Method and device for evaluating performance ability of supplier

Abstract

The invention discloses a method and a device for evaluating the performance ability of a supplier, relates to the technical field of supply chain management, and mainly aims to improve the accuracy of an evaluation result. According to the main technical scheme, the method comprises the steps of obtaining historical contract performance data of a to-be-evaluated supplier under a plurality of index factors, wherein each index factor belongs to a corresponding index item; constructing a first judgment matrix of each index item and a second judgment matrix of a performance capability risk assessment target by adopting a preset quantitative assignment method; calculating a first weight of each index factor based on the first judgment matrix and calculating a second weight of each index item based on the second judgment matrix; after the verification of the first weight and the second weight is passed, for each index item, calculating a weighted comprehensive evaluation value of each index item according to the second weight, the first weight of each index factor and historical contract performance data; and inputting the weighted comprehensive evaluation value of each index item into the trained risk assessment model, and outputting a performance capability risk assessment result.

Description

Technical Field

[0001] This invention relates to the field of supply chain management technology, and in particular to a method and apparatus for assessing a supplier's performance capability. Background Technology

[0002] In the implementation of various projects, suppliers, as key participants in contract performance, directly impact whether the project can be completed on schedule and to a high standard. To ensure the smooth progress of the project, it is necessary to conduct a comprehensive assessment of the supplier's performance capabilities before the contract is signed, serving as an important basis for selecting and determining the final contracting supplier.

[0003] Currently, qualitative assessment methods are mainly used, which rely on expert experience to subjectively evaluate various factors such as the supplier's historical performance, management level, and technical capabilities to obtain an assessment result of the supplier's performance capability. However, because this method is highly dependent on individual experience, the assessment process lacks unified and objective quantitative standards, has a strong subjective nature, and cannot comprehensively and systematically reflect the supplier's true risk situation, which may lead to biased assessment results. Summary of the Invention

[0004] In view of the above problems, the present invention provides a method and apparatus for assessing the performance capability of suppliers, the main purpose of which is to improve the accuracy of the assessment results.

[0005] To solve the above-mentioned technical problems, the present invention proposes the following solution: In a first aspect, the present invention provides a method for assessing a supplier's performance capability, the method comprising: Obtain historical contract performance data of the supplier to be evaluated under multiple indicator factors, where each indicator factor belongs to a corresponding indicator item; A first judgment matrix corresponding to each indicator item and a second judgment matrix corresponding to the performance capability risk assessment target are constructed using a preset quantitative assignment method. The first judgment matrix is ​​used to determine the local importance weight of each indicator factor under the same indicator item relative to the indicator item, and the second judgment matrix is ​​used to determine the global importance weight of each indicator item relative to the performance capability risk assessment target. Calculate the first weight of each indicator factor based on the first judgment matrix and calculate the second weight of each indicator item based on the second judgment matrix. After the consistency verification of the first weight and the second weight is passed, for each indicator item, the weighted comprehensive evaluation value of each indicator item is calculated based on the corresponding second weight, the first weight of each indicator factor under the indicator item, and the historical contract performance data under each indicator factor. The weighted comprehensive evaluation value of each indicator is input into the pre-trained risk assessment model, and the risk assessment result of the performance capability of the supplier to be assessed is output.

[0006] Secondly, the present invention provides a supplier performance capability assessment device, the device comprising: The data acquisition unit is used to acquire historical contract performance data of the supplier to be evaluated under multiple indicator factors, where each indicator factor belongs to a corresponding indicator item. The matrix construction unit is used to construct a first judgment matrix corresponding to each indicator item to which each indicator factor belongs in the data acquisition unit and a second judgment matrix corresponding to the performance capability risk assessment target using a preset quantitative assignment method. The first judgment matrix is ​​used to determine the local importance weight of each indicator factor under the same indicator item relative to the indicator item, and the second judgment matrix is ​​used to determine the global importance weight of each indicator item relative to the performance capability risk assessment target. The weight determination unit is used to calculate the first weight corresponding to each indicator factor based on the first judgment matrix obtained by the matrix construction unit and to calculate the second weight corresponding to each indicator item based on the second judgment matrix. The numerical determination unit is used to calculate the weighted comprehensive evaluation value of each indicator item based on the corresponding second weight, the first weight of each indicator factor under the indicator item, and the historical contract performance data under each indicator factor after the consistency verification of the first weight and the second weight determined by the weight determination unit has passed. The risk assessment unit is used to input the weighted comprehensive evaluation value of each indicator obtained by the numerical determination unit into the pre-trained risk assessment model, and output the risk assessment result of the performance capability of the supplier to be assessed.

[0007] To achieve the above objectives, according to a third aspect of the present invention, a storage medium is provided, the storage medium including a stored program, wherein, when the program is executed, the device on which the storage medium is located executes the supplier performance assessment method of the first aspect.

[0008] To achieve the above objectives, according to a fourth aspect of the present invention, a processor is provided for running a program, wherein the program executes the supplier performance capability assessment method of the first aspect described above.

[0009] By employing the above technical solution, this invention provides a method and apparatus for assessing a supplier's performance capability. First, it acquires historical contract performance data of the supplier under multiple indicator factors, where each indicator factor belongs to a corresponding indicator item. Then, using a preset quantitative assignment method, it constructs a first judgment matrix corresponding to each indicator item and a second judgment matrix corresponding to the performance capability risk assessment target. The first judgment matrix determines the local importance weight of each indicator factor relative to the same indicator item, while the second judgment matrix determines the global importance weight of each indicator item relative to the overall performance capability risk assessment target. Therefore, this method does not rely on expert experience scoring or subjective ranking, but rather constructs judgment matrices through standardized quantitative rules, significantly reducing the subjective arbitrariness caused by human factors from the outset. Next, it calculates the first weight of each indicator factor based on the first judgment matrix and the second weight of each indicator item based on the second judgment matrix. After completing the weight calculation, it performs a consistency check on the judgment matrices corresponding to the first and second weights. Only when the check passes does it proceed to the next calculation process. This mechanism effectively prevents weight distortion caused by data anomalies or inaccurate assignment, thereby ensuring the credibility of the weight allocation. Subsequently, for each indicator, a weighted comprehensive evaluation value is calculated by combining its second weight, the first weight of its respective indicator factors, and the corresponding historical contract performance data. Finally, these weighted comprehensive evaluation values ​​are input into a pre-trained risk assessment model to output the risk assessment results of the supplier's performance capability. Compared to existing technologies, this method employs a combination of pre-defined quantitative assignment and matrix operations in the weight calculation process, ensuring the objectivity and repeatability of the weight generation process. Simultaneously, a consistency verification mechanism further enhances the rationality of the weights. Furthermore, the weighted comprehensive evaluation value of the indicator is calculated jointly by multi-level weights and real historical performance data, fully relying on objective business data and avoiding interference from human impressions or subjective biases. Finally, the data-driven risk assessment model achieves the output results, significantly enhancing the accuracy and objectivity of the assessment results.

[0010] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0011] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart of a supplier performance capability assessment method provided by an embodiment of the present invention is shown; Figure 2 A flowchart of another supplier performance capability assessment method provided by an embodiment of the present invention is shown; Figure 3 This diagram illustrates a block diagram of a supplier performance capability assessment device provided in an embodiment of the present invention. Figure 4 This diagram illustrates a block diagram of another supplier performance capability assessment device provided by an embodiment of the present invention. Detailed Implementation

[0012] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0013] To address the common problems of strong subjectivity and poor repeatability in existing supplier performance evaluation methods, this paper proposes a quantifiable and reproducible supplier performance evaluation method through continuous research and practice. This method abandons the simplistic approach of isolated weighting when determining indicator factors and their weights, instead employing a systematic comparison mechanism: for indicator factors belonging to the same indicator item, pairwise comparisons are used to score their relative importance using a pre-defined quantitative assignment method, thereby constructing a first judgment matrix to scientifically determine the local weight of each factor within that indicator item; simultaneously, for different indicator items, the contribution to achieving the overall performance capability risk assessment goal is comprehensively considered, and a second judgment matrix is ​​constructed through pairwise comparisons and quantitative assignment to determine the global weight of each indicator item. This hierarchical and progressive weight determination mechanism not only fully considers the inherent structural relationships of the indicator system but also reduces the arbitrariness of human judgment through standardized quantitative rules, significantly improving the objectivity, logic, and reproducibility of the evaluation process. Next, combined with... Figure 1 This invention provides a method for assessing a supplier's performance capability, and its specific implementation steps are as follows: Figure 1 As shown, steps 101-105 are included: 101. Obtain historical contract performance data of the supplier to be evaluated under multiple indicator factors, where each indicator factor belongs to the corresponding indicator item.

[0014] In practical applications, a multi-dimensional evaluation indicator system can be constructed based on the specific business needs of supplier performance management. This system includes several primary indicator items, such as "quality performance," "supply schedule," "funding management," "service support," and "confidentiality management." Each indicator item can be further subdivided into multiple indicator factors, and each indicator factor has quantified historical contract performance data to comprehensively reflect the supplier's historical performance capabilities.

[0015] Specifically, the "quality performance" indicators can be further subdivided into: quality management system construction, quality assurance capability, production preparation status, management of purchased parts, process quality control, technical status management, metrological testing level, non-conforming product handling mechanism, pass rate of key parts, pass rate of important parts, and pass rate of general parts. The "Supply Progress" indicator can be further subdivided into: employee absenteeism rate, equipment failure rate, material shortage rate, capital shortage rate, production automation level, production plan priority execution rate, inventory plan matching degree, production balance, on-time delivery rate, and plan early completion rate. The "Fund Management" indicators can be further subdivided into: material utilization rate, equipment utilization rate, working hour utilization rate, capital turnover rate, fund project reduction rate, fund overstatement rate, invoice compliance rate, expense overrun rate and overrun amount. The "Service Support" indicator can be further subdivided into: timeliness of delivery and shipment, speed of on-site technical service response, timeliness and effectiveness of problem handling, standardization of contract change management, satisfaction with dispute and breach of contract handling, completeness of contract information archiving, user satisfaction, and overall satisfaction of the purchaser. The "Confidentiality Management" indicator can be further subdivided into: the validity of confidentiality qualification certificates, other qualification certificates required by the industry, compliance of management of personnel handling confidential information, security level of confidential locations, control measures for confidential carriers, number of leaks, direct economic losses caused by leaks, and the degree of social or business impact of leaks.

[0016] All of the above indicators have corresponding historical data available. This data can be collected through information platforms such as enterprise resource planning systems, supply chain management systems, contract management systems, and quality information management systems, and can cover the actual performance records of the suppliers to be evaluated over a certain period of time (such as the past 12 months or the past three years).

[0017] The data collection scope can include all structured data related to the above-mentioned indicators and factors. After collection, the raw data is cleaned and standardized: obviously erroneous or abnormal data (such as negative values ​​or values ​​outside the reasonable range) are removed, and missing data is filled by means of mean filling, interpolation, or completion based on business rules, and the data units and time granularity are unified (such as unifying all time to "day" and amounts to "ten thousand yuan").

[0018] In addition, for qualitative index factors, they need to be converted into quantitative scores through preset scoring rules to ensure the consistency and operability of subsequent calculations.

[0019] The statement that "for qualitative indicator factors, they need to be converted into quantitative scores through preset scoring rules" can be implemented using the following method: For qualitative indicators such as "the standardization of contract change management," a tiered evaluation standard (e.g., a five-level scoring system: 1–5 points) can be developed, and objective scoring can be based on historical records or system traces. For example: The “Contract Change Management Standardization” can be scored according to the dimensions of whether the change process is complete, whether the approval is timely, and whether the documents are complete. Each item is 1 point, and the total score is obtained by weighting the scores. The "impact of the leak" can be assigned a score of 1, 3, or 5 based on the event level (e.g., general, major, major), and quantified by combining historical security incident reports.

[0020] Of course, all qualitative indicators can be quantified using the methods described above. However, the specific scoring criteria for different indicators need to be set flexibly to ensure that the quantification results truly reflect the supplier's actual performance. Given the variety of indicator factors, the quantification criteria for each one will not be detailed here.

[0021] After processing, all data is categorized into a three-level structure: "Supplier - Indicator Item - Indicator Factor." Each historical performance data point is accurately mapped to its corresponding indicator factor, forming a structured dataset organized by supplier and indicator factor. The corresponding indicator item is clearly labeled, providing a complete, accurate, and computable data foundation for subsequent construction of judgment matrices, weight calculations, and comprehensive risk assessments. This process can be automated using data interfaces to improve the efficiency and consistency of data collection and processing.

[0022] Furthermore, it should be noted that the same indicator factor may be categorized into different indicator items depending on the specific project or business scenario, due to differences in its emphasis. For example, "on-time delivery rate" may be categorized under the "supply schedule" indicator item in a project that emphasizes delivery efficiency, while it may be included under the "service support" indicator item in a scenario that emphasizes customer satisfaction. Therefore, the attribution relationship between indicator factors and indicator items is not absolutely fixed and should be dynamically adjusted according to specific needs.

[0023] 102. Construct the first judgment matrix corresponding to each indicator item and the second judgment matrix corresponding to the performance capability risk assessment target using the pre-set quantitative assignment method.

[0024] The first judgment matrix is ​​used to determine the local importance weight of each indicator factor under the same indicator item relative to the indicator item, and the second judgment matrix is ​​used to determine the global importance weight of each indicator item relative to the performance capability risk assessment target.

[0025] In this step, we will explain the specific application of the pre-defined quantitative assignment method, focusing on the relative importance between primary indicator items and among the subordinate indicator factors of each indicator item. We will further illustrate the complete implementation process of constructing the first and second judgment matrices based on this assignment result. Specifically: First, at the indicator factor level, each indicator factor within the same indicator item is compared pairwise, and values ​​are assigned using a preset quantification rule, thereby constructing a first judgment matrix for determining local weights. Then, at the indicator item level, each first-level indicator item is compared pairwise, and importance is assigned using the same preset quantification assignment method, thereby constructing a second judgment matrix for determining global weights.

[0026] First: The rules for applying the preset quantitative assignment method among indicator items (used to construct the second judgment matrix) are as follows: In the application of the pre-defined quantitative assignment method, corresponding weights can be assigned to each indicator based on the attribution analysis of historical risk events or the rule-based priority setting of project type and strategic orientation.

[0027] The attribution method based on historical risk event causation analysis primarily involves statistically analyzing risk events (such as escalating customer complaints, contract defaults, and project delays) that occurred during the supplier's historical contract performance. This analysis identifies the contribution of each indicator to the risk's causes, thus providing an objective basis for determining its relative importance. Frequency statistics and percentage analysis are recommended: all risk events are categorized by cause, the frequency of each indicator as a primary cause is statistically analyzed, and its proportion of the total risk events is calculated.

[0028] For example, if 100 contract disputes occurred in the past three years, of which 70 were caused by delivery delays (the delivery delay indicator factor belongs to the "supply schedule" indicator), 20 were caused by quality problems (belonging to the "quality performance" indicator), 5 were related to improper fund execution (belonging to the "fund management" indicator), and the remaining 5 were caused by other reasons, then the contributions of "supply schedule", "quality performance" and "fund management" to the causes of risk can be concluded to be 70%, 20% and 5% respectively, showing a significant hierarchical difference.

[0029] Based on this, a pairwise comparison and value assignment were performed on each indicator using the 1–9 scale. Since the risk contribution of "supply schedule" is significantly higher than that of "quality performance," and "quality performance" is significantly higher than that of "funding management," this relative importance relationship revealed by the data can be directly used to construct the judgment matrix: when comparing "supply schedule" and "quality performance," based on the significant difference in their contributions, a value of 5 can be reasonably assigned, indicating that the former is "significantly important" relative to the latter; when comparing "supply schedule" and "funding management," given the even more pronounced difference, a value of 9 can be assigned, indicating "extremely important"; and "quality performance" also has a significant difference relative to "funding management," so a value of 5 can also be assigned. The comparisons between other indicators follow the same principle, ensuring that the overall value assignment results are consistent with the historical risk distribution trend.

[0030] Another approach is to establish rules-based priorities based on project type or strategic orientation. This involves pre-determining the importance ranking rules for each indicator item according to the management priorities or corporate strategic goals of different project types. For example, in cost-control projects, "fund management" is usually considered a core dimension of the performance capability risk assessment indicator and can be assigned a higher value, such as 7 (strongly important), compared to "service support." In projects sensitive to delivery cycles, "supply schedule" is placed with the highest priority.

[0031] After obtaining the values ​​of each indicator relative to other indicators, a second judgment matrix can be formed based on these values, for example: Wherein: each row of the second judgment matrix corresponds to an indicator item, and each element in each row represents the importance ratio between the corresponding indicator item and other indicator items relative to the performance capability risk assessment target. Specifically, in the above example: Supply progress (row) relative to supply progress (column): assigned a value of 1 (importance ratio of 1), indicating that "supply progress" is "equally important" than itself; Supply progress (row) vs. quality performance (column): Assigned a value of 3 (importance ratio of 3), indicating that "supply progress" is "slightly more important" than "quality performance"; So, for Quality Performance (row) relative to Supply Schedule (column): assigning a value of 1 / 3 indicates that Quality Performance is more important than Supply Schedule, and is the reciprocal of the importance of Supply Schedule relative to Quality Performance.

[0032] Supply progress (row) compared to funding management (column): A value of 5 indicates that "supply progress" is "significantly more important" than "funding management".

[0033] Quality Performance (row) vs. Funding Management (column): A value of 2 indicates that Quality Performance is between "equally important" and "slightly important" compared to Funding Management.

[0034] Second: The rules for applying the preset quantitative assignment method among indicator factors (used to construct the first judgment matrix) are as follows: Each indicator item corresponds to a first judgment matrix. That is, within each indicator item, its subordinate indicator factors are compared pairwise to determine their relative importance within that indicator item, thereby constructing the first judgment matrix. This process also employs a pre-defined quantitative assignment method, which can be implemented in the following way: Implementation Method 1: Correlation Analysis Based on Indicator Factors and the Core Objectives of Their Belonging Indicator Items Under this method, the core objective of each indicator can be clearly defined first. For example, the core objective of the "quality performance" indicator can be defined as "customer return rate"; the core objective of the "supply schedule" indicator can be defined as "on-time contract delivery rate".

[0035] In this context, for each indicator, the relevance of every two subordinate indicator factors to achieving the core objective of that indicator can be evaluated, serving as the basis for pairwise comparison and assignment. Specifically: Calculate the Pearson correlation coefficient between the historical data of each indicator factor under this indicator item and the aforementioned core objective. The larger the absolute value of the correlation coefficient, the stronger the influence of the indicator factor on the core objective of the indicator item, and the higher its relative importance.

[0036] For example, under the "Quality Performance" section: The correlation coefficient between "critical component pass rate" and "customer return rate" is -0.85. The correlation coefficient between "general parts pass rate" and "customer return rate" is -0.30.

[0037] This indicates that the "critical component pass rate" has a much greater impact on "quality performance" than the "general component pass rate". Therefore, when making pairwise comparisons in the first judgment matrix, the "critical component pass rate" should be judged as "significantly important" relative to the "general component pass rate", and can be assigned a value of 5 according to the 1-9 scale.

[0038] After assigning values ​​to each indicator factor, the first judgment matrix under the corresponding indicator item can be obtained, for example: Wherein: each row of the first judgment matrix corresponds to an indicator factor, and each element in each row represents the importance ratio between the indicator factor and other indicator factors belonging to the same indicator item relative to the indicator item. Specifically, in the example above: The critical component pass rate (row) compared to the critical component pass rate (column) is assigned a value of 1 (importance ratio is 1), indicating that the "critical component pass rate" is "equally important" to itself; Critical component pass rate (row) compared to general component pass rate (column): A value of 5 indicates that the "critical component pass rate" is "significantly more important" than the "general component pass rate".

[0039] It should be noted that, in both the first and second judgment matrices, each row represents an evaluated indicator factor (first matrix) or indicator item (second matrix), and each element in that row represents the importance ratio of that indicator factor or indicator item relative to the indicator factor or indicator item represented by the corresponding column. In other words, the indicator factor or indicator item corresponding to the column is the reference object for comparison.

[0040] Of course, the above matrix form is just a style shown for ease of understanding; in reality, it can be represented as: (Formula 1) in, This represents the results of pairwise element comparisons.

[0041] 103. Calculate the first weight of each indicator factor based on the first judgment matrix and the second weight of each indicator item based on the second judgment matrix.

[0042] 104. After the consistency verification of the first weight and the second weight is passed, for each indicator item, calculate the weighted comprehensive evaluation value of each indicator item based on the corresponding second weight, the first weight of each indicator factor under the indicator item, and the historical contract performance data under each indicator factor.

[0043] 105. Input the weighted comprehensive evaluation value of each indicator into the pre-trained risk assessment model, and output the risk assessment result of the supplier's performance capability to be evaluated.

[0044] In step 103, the first weight corresponding to each indicator factor can be calculated based on the first judgment matrix, and the second weight corresponding to each indicator item can be calculated based on the second judgment matrix. Then, the consistency of the first and second weights can be verified. After the verification passes, in step 104, for each indicator item, the weighted comprehensive evaluation value can be calculated based on the corresponding second weight, the first weight of each indicator factor under the indicator item, and the historical contract performance data under each indicator factor. Finally, in step 105, the weighted comprehensive evaluation values ​​of all indicator items are organized into a feature vector (e.g., [supply progress evaluation value, quality performance evaluation value, financial management evaluation value, service support evaluation value, confidentiality management evaluation value]) in a preset order, and sent as input data to the pre-trained risk assessment model. The model output can be set according to business needs, and can be "low risk, medium risk, or high risk".

[0045] Based on the above Figure 1 As can be seen from the implementation method, the supplier performance capability assessment method provided by this invention first obtains the historical contract performance data of the supplier to be assessed under multiple indicator factors, where each indicator factor belongs to a corresponding indicator item. Then, using a preset quantitative assignment method, a first judgment matrix corresponding to each indicator item and a second judgment matrix corresponding to the performance capability risk assessment target are constructed respectively: the first judgment matrix is ​​used to determine the local importance weight of each indicator factor relative to the same indicator item, and the second judgment matrix is ​​used to determine the global importance weight of each indicator item relative to the overall performance capability risk assessment target. Therefore, this method does not rely on expert experience scoring or subjective ranking, but constructs the judgment matrix through standardized quantitative rules, significantly reducing the subjective arbitrariness caused by human factors from the source. Next, the first weight of each indicator factor is calculated based on the first judgment matrix, and the second weight of each indicator item is calculated based on the second judgment matrix. After completing the weight calculation, the consistency verification of the judgment matrices corresponding to the first and second weights is performed. Only when the verification passes does the subsequent calculation process proceed; this mechanism effectively prevents weight distortion caused by data anomalies or inaccurate assignment, thereby ensuring the credibility of weight allocation. Subsequently, for each indicator, a weighted comprehensive evaluation value is calculated by combining its second weight, the first weight of its respective indicator factors, and the corresponding historical contract performance data. Finally, these weighted comprehensive evaluation values ​​are input into a pre-trained risk assessment model to output the risk assessment results of the supplier's performance capability. Compared to existing technologies, this method employs a combination of pre-defined quantitative assignment and matrix operations in the weight calculation process, ensuring the objectivity and repeatability of the weight generation process. Simultaneously, a consistency verification mechanism further enhances the rationality of the weights. Furthermore, the weighted comprehensive evaluation value of the indicator is calculated jointly by multi-level weights and real historical performance data, fully relying on objective business data and avoiding interference from human impressions or subjective biases. Finally, the data-driven risk assessment model achieves the output results, significantly enhancing the accuracy and objectivity of the assessment results.

[0046] Furthermore, as a response to Figure 1 Further refinement and extension of the illustrated embodiments, this invention also provides another method for assessing a supplier's performance capabilities, such as... Figure 2 As shown, the specific steps are as follows: 201. Obtain historical contract performance data of the supplier to be evaluated under multiple indicator factors, where each indicator factor belongs to the corresponding indicator item.

[0047] 202. The first judgment matrix corresponding to each indicator item and the second judgment matrix corresponding to the performance capability risk assessment target are constructed by using the pre-set quantitative assignment method.

[0048] The implementation methods of steps 201-202 are the same as those of steps 101-102, and can achieve the same technical effect and solve the same technical problem, so they will not be repeated here.

[0049] 203. Calculate the first weight of each indicator factor based on the first judgment matrix and the second weight of each indicator item based on the second judgment matrix.

[0050] In this step, the execution steps for the first weight corresponding to each indicator factor are as follows: 1. Determine the total importance ratio of each indicator factor in each first judgment matrix. Each row of the first judgment matrix corresponds to one indicator factor, and each element in each row represents the importance ratio between the indicator factor and the indicator factors belonging to the same indicator item relative to the indicator item. 2. Calculate the geometric mean of all importance ratios for each indicator factor to obtain the unnormalized weight of the indicator factor. 3. For each first judgment matrix, the sum of the unnormalized weights of all indicator factors is determined as the first normalization coefficient; 4. For each first judgment matrix, divide the unnormalized weight of each indicator factor by the first normalization coefficient to obtain the first weight of each indicator factor.

[0051] The execution steps for the second weight corresponding to each indicator item are as follows: 1. Determine the total importance ratio of each indicator item in the second judgment matrix. Each row of the second judgment matrix corresponds to one indicator item, and each element in each row represents the importance ratio of the corresponding indicator item to other indicator items relative to the performance capability risk assessment target. 2. Calculate the geometric mean of all importance ratios for each indicator to obtain the unnormalized weight of each indicator. 3. The sum of the unnormalized weights of all indicator items is determined as the second normalization coefficient; 4. Divide the unnormalized weight of each indicator by the second normalization coefficient to obtain the second weight of each indicator.

[0052] For example, consider the following matrix: We can first calculate the total importance ratios for each indicator factor. For example, the total importance ratios for the pass rate of key components include 1, 5, and 3. Then, the geometric mean (unnormalized weights) of these ratios is: (Formula 2) Similarly, the geometric mean of the pass rate for general parts is 0.464, and the geometric mean of the factory inspection completeness rate is 0.874. Therefore, the formula for the geometric mean is: (Formula 3) After obtaining the geometric mean of each indicator factor, the first normalization coefficient can be obtained, which is: 2.466 + 0.464 + 0.874 = 3.804. Based on this, the first weight of each indicator factor can be obtained as follows: The first weight for the pass rate of key components = 2.466 / 3.804 ≈ 0.648 The first weight for the pass rate of general parts = 0.464 / 3.804 ≈ 0.122 The first weight for the factory inspection completeness rate = 0.874 / 3.804 ≈ 0.230 Result verification: 0.648 + 0.122 + 0.230 = 1.000.

[0053] Therefore, the formula for calculating the normalized weights can be obtained as follows: (Formula 4) in, It is the unnormalized geometric mean of the i-th indicator item or indicator factor. It is the sum of all unnormalized weights, also known as the normalization coefficient. It is the normalized weight value of the i-th indicator item or indicator factor.

[0054] 204. Use the consistency test algorithm to verify the first weight of each indicator factor and the second weight of each indicator item.

[0055] In this step, the following process can be executed for each first judgment matrix: 1. Calculate the first weighted summation value corresponding to each indicator factor. The first weighted summation value is obtained by weighting and summing the importance ratio of each indicator factor in the row where the indicator factor is located with the first weight of each target indicator factor being compared. 2. Calculate the first product between the matrix order and the first weight of each index factor; 3. Calculate the first ratio between the first weighted sum and the first product of each indicator factor, and determine the sum of the first ratios of all indicator factors as the first largest eigenvalue.

[0056] The following process can be executed for the second judgment matrix: 1. Calculate the second weighted sum value corresponding to each indicator item. The second weighted sum value is obtained by weighting and summing the importance ratio of each indicator item in the row with the second weight of each target indicator item being compared. 2. Calculate the second product between the matrix order of the second judgment matrix and the second weight of each indicator item; 3. Calculate the second ratio between the second weighted sum and the second product of each indicator item, and determine the sum of the second ratios of all indicator items as the second largest eigenvalue.

[0057] Finally, the following process is executed for both the first and second judgment matrices: The consistency ratio of each first judgment matrix is ​​calculated based on the first largest eigenvalue corresponding to each first judgment matrix; the consistency ratio of the second judgment matrix is ​​calculated based on the second largest eigenvalue.

[0058] The first weight of each indicator factor is verified based on the consistency ratio of each first judgment matrix, and the second weight of each indicator item is verified based on the consistency ratio of the second judgment matrix.

[0059] For example: 1. Based on the above matrix, the first weight of the key component pass rate is approximately 2.466 / 3.804 ≈ 0.648. The first weight for the pass rate of general parts = 0.464 / 3.804 ≈ 0.122 The first weight for the factory inspection completeness rate = 0.874 / 3.804 ≈ 0.230 The first weighted sum of the key component pass rate is: (1×0.648)+(5×0.122)+(3×0.230)=0.648+0.610+0.690=1.948 The first weighted sum of the general component pass rate is: (1 / 5 × 0.648) + (1 × 0.122) + (1 / 2 × 0.230) = 0.1296 + 0.122 + 0.115 = 0.3666 The first weighted sum of the factory inspection completeness rate of the indicator factors is: (1 / 3 × 0.648) + (2 × 0.122) + (1 × 0.230) = 0.216 + 0.244 + 0.230 = 0.690 2. Calculate the first product value (matrix order × first weight). Matrix order n=3 The first product of the pass rate for key components: 3 × 0.648 = 1.944 The first product of the general parts pass rate is: 3 × 0.122 = 0.366 The first product of the factory inspection completeness rate: 3 × 0.230 = 0.690 3. Calculate the first ratio and sum them to obtain the first largest eigenvalue. The first ratio for the pass rate of key components: 1.948 ÷ 1.944 = 1.002 The first ratio for the pass rate of general parts: 0.3666 ÷ 0.366 = 1.002 The first ratio of factory inspection completeness rate: 0.690 ÷ 0.690 = 1.000 The first largest eigenvalue of the first judgment matrix =1.002 + 1.002 + 1.000 = 3.004 Therefore, the formula for calculating the largest eigenvalue can be obtained as follows: (Formula 5) For the weighted sum, It is the product of the matrix order and the weight.

[0060] The steps for calculating the consistency ratio of each first judgment matrix based on the first maximum eigenvalue corresponding to each first judgment matrix, and calculating the consistency ratio of the second judgment matrix based on the second maximum eigenvalue, are as follows: 1. Calculate the first difference between the first largest eigenvalue and the corresponding matrix order of each first judgment matrix, and the second difference between the second largest eigenvalue and the matrix order of the second judgment matrix; 2. Calculate the third difference between the matrix order of each first judgment matrix and the specified value, and the fourth difference between the matrix order of the second judgment matrix and the specified value; 3. The ratio between the first difference and the third difference corresponding to each first judgment matrix is ​​determined as the consistency index corresponding to each first judgment matrix, and the ratio between the second difference and the fourth difference is determined as the consistency index of the second judgment matrix; 4. The ratio between the consistency index of each first judgment matrix and the preset average random consistency index is determined as the consistency ratio of each first judgment matrix. 5. The ratio between the consistency index of the second judgment matrix and the preset average random consistency index is determined as the consistency ratio of the second judgment matrix.

[0061] The formulas involved are as follows: (Formula 6) Where CI is the consistency index, n is the matrix order, and 1 is a specified value.

[0062] The formula for the consistency ratio is as follows: (Formula 7) In the formula, CI is the consistency index, and RI is the preset average random consistency index, which is the average random consistency index obtained from the average random consistency index table and the matrix order.

[0063] Finally, the following method can be used to verify the first weight of each indicator factor based on the consistency ratio of each first judgment matrix and the second weight of each indicator item based on the consistency ratio of the second judgment matrix: For each first judgment matrix and second judgment matrix, determine whether the corresponding consistency ratio is less than a preset threshold (usually set to 0.1); if it is less, determine that the first weight of the corresponding indicator factor or the second weight of the indicator item has passed the verification; if it is not less, re-execute the steps of calculating the first weight and the second weight until the first weight of the indicator factor or the second weight of the indicator item has passed the verification.

[0064] 205. After the consistency verification of the first weight and the second weight is passed, for each indicator item, calculate the weighted comprehensive evaluation value of each indicator item based on the corresponding second weight, the first weight of each indicator factor under the indicator item, and the historical contract performance data under each indicator factor.

[0065] In this step, for each indicator item, the first weight of each indicator factor and the weighted sum of historical contract performance data under each indicator factor can be calculated. The weighted sum and the weighted product of the second weight of the indicator item are used as the weighted comprehensive evaluation value of each indicator item.

[0066] For example, the second weight of the supply schedule indicator is 0.648, and its included indicator factors and corresponding first weights are as follows: on-time delivery rate: 0.750, production plan achievement rate: 0.250. The relevant data are: On-time delivery rate: 92 points; Production plan achievement rate: 88 points. The weighted sum of the supply schedule is: (92 × 0.750) + (88 × 0.250) = 69.0 + 22.0 = 91.0. Furthermore, the weighted comprehensive evaluation value of the supply schedule is: 91.0 × 0.648 = 58.968. 206. Input the weighted comprehensive evaluation value of each indicator into the pre-trained risk assessment model, and output the risk assessment result of the supplier's performance capability to be evaluated.

[0067] In this step, the weighted comprehensive evaluation value of each indicator can be used as input features and fed into the input layer of the model. The number of neurons in the input layer is equal to the number of indicators participating in the evaluation. For example, if the evaluation system includes five indicators: "quality performance," "supply progress," "funding management," "service support," and "confidentiality management," then the input layer has five neurons, each corresponding to one indicator, responsible for receiving and transmitting its weighted comprehensive evaluation value.

[0068] Subsequently, each neuron in the input layer transmits the received value to all neurons in the hidden layer. That is, each neuron in the hidden layer will receive the output signals from all five neurons in the input layer.

[0069] Subsequently, in the hidden layer, the following calculation is performed on each neuron: the received input values ​​are summed with their corresponding connection weights, the neuron's threshold (bias term) is subtracted, and the result is substituted into an activation function (such as Sigmoid, ReLU, etc.) for nonlinear transformation to obtain the neuron's output value. Each neuron in the hidden layer has corresponding connection weights with each neuron in the input layer.

[0070] Furthermore, the outputs of each neuron in the hidden layer can be transmitted to each neuron in the output layer of the risk assessment model, wherein each neuron in the output layer has corresponding connection weights with each neuron in the hidden layer.

[0071] Finally, the output layer neurons also perform a weighted summation of the received signals, subtract their own thresholds, and then process the data through an activation function to generate the model's output. The number of neurons in the output layer is determined by the number of neurons in the input layer and the number of neurons in the hidden layer, while the number of neurons in the hidden layer is determined by the number of neurons in the input layer.

[0072] The output result is the risk assessment result of the supplier's performance capability to be evaluated. Its format can be set according to the model design, for example: The output can be a single numerical value representing a risk score or performance capability index; or a multi-dimensional vector representing the probability distribution of different risk levels (such as the probability of low, medium, and high risk); or a direct classification label.

[0073] The formula for calculating the number of neurons in the hidden layer is: (Formula 8) Where n is the number of neurons in the input layer.

[0074] The formula for calculating the number of neurons in the output layer is: (Formula Nine) The inputs to each neuron in the hidden layer are: (Formula 10) In the formula, The input vector passed to the input layer represents the output of each neuron in the input layer. The connection weights are those from input layer neuron i to hidden layer neuron j. The threshold of the hidden layer neurons. A random number taking values ​​in the range (-1, +1).

[0075] The outputs of each neuron in the hidden layer are: (Formula Eleven) The inputs to each neuron in the output layer are: (Formula 12) In the formula, The connection weights from hidden layer neuron j to output layer neuron t are... The threshold of the output layer neurons. A random number taking values ​​in the range (-1, +1).

[0076] The output of each neuron in the output layer is as follows: (Formula Thirteen) The risk assessment model is trained using the backpropagation algorithm. During training, the model first obtains a predicted output based on the input samples; then, the error between the predicted output and the true label is backpropagated layer by layer along the network. Optimization methods such as gradient descent are used to iteratively adjust the connection weights and neuron thresholds (i.e., bias terms) of each layer based on the error, gradually reducing the prediction error and improving the model's fitting ability and generalization performance. This process is repeated until the model converges, ultimately obtaining stable weight and threshold parameters for subsequent supplier fulfillment risk assessment.

[0077] The calculation model for updating the connection weights from hidden layer neuron j to output layer neuron t is as follows: (Formula Fourteen) In the formula, Here, N represents the connection weights from hidden layer neuron j to output layer neuron t, and N is the number of network training iterations. ; The calculation model is as follows: (Formula Fifteen) The threshold update calculation model for the output layer neurons is as follows: (Formula Sixteen) In the formula, For the threshold of the output layer neurons, For output layer neurons The error signal; The connection weight update calculation model from input layer neuron i to hidden layer neuron j is as follows: (Formula 17) In the formula, Here, N represents the connection weights from input layer neuron i to hidden layer neuron j, and N is the number of network training iterations. ; The error signal of hidden layer neuron j is calculated using the following model: (Formula 18) The threshold update calculation model for the hidden layer neurons is as follows: (Formula 19) In the formula, The threshold of the hidden layer neurons.

[0078] Furthermore, as a response to the above Figure 1 In addition to the implementation of the method shown, this embodiment of the invention also provides a supplier performance capability assessment device for evaluating the above-mentioned... Figure 1 The method shown is implemented accordingly. This device embodiment corresponds to the foregoing method embodiment. For ease of reading, this device embodiment will not repeat the details of the foregoing method embodiment, but it should be clear that the device in this embodiment can implement all the contents of the foregoing method embodiment. Figure 3 As shown, the device includes: The data acquisition unit 301 is used to acquire the historical contract performance data of the supplier to be evaluated under multiple indicator factors, wherein each indicator factor belongs to the corresponding indicator item. The matrix construction unit 302 is used to construct a first judgment matrix corresponding to each indicator item to which each indicator factor belongs in the data acquisition unit 301 and a second judgment matrix corresponding to the performance capability risk assessment target using a preset quantitative assignment method. The first judgment matrix is ​​used to determine the local importance weight of each indicator factor under the same indicator item relative to the indicator item, and the second judgment matrix is ​​used to determine the global importance weight of each indicator item relative to the performance capability risk assessment target. The weight determination unit 303 is used to calculate the first weight corresponding to each indicator factor based on the first judgment matrix obtained by the matrix construction unit 302 and to calculate the second weight corresponding to each indicator item based on the second judgment matrix. The numerical determination unit 304 is used to calculate the weighted comprehensive evaluation value of each indicator item based on the corresponding second weight, the first weight of each indicator factor under the indicator item, and the historical contract performance data under each indicator factor after the consistency verification of the first weight and the second weight determined by the weight determination unit 303 is passed. The risk assessment unit 305 is used to input the weighted comprehensive evaluation value of each indicator obtained by the numerical determination unit 304 into the pre-trained risk assessment model and output the risk assessment result of the performance capability of the supplier to be assessed.

[0079] Furthermore, as a response to the above Figure 2 In addition to the implementation of the method shown, this embodiment of the invention also provides another supplier performance capability assessment device for evaluating the above-mentioned... Figure 2 The method shown is implemented accordingly. This device embodiment corresponds to the foregoing method embodiment. For ease of reading, this device embodiment will not repeat the details of the foregoing method embodiment, but it should be clear that the device in this embodiment can implement all the contents of the foregoing method embodiment. Figure 4 As shown, the device includes: The data acquisition unit 301 is used to acquire the historical contract performance data of the supplier to be evaluated under multiple indicator factors, wherein each indicator factor belongs to the corresponding indicator item. The matrix construction unit 302 is used to construct a first judgment matrix corresponding to each indicator item to which each indicator factor belongs in the data acquisition unit 301 and a second judgment matrix corresponding to the performance capability risk assessment target using a preset quantitative assignment method. The first judgment matrix is ​​used to determine the local importance weight of each indicator factor under the same indicator item relative to the indicator item, and the second judgment matrix is ​​used to determine the global importance weight of each indicator item relative to the performance capability risk assessment target. The weight determination unit 303 is used to calculate the first weight corresponding to each indicator factor based on the first judgment matrix obtained by the matrix construction unit 302 and to calculate the second weight corresponding to each indicator item based on the second judgment matrix. The numerical determination unit 304 is used to calculate the weighted comprehensive evaluation value of each indicator item based on the corresponding second weight, the first weight of each indicator factor under the indicator item, and the historical contract performance data under each indicator factor after the consistency verification of the first weight and the second weight determined by the weight determination unit 303 is passed. The risk assessment unit 305 is used to input the weighted comprehensive evaluation value of each indicator obtained by the numerical determination unit 304 into the pre-trained risk assessment model and output the risk assessment result of the performance capability of the supplier to be assessed.

[0080] In one optional implementation, the weight determination unit 303 is specifically used for: Determine the total importance ratio of each indicator factor in each first judgment matrix and the total importance ratio of each indicator item in the second judgment matrix. Each row in the first judgment matrix corresponds to one indicator factor, and each element in each row represents the importance ratio between the indicator factor and the indicator factors belonging to the same indicator item relative to the indicator item. Each row in the second judgment matrix corresponds to one indicator item, and each element in each row represents the importance ratio between the corresponding indicator item and other indicator items relative to the performance capability risk assessment target. Calculate the geometric mean of all importance ratios for each indicator item and each indicator factor to obtain the unnormalized weights of the indicator factors and each indicator item. For each first judgment matrix, the sum of the unnormalized weights of all indicator factors is determined as the first normalization coefficient, and the sum of the unnormalized weights of all indicator items is determined as the second normalization coefficient. For each first judgment matrix, the unnormalized weight of each indicator factor is divided by the first normalization coefficient to obtain the first weight of each indicator factor. Divide the unnormalized weight of each indicator by the second normalization coefficient to obtain the second weight of each indicator.

[0081] In an optional implementation, after the weight determination unit 303 calculates the first weight corresponding to each indicator factor based on the first judgment matrix and the second weight corresponding to each indicator item based on the second judgment matrix, the device further includes a weight verification unit 306, which is specifically used for: For each first judgment matrix, calculate the first weighted summation value corresponding to each indicator factor. The first weighted summation value is obtained by weighting the importance ratio of each indicator factor in the row where the indicator factor is located with the first weight of each target indicator factor being compared. For each first judgment matrix, calculate the first product between the matrix order and the first weight of each indicator factor; For each first judgment matrix, calculate the first ratio between the first weighted sum and the first product of each indicator factor, and determine the sum of the first ratios corresponding to all indicator factors as the first maximum eigenvalue; For the second judgment matrix, calculate the second weighted summation value corresponding to each indicator item. The second weighted summation value is obtained by weighting the importance ratio of each indicator item in the row with the second weight of each target indicator item being compared. Calculate the second product between the matrix order of the second judgment matrix and the second weight of each indicator item; Calculate the second ratio between the second weighted sum and the second product of each indicator item, and determine the sum of the second ratios corresponding to all indicator items as the second largest eigenvalue; The consistency ratio of each first judgment matrix is ​​calculated based on the first maximum eigenvalue corresponding to each first judgment matrix, and the consistency ratio of the second judgment matrix is ​​calculated based on the second maximum eigenvalue. The first weight of each indicator factor is verified based on the consistency ratio of each first judgment matrix, and the second weight of each indicator item is verified based on the consistency ratio of the second judgment matrix.

[0082] In one optional implementation, when the weight verification unit 306 calculates the consistency ratio of each first judgment matrix based on the first maximum eigenvalue corresponding to each first judgment matrix, and calculates the consistency ratio of the second judgment matrix based on the second maximum eigenvalue, it is specifically used for: Calculate the first difference between the first largest eigenvalue of each first judgment matrix and the corresponding matrix order, and the second difference between the second largest eigenvalue and the matrix order of the second judgment matrix; Calculate the third difference between the matrix order of each first judgment matrix and the specified value, and the fourth difference between the matrix order of the second judgment matrix and the specified value; The ratio between the first difference and the third difference corresponding to each first judgment matrix is ​​determined as the consistency index corresponding to each first judgment matrix, and the ratio between the second difference and the fourth difference is determined as the consistency index of the second judgment matrix. The ratio between the consistency index of each first judgment matrix and the preset average random consistency index is determined as the consistency ratio of each first judgment matrix. The ratio between the consistency index of the second judgment matrix and the preset average random consistency index is determined as the consistency ratio of the second judgment matrix.

[0083] In one optional implementation, the weight verification unit 306, when verifying the first weight of each indicator factor according to the consistency ratio of each first judgment matrix and verifying the second weight of each indicator item according to the consistency ratio of the second judgment matrix, is specifically used for: For each first judgment matrix and second judgment matrix, determine whether the corresponding consistency ratio is less than a preset threshold; If it is less than, then the first weight of the corresponding indicator factor or the second weight of the indicator item is verified as passed; If it is not less than, then the steps of calculating the first weight and the second weight are re-executed until the first weight of the indicator factor or the second weight of the indicator item passes the verification.

[0084] In one optional implementation, the numerical determination unit 304 is specifically used for: For each indicator item, calculate the first weight of each indicator factor and the weighted sum of historical contract performance data under each indicator factor; The weighted summation value and the weighted product of the second weight of the indicator item are used as the weighted comprehensive evaluation value of each indicator item.

[0085] In one optional implementation, the risk assessment unit 305 is specifically used for: The weighted comprehensive evaluation value of each indicator item is input into each neuron in the input layer of the risk assessment model; The weighted comprehensive evaluation values ​​received by each neuron in the input layer are transmitted to each neuron in the hidden layer of the risk assessment model; In the hidden layer, the received weighted comprehensive evaluation values ​​and corresponding connection weights are weighted and summed. The result of the weighted summation is then subtracted from the corresponding neuron threshold and processed by an activation function to obtain the output of each neuron in the hidden layer. Each neuron in the hidden layer has a corresponding connection weight with each neuron in the input layer. The outputs of each neuron in the hidden layer are transmitted to each neuron in the output layer of the risk assessment model, wherein each neuron in the output layer has a corresponding connection weight with each neuron in the hidden layer; In the output layer, the outputs of each neuron in the received hidden layer are weighted and summed with their corresponding connection weights. The weighted sum is then subtracted from the corresponding neuron threshold and processed by an activation function to obtain the risk assessment result of the supplier's performance capability. The number of neurons in the output layer is determined by the number of neurons in the input layer and the number of neurons in the hidden layer, and the number of neurons in the hidden layer is determined by the number of neurons in the input layer.

[0086] Furthermore, embodiments of the present invention also provide a storage medium for storing a computer program, wherein the computer program, when running, controls the device where the storage medium is located to execute the above-described... Figure 1-2 The supplier performance capability assessment method described in the document.

[0087] Furthermore, embodiments of the present invention also provide a processor for running a program, wherein the program executes the above-described... Figure 1-2 The supplier performance capability assessment method described in the document.

[0088] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0089] It is understood that the relevant features in the above methods and apparatus can be referenced interchangeably. Furthermore, the terms "first," "second," etc., in the above embodiments are used to distinguish between embodiments and do not represent the superiority or inferiority of any particular embodiment.

[0090] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0091] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, this invention is not directed to any particular programming language. It should be understood that the contents of the invention described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the invention.

[0092] In addition, the memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0093] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0094] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0095] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0096] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0097] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0098] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0099] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0100] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0101] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0102] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for assessing a supplier's performance capability, characterized in that, The method includes: Obtain historical contract performance data of the supplier to be evaluated under multiple indicator factors, where each indicator factor belongs to a corresponding indicator item; A first judgment matrix corresponding to each indicator item and a second judgment matrix corresponding to the performance capability risk assessment target are constructed using a preset quantitative assignment method. The first judgment matrix is ​​used to determine the local importance weight of each indicator factor under the same indicator item relative to the indicator item, and the second judgment matrix is ​​used to determine the global importance weight of each indicator item relative to the performance capability risk assessment target. Calculate the first weight of each indicator factor based on the first judgment matrix and calculate the second weight of each indicator item based on the second judgment matrix. After the consistency verification of the first weight and the second weight is passed, for each indicator item, the weighted comprehensive evaluation value of each indicator item is calculated based on the corresponding second weight, the first weight of each indicator factor under the indicator item, and the historical contract performance data under each indicator factor. The weighted comprehensive evaluation value of each indicator is input into the pre-trained risk assessment model, and the risk assessment result of the performance capability of the supplier to be assessed is output.

2. The method according to claim 1, characterized in that, The calculation of the first weight corresponding to each indicator factor based on the first judgment matrix and the calculation of the second weight corresponding to each indicator item based on the second judgment matrix include: Determine the total importance ratio of each indicator factor in each first judgment matrix and the total importance ratio of each indicator item in the second judgment matrix. Each row in the first judgment matrix corresponds to one indicator factor, and each element in each row represents the importance ratio between the indicator factor and the indicator factors belonging to the same indicator item relative to the indicator item. Each row in the second judgment matrix corresponds to one indicator item, and each element in each row represents the importance ratio between the corresponding indicator item and other indicator items relative to the performance capability risk assessment target. Calculate the geometric mean of all importance ratios for each indicator item and each indicator factor to obtain the unnormalized weights of the indicator factors and each indicator item. For each first judgment matrix, the sum of the unnormalized weights of all indicator factors is determined as the first normalization coefficient, and the sum of the unnormalized weights of all indicator items is determined as the second normalization coefficient. For each first judgment matrix, the unnormalized weight of each indicator factor is divided by the first normalization coefficient to obtain the first weight of each indicator factor. Divide the unnormalized weight of each indicator by the second normalization coefficient to obtain the second weight of each indicator.

3. The method according to claim 1, characterized in that, After calculating the first weight corresponding to each indicator factor based on the first judgment matrix and calculating the second weight corresponding to each indicator item based on the second judgment matrix, the method further includes: For each first judgment matrix, calculate the first weighted summation value corresponding to each indicator factor. The first weighted summation value is obtained by weighting the importance ratio of each indicator factor in the row where the indicator factor is located with the first weight of each target indicator factor being compared. For each first judgment matrix, calculate the first product between the matrix order and the first weight of each indicator factor; For each first judgment matrix, calculate the first ratio between the first weighted sum and the first product of each indicator factor, and determine the sum of the first ratios corresponding to all indicator factors as the first maximum eigenvalue; For the second judgment matrix, calculate the second weighted summation value corresponding to each indicator item. The second weighted summation value is obtained by weighting the importance ratio of each indicator item in the row with the second weight of each target indicator item being compared. Calculate the second product between the matrix order of the second judgment matrix and the second weight of each indicator item; Calculate the second ratio between the second weighted sum and the second product of each indicator item, and determine the sum of the second ratios of all indicator items as the second largest eigenvalue; The consistency ratio of each first judgment matrix is ​​calculated based on the first maximum eigenvalue corresponding to each first judgment matrix, and the consistency ratio of the second judgment matrix is ​​calculated based on the second maximum eigenvalue. The first weight of each indicator factor is verified based on the consistency ratio of each first judgment matrix, and the second weight of each indicator item is verified based on the consistency ratio of the second judgment matrix.

4. The method according to claim 3, characterized in that, The consistency ratio of each first judgment matrix is ​​calculated based on the first maximum eigenvalue corresponding to each first judgment matrix, and the consistency ratio of the second judgment matrix is ​​calculated based on the second maximum eigenvalue, including: Calculate the first difference between the first largest eigenvalue of each first judgment matrix and the corresponding matrix order, and the second difference between the second largest eigenvalue and the matrix order of the second judgment matrix; Calculate the third difference between the matrix order of each first judgment matrix and the specified value, and the fourth difference between the matrix order of the second judgment matrix and the specified value; The ratio between the first difference and the third difference corresponding to each first judgment matrix is ​​determined as the consistency index corresponding to each first judgment matrix, and the ratio between the second difference and the fourth difference is determined as the consistency index of the second judgment matrix. The ratio between the consistency index of each first judgment matrix and the preset average random consistency index is determined as the consistency ratio of each first judgment matrix. The ratio between the consistency index of the second judgment matrix and the preset average random consistency index is determined as the consistency ratio of the second judgment matrix.

5. The method according to claim 3, characterized in that, The first weight of each indicator factor is verified based on the consistency ratio of each first judgment matrix, and the second weight of each indicator item is verified based on the consistency ratio of the second judgment matrix, including: For each first judgment matrix and second judgment matrix, determine whether the corresponding consistency ratio is less than a preset threshold; If it is less than, then the first weight of the corresponding indicator factor or the second weight of the indicator item is verified as passed; If it is not less than, then the steps of calculating the first weight and the second weight are re-executed until the first weight of the indicator factor or the second weight of the indicator item passes the verification.

6. The method according to claim 1, characterized in that, For each indicator item, a weighted comprehensive evaluation value is calculated based on the corresponding second weight, the first weight of each indicator factor under the indicator item, and the historical contract performance data under each indicator factor, including: For each indicator item, calculate the first weight of each indicator factor and the weighted sum of historical contract performance data under each indicator factor; The weighted summation value and the weighted product of the second weight of the indicator item are used as the weighted comprehensive evaluation value of each indicator item.

7. The method according to claim 1, characterized in that, The weighted comprehensive evaluation value of each indicator is input into a pre-trained risk assessment model, which outputs the risk assessment results of the supplier's performance capability, including: The weighted comprehensive evaluation value of each indicator item is input into each neuron in the input layer of the risk assessment model; The weighted comprehensive evaluation values ​​received by each neuron in the input layer are transmitted to each neuron in the hidden layer of the risk assessment model; In the hidden layer, the received weighted comprehensive evaluation values ​​and corresponding connection weights are weighted and summed. The result of the weighted summation is then subtracted from the corresponding neuron threshold and processed by an activation function to obtain the output of each neuron in the hidden layer. Each neuron in the hidden layer has a corresponding connection weight with each neuron in the input layer. The outputs of each neuron in the hidden layer are transmitted to each neuron in the output layer of the risk assessment model, wherein each neuron in the output layer has a corresponding connection weight with each neuron in the hidden layer; In the output layer, the outputs of each neuron in the received hidden layer are weighted and summed with their corresponding connection weights. The weighted sum is then subtracted from the corresponding neuron threshold and processed by an activation function to obtain the risk assessment result of the supplier's performance capability. The number of neurons in the output layer is determined by the number of neurons in the input layer and the number of neurons in the hidden layer, and the number of neurons in the hidden layer is determined by the number of neurons in the input layer.

8. A supplier performance capability assessment device, characterized in that, The device includes: The data acquisition unit is used to acquire historical contract performance data of the supplier to be evaluated under multiple indicator factors, where each indicator factor belongs to a corresponding indicator item. The matrix construction unit is used to construct a first judgment matrix corresponding to each indicator item to which each indicator factor belongs in the data acquisition unit and a second judgment matrix corresponding to the performance capability risk assessment target using a preset quantitative assignment method. The first judgment matrix is ​​used to determine the local importance weight of each indicator factor under the same indicator item relative to the indicator item, and the second judgment matrix is ​​used to determine the global importance weight of each indicator item relative to the performance capability risk assessment target. The weight determination unit is used to calculate the first weight corresponding to each indicator factor based on the first judgment matrix obtained by the matrix construction unit and to calculate the second weight corresponding to each indicator item based on the second judgment matrix. The numerical determination unit is used to calculate the weighted comprehensive evaluation value of each indicator item based on the corresponding second weight, the first weight of each indicator factor under the indicator item, and the historical contract performance data under each indicator factor after the consistency verification of the first weight and the second weight determined by the weight determination unit has passed. The risk assessment unit is used to input the weighted comprehensive evaluation value of each indicator obtained by the numerical determination unit into the pre-trained risk assessment model, and output the risk assessment result of the performance capability of the supplier to be assessed.

9. A storage medium, characterized in that, The storage medium includes a stored program, wherein, when the program is executed, it controls the device where the storage medium is located to perform the supplier performance assessment method as described in any one of claims 1 to 7.

10. A processor, characterized in that, The processor is used to run a program, wherein the program executes the supplier performance capability assessment method as described in any one of claims 1 to 7.

Images