Supplier determination method, apparatus, and system

Abstract

The application relates to a supplier determination method, device and system. The method comprises the following steps: analyzing order information, obtaining a supplier identifier and an order commodity list; mapping each commodity in the order commodity list to a leaf node of a supply capability topology to obtain an order capability node set; obtaining an authentication record of the supplier according to the supplier identifier, mapping each authentication item in the authentication record to a leaf node of the supply capability topology to obtain an effective authentication node set; judging whether the order capability node set matches the effective authentication node set; obtaining a credit limit and an occupied limit of the supplier, calculating an available undertaking limit of the supplier based on the credit limit and the occupied limit; evaluating the qualification risk of the supplier according to the judgment result, determining a comprehensive risk value by combining the qualification risk, the available undertaking limit and the order amount, and determining whether to enable the supplier corresponding to the order information according to the comprehensive risk value.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a supplier selection method, apparatus and system. Background Technology

[0002] In existing supplier selection schemes, category matching and credit limit assessment are independent of each other. Category matching often uses a tree-structured intersection judgment logic, which cannot guarantee that the supplier has the capability for all minimum categories involved in the order. Credit limit assessment is based on a static credit granting model, which is difficult to reflect the supplier's real-time available credit limit at the time of order submission, resulting in insufficient accuracy in supplier selection. Summary of the Invention

[0003] Therefore, it is necessary to provide a supplier selection method, device, and system that can improve the accuracy of supplier selection in response to the above-mentioned technical problems.

[0004] To address the aforementioned technical problems, firstly, a supplier selection method is provided, the method comprising: Parse the order information to obtain the supplier identifier and the list of items in the order; Map each product in the order product list to a leaf node of the supply capacity topology to obtain the order capacity node set; Obtain the supplier's certification record based on the supplier identifier, map each certification item in the certification record to the leaf node of the supply capability topology, and obtain the set of valid certification nodes. The leaf node represents the smallest category node that can be referenced by the product and can be covered by the certification record. Determine whether the set of order capability nodes matches the set of valid authentication nodes, and obtain the determination result; Obtain the supplier's credit limit and the amount already used, and calculate the supplier's available credit limit based on the credit limit and the amount already used; The supplier's qualification risk is assessed based on the judgment results. A comprehensive risk value is determined by combining the qualification risk, available credit limit, and order amount. Based on the comprehensive risk value, it is determined whether to activate the supplier for the order corresponding to the order information.

[0005] In one embodiment, each product in the order product list is mapped to a leaf node of the supply capacity topology, resulting in a set of order capacity nodes including: Iterate through the product identifiers of each product in the order product list and search for nodes that match the product identifiers in the supply capacity topology; If a node matching the product identifier is found and the matching node is a leaf node, then the matching node will be used as the order capability node corresponding to the product. If a node matching the product identifier is found and the matching node is not a leaf node, then the child node of the matching node will be used as the order capability node corresponding to the product. If no node matching the product identifier is found, an error identifier is generated.

[0006] In one embodiment, finding a node matching the product identifier in the supply capacity topology includes: Extract the semantic feature vector of the product identifier. The semantic feature vector includes at least one of the core word vector, attribute word vector, and relation word vector. Starting from the leaf node level of the supply capacity topology, the semantic feature vectors of each node at the current level are obtained step by step, and the similarity between the semantic feature vector of the product identifier and the semantic feature vectors of each node at the current level is calculated. If there are nodes at the current level with a similarity greater than or equal to a preset threshold, then the node with the highest similarity will be selected as the order capability node. If there are no nodes with a similarity greater than or equal to the preset threshold at the current level, then trace back to the parent level of the current level, re-acquire the semantic feature vectors of each node in the parent level and calculate the similarity, until tracing back to the root node of the supply capacity topology.

[0007] In one embodiment, re-acquiring the semantic feature vectors of each node in the parent hierarchy and calculating the similarity includes: If a node with a similarity greater than or equal to a preset threshold is found in the parent level, the child node of the found node will be used as the order capability node, and the depth of the upstream level will be recorded. If the depth of the upstream hierarchy exceeds the preset depth threshold, a supply capacity topology update will be triggered.

[0008] In one embodiment, the supply capacity topology update includes: Obtain each existing node in the supply capacity topology, and calculate the global semantic distance of the supply capacity topology after inserting the product as a new node under each existing node; Calculate the difference between the global semantic distance of the supply capacity topology before insertion and the global semantic distance after insertion, and determine the structural semantic offset corresponding to each insertion position based on the ratio of each difference to the global semantic distance before insertion. The global semantic distance is the sum of the semantic distances between nodes in the supply capacity topology; the semantic distance is determined based on the hierarchical path length between nodes and the name similarity between node names. Select the insertion position with the smallest structural semantic offset, create a new node at the insertion position, and use the new node as the order capability node.

[0009] In one embodiment, each authentication item in the authentication record is mapped to a leaf node of the provisioning capability topology to obtain a set of valid authentication nodes, including: Certifications with a status of "in effect", a validity period including the current time, and an applicable scope including the procurement organization and delivery area are selected from the certification records and designated as valid certifications. Determine the leaf node corresponding to each valid authentication item in the supply capacity topology, and form a set of valid authentication nodes from the determined leaf nodes.

[0010] In one embodiment, determining whether the set of order capability nodes matches the set of valid authentication nodes yields the following result: If all order capability nodes in the order capability node set exist in the valid authentication node set, the judgment result is passed; If at least one order capability node does not exist in the set of valid authentication nodes, the result is "failed" and the set of failed nodes is output. The set of failed nodes consists of all nodes that do not exist in the set of valid authentication nodes.

[0011] In one embodiment, obtaining the supplier's credit limit and used credit limit, and calculating the supplier's available credit limit based on the credit limit and used credit limit includes: Obtain multidimensional metrics from suppliers, including at least two of the following: delivery metrics, quality metrics, financial metrics, capacity metrics, and collaboration metrics. The comprehensive ability score is obtained by normalizing the multidimensional indicators and then summing them by weight. The supplier's basic credit limit is adjusted based on the comprehensive capability score to obtain a dynamic credit limit; The available credit limit is obtained by subtracting the occupied credit limit from the dynamic credit limit. The occupied credit limit includes the amount of unsettled amount in transit and / or the amount of frozen credit.

[0012] In one embodiment, the multidimensional indicators are normalized and then weighted and summed to obtain the comprehensive capability score, which includes: Dimensional indicators that are positively correlated with supplier capabilities are normalized using the first method. For dimensional indicators whose numerical values ​​are negatively correlated with supplier capabilities, normalization is performed using the second method. The collaboration index is obtained by weighted summation of order confirmation timeliness rate, change response timeliness rate, exception closure timeliness rate, and data completeness rate. The weighted sum of the normalized indicators across all dimensions yields the comprehensive capability score.

[0013] In one embodiment, the supplier's basic credit limit is adjusted based on the comprehensive capability score to obtain a dynamic credit limit. The available credit limit is obtained by subtracting the already used limit from the dynamic credit limit, including: Obtain the supplier's basic credit limit, benchmark capability score, capability adjustment coefficient, category risk coefficient, and data timeliness decay coefficient; The basic credit limit is adjusted based on the difference between the comprehensive ability score and the benchmark ability score, combined with the ability adjustment coefficient, to obtain the ability-adjusted credit limit. The adjusted credit limit is multiplied by the category risk coefficient and the data timeliness decay coefficient to obtain the dynamic credit limit. Obtain the supplier's used credit limit, which includes outstanding credit limit in transit and frozen credit limit; Subtracting the already used amount from the dynamic credit limit yields the available credit limit.

[0014] In one embodiment, the supplier's qualification risk is assessed based on the judgment result. A comprehensive risk value is determined by combining the qualification risk, available credit limit, and order amount. The determination of whether to enable the supplier for the order corresponding to the order information is based on the comprehensive risk value includes: When the judgment result is unsatisfactory, the supplier's qualification risk is assessed as the first risk value; When the judgment result is "pass", the supplier's qualification risk is assessed as the second risk value, and the first risk value is higher than the second risk value. Determine credit risk based on available credit limit and order amount; The overall risk value is obtained by weighting and summing the qualification risk and the credit limit risk. Whether to enable the supplier for the order corresponding to the order information is determined based on the comparison result between the comprehensive risk value and the preset threshold.

[0015] In one embodiment, before determining whether to enable a supplier for an order corresponding to the order information based on a comparison between a comprehensive risk value and a preset threshold, the method further includes: Determine whether the constraints are met. The constraints include at least one of the following: the supplier is suspended or blacklisted; the result is failure and the nodes in the failure node set fail because the certification has expired or the supplier certification is suspended; an abnormal mark is generated; the key certification is invalid; the order amount exceeds the single absolute limit; the available acceptance limit is zero and the order amount is greater than zero. If the constraints are met, the supplier will not be allowed to accept orders.

[0016] In one embodiment, determining whether to enable the supplier for an order corresponding to the order information based on a comparison between the comprehensive risk value and a preset threshold includes: If the constraints are not met and the overall risk value is less than the first threshold, then the supplier is allowed to accept the order. If the constraints are not met and the overall risk value is greater than or equal to the first threshold and less than the second threshold, then the supplier will not be allowed to accept the order and will proceed to the review process. If the constraints are not met and the overall risk value is greater than or equal to the second threshold, then the supplier will not be allowed to accept orders.

[0017] To address the aforementioned technical problems, a second aspect provides a supplier-determined device, comprising a memory for storing a computer program; a processor connected to the memory for executing the computer program; and a communication interface connected to the processor for data interaction with an external device; wherein the processor, when executing the computer program, implements the steps of the method described in the first aspect.

[0018] To address the aforementioned technical problems, in a third aspect, this application provides a supplier determination system, which includes: the supplier determination device described in the second aspect above, and external devices connected to the supplier determination device, including an order system, a data source system, and a configuration center.

[0019] Unlike related technologies, this application maps each product in the order's product list and the supplier's certification record to leaf nodes of the same supply capacity topology, obtaining a set of order capacity nodes and a set of valid certification nodes. This aligns order requirements and supplier certification scope at the finest category level. By determining whether the set of order capacity nodes matches the set of valid certification nodes, every smallest category involved in the order is included in the verification process, preventing partial category matching from leading to selection. By obtaining the supplier's credit limit and used credit limit, the available capacity of the supplier is calculated based on the credit limit and used credit limit, allowing the capacity to be dynamically updated as the supplier uses it in real time, overcoming the deficiency that fixed credit limits cannot reflect the supplier's current actual capacity. By assessing the supplier's qualification risk based on the judgment results, and combining qualification risk, available capacity, and order amount to determine a comprehensive risk value, the supplier selection is constrained by both category coverage capability and available capacity, avoiding selection based on only one aspect. By determining whether to activate the supplier for the order corresponding to the order information based on the comprehensive risk value, a differentiated supplier selection strategy is implemented, thereby improving the accuracy and reliability of supplier selection. Attached Figure Description

[0020] To more clearly illustrate the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the supplier determination system in one embodiment; Figure 2 This is a schematic diagram of the structure of the equipment determined by the supplier in one embodiment; Figure 3 This is a flowchart illustrating a supplier determination method in one embodiment; Figure 4 This is a schematic diagram of the supply capacity topology in one embodiment; Figure 5 This is a flowchart illustrating the supplier determination method in another embodiment; Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0023] Supplier onboarding and procurement scope control are important components of supply chain management. Related technologies typically maintain separate records for supplier basic information, qualification certificates, supplier ratings, procureable product categories, and credit limits. When creating a purchase order, procurement personnel select suppliers and products, and decide whether to allow submission based on general rules such as whether the supplier is enabled and whether the order fields are complete.

[0024] In procurement scenarios with multi-level product category systems, a supplier may only possess the production, processing, or supply capabilities for certain specific product categories. For example, a supplier's ability to process ordinary turned parts does not necessarily imply that it has the technological capabilities for precision shafts or high-precision gears. Matching only within supplier directories or broad product category levels cannot accurately determine whether each smallest product category unit in an order falls within the supplier's valid certification scope.

[0025] Credit limits are used to constrain the amount of purchases a supplier can undertake within a certain period. Related technologies use a fixed value from the supplier's master data as the credit limit, comparing only the amount of a single order with the fixed limit during verification, without considering recent deliveries, quality, financial soundness, capacity load, category risk, certification timeliness, outstanding amounts in transit, and frozen credit. This approach fails to reflect the supplier's actual performance and financial capacity at the current point in time. Therefore, whether a supplier possesses the capability for the target category and the current order amount they can undertake are two related but technically different issues. The former requires structured set operations on order categories and certification scope, while the latter requires standardization of heterogeneous indicators, dynamic credit limit calculation, and real-time occupancy control, resulting in a unified handling outcome within the same order control chain.

[0026] Supplier certificates, capability labels, and product categories may use different levels or codes, relying solely on broad category names or manual experience for matching. This can easily lead to precision parts orders being awarded to suppliers with only basic processing capabilities. Determining supplier eligibility by comparing order amounts and fixed credit lines fails to simultaneously assess a supplier's product category capabilities and current capacity. Furthermore, fixed credit lines do not reflect changes in supplier status. Fixed credit lines may remain unchanged even with recent delivery delays, quality issues, financial risks, capacity constraints, or certification expiration; orders in transit and frozen credit lines may also be reused. A binary "pass" or "fail" outcome fails to differentiate between medium-risk cases requiring manual review and eligibility-related risks that necessitate blocking, potentially leading to over-granting or over-blocking. Furthermore, simply saving the final status fails to identify which certification nodes, indicator values, weights, thresholds, and rule versions were used, hindering subsequent auditing, review, and optimization.

[0027] like Figure 1 As shown, to solve the above-mentioned technical problems, this application provides a supplier determination system. The system includes a supplier determination device and an external device connected to the supplier determination device. The supplier determination device is used to execute a supplier determination method, and specifically may include a communication interface, a memory, and a processor. The communication interface is used to interact with the external device to obtain order information and supplier data; the memory is used to store computer programs; and the processor is used to execute the computer programs to implement the supplier determination method.

[0028] External devices may include an order system, a data source system, and a configuration center. The order system is used to submit order information to the supplier to determine the device. The order information includes the supplier and the list of ordered products, and receives the supplier determination result returned by the supplier to determine the device.

[0029] The data source system is used to determine the certification records and multi-dimensional indicators of the equipment suppliers. The certification records include the supplier's certification node, certification status, validity period, applicable procurement organization, and applicable delivery region; the multi-dimensional indicators include the supplier's delivery indicators, quality indicators, financial indicators, capacity indicators, and collaboration indicators.

[0030] The configuration center is used to provide equipment configuration information to suppliers. This configuration information includes rule versions, credit calculation parameters, and supply capacity topology. Rule versions include certification node expansion rules, risk weights, thresholds, and a list of hard constraints. Credit calculation parameters include basic credit limit, baseline capability score, capability adjustment coefficient, category risk coefficient, data timeliness decay coefficient, upper credit limit, and lower credit limit. The supply capacity topology provides a unified leaf node mapping benchmark for order products and certification records.

[0031] Each external device connects to the supplier-designated device and obtains order information, supplier data, and configuration information through data interaction to support the supplier-designated device in executing the supplier-designated method.

[0032] The connection relationships between external devices can be adjusted according to the actual deployment. For example, the data source system can consist of multiple independent business systems, such as the supplier master data system, fulfillment system, quality system, financial system, and capacity system, with each system connecting to the supplier's devices separately.

[0033] like Figure 2 As shown, the supplier determination equipment provided in this application may include: a data acquisition module, a tag tree and node mapping module, a qualification breadth verification module, a dynamic credit calculation module, a comprehensive risk and rule judgment module, an order control interface, and a verification record module.

[0034] The data acquisition module is used to obtain order information from the order system, supplier certification records and multi-dimensional indicators from the data source system, and configuration information from the configuration center. After unifying the format of the acquired data, it outputs it to the tag tree and node mapping module, the qualification breadth verification module, and the dynamic credit calculation module.

[0035] The tag tree and node mapping module receives the order product list and authentication records from the data acquisition module, maps each product in the order product list to the leaf node of the supply capability topology to obtain the order capability node set, maps each authentication item in the authentication record to the leaf node of the supply capability topology to obtain the valid authentication node set, and outputs the order capability node set and the valid authentication node set to the qualification breadth verification module.

[0036] The qualification breadth verification module is used to receive the set of order capability nodes and the set of valid authentication nodes from the tag tree and node mapping module, determine whether the set of order capability nodes matches the set of valid authentication nodes, obtain the judgment result, and output the judgment result to the comprehensive risk and rule judgment module.

[0037] The dynamic credit calculation module receives multi-dimensional indicators from the supplier from the data acquisition module, obtains the supplier's credit limit and used credit limit based on the multi-dimensional indicators, calculates the supplier's available credit limit based on the credit limit and used credit limit, and outputs the available credit limit to the comprehensive risk and rule determination module.

[0038] The comprehensive risk and rule determination module receives the judgment results from the qualification breadth verification module and the available credit limit from the dynamic credit calculation module. It assesses the qualification risk based on the judgment results, determines the comprehensive risk value by combining the qualification risk, available credit limit, and order amount, and determines whether the supplier is enabled for the order corresponding to the order information based on the comprehensive risk value. In other words, it obtains the supplier determination structure and outputs the supplier determination result to the order control interface.

[0039] The order control interface is used to receive supplier confirmation results from the comprehensive risk and rule determination module. The supplier's decision result includes three options: enabled, temporarily disabled, or disabled. The order control interface translates the decision result into a corresponding control action: if the result is enabled, it allows submission; if the result is temporarily disabled, it suspends review; and if the result is disabled, it blocks submission. The order control interface then returns the translated control action to the order system, enabling the system to execute the corresponding order processing flow based on the action.

[0040] The verification record module is used to obtain input snapshots and intermediate calculations from the data acquisition module, tag tree and node mapping module, qualification breadth verification module, dynamic credit calculation module and comprehensive risk and rule judgment module, and save the final results for subsequent auditing and review.

[0041] The supplier identification equipment provided in this application maps order products and supplier certification records to leaf nodes of the same supply capacity topology through a tag tree and node mapping module, aligning order demand and supplier capabilities at the finest granular level. A qualification breadth verification module performs full-set coverage verification, ensuring all minimum categories involved in the order are included in the verification, avoiding misjudgments based on partial category matching. A dynamic credit calculation module dynamically adjusts the credit limit based on multi-dimensional supplier indicators and deducts the already used limit, ensuring the credit limit is dynamically updated according to changes in supplier capabilities and real-time usage, solving the problem that a fixed credit limit cannot reflect the supplier's current true capacity. The system uses a comprehensive risk and rule-based judgment module to weight the qualification risk corresponding to the category matching result and the limit risk corresponding to the real-time acceptance limit into a comprehensive risk value, ensuring that supplier selection is constrained by both category capability and available acceptance limit. The order control interface converts the selection result into control actions such as allowing submission, pausing review, or blocking submission, enabling differentiated supplier selection strategies. The verification record module saves input snapshots and intermediate calculations, making the selection result traceable and verifiable. These modules work together to achieve precise matching between suppliers and orders in terms of category capability and acceptance limit, improving the accuracy of supplier selection and the reliability of supply chain management.

[0042] In one embodiment, such asFigure 3 As shown, this application provides a supplier determination method, which includes the following steps: Step 101: Parse the order information to obtain the supplier identifier and the list of order items.

[0043] Specifically, the supplier determines that the equipment interacts with the order system through a communication interface to obtain order information, which includes at least the supplier identifier and a list of ordered items. The supplier identifier uniquely identifies a supplier and serves as a data index for subsequently obtaining the supplier's certification records and multi-dimensional indicators. The supplier identifier can be a supplier number, a unified social credit code, or a unique identifier within the system, etc.

[0044] The order product list is a collection of all product entries contained in the order. Each product entry includes at least a product code. The product code is used to identify the product category and serves as the basis for subsequent mapping to the leaf nodes of the supply capacity topology. The product code can be an internal category code of the enterprise, an industry standard category code (such as UNSPSC), or other coding forms that can uniquely identify the product category.

[0045] Each product entry in the order product list may also include at least one of the following: product name, specifications, quantity, unit price, process requirements, or precision requirements. These are used to assist in determining the mapping result when the product code cannot be uniquely mapped to a leaf node of the supply capacity topology. For example, when the product code is missing or only the product name is provided, the leaf node corresponding to the product can be determined by combining the specifications and process requirements.

[0046] When order information involves multiple currencies, the order amount in each currency can be uniformly converted into the preset credit limit amount based on the exchange rate snapshot in the rule version, so as to facilitate subsequent comparison and calculation of credit limits.

[0047] If the order product list in the order information is empty or the supplier identifier is missing, a data anomaly flag is generated, the subsequent process is terminated, and the anomaly information is returned to the order. This method provides a complete and standardized data foundation for subsequent category matching and quota calculation.

[0048] Step 102: Map each product in the order product list to a leaf node of the supply capacity topology to obtain a set of order capacity nodes.

[0049] Iterate through the product identifiers of each product in the order product list and search for a node in the supply capacity topology that matches the product identifier. If a matching node is found and it is a leaf node, then the matching node is used as the order capacity node corresponding to the product. If a matching node is found and it is not a leaf node, then the child node of the matching node is used as the order capacity node corresponding to the product. If no matching node is found, an exception flag is generated.

[0050] Specifically, suppliers determine the pre-built supply capability topology for equipment acquisition. This topology is used to align and compare product categories and certification scopes at different granularities down to the leaf node level. The supply capability topology is maintained and distributed centrally by the configuration center and can be constructed from category master data, process dictionaries, certification standards, and procurement organization configurations. In actual deployment, the supply capability topology can be any of the following structures: a multi-level tree structure, a rule matrix, or a directed acyclic graph.

[0051] This explanation uses a multi-level tree structure as an example to illustrate the supply capacity topology. Please refer to [link / reference]. Figure 4 The multi-level tree structure uses a capability domain as the root node, which represents the highest-level classification of the procurement category. Under the capability domain, the first-level subdivision is carried out according to the major procurement category. For example, mechanical processing and electronic components are the major procurement category nodes, representing the mechanical processing field and the electronic components field, respectively.

[0052] Starting from the major category node, drill down level by level according to at least one of the following: material, process route, equipment type, specification range, and precision level, forming a progressively refined path from the major category to the sub-categories, eventually reaching the leaf node. The leaf node represents the smallest category capability unit, meaning that this node has no lower-level child nodes in the supply capability topology.

[0053] For example, under the machining branch, the category node "Machining" is further refined level by level according to materials, processes, and precision, ultimately reaching the leaf nodes L-211 Precision Shafts and L-212 High-Precision Gears. Under the electronic components branch, the category node "Electronic Components" is further refined level by level according to equipment type and specifications, ultimately reaching the leaf nodes L-311 Control Boards and L-312 Sensors. The L-211 Precision Shafts leaf node represents the supplier-certified node, and the L-212 High-Precision Gears leaf node represents the order-pending verification node. By mapping order products and supplier certifications to leaf nodes within the same multi-level tree structure, item-by-item coverage comparisons can be performed at the same granularity level.

[0054] Each node is labeled with a standard code, such as L-211, L-212, a standard name (e.g., precision shafts, high-precision gears), an alias set, and a tree version number. The standard code uniquely identifies the node; the standard name standardizes the category description; the alias set supports precise matching of product identifiers; and the tree version number identifies the tree structure version to which the node belongs, ensuring that orders and certification records are mapped and compared within the same version. Certification information is normalized into a set of leaf nodes before verification; that is, regardless of whether the supplier's certification record corresponds to a major category node or a sub-category node, it is expanded to all the leaf nodes it covers. Order products are also mapped to leaf nodes, forming a set of order nodes to be verified.

[0055] The supplier determines the product identifier for each item in the order product list and searches for a matching node in the supply capacity topology. The product identifier includes the product code, product name, specifications, etc. Matching in this application can include both exact matching and semantic matching.

[0056] First, an exact match can be performed. An exact match means that the product code or name in the product identifier is exactly the same as the standard code or standard name of a node in the supply capacity topology, or the product code or name in the product identifier falls into the alias set of a node in the supply capacity topology, i.e., the synonym list of that node. If an exact match finds a node that matches the product identifier and the matching node is a leaf node, then the matching node is used as the order capacity node. For example, if the product in the order is a precision shaft, and there is a leaf node L-211 in the supply capacity topology with the standard name "precision shaft", then that leaf node is directly used as the node corresponding to that product; or, if the alias set of the precision shaft node in the supply capacity topology contains precision turning shaft, and the order product identifier is precision turning shaft, then that node is matched through alias matching.

[0057] If an exact match finds a node that matches the product identifier and that node is not a leaf node, then all child nodes under the matching node are considered as order capability nodes. Here, child nodes refer to all subordinate nodes of the matching node. For example, if the product in the order is gear processing, and there is a non-leaf node in the supply capability topology with the standard name "gear processing," and this node contains two leaf nodes: spur gears and helical gears, then all leaf nodes corresponding to this product are spur gears and helical gears, and both are added to the order capability node set. By expanding a non-leaf node to all its leaf nodes, order requirements can be aligned and compared with the supplier certification scope at the finest granularity.

[0058] When the product identifier cannot be precisely matched with the standard code, standard name, and alias set of any node in the supply capacity topology, the matching node is determined by calculating the semantic similarity between the product identifier and the node.

[0059] Specifically, the supplier determines that the equipment performs word segmentation on the product identifier, resulting in multiple words. These segmented words are then input into a pre-trained word embedding model, which outputs a word vector for each word. The word embedding model is pre-trained based on industry corpora and can map words to vector representations in semantic space.

[0060] The supplier categorizes the vocabulary used in the equipment identification into core words, attribute words, and relational words. Core words are those that characterize the essence of the product category; attribute words are those that characterize materials, processes, precision, or dimensions; and relational words are those that characterize the product's category or its relationship with similar products. For multiple words within the same category, semantic feature vectors for that category are obtained through vector average pooling or weighted average pooling, namely, core word vectors, attribute word vectors, and relational word vectors.

[0061] Each node in the supply capacity topology has pre-stored standard core word vectors, standard attribute word vectors, and standard relation word vectors, which are generated in the same way based on the node's standard category name, attribute tag, upper-level node information, and same-level node information.

[0062] Suppliers determine equipment by performing semantic matching level by level, starting from the leaf node level of the supply capacity topology. At each level, the core similarity between the core word vector and the standard core word vector of the node, the attribute similarity between the attribute word vector and the standard attribute word vector of the node, and the relationship similarity between the relation word vector and the standard relation word vector of the node are calculated. The core similarity, attribute similarity, and relation similarity are then weighted and summed according to preset weights to obtain the comprehensive similarity. The weight of core similarity is greater than the weight of attribute similarity, and the weight of attribute similarity is greater than the weight of relation similarity.

[0063] If there are nodes at the current level with a comprehensive similarity greater than or equal to a preset threshold, the node with the highest comprehensive similarity will be selected as the order capability node. The comprehensive similarity is used to measure the semantic matching degree between the product identifier and the node; the higher the comprehensive similarity, the higher the matching degree. The preset threshold is used to limit the minimum acceptable level of matching; nodes below this threshold are not considered as matches.

[0064] If there is no node with a comprehensive similarity greater than or equal to the preset threshold at the current level, then trace back to the parent level of the current level, that is, the level above the current level, re-acquire the standard semantic feature vectors of each node in the parent level and calculate the comprehensive similarity, until tracing back to the root node of the supply capability topology.

[0065] If a node with a comprehensive similarity greater than or equal to a preset threshold is found at a certain level, its child nodes are designated as order capability nodes. Simultaneously, the upward tracing depth is recorded. Here, child nodes refer to all subordinate nodes of the found node. The upward tracing depth represents the number of levels traced up from the leaf node level. The depth is 0 when a matching node is found directly at the leaf node level, 1 when a matching node is found after tracing up one level, and so on. If the upward tracing depth exceeds the preset depth threshold, it indicates that the level of the matching node is too far from the leaf node level, resulting in coarse matching granularity. In this case, a supply capability topology update is triggered.

[0066] If no node with a comprehensive similarity greater than or equal to a preset threshold is found across all levels, an anomaly flag is generated, and no node corresponding to that product is included in the order capability node set. The anomaly flag will be passed to subsequent steps to trigger hard constraints or review processes, ensuring that unmapped products are not ignored or allowed by default.

[0067] In one embodiment of this application, the supply capacity topology update includes: the supplier determining device acquiring each existing node in the supply capacity topology. An existing node refers to a node in the current supply capacity topology. For each existing node, it is assumed that a product is inserted as a new node below that existing node, as a child node of that existing node. The supplier determining device calculates the global semantic distance of the supply capacity topology after insertion under this assumption and compares the global semantic distance after insertion with the global semantic distance of the supply capacity topology before insertion.

[0068] Specifically, the semantic distance is calculated for each pair of nodes in the supply capacity topology by traversing all nodes. The semantic distances of all node pairs are then summed to obtain the global semantic distance. The semantic distance is determined based on the hierarchical path length between nodes and the name similarity between node names. The hierarchical path length measures the distance between two nodes in the topological structure. The hierarchical path length is the total number of edges traversed from a node, traversing along the parent node's direction to a common parent node, and then traversing down to the target node. A parent node and a child node are connected by a parent-child edge, and the hierarchical path length between them is 1, indicating that the two nodes are directly adjacent in the topology.

[0069] For nodes at the same level and sharing a common parent node, the path length is 2, meaning they are connected topologically through a common parent node. A longer path length indicates a greater topological distance and lower semantic association between the two nodes; conversely, a shorter path length indicates closer topological proximity and higher semantic association.

[0070] Name similarity is determined based on the semantic closeness between the standard names of nodes, and its value ranges from 0 to 1. When two node names are exactly the same, the semantic similarity is 1, and when two node names are completely unrelated semantically, the semantic similarity approaches 0.

[0071] Specifically, semantic distance can be calculated as: hierarchical path length × (1 / semantic similarity factor), or simplified to semantic distance = hierarchical path length / semantic similarity factor. When the semantic similarity factor approaches 1, the semantic distance is approximately equal to the hierarchical path length, indicating that the two nodes are highly related semantically, and their distance is mainly determined by the hierarchical span. When the semantic similarity factor approaches 0, the semantic distance approaches infinity, indicating that the two nodes are almost unrelated semantically, and even if the hierarchical path is short, they should be considered far apart. In this way, semantic distance can simultaneously reflect the positional relationship (hierarchical path length) and semantic association (semantic similarity) of two nodes in the supply capacity topology, thus accurately characterizing the semantic proximity between nodes.

[0072] The structural semantic offset represents the change in the global semantic distance of the supply capability topology after insertion relative to the global semantic distance of the supply capability topology before insertion. The supplier determines the equipment by calculating the difference between the global semantic distance before insertion and each global semantic distance after insertion, and then dividing each difference by the global semantic distance before insertion to obtain the structural semantic offset corresponding to each insertion position. For example, if the global semantic distance before insertion is 1000, and the global semantic distance after insertion at a certain position is 1100, then the difference is 100, and the structural semantic offset is 100 / 1000 = 0.1, indicating that the global semantic distance increased by 10% after inserting the node. The smaller the structural semantic offset, the higher the compatibility of the new node with the existing supply capability topology, the smaller the disturbance to the overall semantics of the supply capability topology after insertion, and the more reasonable the insertion position.

[0073] After determining the structural semantic offset of each existing node as a candidate insertion position, the supplier selects the insertion position with the smallest structural semantic offset, establishes a new node at the insertion position, and uses the new node as the order capability node. In this way, the new node is inserted into the position with the least impact on the semantics of the existing supply capability topology, ensuring that the new category node can be correctly mapped and maintaining the overall semantic stability of the supply capability topology to the greatest extent.

[0074] Through the above-mentioned supply capacity topology update method, the supply capacity topology can dynamically evolve with the continuous emergence of new product categories, without the need for frequent manual maintenance.

[0075] After performing the above mapping operation on each product in the order product list, the resulting order capability nodes form an order capability node set. This method allows order requirements to be precisely broken down into a set of the smallest capability units comparable to supplier certification, providing an accurate matching benchmark for subsequent qualification breadth verification.

[0076] Step 103: Obtain the supplier's certification record based on the supplier identifier, and map each certification item in the certification record to the leaf node of the supply capability topology to obtain a set of valid certification nodes. The leaf node represents the smallest category node that can be referenced by the product and can be covered by the certification record.

[0077] Filter certification items from the certification records that are in effect, have a validity period including the current time, and are applicable to both the procurement organization and the delivery area, and designate them as valid certification items; determine the leaf nodes corresponding to each valid certification item in the supply capacity topology, and form a set of valid certification nodes from the determined leaf nodes.

[0078] The supplier identifies the equipment and retrieves all of that supplier's certification records from the data source system, indexed by the supplier's identifier. Certification records are qualification documents or data records obtained by the supplier within the procuring organization or from a third-party certification body, used to demonstrate its capability to produce or supply specific product categories. Each certification record includes at least the following attributes: certification item, certification status, validity period, applicable procuring organization, and applicable delivery region.

[0079] The certification item represents the category or capability scope covered by the certification record, which can be a leaf node or a non-leaf node in the supply capability topology. The certification status indicates whether the certification is currently effective, including active, expired, suspended, or revoked states. The validity period includes the start and end times, used to determine the time validity of the certification. Applicable procurement organizations indicate which procurement organizations the certification applies to, and applicable delivery regions indicate which delivery regions the certification applies to.

[0080] In order to obtain valid certification records in terms of time, space, and status, the supplier determines that the equipment should be selected from all certification records as valid certification items if the certification status is active, the validity period includes the current time, and the applicable scope includes the procurement organization and delivery area of ​​the current order.

[0081] If the certification status is "in effect," then certification records in ineffective states such as expired, suspended, or revoked are excluded. The validity period must include the current time; that is, the start time of the validity period is no later than the current time and the end time is no earlier than the current time. The applicable scope must include the purchasing organization and delivery region of the current order. This can be because the applicable purchasing organization set in the certification record includes the purchasing organization of the current order, and the applicable delivery region set includes the delivery region of the current order. If the certification record is marked as applicable to all organizations or all regions, it is considered to meet the corresponding scope conditions. Through the above multi-level filtering, valid certification items are obtained. Certification items that fail the filtering are not included in subsequent processing, and the corresponding exclusion reasons are recorded for display to the user when needed.

[0082] Subsequently, the supplier determines the leaf nodes corresponding to each valid certification item in the supply capability topology. If a valid certification item is itself a leaf node, then that node is directly added to the set of valid certification nodes; if a valid certification item is a non-leaf node, then all child nodes under that non-leaf node are added to the set of valid certification nodes. The set of valid certification nodes represents all the smallest category capability units that the supplier has been validly certified under the current time, current procurement organization, and current delivery region.

[0083] By using the above method, the scope of supplier certification can be unified from certification items of different granularities to the leaf node granularity, providing an accurate capability set for subsequent full-coverage verification on the same benchmark as the order capability node set.

[0084] Step 104: Determine whether the set of order capability nodes matches the set of valid authentication nodes, and obtain the determination result.

[0085] If all order capability nodes in the order capability node set exist in the valid authentication node set, the judgment result is passed; If at least one order capability node does not exist in the set of valid authentication nodes, the result is "failed" and the set of failed nodes is output. The set of failed nodes consists of all nodes that do not exist in the set of valid authentication nodes.

[0086] The supplier determines that each node in the order capability node set O is compared with each node in the valid authentication node set C_eff(s,t). If every node in O has a corresponding node in C_eff(s,t), i.e., they belong to the same leaf node, the result is "pass". If at least one node in O has no corresponding node in C_eff(s,t), the result is "fail".

[0087] If all nodes in the order capability node set O exist in the valid authentication node set C_eff(s,t), then O If C_eff(s,t) is executed, the result is considered successful. This indicates that the supplier possesses valid certification qualifications for all the smallest product categories involved in the current order, and the process then proceeds to the credit limit calculation and comprehensive risk assessment stages.

[0088] If at least one node in the order capability node set O is not present in the valid authentication node set C_eff(s,t), the judgment result is failure, and the failure node set M is output. The failure node set M consists of all nodes in the order capability node set O that do not have a matching node in the valid authentication node set C_eff(s,t), i.e., M=O\C_eff(s,t).

[0089] For example, suppose the supply capability topology has a capability domain as the root node, and the capability domain has a category node called Machining. Machining is further divided into two leaf nodes: General Turning Parts and Precision Shafts, as well as a non-leaf node called Gear Machining. Gear Machining is further divided into two leaf nodes: Spur Gears and Helical Gears.

[0090] An order contains two items: precision shafts and gear assemblies. Precision shafts are leaf nodes, directly mapped to L-102 in the order's capability node set. Gear assemblies are non-leaf nodes, expanding to correspond to two leaf nodes: spur gears (L-201) and helical gears (L-202). Therefore, the order's capability node set O contains three nodes: L-102, L-201, and L-202.

[0091] The supplier's certification records include precision shaft certification and gear machining certification. Precision shaft certification directly corresponds to leaf node L-102; gear machining certification corresponds to the non-leaf node gear machining, which, when expanded, corresponds to two leaf nodes: spur gear (L-201) and helical gear (L-202). Therefore, the valid certification node set C_eff also contains three nodes: L-102, L-201, and L-202. The supplier determines that the equipment compares each of the three nodes in the order capability node set O with the nodes in the valid certification node set C_eff, finding that all three nodes can be found in C_eff with corresponding identical nodes, i.e., O... If the result of C_eff is "pass", it means that the supplier has the capability to handle all product categories of the current order.

[0092] If the supplier only has precision shaft certification, then the valid certified node set C_eff contains only one node, L-102. The supplier determines that the equipment compares each of the three nodes in O with the nodes in C_eff. If L-201 and L-202 do not have any identical nodes in C_eff, the result is a failure, and the set of failed nodes M={L-201,L-202} is output, indicating that the supplier does not have the certification qualifications for spur gears and helical gears and cannot accept the order.

[0093] For each node not found in the node set M, the supplier determines that the device is also associated with the corresponding order product information (such as product line number and product name) and the reason for failure. Reasons for failure include, but are not limited to: the product category corresponding to this node has never appeared in the supplier's entire certification record (uncertified), certification has expired, certification is not effective, certification status is suspended or revoked (abnormal status), certification record is not applicable to the current purchasing organization (purchasing organization not applicable), and certification record is not applicable to the current delivery region (delivery region not applicable). By associating each missing node with a specific order product line number and reason for failure, precise evidence can be provided for subsequent review or rejection, enabling purchasing personnel to quickly locate the problem.

[0094] Furthermore, when there are missing nodes in the failure node set M due to expired certification or suspended supplier certification, these nodes are marked as critical missing nodes. Critical missing nodes indicate that the missing node is a qualification risk inherent to the supplier itself, rather than a negotiable scope configuration issue. When a judgment result containing a critical missing node is a failure, a hard constraint will be directly triggered, preventing the supplier from accepting orders.

[0095] When the judgment result is "not approved" but the reason for the missing information is only that the procurement organization is not applicable or the delivery area is not applicable, the missing information belongs to the scope configuration problem that can be reviewed. It does not directly trigger a hard block, but is included in the comprehensive risk value through the qualification risk component, and the threshold determines whether to enter the review process.

[0096] The above method achieves full coverage verification of order requirements and supplier certification capabilities at the leaf node granularity, and can accurately output uncovered nodes and their associated order products and failure reasons, providing an accurate data foundation for subsequent risk assessment and differentiated handling.

[0097] This application sets up a system where the products and certification records of an order are mapped to the same level (leaf node level) within the same supply capability topology, resulting in a set of order capability nodes and a set of valid certification nodes. The nodes in the order capability node set are the nodes corresponding to the products of each order at the leaf node level, representing the smallest category granularity of the order demand. The nodes in the valid certification node set are the nodes corresponding to each certification item at the leaf node level, representing the smallest category granularity of the supplier's certification capabilities. Since the nodes in both sets are located at the same level within the same topology, precise alignment between order demand and supplier capabilities is achieved by determining whether the order capability node set matches the valid certification node set.

[0098] Step 105: Obtain the supplier's credit limit and the amount already used, and calculate the supplier's available credit limit based on the credit limit and the amount already used.

[0099] Obtain multi-dimensional indicators from suppliers, including at least two of the following: delivery indicators, quality indicators, financial indicators, capacity indicators, and collaboration indicators. For dimensional indicators whose values ​​are positively correlated with supplier capabilities, normalize them using a first method. For dimensional indicators whose values ​​are negatively correlated with supplier capabilities, normalize them using a second method. Obtain the collaboration indicator by weighted summation based on order confirmation timeliness, change response timeliness, anomaly closure timeliness, and data completeness. Finally, weighted summation of the normalized dimensional indicators yields the comprehensive capability score.

[0100] Specifically, the supplier determines the equipment and obtains multi-dimensional metrics from the data source. These multi-dimensional metrics are used to evaluate the supplier's overall capabilities within the current evaluation window and include at least two of the following: delivery metrics, quality metrics, financial metrics, capacity metrics, and collaboration metrics. Delivery metrics reflect the supplier's order confirmation timeliness and on-time delivery rate; quality metrics reflect the supplier's product qualification rate and defect rate; financial metrics reflect the supplier's accounts payable balance and payment stability; capacity metrics reflect the supplier's capacity utilization rate and remaining capacity; and collaboration metrics reflect the supplier's responsiveness and closed-loop capability in the procurement collaboration process.

[0101] Multidimensional metrics are linked according to supplier identifier, order identifier, category code, procurement organization, delivery region, and event time, and a unified snapshot is generated from different business units by the data acquisition module: Supplier master data provides supplier status, level, basic credit, and organizational relationship; Fulfillment provides order confirmation, shipment, arrival, and delay events; Quality provides inspection pass rate, defect rate, returns, and quality anomaly events; Finance provides accounts payable balance, overdue amount, payment stability, and frozen amount; Capacity provides rated capacity, planned capacity, scheduled output, and capacity utilization rate. Each metric record saves the source, collection time, validity period, evaluation window, and data version, and is associated by default with a combination key of supplier identifier and category, organization, or order; in case of missing, delayed, or conflicting data, the conservative default value in the rule version is used and the reason code is recorded.

[0102] The collaborative service indicator represents the supplier's responsiveness and closed-loop capability in the procurement collaboration process. It is calculated by weighting order confirmation timeliness, change response timeliness, anomaly closure timeliness, and data completeness rate according to rules. The order confirmation timeliness rate is the ratio of the number of events in which the supplier confirms an order within the agreed timeframe to the total number of order events that should be confirmed within the evaluation window. The change response timeliness rate is the ratio of the number of events in which the supplier responds to change requests within the agreed timeframe to the total number of change request events within the evaluation window. The anomaly closure timeliness rate is the ratio of the number of events in which the supplier completes anomaly closure within the agreed timeframe to the total number of anomalies within the evaluation window. The data completeness rate is the ratio of the number of events in the evaluation window where the supplier submitted complete order-related data to the total number of events where data should be submitted. Each sub-indicator is calculated as the ratio of the number of events that meet the agreed service timeframe to the number of events that should be completed within the evaluation window, and is uniformly normalized to 0 to 1. When there are no events that should be completed or event data is missing within the evaluation window, the configured conservative default value is used, and the data missing reason code is recorded.

[0103] Because the dimensions of the indicators differ, and some indicators are positively correlated with supplier capabilities (higher values ​​indicate stronger supplier capabilities, such as on-time delivery rate), while others are negatively correlated (higher values ​​indicate weaker supplier capabilities, such as defect rate), it is necessary to normalize all indicators to the range of 0 to 1. For positively correlated indicators, normalization is performed using the first method, so that the normalized value increases as the original value increases. For negatively correlated indicators, normalization is performed using the second method, so that the normalized value decreases as the original value increases. Taking on-time delivery rate as an example, a higher value indicates stronger supplier delivery capabilities, making it a positively correlated indicator. After normalization using the first method, a higher on-time delivery rate results in a higher normalized value. Similarly, for defect rate, a higher value indicates weaker supplier quality capabilities, making it a negatively correlated indicator. After normalization using the second method, a higher defect rate results in a lower normalized value.

[0104] For positively correlated indicators, normalization is performed using the following formula: ; For negatively correlated indicators, first perform a directional reversal: ; Among them, v j a is the original value of the j-th dimension indicator; j b is the lower bound of the j-th dimension indicator; j x is the upper limit of the j-th dimension indicator; j Let be the normalized value of the j-th dimension indicator, ranging from 0 to 1; the clip function limits the result to the range of 0 to 1 to reduce the abnormal impact of extreme values ​​on credit granting. When b j =a j If data is missing, use the conservative default value in the rule version and generate a data missing reason code.

[0105] Forward normalization maps the original value to the range of 0 to 1. The mapping result increases as the original value increases, making it suitable for indicators where larger values ​​indicate stronger supplier capabilities. Inverse normalization maps the original value to the range of 0 to 1 and then subtracts the mapping result from 1, causing the final value to decrease as the original value increases. This is suitable for indicators where larger values ​​indicate weaker supplier capabilities.

[0106] The weighted sum of the normalized indicators across all dimensions yields the overall capability score S: S=w d ×x d +w q ×x q +w f ×x f +w c ×x c +w s ×x s ; Where, x d x q x f x c and x s These represent the normalized scores for delivery metrics, quality metrics, financial metrics, capacity metrics, and collaboration metrics, respectively; w d w q w f w c and w s These are weights assigned to delivery, quality, finance, capacity, and collaboration metrics, respectively. Each weight is non-negative, and the sum of the weights is 1. Weights can be configured separately according to product category, supplier level, or purchasing organization.

[0107] Then, the supplier determines the basic credit line L for the equipment. baseBaseline capability score S0, capability adjustment coefficient α, category risk coefficient k category and data time-related decay coefficient k time The basic credit limit is the preset credit benchmark value in the supplier master data; the benchmark capability score is the initial comprehensive score at the time of supplier access or the comprehensive score of the previous evaluation period, with the specific value used determined by the configuration in the credit version number; the capability adjustment coefficient is used to control the adjustment range of the credit limit due to capability changes; the capability adjustment coefficient is not less than 0; the category risk coefficient reflects the process complexity, quality sensitivity, and substitution difficulty of the order's category, with a smaller coefficient for higher category risk; the data timeliness decay coefficient reflects the timeliness of capability indicator data, with a smaller coefficient for older data or certification closer to the expiration date, and the supplier status risk can be increased simultaneously when certification is nearing expiration. The category risk coefficient and the data timeliness decay coefficient are greater than 0 and less than or equal to 1.

[0108] Based on the deviation between the comprehensive capability score S and the baseline capability score S0, and combined with the capability adjustment coefficient α, the basic credit limit is adjusted to obtain the dynamic credit limit L. dyn : L dyn =clip(L base ×[1+α×(S-S0)×k category ×k time ,L min ,L max ); Among them, L min and L max These are the lower and upper limits of credit, respectively, where 0 ≤ L. min ≤L max The `clip` function ensures that the dynamically adjusted credit limit will not exceed the preset absolute upper and lower bounds. When the overall capability score is higher than the benchmark capability score, the credit limit is increased; when the overall capability score is lower than the benchmark capability score, the credit limit is decreased. The higher the category risk, the higher the credit limit. category The smaller the value, the lower the dynamic credit limit; the more outdated the capability data or the closer the certification is to its expiration date, the lower the credit limit. time The smaller the value, the lower the dynamic credit limit.

[0109] Obtain the supplier's used credit limit, which includes outstanding balances in transit (E). open And the frozen amount E lock The outstanding balance in transit represents the total amount of all unfulfilled orders from the supplier, reflecting the amount committed but not yet settled. The frozen balance is temporarily frozen due to quality disputes, risk warnings, etc., and cannot be used for new orders. Subtracting the occupied balance from the dynamic credit limit yields the available credit limit L. available : L available=max(0,L dyn -E open -E lock ); Available credit limit represents the maximum amount a supplier can currently use to accept new orders. When the available credit limit is zero and the order amount is greater than zero, a hard constraint on the credit limit is triggered, blocking the order. To avoid multiple concurrent orders using the same available credit limit, the calculation of dynamic credit limit and order credit limit allocation are completed in the same database transaction, or the balance is updated through a version-numbered comparison and exchange mechanism to ensure the atomicity and consistency of credit limit allocation. When an order is canceled, settled, or authorized for release, the credit limit is returned according to the corresponding allocation record.

[0110] In one specific implementation, the supplier determination method provided in this application is illustrated using a purchase order for precision shaft products as an example. The following data are merely example parameters set in the calculation link and do not represent the actual weights, thresholds, or credit data used in the production system.

[0111] Assume the order includes precision shaft products, corresponding to leaf node L-211 in the supply capacity topology, and the order amount is 520,000 yuan. The valid certification set for supplier S-001 is {L-211, L-212}, and the relevant certifications are neither expired nor suspended. Therefore, the order capacity node set O = {L-211}, satisfying O... C_eff(S-001,t) means that all nodes in the order capability node set exist in the valid authentication node set, and the qualification breadth verification is passed.

[0112] In this embodiment, when a precision shaft is entered on the order page, an exact match is first attempted. If no node with the standard name or alias "precision shaft" is found in the supply capacity topology, the semantic match process is initiated. Two candidate names, "precision turned shaft" and "precision turned shaft category," are retrieved from the alias table. Then, the leaf node L-211 is confirmed using the material (alloy steel), turning process, IT6 to IT7 precision, and diameter range (20 to 40 mm) attributes. If L-211 exists in the valid certification set and the certification status, validity period, procurement organization, and delivery region all meet the conditions, the order node is included in the coverage set; otherwise, a corresponding failure reason is generated.

[0113] Since O = {L - 211} satisfies O C_eff(S-001,t), qualification breadth verification passed, qualification risk r auth =0.

[0114] The supplier determines the equipment and obtains supplier S-001's multi-dimensional indicators within the current evaluation window. Let the delivery, quality, financial, capacity, and collaboration indicators be 0.92, 0.86, 0.78, 0.88, and 0.82 respectively, with corresponding weights of 0.25, 0.25, 0.20, 0.20, and 0.10. Then the comprehensive capability score S is: S=0.25×0.92+0.25×0.86+0.20×0.78+0.20×0.88+0.10×0.82=0.8590; Suppliers determine the basic credit line L for equipment. base =1,000,000 yuan, baseline capability score S0=0.75, capability adjustment coefficient α=0.80, category risk coefficient k category =0.90, data timeliness attenuation coefficient k time =0.98, lower limit of credit L min =500,000 yuan, credit limit L max =1,500,000 yuan, then the dynamic credit limit L dyn for: L dyn =1,000,000×[1+0.80×(0.8590-0.75)]×0.90×0.98=958,910.40 yuan; Let E be the amount of unsettled funds in transit. open =300,000 yuan, frozen amount E lock =80,000 yuan, then the available credit limit L available for: L available =958,910.40-300,000-80,000=578,910.40 yuan.

[0115] The ratio of order amount to available credit limit is used to determine the credit limit risk (r). credit =0.8982. Let the category risk be r. category =0.30, Supplier Status Risk r supplier =0.20, Order Abnormal Risk r order =0.25, and the risk weights are 0.30, 0.30, 0.15, 0.15, and 0.10 respectively. Therefore, the overall risk value R is: R = 0.30 × 0 + 0.30 × 0.8982 + 0.15 × 0.30 + 0.15 × 0.20 + 0.10 × 0.25 = 0.3695. In this embodiment, the hard constraint was not hit. When the soft interception threshold is 0.35 and the hard interception threshold is 0.65, since 0.35 ≤ R < 0.65, soft interception is output, the supplier is temporarily not allowed to accept the order, and the process enters the review process, with a prompt indicating that the quota usage ratio is high. After the authorized personnel review and approve the order, the approver, approval basis, and approval time are recorded, and the corresponding quota is used. The order then enters the subsequent process.

[0116] By incorporating the supplier's multi-dimensional capability indicators and real-time occupancy status into the calculation of available acceptance quota, the quota assessment results can dynamically reflect the supplier's true acceptance capacity at the current point in time.

[0117] Step 106: Assess the supplier's qualification risk based on the judgment result, determine the comprehensive risk value by combining the qualification risk, available acceptance limit and order amount, and determine whether to enable the supplier for the order corresponding to the order information based on the comprehensive risk value.

[0118] When the judgment result is "not passed", the qualification risk is assessed as the first risk value; when the judgment result is "passed", the qualification risk is assessed as the second risk value, and the first risk value is higher than the second risk value; the credit limit risk is determined based on the available acceptance limit and the order amount; the qualification risk and the credit limit risk are weighted and summed to obtain the comprehensive risk value; the supplier of the order is determined based on the comparison result of the comprehensive risk value and the preset threshold.

[0119] Suppliers assess qualification risk based on the evaluation results. When the evaluation result is "fail," meaning at least one node in the order's capability node set is not covered by the valid certification node set, it indicates that the supplier's certification scope cannot cover all the smallest category nodes of the order. In this case, the qualification risk is high, and the assessed qualification risk is the first risk value. When the evaluation result is "pass," meaning all nodes in the order's capability node set are covered by the valid certification node set, it indicates that the supplier's certification scope covers all the smallest category nodes of the order. In this case, the qualification risk is low, and the assessed qualification risk is the second risk value. The first risk value is higher than the second risk value.

[0120] Suppliers determine credit limit risk based on their available credit limit and the order amount. Credit limit risk represents the adequacy of the supplier's current available credit limit relative to the order amount. The closer the available credit limit is to the order amount, the higher the credit limit risk; the lower the available credit limit, the higher the credit limit risk. For example, when the order amount is 1 million yuan and the available credit limit is 1.2 million yuan, the credit limit risk is low; when the order amount is 1 million yuan and the available credit limit is 200,000 yuan, the credit limit risk is high.

[0121] The supplier determines the overall risk value by weighting and summing the risks associated with qualification and credit limits. This overall risk value is a quantified result that combines the risks of both product category matching and available credit limits.

[0122] Specifically, suppliers determine equipment based on qualification risk. auth Order amount A order With available acceptance limit L available Proportional calculation of credit limit risk r credit : ; Among them, A order For order amount; L available The available credit limit; ε is a very small constant greater than 0, used to prevent L... available A division by zero occurs when the value is 0. Additionally, the excess ratio r can be calculated. over This is used to determine whether the severe over-limit threshold that necessitates hard blocking has been reached: ; The comprehensive risk model also includes category risk r category Supplier status risk supplier and order anomaly risk order Product category risk is reflected by process complexity, quality sensitivity, and difficulty of substitution; supplier status risk reflects rating downgrades, major anomalies, certification nearing expiration, and sudden changes in indicators; order anomaly risk reflects deviations in amount from historical ranges, short-term concentrated orders, or the result of splitting and aggregating orders.

[0123] Risk of qualification auth Credit limit risk credit Category risk category Supplier status risk supplier and order anomaly risk order Weighted summation yields the overall risk value R: R=λ1×r auth +λ2×r credit +λ3×r category +λ4×r supplier +λ5×r order ; Where λ1, λ2, λ3, λ4, and λ5 are the weights corresponding to qualification risk, credit limit risk, category risk, supplier status risk, and order anomaly risk, respectively, and each risk weight is non-negative and their sum is 1. The system also outputs λ. i ×r i As the contribution value of each risk component, and generate reason codes according to the contribution level such as high quota occupancy ratio, certification nearing expiration, and order amount significantly deviating from the historical range, for use by soft interception review and historical audit.

[0124] The supplier selection process determines the supplier for an order based on a comparison of the overall risk value and a preset threshold. Before making this determination, the supplier selection process also checks whether constraints are met. These constraints are unavoidable hard restrictions and include at least one of the following: the supplier is suspended or blacklisted; the judgment result is unsuccessful and the set of unsuccessful nodes contains nodes that failed due to expired certification or suspended supplier certification (i.e., critical missing nodes); abnormal identification generated by product mapping; critical certification failure; order amount exceeding the single absolute limit; or available acceptance limit being zero and order amount being greater than zero. If any of the above constraints are met, the supplier is directly determined not to accept the order, and the order submission is blocked without threshold comparison.

[0125] If the constraints are not met, the supplier for the order is determined based on a comparison between the comprehensive risk value and a preset threshold. When the comprehensive risk value is less than the first threshold, the order risk is low, the supplier is deemed capable of accepting the order, and the order is allowed to be submitted. When the comprehensive risk value is greater than or equal to the first threshold but less than the second threshold, the order carries a medium risk; the supplier is temporarily not allowed to accept the order, the order submission is suspended, and the process enters a review phase where authorized personnel review the order and decide whether to release it. When the comprehensive risk value is greater than or equal to the second threshold, the order risk is high, the supplier is deemed unsuitable to accept the order, and the supplier is not allowed to accept the order.

[0126] By employing the above methods, supplier selection is constrained by both product category coverage and available contract value, avoiding selection based on only one aspect and achieving a differentiated supplier selection strategy.

[0127] This application overcomes the problem of traditional solutions that only judge at the supplier or category level and have difficulty identifying differences in subdivided categories by unifying order categories and certification scope into a set of category leaf nodes and performing full coverage verification. It overcomes the problem of fixed credit fields not reflecting changes in supplier capabilities and usage in a timely manner by calculating dynamic available credit through comprehensive capability scores, category risk, data timeliness, in-transit usage, and frozen credit limits. It overcomes the problem of separate verification of qualifications and credit limits and the lack of unified risk results by combining qualifications, credit limits, categories, supplier status, and order anomalies into an interpretable comprehensive risk value. It overcomes the limitation of traditional solutions that only offer two results: allow or prohibit, by prioritizing hard constraints and outputting allowances, soft interceptions, or hard interceptions with dual thresholds. It overcomes the problem of difficulty in recalculating results by saving input snapshots, risk contributions, reason codes, tag tree versions, and rule versions. It overcomes the problem of only recording final pass or failure and the difficulty in recalculation by completing verification and credit limit usage within the same transaction or versioned comparison exchange process, thereby improving the accuracy and reliability of supplier determination.

[0128] Please see Figure 5 In one specific embodiment, the supplier determination method provided in this application includes: Step 201: Obtain order information.

[0129] Step 202: Map each product in the order product list to a leaf node of the supply capacity topology to obtain a set of order capacity nodes.

[0130] Step 203: Obtain the supplier's certification record based on the supplier identifier, and map each certification item in the certification record to the leaf node of the supply capability topology to obtain a set of valid certification nodes.

[0131] Step 204: Determine whether the set of order capability nodes matches the set of valid authentication nodes, and obtain the determination result.

[0132] Step 205: Normalize the multidimensional metrics of suppliers.

[0133] Step 206: Obtain the supplier's credit limit based on the normalized multidimensional indicators, and calculate the supplier's available credit limit based on the adjusted credit limit and the already used credit limit.

[0134] Step 207: Assess the supplier's qualification risk based on the judgment results, and determine the comprehensive risk value by combining qualification risk, available acceptance limit and order amount.

[0135] Step 208: Tiered handling, allowing / soft blocking / hard blocking.

[0136] The supplier first determines whether the equipment meets the hard constraints (constraint conditions). Hardware constraints include supplier shutdown or blacklisting, failure to pass the judgment and existence of key missing items, abnormal product mapping, invalid key certification, order amount exceeding the limit, and available acceptance quota of zero and order amount greater than zero. If the conditions are met, the supplier is not allowed to accept orders. If the conditions are not met, the supplier is handled in a tiered manner based on the comparison between the comprehensive risk value and the preset threshold: if the risk value is less than the first threshold, the supplier is allowed to accept orders; if the risk value is between the first and second thresholds, the supplier is temporarily not allowed to accept orders and enters the review process; if the risk value is greater than or equal to the second threshold, the supplier is not allowed to accept orders, thus achieving differentiated risk control.

[0137] Step 209: Obtain input snapshots and intermediate calculations, and save the final results for subsequent auditing and review.

[0138] The supplier ensures that the equipment retains a complete record of the determination, including order information, the set of each node, the judgment result, the comprehensive capability score, the dynamic credit limit, the available acceptance limit, the comprehensive risk value, the final disposal result and version information, so that the determination result is traceable, recalculated and auditable, which facilitates subsequent dispute positioning and parameter iteration.

[0139] In one embodiment, the method provided in this application further includes rule versioning management. Specifically, weights, thresholds, hard constraints, upper and lower bounds of indicators, conservative default values, and node expansion rules are versioned by the configuration center according to the procurement organization, product category, and supplier level. Each order determines the version number of the supply capability topology, the rule version number, and the credit version number, and saves the order context, the order capability node set, the effective authentication node set, the indicator vector, the comprehensive capability score, the dynamic credit limit, the available acceptance limit, the risk vector, the comprehensive risk value, the hit reason, and the final action. Rule updates do not retroactively change existing records, but when the order is resubmitted, the new version is used for re-determination. When the product category cannot be uniquely mapped, the authentication status conflicts, key indicators are missing, or the data source is temporarily unavailable, a conservative state is adopted and the data quality reason code is recorded. Higher credit or default release cannot be obtained due to missing data. When the available acceptance limit is zero and the order amount is greater than zero, the limit risk is taken as the upper limit and the hard constraint of the limit can be hit. In this way, rule versioning management and full-process traceability are achieved. Each time the supply capacity topology version number, rule version number, and credit version number are determined, the order context, node set, indicator vector, comprehensive capability score, dynamic credit limit, available capacity limit, risk vector, comprehensive risk value, hit reason, and final action are saved, making the determination results traceable and auditable, and rule updates do not affect historical records.

[0140] This application improves supplier qualification matching accuracy by mapping product categories and supplier certification scope to the leaf node granularity of the same supply capacity topology and performing full set coverage verification on the order capacity node set, ensuring that all the smallest categories involved in the order are included in the verification. This effectively avoids the qualification misjudgment caused by traditional solutions that only judge at the supplier or category level. By calculating a comprehensive capability score based on multi-dimensional supplier indicators and dynamically adjusting the credit limit in combination with category risk and data timeliness, and then deducting the credit limit in transit and frozen, the credit limit is dynamically updated according to changes in supplier capability and real-time occupancy. This overcomes the defect that fixed credit limits cannot reflect the supplier's current true capacity. By combining the qualification risk corresponding to the qualification breadth verification result with the credit limit risk corresponding to dynamic credit and other risks, the application further improves the accuracy of supplier qualification matching. The quantitative analysis is transformed into an interpretable comprehensive risk value, ensuring that supplier selection is constrained by both category coverage and available credit limits, preventing selection based on only one aspect. Hard constraints prioritize blocking eligibility risks and severe over-limits, while dual thresholds differentiate between low-risk (automatic release), medium-risk (soft interception and review), and high-risk (hard interception), achieving differentiated risk control intensity and avoiding the limitation of only allowing or prohibiting outcomes. Saving input snapshots, node sets, indicators, formula parameters, risk contributions, and rule versions makes the judgment results traceable, recalculated, and auditable, facilitating subsequent dispute localization and parameter iteration. Completing verification and credit limit occupancy within the same transaction or atomic update process with versioning prevents multiple concurrent orders from repeatedly consuming the same available balance, reducing the risk of credit limit overuse. This application transforms the process of supplier qualification verification, order limit determination, and risk management from scattered manual inquiries to automatic system calculations. This reduces the time that procurement personnel spend repeatedly reviewing certificates, ratings, and historical orders. By identifying category mismatches, insufficient limits, and abnormal orders in advance, it lowers the costs of reassignment, cancellation, supplementary review, and handling of performance anomalies after erroneous orders are placed, thereby improving the accuracy of supplier selection and the reliability of supply chain management.

[0141] It should be understood that, although Figure 3 , Figure 5 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 3 , Figure 5 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0142] In one embodiment, this application also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, and when the program instructions are executed by a computer, the computer is able to perform the supplier determination method provided by the above methods.

[0143] In one embodiment, the supplier determines that the device is located within a computer device, which may be a server, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, communication interface, and database connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores data used in the supplier determination method. The network interface of the computer device communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements a supplier determination method.

[0144] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0145] In one embodiment, a supplier-determined device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: Step 101: Parse the order information to obtain the supplier identifier and the list of order items.

[0146] Step 102: Map each product in the order product list to a leaf node of the supply capacity topology to obtain a set of order capacity nodes.

[0147] Step 103: Obtain the supplier's certification record based on the supplier identifier, and map each certification item in the certification record to the leaf node of the supply capability topology to obtain a set of valid certification nodes. The leaf node represents the smallest category node that can be referenced by the product and can be covered by the certification record.

[0148] Step 104: Determine whether the set of order capability nodes matches the set of valid authentication nodes, and obtain the determination result.

[0149] Step 105: Obtain the supplier's credit limit and the amount already used, and calculate the supplier's available credit limit based on the credit limit and the amount already used.

[0150] Step 106: Assess the supplier's qualification risk based on the judgment result, determine the comprehensive risk value by combining the qualification risk, available acceptance limit and order amount, and determine whether to enable the supplier for the order corresponding to the order information based on the comprehensive risk value.

[0151] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Step 101: Parse the order information to obtain the supplier identifier and the list of order items.

[0152] Step 102: Map each product in the order product list to a leaf node of the supply capacity topology to obtain a set of order capacity nodes.

[0153] Step 103: Obtain the supplier's certification record based on the supplier identifier, and map each certification item in the certification record to the leaf node of the supply capability topology to obtain a set of valid certification nodes. The leaf node represents the smallest category node that can be referenced by the product and can be covered by the certification record.

[0154] Step 104: Determine whether the set of order capability nodes matches the set of valid authentication nodes, and obtain the determination result.

[0155] Step 105: Obtain the supplier's credit limit and the amount already used, and calculate the supplier's available credit limit based on the credit limit and the amount already used.

[0156] Step 106: Assess the supplier's qualification risk based on the judgment result, determine the comprehensive risk value by combining the qualification risk, available acceptance limit and order amount, and determine whether to enable the supplier for the order corresponding to the order information based on the comprehensive risk value.

[0157] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0158] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be determined to be within the scope of this specification.

[0159] The above embodiments merely illustrate several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A supplier selection method, characterized in that, The method includes: Parse the order information to obtain the supplier identifier and the list of items in the order; Map each product in the order product list to a leaf node of the supply capacity topology to obtain a set of order capacity nodes; The supplier's certification record is obtained based on the supplier identifier. Each certification item in the certification record is mapped to a leaf node of the supply capability topology to obtain a set of valid certification nodes. The leaf node represents the smallest category node that can be referenced by the product and covered by the certification record. Determine whether the set of order capability nodes matches the set of valid authentication nodes, and obtain the determination result; Obtain the supplier's credit limit and the amount already used, and calculate the supplier's available credit limit based on the credit limit and the amount already used; The supplier's qualification risk is assessed based on the judgment result. A comprehensive risk value is determined by combining the qualification risk, the available acceptance limit, and the order amount. Based on the comprehensive risk value, it is determined whether the supplier is enabled for the order corresponding to the order information.

2. The method according to claim 1, characterized in that, The step of mapping each product in the order product list to a leaf node of the supply capacity topology to obtain the order capacity node set includes: Iterate through the product identifiers of each product in the order product list and search for a node in the supply capacity topology that matches the product identifier; If a node matching the product identifier is found and the matching node is a leaf node, then the matching node will be used as the order capability node corresponding to the product. If a node matching the product identifier is found and the matching node is a non-leaf node, then the child node of the matching node is taken as the order capability node corresponding to the product. If no node matching the product identifier is found, an error identifier is generated.

3. The method according to claim 2, characterized in that, The step of searching for a node matching the product identifier in the supply capacity topology includes: Extract the semantic feature vector of the product identifier, wherein the semantic feature vector includes at least one of the core word vector, attribute word vector, and relation word vector; Starting from the leaf node level of the supply capacity topology, the semantic feature vectors of each node at the current level are obtained level by level, and the similarity between the semantic feature vector of the product identifier and the semantic feature vectors of each node at the current level is calculated. If there are nodes at the current level with a similarity greater than or equal to a preset threshold, then the node with the highest similarity will be selected as the order capability node. If there are no nodes with a similarity greater than or equal to the preset threshold at the current level, then the process traces back to the parent level of the current level, re-acquires the semantic feature vectors of each node in the parent level and calculates the similarity, until the root node of the supply capability topology is reached.

4. The method according to claim 3, characterized in that, The step of re-acquiring the semantic feature vectors of each node in the parent hierarchy and calculating the similarity includes: If a node with a similarity greater than or equal to the preset threshold is found in the parent level, the child node of the found node will be used as the order capability node, and the depth of the upstream level will be recorded. If the depth of the upstream level exceeds a preset depth threshold, a supply capacity topology update is triggered.

5. The method according to claim 4, characterized in that, The supply capacity topology update includes: Obtain each existing node in the supply capacity topology, and calculate the global semantic distance of the supply capacity topology after inserting the product as a new node below each existing node; Calculate the difference between the global semantic distance of the supply capability topology before insertion and the global semantic distance after insertion, and determine the structural semantic offset corresponding to each insertion position based on the ratio of each difference to the global semantic distance before insertion. Wherein, the global semantic distance is the sum of the semantic distances between nodes in the supply capacity topology; the semantic distance is determined based on the hierarchical path length between nodes and the name similarity between node names; Select the insertion position with the smallest semantic offset of the structure, establish the new node at the insertion position, and use the new node as the order capability node.

6. The method according to claim 1, characterized in that, The step of mapping each authentication item in the authentication record to the leaf node of the supply capacity topology to obtain a set of valid authentication nodes includes: Valid certification items are selected from the certification records that are in effect, have a validity period that includes the current time, and have an applicable scope that includes the procurement organization and the delivery area. Determine the leaf node corresponding to each valid authentication item in the supply capacity topology, and form a set of valid authentication nodes from the determined leaf nodes.

7. The method according to claim 1, characterized in that, The determination of whether the order capability node set matches the valid authentication node set includes the following: If all order capability nodes in the order capability node set exist in the valid authentication node set, then the judgment result is passed; If at least one order capability node does not exist in the set of valid authentication nodes, the judgment result is failure, and a set of failure nodes is output. The set of failure nodes consists of all nodes that do not exist in the set of valid authentication nodes.

8. The method according to claim 1, characterized in that, The step of obtaining the supplier's credit limit and the used credit limit, and calculating the supplier's available credit limit based on the credit limit and the used credit limit includes: Obtain the supplier's multidimensional metrics, which include at least two of the following: delivery metrics, quality metrics, financial metrics, capacity metrics, and collaboration metrics. The comprehensive ability score is obtained by normalizing the multidimensional indicators and then summing them by weight. The supplier's basic credit limit is adjusted based on the comprehensive capability score to obtain a dynamic credit limit; The available credit limit is obtained by subtracting the occupied credit limit from the dynamic credit limit. The occupied credit limit includes the unsettled amount in transit and / or the frozen amount.

9. The method according to any one of claims 1 to 8, characterized in that, The step of assessing the supplier's qualification risk based on the judgment result, determining a comprehensive risk value by combining the qualification risk, the available acceptance limit, and the order amount, and determining whether to enable the supplier for the order corresponding to the order information based on the comprehensive risk value includes: When the judgment result is unsuccessful, the supplier's qualification risk is assessed as the first risk value; When the judgment result is "pass", the supplier's qualification risk is assessed as a second risk value, and the first risk value is higher than the second risk value; The credit limit risk is determined based on the available credit limit and the order amount. The comprehensive risk value is obtained by weighted summing of the qualification risk and the credit limit risk. The supplier is determined based on the comparison between the comprehensive risk value and the preset threshold.

10. A supplier-selected equipment, characterized in that, The supplier-identified equipment includes: Memory, used to store computer programs; A processor, connected to the memory, is used to execute the computer program; A communication interface, connected to the processor, is used for data interaction with external devices; When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 9.

11. A supplier determination system, characterized in that, The supplier determination system includes: Supplier-selected equipment as described in claim 10; External devices connected to the supplier's designated equipment, including an order system, a data source system, and a configuration center.

Images