding to the supplier master identifier based on the quantity consistency parameter, timeliness consistency parameter, quality consistency parameter, platform data credibility, and cross-platform selective performance risk marker; S9: Based on the supplier's credit rating, cross-platform selective performance risk markers, and the performance node of the corresponding order in the performance event chain, output risk control instructions to the order processing module, sampling task generation module, and settlement review module of the corresponding agricultural product trading platform; S10: After the fulfillment of the contract by the subsequent batch of fulfillment units, the corresponding actual fulfillment results will be fed back to the batch fulfillment units, and the platform data credibility and supplier credit rating will be updated based on the feedback results.

[0022] The method is executed by a server. The server communicates with multiple agricultural product trading platforms, logistics platforms, warehousing platforms, testing platforms, and settlement platforms. The server is configured with a data access module, a field standardization module, a supplier merging module, a batch fulfillment unit generation module, a fulfillment event chain generation module, a consistency parameter calculation module, a platform data credibility update module, a selective fulfillment risk identification module, a credit rating generation module, a risk control instruction output module, and a fulfillment feedback update module.

[0023] The agricultural product trading platform provides order data, supplier account data, agricultural product transaction status data, after-sales data, and transaction settlement status data; the logistics platform provides tracking numbers, pickup time, transit status, receipt time, and transport quantity; the warehousing platform provides warehousing batch numbers, warehousing time, outbound time, outbound quantity, and warehousing circulation status; the testing platform provides testing certificate numbers, testing time, testing item codes, testing conclusions, testing validity periods, and the number of testing samples; and the settlement platform provides settlement transaction numbers, settlement time, deduction reason codes, deduction amounts, and review status.

[0024] This embodiment does not directly use supplier ratings from a single agricultural product trading platform, nor does it simply weight transaction data from multiple platforms. Instead, it first merges data from different platforms under the same supplier master identifier and the same batch fulfillment unit. Then, it forms a fulfillment event chain based on the batch fulfillment unit, generating quantity consistency parameters, timeliness consistency parameters, and quality consistency parameters on the fulfillment event chain. Subsequently, the server uses data from the logistics platform, warehousing platform, testing platform, and settlement platform to reverse-calibrate the data credibility of the agricultural product trading platform. Under the same supplier master identifier, the same agricultural product category, the same production area, and the same harvest cycle, it identifies cross-platform selective fulfillment risks. Finally, it outputs risk control instructions to the order processing module, the sampling task generation module, and the settlement review module according to the fulfillment node of the order.

[0025] Through the above processing, this embodiment can avoid the distortion of credit assessment caused by suppliers using different accounts on different platforms, the same agricultural product batch being split and recorded, platform data not being synchronized, and the test results not matching the actual delivery cycle.

[0026] Acquisition and standardization of cross-platform fulfillment data: The server obtains cross-platform fulfillment data through API calls, data file imports, or message queue reception. This cross-platform fulfillment data includes supplier identity data, order data, agricultural product batch data, harvesting data, warehousing and inventory data, logistics status data, testing result data, receipt data, after-sales data, and settlement data.

[0027] Supplier identity data includes the Unified Social Credit Code, business entity name, platform supplier account, payment account, shipping address, place of origin registration information, and contact person information. Order data includes order number, order quantity, order amount, order confirmation time, promised delivery time, agricultural product category, and platform. Agricultural product batch data includes agricultural product category, production area, harvest time, packaging specifications, quantity shipped, tracking number, testing certificate number, and storage batch number. Warehousing inbound / outbound data includes storage batch number, inbound time, outbound time, outbound quantity, and outbound status. Logistics status data includes tracking number, logistics pickup time, transported quantity, transit status, receipt time, and logistics receipt quantity. Testing result data includes testing certificate number, testing item code, testing time, testing conclusion, testing validity period, and number of tested samples. After-sales data includes after-sales occurrence time, after-sales reason code, return reason code, complaint target, and after-sales processing status. Settlement data includes settlement transaction number, settlement time, settlement status, deduction reason code, deduction amount, and review status.

[0028] The field standardization module unifies the supplier account field from different platforms into the platform supplier account, the agricultural product name field into the agricultural product category code, the place of origin field into the production area code, and writes the order number, logistics tracking number, testing certificate number, warehousing batch number, and settlement transaction number into their respective unique identifier fields.

[0029] The server also performs unified processing on time fields. Order confirmation time, receiving time, outbound time, inspection time, shipping time, receipt time, after-sales occurrence time, and settlement time are all converted to event times under the same time base. For situations where different platforms have differences in time zones, time formats, or time precision, the server unifies them into timestamps of the same format according to pre-configured time conversion rules.

[0030] If data records with the same unique identifier field and the same event time exist within the same platform, the server will merge these data records and retain the data source identifier. The data source identifier is used to record whether the data comes from an agricultural product trading platform, logistics platform, warehousing platform, testing platform, or settlement platform. In the event of subsequent event status conflicts, the server can determine the priority data source based on the data source identifier and write the unused data record into the event conflict record.

[0031] Generation of the supplier's main identifier: The supplier consolidation module generates a primary supplier identifier based on supplier identity data. This primary identifier is used to consolidate supplier accounts belonging to the same agricultural product supplier across different platforms.

[0032] When supplier identity data across different platforms shares the same Unified Social Credit Code, the server merges the corresponding supplier accounts under the same supplier master identifier. When supplier identity data across different platforms lacks a Unified Social Credit Code, the server compares the business entity name, payment account, shipping address, and place of origin registration information. If the business entity name is identical, and two of the payment account, shipping address, and place of origin registration information are also identical, the server merges the corresponding supplier accounts under the same supplier master identifier.

[0033] When there are conflicting fields among the business entity name, payment account, shipping address, and place of origin registration information, the server does not forcibly merge the corresponding supplier accounts. Instead, it generates different primary supplier identifiers for each field and writes the conflicting fields into the supplier identity conflict record. The supplier identity conflict record includes the platform supplier account involved in the conflict, the name of the conflicting field, the value of the conflicting field, the data source identifier, and the generation time.

[0034] The above processing can prevent the incorrect merging of different suppliers based solely on similar names or the same contact person, and can also prevent the same supplier from using different accounts on different platforms, which would cause the performance data to be split.

[0035] Generation of batch fulfillment units: The batch fulfillment unit generation module generates batch fulfillment units based on the supplier's main identifier and the agricultural product batch attributes. The agricultural product batch attributes include agricultural product category, production area, harvest time, packaging specifications, shipment quantity, order time, tracking number, testing certificate number, and storage batch number.

[0036] The server first uses the supplier's main identifier, agricultural product category, and production area as batch aggregation fields to initially aggregate order data, warehouse inbound / outbound data, logistics status data, test result data, receipt data, after-sales data, and settlement data. After completing the initial aggregation, the server then determines whether data records belong to the same batch fulfillment unit based on the harvest time, delivery time, test certificate number, and warehouse batch number.

[0037] When data records correspond to the same supplier main identifier, the same agricultural product category, and the same production area, and the harvest time difference is no greater than the first time threshold and the delivery time difference is no greater than the second time threshold, and the detection voucher number or the storage batch number is the same, the server will classify the data record into the same batch fulfillment unit.

[0038] When both the testing certificate number and the storage batch number are missing, the server further compares the agricultural product category, production area, packaging specifications, tracking number, and shipment quantity. If the agricultural product category, production area, packaging specifications, tracking number, and shipment quantity are all consistent, and the harvest time difference is no greater than the first time threshold, the corresponding data record is grouped into the same batch fulfillment unit. If the above fields still cannot meet the grouping conditions, the server treats the corresponding data record as a batch record awaiting confirmation. This batch record awaiting confirmation will not participate in the supplier credit rating calculation until the testing certificate number, storage batch number, or tracking number is subsequently supplemented before being grouped.

[0039] The first and second time thresholds are pre-configured in the category-specific timeliness parameter table according to the agricultural product category. This table, pre-established by the server management, records the harvest-to-warehouse restrictions, delivery time restrictions, testing validity rules, shelf-life thresholds, and allowed delivery delays for different agricultural product categories. For agricultural products with short post-harvest shelf-life, the first and second time thresholds are set to shorter times; for agricultural products with longer shelf-life, they are set to longer times. Specific values ​​are configured by the implementation platform based on the agricultural product category, production area, shelf-life, and fulfillment management rules.

[0040] This embodiment uses batch fulfillment units as the minimum processing object for subsequent credit assessment, rather than directly using the supplier's entire historical order history as the processing object. This avoids masking the fulfillment differences of the same supplier across different harvesting batches, different production areas, or different platforms by averaging them all together.

[0041] Determining the harvesting cycle: The server determines the harvesting cycle based on the harvesting cycle parameter table. The harvesting cycle parameter table includes the agricultural product category, production area, harvesting cycle start date, harvesting cycle end date, and a fourth time threshold.

[0042] When the agricultural product category and production area have corresponding records in the harvest cycle parameter table, the server reads the start date and end date of the harvest cycle and classifies batch fulfillment units whose harvest time is between the start date and end date of the same harvest cycle into the same harvest cycle.

[0043] When there is no corresponding record for the agricultural product category and production area in the harvest cycle parameter table, the server uses the earliest harvest time under the same supplier main identifier, the same agricultural product category, and the same production area as the cycle start time, and uses the fourth time threshold after the cycle start time as the cycle end time. The fourth time threshold is pre-configured in the harvest cycle parameter table according to the agricultural product category.

[0044] The harvesting cycle is used to subsequently determine whether the same supplier has selective fulfillment risks across platforms for the same agricultural product category and production area. Limiting risk identification to the same harvesting cycle can avoid interference from differences in supply capacity across different seasons and production areas.

[0045] Generation of the performance event chain: The fulfillment event chain generation module generates a fulfillment event chain based on the occurrence time of fulfillment events within the same batch of fulfillment units. Fulfillment events include order confirmation events, inventory preparation events, outbound events, inspection events, shipment events, in-transit status events, receipt events, after-sales events, and settlement events.

[0046] The server generates an event record for each fulfillment event. The event record includes the event type, event object identifier, event time, data source identifier, and event status value. The event type indicates order confirmation, preparation, outbound shipment, inspection, dispatch, in transit, receipt, after-sales service, or settlement. The event object identifier indicates the order number, tracking number, inspection certificate number, warehousing batch number, or settlement transaction number. The event time is a standardized timestamp. The data source identifier marks the event as originating from an agricultural product trading platform, logistics platform, warehousing platform, inspection platform, or settlement platform. The event status value indicates the specific status of the event, including confirmed, outbound, received, inspection passed, inspection failed, received, returned, settled, and pending review.

[0047] The server arranges event records from earliest to latest according to event time, obtaining a fulfillment event chain. When duplicate event records with the same event type and identical event status value exist under the same event object identifier, the server retains the event record with the latest event time. When event records with the same event type but different event status values ​​exist under the same event object identifier, the server determines the priority data source in the order of detection platform, warehousing platform, logistics platform, settlement platform, and agricultural product trading platform, and writes the unused event records into the event conflict record.

[0048] The reason for adopting the above priority order is that the testing platform, warehousing platform, logistics platform, and settlement platform typically correspond to specific testing certificates, warehousing inbound and outbound records, logistics status, and settlement flow, which better reflect the objective fulfillment process compared to the order display status on agricultural product trading platforms. The server does not delete unused event records but instead writes them into the event conflict log for subsequent updates to the platform's data credibility.

[0049] Generation of quantity consistency parameters: The consistency parameter calculation module generates quantity consistency parameters based on the correspondence between order quantity, outbound quantity, transportation quantity, inspection sample quantity, and signed quantity in the fulfillment event chain.

[0050] Before calculating quantities, the server first determines the order quantity, outbound quantity, transported quantity, and signed-for quantity. For the number of test samples, the server converts the number of test samples into the test coverage quantity according to the sampling rules corresponding to the agricultural product category. If the testing platform has already provided the test coverage quantity, it is used directly; if the testing platform only provides the sample quantity, it is converted into the test coverage quantity according to the sampling ratio or category sampling rules in the testing platform. The conversion rules are pre-stored in the category timeliness parameter table or the testing rule parameter table.

[0051] The quantity difference is calculated as follows: ; in, Indicates the percentage difference in batch quantity; Indicates the order quantity; Indicates the quantity shipped out; Indicates the quantity transported; Indicates the number of items received; Indicates the number of tests covered; This indicates that the maximum value among the multiple values ​​within the parentheses is selected. It represents the absolute value.

[0052] The server will determine the batch quantity difference ratio. Mapped to quantity consistency parameters Specifically: ; in, Indicates quantity consistency parameter; Indicates the percentage difference in batch quantity; This indicates taking the smaller value within the parentheses.

[0053] When the order quantity is zero or the order quantity field is missing, the server marks the batch of fulfillment units as having an unavailable quantity parameter and stops using that batch of fulfillment units to calculate the supplier credit rating. The batch of fulfillment units is still stored on the server, but is only used as a record to be supplemented.

[0054] This calculation method allows for the verification of order quantity, warehouse outbound quantity, logistics transportation quantity, inspection coverage quantity, and signed receipt quantity within the same batch of fulfillment units, rather than simply judging whether the quantities are consistent within a single platform.

[0055] Generation of timeliness consistency parameters: The consistency parameter calculation module generates timeliness consistency parameters based on the correspondence between the order's promised delivery time, actual outbound time, logistics pickup time, signing time, harvesting time, testing time, and testing validity period.

[0056] The server determines the time from harvesting to delivery based on the harvesting time and the actual delivery time, the time from delivery to pickup based on the actual delivery time and the logistics pickup time, the transportation time based on the logistics pickup time and the signing time, and the effective expiration time of the test based on the testing time and the test validity period.

[0057] The effective deadline for testing shall be determined as follows: ; in, Indicates the valid deadline for testing; Indicates the detection time; This indicates the validity period of the inspection. When the receipt time is later than the valid inspection deadline, the server generates a conflict flag for the inspection validity period. When the transportation time exceeds the shelf-life threshold corresponding to the agricultural product category, the server generates a transportation timeliness conflict flag. The shelf-life threshold is pre-configured in the category timeliness parameter table according to the agricultural product category. Timeliness consistency parameters are determined as follows: ; in, Indicates parameters for consistency in timeliness; This indicates a conflict flag for the validity period of the test. The value is 1 when the receipt time is later than the validity period deadline of the test, and 0 otherwise. This indicates a transportation timeliness conflict flag. The value is 1 when the transportation time exceeds the shelf life threshold, and 0 otherwise. This indicates the percentage delay between the promised delivery time and the actual shipment time. , and These represent the weights corresponding to conflicts in the validity period of testing, conflicts in transportation timeliness, and delays in delivery. The weights are configured in the category timeliness parameter table according to the agricultural product category.

[0058] in, The delay ratio is determined as follows: when the actual outbound time is later than the promised delivery time of the order, the server determines the delay ratio based on the time difference between the two and the allowable delivery delay duration corresponding to the agricultural product category; when the actual outbound time is not later than the promised delivery time of the order, the delay ratio is zero. The allowable delivery delay duration is configured in the category timeliness parameter table.

[0059] The weights and thresholds in the above formula are not fixed values, but are configured by the implementation platform based on the agricultural product category, production area, shelf life, and fulfillment management rules. The server reads the corresponding parameters from the category timeliness parameter table during calculation.

[0060] Generation of quality consistency parameters: The consistency parameter calculation module generates quality consistency parameters based on the correspondence between test results, after-sales reasons, return reasons, and settlement deduction reasons.

[0061] The server writes the test item code and test conclusion from the test result data into the test result field, writes the after-sales reason code, return reason code and complaint object from the after-sales data into the after-sales result field, and writes the deduction reason code, deduction amount and review status from the settlement data into the settlement result field.

[0062] When the test result is unqualified, and the after-sales reason code, return reason code, or deduction reason code points to a quality reason within the same agricultural product category, the server generates a quality result consistency marker and saves this marker as a quality anomaly in the batch risk record. When the test result is qualified, but the after-sales reason code, return reason code, or deduction reason code points to a quality reason within the same agricultural product category, the server generates a quality result conflict marker. When test result data is missing in a batch fulfillment unit, and an after-sales reason code, return reason code, or deduction reason code exists in the same batch fulfillment unit, the server generates a quality data missing marker.

[0063] The server determines the quality consistency parameter based on quality result conflict markers and quality data missing markers. When a quality result conflict marker exists, the quality consistency parameter is determined to be the non-compliant range; when a quality data missing marker exists but no quality result conflict marker exists, the quality consistency parameter is determined to be the observation range; when neither a quality result conflict marker nor a quality data missing marker exists, the quality consistency parameter is determined to be the compliant range. The non-compliant range, observation range, and compliant range are pre-configured in the credit rating parameter table according to the agricultural product category.

[0064] In this embodiment, the quality consistency parameter is not determined solely by whether the test is qualified or unqualified, but rather by combining the test results, after-sales reasons, reasons for returns, and reasons for settlement deductions. This can identify situations where the test results are inconsistent with the actual after-sales anomalies, preventing suppliers from using a single test result to cover up subsequent quality disputes.

[0065] Generation and updating of platform data credibility: The platform data credibility update module compares the order data, after-sales data, and settlement status data provided by the agricultural product trading platform with the logistics status data provided by the logistics platform, the warehousing inbound and outbound data provided by the warehousing platform, the testing result data provided by the testing platform, and the settlement data provided by the settlement platform, and generates the platform data credibility based on the comparison results.

[0066] The server uses agricultural product trading platforms, supplier main identifiers, and agricultural product categories as the objects of credibility calculation, and counts the number of data missing and data conflict times of these objects in consecutively completed batch fulfillment units.

[0067] Data missing counts include missing logistics status data, missing test result data, unsynchronized after-sales data, and unreported reasons for settlement deductions. Data conflict counts include discrepancies between order quantity and outbound quantity, discrepancies between delivery time and logistics delivery time, and discrepancies between test conclusions and after-sales results.

[0068] The platform's data credibility is divided into four levels: Level 1, Level 2, Level 3, and Level 4. Level 1 credibility is higher than Level 2, Level 2 credibility is higher than Level 3, and Level 3 credibility is higher than Level 4. When a server first connects to the platform, the platform's data credibility is set to Level 2. Subsequent levels are adjusted upwards or downwards based on the number of missing data points and data conflicts.

[0069] When the same agricultural product trading platform experiences multiple fulfillment units where order data is complete but logistics status data, testing result data, after-sales data is not synchronized, or settlement deduction reasons are not reported, and the number of times the corresponding data is missing reaches the first threshold, the server will downgrade the platform's data credibility by one level under the corresponding supplier's main identifier and the corresponding agricultural product category. When the number of data conflicts reaches the second threshold, the server will downgrade the platform's data credibility by one level under the corresponding supplier's main identifier and the corresponding agricultural product category. When no data is missing or data conflicts occur in the third consecutive fulfillment unit, the server will upgrade the platform's data credibility by one level.

[0070] The first, second, and third count thresholds are pre-configured in the credibility update parameter table according to the agricultural product category. A continuously completed batch fulfillment unit refers to a batch fulfillment unit that has generated a receipt event and a settlement event, and whose detection event and after-sales event status have been written into the fulfillment event chain.

[0071] This processing method means that the server does not default to accepting the data from the agricultural product trading platform. Instead, it uses relatively objective performance data provided by the logistics platform, warehousing platform, testing platform, and settlement platform to dynamically update the data credibility of the agricultural product trading platform for specific suppliers and specific agricultural product categories.

[0072] Identification of cross-platform selective performance risks: The selective performance risk identification module divides the batch performance units corresponding to different agricultural product trading platforms into the first platform batch group and the second platform batch group under the same supplier main identifier, the same agricultural product category, the same production area and the same harvest cycle.

[0073] When any of the following flags appear in the first platform batch group: unqualified inspection, delayed receipt, return, after-sales complaint, or settlement deduction, and any of the following flags appear in the second platform batch group: missing inspection certificate, batch change, order quantity decrease, or early settlement, the server generates a cross-platform selective performance risk flag.

[0074] The order quantity decrease flag is determined as follows: ; in, Indicates the percentage change in order quantity; This indicates the number of orders for the second platform batch group within the current harvesting cycle; , and These represent the number of orders placed by the supplier's main identifier on the corresponding platform within the first three completed harvesting cycles for the same agricultural product category and the same production area.

[0075] when When the number of orders falls below a first proportion threshold, the server generates a flag indicating a decrease in the number of orders. The first proportion threshold is pre-configured in the risk identification parameter table according to the agricultural product category. If there are missing periods within the first three completed harvesting periods, the server selects completed harvesting periods from which orders can be obtained to calculate the average number of orders; if there are fewer than two completed harvesting periods from which orders can be obtained, the server does not generate a flag indicating a decrease in the number of orders and writes the batch group of that platform into the insufficient sample record.

[0076] The early settlement marker is determined based on the settlement time and the receipt time or the time the test result data is written in the performance event chain. The server generates an early settlement marker when the settlement time is earlier than the receipt time or the settlement time is earlier than the time the test result data is written.

[0077] Batch change markings are determined based on the testing certificate number, warehouse batch number, logistics tracking number, and harvesting time. Normal batch switching rules include at least one of the following situations: the harvesting time of the current batch fulfillment unit belongs to the next harvesting cycle; the testing certificate number, warehouse batch number, and logistics tracking number of the current batch fulfillment unit all correspond to the new order; the quantity shipped from the current batch fulfillment unit is consistent with the quantity of the new order, and there are no markings indicating non-compliance, return, after-sales complaint, or settlement deduction.

[0078] When a batch change occurs within the same harvesting cycle, and the batch fulfillment unit before the change already has a test failure mark, return mark, after-sales complaint mark, or settlement deduction mark, and the batch fulfillment unit after the change lacks any of the following: test certificate number, warehouse batch number, or logistics tracking number, the server generates a batch change mark.

[0079] This embodiment identifies selective performance risk across platforms, rather than default risk within a single platform. The determination of selective performance risk must meet four conditions: the same supplier's main identifier, the same agricultural product category, the same production area, and the same harvest cycle. This avoids misjudgments caused by differences in supply capacity due to different product categories, production areas, or seasons.

[0080] The credit rating generation module generates a batch risk record for each batch of performance units. The batch risk record includes quantity consistency parameters, timeliness consistency parameters, quality consistency parameters, platform data credibility, and cross-platform selective performance risk markers.

[0081] The server selects the most recently completed fifth-time threshold risk records under the same supplier main identifier and the same agricultural product category as the sample for credit rating calculation. The fifth-time threshold is pre-configured in the credit rating parameter table according to the agricultural product category.

[0082] When there are cross-platform selective performance risk markers in the credit rating calculation sample, or when there are quality result conflict markers and the corresponding platform data credibility is at the third or fourth credibility level, the server will determine the supplier's credit rating as a restricted level.

[0083] When there are no cross-platform selective performance risk markers in the credit rating calculation sample, and the quantity consistency parameter, timeliness consistency parameter, and quality consistency parameter are all within the corresponding qualified range, and the corresponding platform data credibility is at the first credibility level or the second credibility level, the server will determine the supplier's credit rating as normal.

[0084] In all other cases, the server will set the supplier's credit rating to the observation level.

[0085] This embodiment employs a rule-based credit rating determination method, rather than a black-box model. Quantity consistency parameters, timeliness consistency parameters, quality consistency parameters, platform data credibility, and cross-platform selective performance risk markers can all be directly determined from batch risk records stored on the server. This ensures that supplier credit ratings have clear data sources and calculation bases.

[0086] The risk control instruction output module determines the fulfillment node of the corresponding order based on the latest event record in the fulfillment event chain.

[0087] When the latest event record is an order confirmation event or a stock preparation event, the server will determine the corresponding order as an unshipped node. When the latest event record is a shipment event or a transit status event, the server will determine the corresponding order as a transit node. When the latest event record is a receipt event and there is no settlement event, the server will determine the corresponding order as a received but unsettled node.

[0088] If both a receipt event and a settlement event exist in the fulfillment event chain, the server will designate the corresponding order as a settled node. Settled nodes do not trigger a delayed settlement instruction, but their after-sales results, return reasons, and deduction reasons are still written into the batch risk record for subsequent updates to platform data credibility and supplier credit rating.

[0089] By identifying fulfillment nodes, the server can execute different risk control instructions for orders in different statuses, avoiding the use of the same risk handling method for all orders.

[0090] Based on the supplier's credit rating, cross-platform selective performance risk markers, and the performance node of the order, the server outputs risk control instructions to the order processing module, sampling task generation module, and settlement review module of the corresponding agricultural product trading platform.

[0091] Risk control instructions include instruction number, supplier master identifier, order number, agricultural product category code, batch fulfillment unit identifier, instruction type, target module, execution field, execution value, generation time, and receipt status.

[0092] When a corresponding order is in the unshipped stage and has a cross-platform selective fulfillment risk marker, the server outputs a new order limit instruction to the order processing module and a supplementary inspection certificate instruction to the sampling task generation module. The new order limit instruction includes a limit value. The server determines the limit value based on the supplier's credit rating, the basic order limit, and the limit control coefficient. The basic order limit is determined based on the number of orders corresponding to the fifth most recent batch fulfillment unit completed under the same agricultural product category by the supplier's main identifier, or it can be pre-configured by the agricultural product trading platform; the limit control coefficient is pre-configured in the credit rating parameter table according to the supplier's credit rating. After receiving the new order limit instruction, the order processing module updates the new order limit field to the limit value. After receiving the supplementary inspection certificate instruction, the sampling task generation module updates the inspection certificate status field of the corresponding order to "pending supplementation."

[0093] When a corresponding order is in transit and there are conflicting indicators for inspection validity or transportation timeliness, the server outputs an instruction to increase the arrival sampling ratio and a supplementary cold chain transportation certificate instruction to the sampling task generation module. The server determines the target sampling ratio based on the basic sampling ratio, risk enhancement coefficient, and maximum sampling ratio corresponding to the agricultural product category. These three ratios are pre-configured in the credit rating parameter table or risk identification parameter table. Upon receiving the instruction to increase the arrival sampling ratio, the sampling task generation module updates the sampling ratio field of the corresponding order to the target sampling ratio.

[0094] In this embodiment, cold chain transportation credentials include cold chain logistics documents, temperature records, transportation equipment numbers, temperature curves during transportation periods, or temperature-controlled transportation credentials recognized by the platform. For agricultural product categories marked as requiring cold chain transportation in the category timeliness parameter table, the server requires supplementary cold chain logistics documents and temperature records during transportation periods; for agricultural product categories not marked as requiring cold chain transportation in the category timeliness parameter table but with transportation timeliness conflict markers, the server requires supplementary transportation timeliness credentials and writes these credentials into the extended record of the cold chain transportation credential field.

[0095] When an order is in the "received but not settled" stage and has a quality result conflict flag or a settlement deduction flag, the server outputs a "delay settlement" instruction and a "settlement review" instruction to the settlement review module. Upon receiving the "delay settlement" instruction, the settlement review module updates the settlement status field of the corresponding order from "pending settlement" to "pending review" and updates the "settlement release" field to "prohibited release." After the settlement review task is completed, the settlement review module updates the settlement status field to "allow settlement" or maintains the "pending review" status based on the review results. The review results are written to the batch risk record and serve as the data source for subsequent updates to platform data credibility and supplier credit rating.

[0096] When the number of times the same supplier's main identifier appears with a cross-platform selective fulfillment risk marker under the same agricultural product category reaches a preset consecutive count threshold, the server outputs a suspension instruction to the order processing module. Upon receiving this instruction, the order processing module updates the supply status field of the supplier's main identifier under the corresponding agricultural product category to a suspended status. The preset consecutive count threshold is pre-configured in the risk identification parameter table according to the agricultural product category.

[0097] In this embodiment, the risk control instructions are not simply manual management suggestions, but executable data instructions sent by the server to the corresponding platform module. Control is achieved by updating the order amount field, inspection voucher status field, sampling ratio field, cold chain transportation voucher field, settlement status field, settlement release field, and supply status field.

[0098] After a subsequent batch of fulfillment units completes its fulfillment, the fulfillment feedback update module reads the actual quantity received, actual receipt time, inspection results, after-sales reason code, return reason code, deduction reason code, and settlement status of that subsequent batch of fulfillment units.

[0099] In this embodiment, "fulfillment of fulfillment in a subsequent batch" means that a receipt event and a settlement event have been generated in that subsequent batch, and the status of the detection event and after-sales event has been written into the fulfillment event chain. When no after-sales event occurs in the subsequent batch, the server records the after-sales event status as "no after-sales"; when no return event occurs in the subsequent batch, the server records the return reason code as a null value and treats this null value as a "no return" status. If a receipt event, detection event, or settlement event is missing, the subsequent batch will not be considered a completed batch fulfillment unit and will not participate in the recalculation of the supplier credit rating.

[0100] The server writes the actual quantity received, actual receipt time, inspection results, after-sales reason code, return reason code, deduction reason code, and settlement status into the corresponding batch risk record. After the batch risk record is updated, the server recalculates the platform data credibility corresponding to the subsequent batch fulfillment unit, and re-determines the supplier's credit rating based on the recalculated platform data credibility and the fifth most recently completed threshold batch risk record under the supplier's main identifier.

[0101] When subsequent performance results show that the risk corresponding to the previously generated risk marker has been eliminated, the server updates the platform data credibility according to the corresponding platform data credibility adjustment rules. When subsequent performance results show that after-sales anomalies, quality conflicts, or settlement deductions that were not previously synchronized by the trading platform have occurred, the server increases the corresponding platform's data missing count or data conflict count and adjusts the platform data credibility according to the credibility update parameter table.

[0102] Through this feedback process, the supplier credit rating is not a static result generated all at once, but is continuously updated based on the performance results of subsequent batches.

[0103] In this embodiment, the first time threshold, second time threshold, fourth time threshold, shelf life threshold, first number threshold, second number threshold, third number threshold, first proportion threshold, fifth number threshold, preset consecutive number threshold, weight, basic sampling ratio, risk enhancement coefficient, maximum sampling ratio, quota control coefficient, and qualified range are respectively configured in the category timeliness parameter table, harvesting cycle parameter table, credibility update parameter table, risk identification parameter table, and credit rating parameter table.

[0104] The aforementioned parameter table is pre-established by the server management system. The initial records in the parameter table are generated based on agricultural product category, production area, platform fulfillment rules, testing validity rules issued by testing institutions, logistics service timeliness rules, warehousing and outbound management rules, and completed batch fulfillment records. For agricultural product categories with stable fulfillment records, the server uses the harvest time, outbound time, logistics pickup time, signing time, and testing validity period of completed batch fulfillment units within the same production area as the statistical basis to determine the first time threshold, second time threshold, fourth time threshold, and shelf-life threshold. For agricultural product categories lacking historical fulfillment records, the server uses the initial parameters configured for that agricultural product category by the platform management system and updates the corresponding parameter table based on the actual fulfillment results after subsequent batch fulfillment units complete signing, testing, after-sales service, and settlement.

[0105] The thresholds in the parameter table are not generated by the server during a single risk assessment, but are read in advance before batch performance unit generation, timeliness consistency calculation, platform data credibility update, selective performance risk identification, and credit rating calculation. The server records the parameter table version number used in each calculation, so that the calculation process corresponding to the same batch of performance units can be traced.

[0106] The parameter table above does not require different agricultural product categories to use the same parameters. Different agricultural product categories, different production areas, and different fulfillment modes can correspond to different parameter records. Through parameter table configuration, this embodiment can be adapted to different agricultural product supply scenarios without changing the core processing flow.

[0107] Conventional supplier credit ratings typically use supplier company information, historical order volume, number of complaints, fulfillment rate, or financial data as the basis for evaluation, mainly outputting a supplier credit score or risk level. This type of method usually evaluates the supplier as a whole, making it difficult to identify situations where the same supplier selectively discloses abnormal fulfillment data, transfers abnormal batches, or conceals quality conflicts across different platforms.

[0108] This embodiment uses batch fulfillment units as the smallest processing object and unifies order, warehousing, logistics, testing, receipt, after-sales service, and settlement into a fulfillment event chain. The server first determines the consistency of a single batch in terms of quantity, timeliness, and quality. Then, it uses data from the logistics platform, warehousing platform, testing platform, and settlement platform to reverse-calibrate the data credibility of the agricultural product trading platform. Finally, it identifies cross-platform selective fulfillment risks under the same supplier main identifier, the same agricultural product category, the same production area, and the same harvest cycle.

[0109] The risk control instructions output in this embodiment are not simply risk warnings, but rather executable instructions targeting the order processing module, the sampling task generation module, and the settlement review module. Control is achieved through updates to the order amount field, sampling ratio field, inspection certificate status field, cold chain transportation certificate field, settlement status field, settlement release field, and supply status field. Therefore, this embodiment has clearly defined data processing objects, data processing flows, and module execution relationships.

[0110] After the server accesses cross-platform fulfillment data from multiple platforms, it first standardizes the data fields and timestamps from different platforms. The standardized data then enters the supplier merging module, which generates a primary supplier identifier based on the unified social credit code, business entity name, payment account, shipping address, and place of origin registration information. Next, the batch fulfillment unit generation module generates batch fulfillment units based on the primary supplier identifier, agricultural product category, production area, harvest time, packaging specifications, shipping quantity, order time, tracking number, testing certificate number, and warehouse batch number.

[0111] After the same batch of fulfillment units is generated, the fulfillment event chain generation module forms a fulfillment event chain based on the occurrence time of order confirmation events, inventory preparation events, outbound events, inspection events, shipment events, in-transit status events, receipt events, after-sales events, and settlement events. The consistency parameter calculation module calculates quantity consistency parameters, timeliness consistency parameters, and quality consistency parameters on the fulfillment event chain. The platform data credibility update module compares the data provided by the agricultural product trading platform with the data provided by the logistics platform, warehousing platform, inspection platform, and settlement platform to generate and update the platform data credibility.

[0112] The selective performance risk identification module compares batch performance units across different agricultural product trading platforms under the same supplier main identifier, same agricultural product category, same production area, and same harvest cycle. When one platform experiences quality anomalies, timeliness anomalies, or settlement anomalies, and another platform experiences missing testing certificates, batch changes, decreased order quantities, or early settlement, the server generates a cross-platform selective performance risk marker.

[0113] The credit rating generation module determines the supplier's credit rating based on quantity consistency parameters, timeliness consistency parameters, quality consistency parameters, platform data credibility, and cross-platform selective performance risk markers. The risk control instruction output module outputs corresponding risk control instructions to the order processing module, sampling task generation module, and settlement review module based on the supplier's credit rating, cross-platform selective performance risk markers, and the order's current performance stage. After subsequent batch performance is completed, the server updates the batch risk record based on the actual signed quantity, actual signed time, inspection results, after-sales reason code, return reason code, deduction reason code, and settlement status, and recalculates the platform data credibility and supplier credit rating.

[0114] Through the above-described process, this embodiment can transform supplier credit assessment from a static evaluation into a closed-loop process based on batch fulfillment chains, platform credibility, and risk control instructions in cross-platform agricultural product supply scenarios.

[0115] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for cross-platform agricultural product supplier credit assessment and risk control, characterized in that, Includes the following steps: S1: Obtain cross-platform fulfillment data from multiple agricultural product trading platforms, logistics platforms, warehousing platforms, testing platforms, and settlement platforms, and perform field standardization and timestamp unification processing on the cross-platform fulfillment data; S2: Generate a supplier master identifier based on the supplier identity data in the cross-platform fulfillment data, so that supplier accounts corresponding to the same agricultural product supplier on different platforms are merged into the same supplier master identifier; S3: Generate batch fulfillment units based on the supplier's main identifier and agricultural product batch attributes, and group data records that meet the batch association conditions into the same batch fulfillment unit; S4: Generate a performance event chain based on the occurrence time of performance events in the same batch of performance units; S5: Generate quantity consistency parameters, timeliness consistency parameters, and quality consistency parameters based on the performance event chain; S6: Compare the data provided by the agricultural product trading platform with the data provided by the logistics platform, warehousing platform, testing platform and settlement platform, and generate the platform data credibility based on the comparison results; S7: Under the same supplier main identifier, the same agricultural product category, the same production area and the same harvest cycle, based on the quantity consistency parameters, timeliness consistency parameters, quality consistency parameters and platform data credibility corresponding to different agricultural product trading platforms, identify cross-platform selective performance risks and generate cross-platform selective performance risk markers; S8: Generate the supplier credit rating corresponding to the supplier master identifier based on the quantity consistency parameter, timeliness consistency parameter, quality consistency parameter, platform data credibility, and cross-platform selective performance risk marker; S9: Based on the supplier's credit rating, cross-platform selective performance risk markers, and the performance node of the corresponding order in the performance event chain, output risk control instructions to the order processing module, sampling task generation module, and settlement review module of the corresponding agricultural product trading platform; S10: After the fulfillment of the contract by the subsequent batch of fulfillment units, the corresponding actual fulfillment results will be fed back to the batch fulfillment units, and the platform data credibility and supplier credit rating will be updated based on the feedback results.

2. The method for cross-platform agricultural product supplier credit assessment and risk control according to claim 1, characterized in that, S1's cross-platform fulfillment data includes supplier identity data, order data, agricultural product batch data, harvesting data, warehousing inbound and outbound data, logistics status data, testing result data, receipt data, after-sales data, and settlement data for agricultural product suppliers. The field standardization and timestamp unification processing includes: The supplier account field in different platforms is unified as platform supplier account, the agricultural product name field is unified as agricultural product category code, the place of origin field is unified as production area code, and the order number, logistics number, testing certificate number, storage batch number and settlement flow number are written into the corresponding unique identifier field respectively. Order confirmation time, harvesting time, outbound time, testing time, shipping time, receipt time, after-sales occurrence time, and settlement time are all converted into event times under the same time base. Data records with the same unique identifier field and the same event time within the same platform are merged, while retaining the data source identifier of the data record.

3. The method for cross-platform agricultural product supplier credit assessment and risk control according to claim 2, characterized in that, Supplier identity data in S2 includes two or more of the following: unified social credit code, business entity name, payment account, shipping address, place of origin registration information, and contact person information; The generated supplier master identifier includes: When supplier identity data on different platforms have the same unified social credit code, the corresponding supplier accounts will be merged into the same supplier master identifier. When supplier identity data on different platforms lacks a unified social credit code, the business entity name, payment account, shipping address, and place of origin registration information are compared. When the business entity name is the same, and two of the following information are consistent: the receiving account, the shipping address, and the place of origin registration information, the corresponding supplier accounts will be merged into the same supplier master identifier. When there are conflicting fields between the business entity name, payment account, shipping address and place of origin registration information, different supplier master identifiers will be generated for the corresponding supplier accounts, and the conflicting fields will be written into the supplier identity conflict record.

4. The method for cross-platform agricultural product supplier credit assessment and risk control according to claim 3, characterized in that, The batch attributes of agricultural products in S3 include agricultural product category, production area, harvest time, packaging specifications, shipment quantity, order time, logistics tracking number, testing certificate number, and storage batch number; The generation of batch fulfillment units includes: First, the supplier's main identifier, agricultural product category, and production area are used as batch aggregation fields to initially aggregate order data, warehouse inbound and outbound data, logistics status data, test result data, receipt data, after-sales data, and settlement data; Then, data records with harvest time differences not exceeding the first time threshold, delivery time differences not exceeding the second time threshold, and identical test certificate numbers or storage batch numbers are grouped into the same batch fulfillment unit. When both the testing certificate number and the storage batch number are missing, data records with consistent agricultural product category, production area, packaging specifications, logistics tracking number, and shipment quantity, and with a harvest time difference not exceeding the first time threshold, are classified into the same batch fulfillment unit. The first and second time thresholds are pre-configured in the category timeliness parameter table according to the agricultural product category; The harvesting cycle is determined as follows: Read the start date and end date of the harvest cycle from the harvest cycle parameter table according to the agricultural product category and production area; Batch fulfillment units whose harvesting time falls between the start date and end date of the same harvesting cycle are grouped into the same harvesting cycle; When there is no corresponding record for agricultural product category and production area in the harvest cycle parameter table, the earliest harvest time under the same supplier main identifier, the same agricultural product category and the same production area shall be used as the cycle start time, and the fourth time threshold after the cycle start time shall be used as the cycle end time. The fourth time threshold is pre-configured in the harvest cycle parameter table according to the agricultural product category.

5. The method for cross-platform agricultural product supplier credit assessment and risk control according to claim 4, characterized in that, The chain of performance events generated in S4 includes: Order confirmation events, stock preparation events, outbound events, inspection events, shipment events, in-transit status events, receipt events, after-sales events, and settlement events within the same batch of fulfillment units are identified as fulfillment events; For each performance event, an event record is generated, which includes the event type, event object identifier, event time, data source identifier, and event status value; Event records are arranged from earliest to latest according to the event time; When there are duplicate event records with the same event type and the same event status value under the same event object identifier, the event record with the latest event time is retained; When there are event records with the same event type but different event status values ​​under the same event object identifier, the priority data source is determined according to the order of detection platform, warehousing platform, logistics platform, settlement platform, and agricultural product trading platform, and the event records that are not adopted are written into the event conflict record.

6. The method for cross-platform agricultural product supplier credit assessment and risk control according to claim 5, characterized in that, The quantity consistency parameters, timeliness consistency parameters, and quality consistency parameters generated in S5 include: Quantity consistency parameters are generated based on the correspondence between order quantity, outbound quantity, transportation quantity, test sample quantity, and signed quantity in the fulfillment event chain; Calculate the first quantity difference between the order quantity and the outbound quantity, the second quantity difference between the order quantity and the transportation quantity, the third quantity difference between the order quantity and the signed quantity, and the fourth quantity difference between the outbound quantity and the number of test samples. The maximum value among the first, second, third, and fourth quantity difference values ​​is taken as the batch quantity difference value, and the quantity consistency parameter is determined based on the proportion of the batch quantity difference value to the order quantity. When the order quantity is zero or the order quantity field is missing, the batch fulfillment unit is marked as having an unavailable quantity parameter, and the calculation of supplier credit rating is stopped using the batch fulfillment unit. Generate timeliness consistency parameters based on the correspondence between the promised delivery time, actual outbound time, logistics pickup time, signing time, harvesting time, testing time, and testing validity period; The time from harvesting to delivery is determined based on the harvesting time and the actual delivery time; the time from delivery to pickup is determined based on the actual delivery time and the logistics pickup time; the transportation time is determined based on the logistics pickup time and the signing time; and the valid expiration time of testing is determined based on the testing time and the testing validity period. When the receipt time is later than the valid deadline for testing, a test validity conflict flag is generated; When the transportation time exceeds the shelf life threshold corresponding to the agricultural product category, a transportation timeliness conflict marker is generated. Based on the conflict markers of the test validity period, the conflict markers of the transportation timeliness, and the time difference between the promised delivery time and the actual delivery time, the timeliness consistency parameters are determined. Quality consistency parameters are generated based on the correspondence between test results, after-sales reasons, return reasons, and settlement deduction reasons. Write the test item code and test conclusion from the test result data into the test result field; write the after-sales reason code, return reason code and complaint object from the after-sales data into the after-sales result field; write the deduction reason code, deduction amount and review status from the settlement data into the settlement result field. When the test result is unqualified, and the after-sales reason code, return reason code, or deduction reason code points to the same quality reason of agricultural product category, a quality result consistency mark is generated; When the test result is qualified, but the after-sales reason code, return reason code, or deduction reason code points to the same quality reason of agricultural product category, a quality result conflict mark is generated. The quality consistency parameters are determined based on the quality result consistency markers and quality result conflict markers. The shelf life threshold is pre-configured in the category timeliness parameter table according to the agricultural product category.

7. The method for cross-platform agricultural product supplier credit assessment and risk control according to claim 6, characterized in that, The credibility of platform data generated in S6 includes: The order data, after-sales data, and settlement status data provided by the agricultural product trading platform are compared with the logistics status data provided by the logistics platform, the warehousing inbound and outbound data provided by the warehousing platform, the testing result data provided by the testing platform, and the settlement data provided by the settlement platform. Using agricultural product trading platforms, supplier main identifiers, and agricultural product categories as the objects of credibility calculation, the number of times data is missing and the number of times data conflicts occur in consecutively completed batch fulfillment units for these objects of credibility calculation. The number of missing data entries includes the number of times logistics status data is missing, the number of times test result data is missing, the number of times after-sales data is not synchronized, and the number of times the reason for settlement deduction is not reported. The number of data conflicts includes the number of times the order quantity is inconsistent with the outbound quantity, the number of times the signing time is inconsistent with the logistics signing time, and the number of times the test conclusion and after-sales result fields are inconsistent. The credibility of the platform data includes a first credibility level, a second credibility level, a third credibility level, and a fourth credibility level. The first credibility level is higher than the second credibility level, the second credibility level is higher than the third credibility level, and the third credibility level is higher than the fourth credibility level. When the same agricultural product trading platform has complete order data but missing logistics status data, missing test result data, unsynchronized after-sales data, or no feedback on the reason for settlement deduction in multiple batch fulfillment units, and the number of missing data reaches the first threshold, the platform data credibility of the corresponding agricultural product trading platform under the corresponding supplier main identifier and the corresponding agricultural product category will be downgraded by one level. When the number of data conflicts reaches the second threshold, the credibility of the platform data of the corresponding agricultural product trading platform under the corresponding supplier main identifier and the corresponding agricultural product category will be downgraded by one level; When no data loss or data conflict occurs in the third consecutive batch of fulfillment units that meet the threshold number of consecutive completions, the corresponding platform data credibility will be upgraded by one level. The first number threshold, the second number threshold, and the third number threshold are pre-configured in the credibility update parameter table according to the agricultural product category.

8. The method for cross-platform agricultural product supplier credit assessment and risk control according to claim 7, characterized in that, Identifying cross-platform selective performance risks in S7 includes: Under the same supplier main identifier, the same agricultural product category, the same production area and the same harvest cycle, the batch fulfillment units corresponding to different agricultural product trading platforms are divided into the first platform batch group and the second platform batch group; When any one of the following appears in the first platform batch group: inspection failure mark, delayed receipt mark, return mark, after-sales complaint mark, or settlement deduction mark, and any one of the following appears in the second platform batch group: missing inspection certificate mark, batch change mark, order quantity decrease mark, or early settlement mark, a cross-platform selective performance risk mark is generated. The decrease in order quantity is marked as the number of orders in the second platform batch group during the current harvest cycle being lower than the product of the average number of orders in the first three completed harvest cycles of the same agricultural product category and the same production area of ​​the supplier main identifier and the first proportion threshold. The advance settlement marker is defined as a settlement time earlier than the receipt time, or a settlement time earlier than the time the test result data is written. The first proportional threshold and the preset consecutive number threshold are pre-configured in the risk identification parameter table according to the agricultural product category.

9. The method for cross-platform agricultural product supplier credit assessment and risk control according to claim 8, characterized in that, The generation of supplier credit ratings and the output of risk control instructions in S8 and S9 include: A batch risk record is generated for each batch of performance units. The batch risk record includes quantity consistency parameters, timeliness consistency parameters, quality consistency parameters, platform data credibility, and cross-platform selective performance risk markers. The risk records of the fifth most recent batch under the same supplier main identifier and the same agricultural product category were selected as the samples for credit rating calculation. When there are cross-platform selective performance risk markers in the credit rating calculation sample, or when there are quality result conflict markers and the corresponding platform data credibility is at the third or fourth credibility level, the supplier credit rating will be determined as a restricted level. When there are no cross-platform selective performance risk markers in the credit rating calculation sample, and the quantity consistency parameter, timeliness consistency parameter, and quality consistency parameter are all within the corresponding qualified range, and the data credibility of the corresponding platform is at the first or second credibility level, the supplier's credit rating will be determined as normal. In all other cases, the supplier's credit rating will be set to the observation level; The fulfillment node of the corresponding order is determined based on the latest event record in the fulfillment event chain. When the latest event record is an order confirmation event or a stock preparation event, the corresponding order is determined to be an unshipped node; when the latest event record is a shipment event or a transit status event, the corresponding order is determined to be a transit node; when the latest event record is a receipt event and there is no settlement event, the corresponding order is determined to be a signed but unsettled node. When the corresponding order is in the unshipped stage and there is a cross-platform selective fulfillment risk mark, output a new order limit instruction to the order processing module and a supplementary inspection certificate instruction to the sampling task generation module. When the corresponding order is in transit and there are conflict marks for the validity period of the inspection or the transportation timeliness, the sampling task generation module is output with instructions to increase the sampling ratio upon arrival and to supplement the cold chain transportation certificate, and the sampling ratio field of the corresponding order is updated to a target sampling ratio that is higher than the basic sampling ratio of the agricultural product category. When the corresponding order is in the "signed but not settled" stage and there is a quality result conflict mark or a settlement deduction mark, output a "temporary settlement" instruction and a "settlement review" instruction to the settlement review module, and update the settlement status field of the corresponding order to "pending review" status; When the number of times the same supplier's main identifier appears with the cross-platform selective fulfillment risk mark under the same agricultural product category reaches the preset consecutive number threshold, an instruction to suspend the supply of that agricultural product category will be output to the order processing module. The fifth threshold and qualified range are pre-configured in the credit rating parameter table according to the agricultural product category.

10. The method for cross-platform agricultural product supplier credit assessment and risk control according to claim 9, characterized in that, The updated platform data credibility and supplier credit rating in S10 include: After subsequent batches of fulfillment units complete signing, inspection, after-sales service, and settlement, read the actual signed quantity, actual signing time, inspection conclusion, after-sales reason code, return reason code, deduction reason code, and settlement status of that subsequent batch of fulfillment units; Write the actual quantity received, actual receipt time, test results, after-sales reason code, return reason code, deduction reason code, and settlement status into the corresponding batch risk record; The platform data credibility corresponding to the subsequent batch fulfillment unit is recalculated based on the updated batch risk record. The supplier's credit rating is reassessed based on the recalculated platform data credibility and the fifth most recent batch risk record under the supplier's main identifier.

Images