Supplier analysis recommendation method and system, computing device and readable storage medium

Abstract

The invention provides a supplier analysis recommendation method and system, a computing device and a readable storage medium. The method comprises the steps of obtaining supplier influence factors and supplier expression trend embedding; constructing an influence factor adjacency matrix according to the supplier influence factors; based on the supplier performance trend embedding and the influence factor adjacency matrix, using a graph convolutional network to analyze supplier performance scores; and performing clustering analysis calculation on the supplier performance score through an LSTM algorithm, and determining a supplier analysis result and a supplier recommendation scheme according to a calculation result. According to the method, scientific and efficient supplier selection decision support can be provided for an enterprise based on accurate analysis of deep association between time sequence information and factors, and optimization and upgrading of a purchasing strategy are ensured.

Description

Technical Field

[0001] This invention relates to the field of supply chain technology, and in particular to a supplier analysis and recommendation method, system, computing device, and readable storage medium. Background Technology

[0002] In the context of current economic globalization and rapid development of information technology, the importance of enterprise procurement management is becoming increasingly prominent, especially multi-supplier data analysis. It is essential to establish a comprehensive and multi-dimensional supplier evaluation system, including but not limited to factors such as price, quality, delivery time, after-sales service, and technological innovation capabilities, to ensure that the selected suppliers can not only meet current needs but also match the company's long-term development.

[0003] Supplier analysis models are an important tool for enterprises to enhance their competitiveness in the context of economic globalization. Through continuous optimization and innovation, enterprises can manage procurement activities more efficiently and accurately, thereby gaining a favorable position in fierce market competition. Existing enterprise supplier analysis and forecasting typically use statically defined procurement data based on prior knowledge (such as industry feedback and supply relationships), making it difficult to capture supplier influencing factors that change over time.

[0004] Therefore, there is an urgent need for a supplier analysis and recommendation method, system, computing device, and readable storage medium to solve the above problems. Summary of the Invention

[0005] The technical problem solved by this invention is to provide a supplier analysis and recommendation method, system, computing device and readable storage medium. While deeply considering the multi-dimensional influencing factors such as supplier supply quality, timeliness, supply capacity and price, it also cleverly captures the dynamic changes and mutual influences of these influencing factors in different time series, which significantly improves the accuracy and efficiency of the decision-making process.

[0006] The technical problem solved by this invention is achieved by the following technical solution: A supplier analysis and recommendation method includes: obtaining supplier influence factors and supplier performance trend embeddings; constructing an adjacency matrix of the influence factors based on the supplier influence factors; analyzing supplier performance scores using a graph convolutional network based on the supplier performance trend embeddings and the influence factor adjacency matrix; performing cluster analysis on the supplier performance scores using an LSTM algorithm; and determining supplier analysis results and supplier recommendation schemes based on the calculation results.

[0007] In a preferred embodiment of the present invention, the step of constructing the adjacency matrix of influencing factors based on the supplier influencing factors includes: determining the correlation between the supplier influencing factors based on each supplier influencing factor; generating an influencing factor similarity graph based on the correlation; and determining the influencing factor adjacency matrix based on the influencing factor similarity graph.

[0008] In a preferred embodiment of the present invention, the step of obtaining the supplier performance trend embedding includes: determining the historical supplier influence factor and the number of historical suppliers within a preset time period; and determining the supplier performance trend embedding based on the historical supplier influence factor and the number of historical suppliers.

[0009] In a preferred embodiment of the present invention, the step of determining the adjacency matrix of the influencing factors based on the similarity relationship graph of the influencing factors includes: determining the correlation matrix between the supplier influencing factors based on the similarity relationship graph of the influencing factors; confirming the unit matrix based on the number of the influencing factors; and determining the adjacency matrix of the influencing factors based on the unit matrix and the correlation matrix.

[0010] In a preferred embodiment of the present invention, the supplier analysis results include strategic, cooperative, and elimination types. The step of determining the supplier analysis results and supplier recommendation scheme based on the calculation results further includes: obtaining a comprehensive score for each supplier based on the calculation results, and classifying the supplier recommendation level based on the comprehensive score, so as to identify suppliers whose comprehensive score exceeds a first preset value as strategic types, suppliers whose comprehensive score is within the range of a second preset value to the first preset value as cooperative types, and suppliers whose comprehensive score does not exceed the second preset value as elimination types; and generating the supplier recommendation scheme based on the recommendation level.

[0011] In a preferred embodiment of the present invention, after the steps of performing cluster analysis calculation on the supplier performance score using the LSTM algorithm and determining the supplier analysis and supplier recommendation scheme based on the calculation results, the method further includes: outputting the supplier analysis results and supplier recommendation scheme to the enterprise SRM system through an API interface, wherein the enterprise SRM system supports multi-terminal visualization display.

[0012] In a preferred embodiment of the present invention, after the steps of performing cluster analysis calculation on the supplier performance score using the LSTM algorithm and determining the supplier analysis and supplier recommendation scheme based on the calculation results, the method further includes: assessing the supply chain disruption risk using Monte Carlo simulation based on the supplier recommendation scheme; and generating a reminder message when the supply chain disruption risk exceeds a threshold to remind the user to introduce a new supplier.

[0013] A supplier analysis and recommendation system includes: a data acquisition module for acquiring supplier influencing factors and supplier performance trend embeddings; a data processing module for constructing the influencing factor adjacency matrix based on the supplier influencing factors; and a supplier analysis and recommendation module for analyzing supplier performance scores using a graph convolutional network based on the supplier performance trend embeddings and the influencing factor adjacency matrix, performing cluster analysis on the supplier performance scores using an LSTM algorithm, and determining supplier analysis results and supplier recommendation schemes based on the calculation results.

[0014] A computing device includes a memory and a processor, wherein the memory stores a supplier analysis and recommendation program, and when executed by the processor, the supplier analysis and recommendation program implements the steps of the supplier analysis and recommendation method as described in any one of the preceding descriptions.

[0015] A readable storage medium storing a computer program that, when executed by a processor, implements the steps of the supplier analysis and recommendation method as described in any one of the preceding descriptions.

[0016] The technical effects achieved by the above-mentioned technical solution of this invention are as follows: by using graph attention networks to analyze supplier performance scores, and finally using LSTM algorithms to cluster the data, the best supplier analysis and recommendation scheme is determined based on the results; by combining long short-term memory networks and graph convolutional networks, and utilizing the correlation between various influencing factors, multiple relationship graphs are effectively aggregated, thereby forming an effective relationship embedding for the analysis of the next time series of suppliers, which significantly improves the accuracy and efficiency of the decision-making process.

[0017] 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 to make the above and other objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the steps of a supplier analysis and recommendation method according to the present invention.

[0019] Figure 2 This invention illustrates a similarity relationship diagram of influencing factors.

[0020] Figure 3 This is a block diagram illustrating a supplier analysis and recommendation system according to the present invention.

[0021] Figure 4 This is a schematic diagram of the structure of a computing device according to the present invention. Detailed Implementation

[0022] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the embodiments of the present invention. Through the description of the specific embodiments, a more in-depth and specific understanding of the technical means and effects adopted by the present invention to achieve its intended purpose can be obtained. Moreover, the accompanying drawings are only provided for reference and illustration and are not intended to limit the present invention.

[0023] Please see Figure 1 , Figure 1 This is a flowchart illustrating the steps of a supplier analysis and recommendation method according to the present invention.

[0024] like Figure 1 As shown, a supplier analysis and recommendation method provided by an embodiment of the present invention includes the following steps: S11: Obtain supplier influence factors and supplier performance trend embeddings.

[0025] By clearly defining the company's specific needs and standards, including product quality, price, delivery time, and service level, these standards will then serve as the benchmark for selecting suppliers. These standards are known as supplier influencing factors.

[0026] , represents the embedded performance trend of all suppliers; that is, the score matrix of S suppliers corresponding to K influencing factors (the score matrix of the Sth supplier to the k influencing factors at time t), which can illustrate the historical scores of each supplier.

[0027] Specifically, to analyze suppliers, it is first necessary to scientifically construct a supplier time-series analysis system. In some embodiments, the method includes: identifying supplier influencing factors (i.e., supplier influencing factors and evaluation indicators) based on supplier evaluation requirements; wherein, supplier influencing factors include, but are not limited to, price, quality, delivery time (such as timeliness and supply capacity), after-sales service, and technological innovation capabilities. For example, the method also includes: periodically collecting data on suppliers and supplier influencing factors. The collection period can be automatically updated daily, monthly, or every three months, and the collection period is determined according to the recommended requirements, without any limitation here.

[0028] Optionally, the step of obtaining the supplier performance trend embedding includes: determining the historical supplier influence factor and the number of historical suppliers within a preset time period; and determining the supplier performance trend embedding based on the historical supplier influence factor and the number of historical suppliers.

[0029] For example, the preset time period can be customized and adjusted according to needs, and it can be ten days, one month, half a year, one year or two years, etc.

[0030] In this embodiment, four influencing factors—price, quality, after-sales service, and technological innovation—are used as examples for explanation.

[0031] S12: Construct an adjacency matrix of influencing factors based on supplier influence factors.

[0032] Optionally, the step of constructing the adjacency matrix of influencing factors based on the supplier influencing factors includes: determining the correlation between the supplier influencing factors based on each supplier influencing factor; generating an influencing factor similarity graph based on the correlation; and determining the influencing factor adjacency matrix based on the influencing factor similarity graph.

[0033] Optionally, the step of determining the adjacency matrix of the influencing factors based on the similarity graph of the influencing factors includes: determining the correlation matrix between the supplier influencing factors based on the similarity graph of the influencing factors; confirming the unit matrix based on the number of influencing factors; and determining the adjacency matrix of the influencing factors based on the unit matrix and the correlation matrix.

[0034] Specifically, the adjacency matrix of influencing factors is determined by the similarity relationship diagram of the influencing factors of suppliers (also known as the influencing factor relationship diagram, such as...). Figure 2 As shown below; the following explanation uses four influencing factors—price, quality, after-sales service, and technological innovation—as examples; the similarity relationship diagram of influencing factors (i.e., the relationship between the four influencing factors—price, quality, after-sales service, and technological innovation) and the adjacency matrix of influencing factors are also provided. +I.

[0035] like Figure 2 The diagram shows the similarity relationship between influencing factors. The lines connecting the various influencing factors represent the correlation between them. Each line represents a correlation, and the more lines there are, the more correlations there are (which can be represented by the number 1 in the correlation matrix). Figure 2 The labels next to each impact factor can be used to indicate the order of impact factors, and they can be reflected in the identity matrix and the correlation matrix.

[0036] In such Figure 2 The similarity relationship diagram of the influencing factors is shown below: A= ;I=

[0037] Where I is the identity matrix and A is the correlation matrix between the influencing factors.

[0038] It should be noted that the number of permutations in both the correlation matrix and the identity matrix is ​​the same as the number of influencing factors. The more influencing factors there are, the more permutations are required in these two matrices. For example, with 3 influencing factors, the identity matrix and correlation matrix would have 3 rows and 3 columns; with 5 influencing factors, the identity matrix and correlation matrix would have 5 rows and 5 columns. The number of influencing factors for suppliers is determined based on the actual scenario and is not limited here; for example, when suppliers are involved in cross-border transactions, exchange rate volatility can also be considered as an additional influencing factor.

[0039] For example, in the correlation matrix A shown in this application, the first row is used to represent the impact factor 1 (technological innovation, in Figure 2 The correlation between the index 1 and the influence factor 2 (price) is only related to the influence factor 2. Figure 2 The factor with index 2 is associated, so the second column of the first row in the matrix is ​​set to 1; the second row is used to represent the association of factor 2 (price), which is associated with all three other factors, so the three columns excluding the second column of the second row are all set to 1; the third row is used to represent the association of factor 3 (quality, in Figure 2 The relationship between the index number 3 and its impact factors 2 (price) and 4 (after-sales service) is discussed. Figure 2 The index number is 4), so the second and fourth columns in the third row are set to 1; the fourth row is used to represent the correlation of impact factor 4 (after-sales), which is associated with impact factor 2 and impact factor 3, so the second and third columns in the fourth row are set to 1.

[0040] Specifically, the adjacency matrix of influencing factors is obtained based on the similarity relationship graph of influencing factors. Also known as the edge weight of a graph.

[0041] S13: Analyze supplier performance scores using graph convolutional networks based on supplier performance trend embedding and influencing factor adjacency matrix.

[0042] Graph Convolutional Networks (GCNs) are deep learning models based on graph-structured data. They extend convolutional operations to a non-Euclidean space, enabling the aggregation and propagation of node features. By learning the relationships between nodes and their neighbors, GCNs are widely used in knowledge graphs, social network analysis, and action recognition. A typical GCN includes a spatiotemporal attention mechanism; the spatial attention layer calculates the correlation between supplier nodes, while the temporal attention layer processes the periodic features of historical performance data.

[0043] For example, the inputs to GCN include the following two aspects: supplier performance trend embedding H (also known as the node features of the graph) and the adjacency matrix of the similarity relationship graph of influencing factors. (Edge weights of the graph). Details are as follows: ; Where ReLU is a non-linear activation function. +I, where I is the identity matrix and D is... The degree matrix, This represents the weight matrix for time series t, with its initial values ​​(i.e., the scores of the S suppliers and K influence factors at time t). The input is based on the supplier's actual performance (i.e., obtained offline).

[0044] Specifically, by utilizing the connections in the supplier influencing factor relationship graph, the features of adjacent factor nodes (adjacent factor nodes are nodes directly connected by an edge in the influencing factor relationship graph) are aggregated through graph convolution operations; that is, the information of each supplier factor node (influencing factor) is distributed to its neighboring nodes, and a new representation is generated for each node, thereby learning the cross-influence between influencing factors.

[0045] S14: Perform cluster analysis on supplier performance scores using the LSTM algorithm, and determine the supplier analysis results and supplier recommendation scheme based on the calculation results.

[0046] Long Short-Term Memory (LSTM) is a special type of Recurrent Neural Network (RNN). LSTM was designed to address the vanishing or exploding gradient problems that traditional RNNs encounter when processing long sequences of data, enabling the network to learn long-term dependencies.

[0047] By combining graph convolutional networks and LSTM algorithms, it is possible to simultaneously capture the local correlations between influencing factors (CNN) and the long-term rating trend (LSTM).

[0048] Specifically, the convolutional results from all time steps are concatenated (i.e., supplier performance scores are clustered) to form a comprehensive time-series feature representation. An LSTM layer is used to receive this time-series feature representation, and its gating mechanism captures the time-dependent and dynamic changes of node attributes. The calculation formula between LSTM layers is shown below: [ ); [ ); [ ); + * ; [ ); ; Where t represents the time series corresponding to the supplier's performance (e.g., first year, second year, etc.), σ is the sigmoid activation function, k represents the input gate output, q represents the forget gate output, v represents the cell state, o represents the output gate output, h represents the hidden state, W represents the weight matrix, b represents the bias term, Wq represents the weight of the forget gate, and bq represents the bias term of the forget gate.

[0049] Specifically, fully connected layers are used to map the complex and high-dimensional features output by the LSTM layers. By using activation functions, the non-linear relationships between features are learned. The continuous temporal supplier performance ratings (initial historical data) obtained through the graph convolutional network are then input into the model to predict the supplier performance in the next time series (which can be used to describe the supplier's future rating). The S suppliers can be calculated using LSTM. The final recommended supplier performance analysis values ​​are then used to obtain the supplier analysis results and supplier recommendation schemes.

[0050] In this embodiment, the convolution result is obtained through a graph convolutional network, and then input into the LSTM algorithm to obtain the final result. The following scheme can be referred to: Convolutional Feature Extraction: Apply convolution to the score matrix (e.g., 10 suppliers × 5 years × 5 indicators), using a 3×3 kernel to extract local temporal patterns. For example: Input: [10, 5, 5] Convolution Output: [10, 3, 8] (8 feature maps, time steps reduced due to convolution) Dimensional transformation: Adjust the convolution output to the LSTM input format: [10, 3, 8] (8-dimensional features are used as input to the new time step). LSTM Temporal Modeling: LSTM processes convolutional features step by step over time and outputs the final temporal representation (e.g., [10, 36]) to predict supplier risk levels.

[0051] Optionally, the supplier analysis results include strategic, cooperative, and elimination categories. The step of determining the supplier analysis results and supplier recommendation scheme based on the calculation results further includes: obtaining a comprehensive score for each supplier based on the calculation results, and classifying the supplier recommendation level based on the comprehensive score, so as to identify suppliers whose comprehensive score exceeds a first preset value as strategic, suppliers whose comprehensive score is within the range of a second preset value to the first preset value as cooperative, and suppliers whose comprehensive score does not exceed the second preset value as elimination; and generating the supplier recommendation scheme based on the recommendation level.

[0052] Based on the predicted supplier performance (score) for the next timeframe, suppliers can be categorized into three levels: Strategic (cooperation priority ≥ 90 points), Leverage (70-89 points), and Elimination (< 70 points). Corresponding strategies are then implemented based on each supplier's level, including supplier recommendations. When recommending suppliers, those with higher scores (Strategic) are prioritized for similar products.

[0053] Supplier risk assessments are conducted based on predicted supplier performance (scores) for the next timeframe, categorized as high-risk, medium-risk, and low-risk. For high-risk suppliers, a deadline for rectification or replacement can be set, and a special quality meeting will be held. For medium- and low-risk suppliers, risk can be mitigated through a "1+1+N" system (primary supplier + alternative supplier + regional substitute).

[0054] Optionally, after performing cluster analysis on the supplier performance score using the LSTM algorithm and determining the supplier analysis and recommendation scheme based on the calculation results, the method further includes: using Monte Carlo simulation to assess the supply chain disruption risk based on the supplier recommendation scheme; and generating a reminder message when the supply chain disruption risk exceeds a threshold to remind the user to introduce a new supplier.

[0055] Based on the supplier recommendation scheme, Monte Carlo simulation is used to assess the risk of supply chain disruption, including: using Monte Carlo simulation to assess and confirm the supply chain disruption risk of all the company's current supplier systems; when the supply chain disruption risk exceeds a preset threshold, a reminder message can be sent to remind users to introduce new suppliers to reduce or diversify the risk.

[0056] Optionally, after performing cluster analysis calculation on the supplier performance score using the LSTM algorithm and determining the supplier analysis and supplier recommendation scheme based on the calculation results, the process includes: outputting the supplier analysis results and supplier recommendation scheme to the enterprise SRM system through an API interface, wherein the enterprise SRM system supports multi-terminal visualization display.

[0057] After obtaining the supplier analysis results and supplier recommendations, the results are output to the enterprise SRM system via API interface for display. This allows users to quickly understand the performance of all the company's suppliers through a visual interface, providing the enterprise with scientific and efficient supplier selection decision support, ensuring the optimization and upgrading of procurement strategies, and significantly improving the accuracy and efficiency of the decision-making process.

[0058] The supplier analysis and recommendation provided in this implementation not only deeply considers multiple dimensions of supplier performance, such as supply quality, timeliness, supply capacity, and price, but also cleverly captures the dynamic changes and mutual influences of these indicators across different time series. This enables comprehensive and detailed performance tracking and evaluation of enterprise suppliers. Furthermore, through the powerful capabilities of deep learning, especially the multi-attention mechanism of graph attention networks, it can intelligently analyze and interpret complex data relationships, efficiently classifying and managing information across multiple time series. Ultimately, based on the precise analysis of these time-series information and the deep-seated relationships between factors, the model can provide enterprises with scientific and efficient supplier selection decision support, ensuring the optimization and upgrading of procurement strategies and significantly improving the accuracy and efficiency of the decision-making process.

[0059] Figure 3 This is a block diagram illustrating a supplier analysis and recommendation system according to the present invention.

[0060] like Figure 3 As shown, the present invention also provides a supplier analysis and recommendation system, comprising: a data acquisition module for acquiring supplier influencing factors and supplier performance trend embeddings; a data processing module for constructing the influencing factor adjacency matrix based on the supplier influencing factors; and a supplier analysis and recommendation module for analyzing supplier performance scores using a graph convolutional network based on the supplier performance trend embeddings and the influencing factor adjacency matrix, performing cluster analysis calculations on the supplier performance scores using an LSTM algorithm, and determining supplier analysis results and supplier recommendation schemes based on the calculation results.

[0061] The supplier analysis and recommendation system described in this embodiment can be implemented using the steps of the supplier analysis and recommendation method described in any one of the above embodiments.

[0062] Based on the same inventive concept as the foregoing embodiments, this embodiment of the invention provides a computing device, such as... Figure 4 As shown, the computing device includes: a processor 610 and a memory 611 storing computer programs; wherein, Figure 4 The processor 610 shown in the diagram does not indicate that there is only one processor 610, but only indicates the positional relationship of the processor 610 relative to other devices. In practical applications, there can be one or more processors 610; similarly, Figure 4The memory 611 illustrated in the diagram has the same meaning, that is, it is only used to indicate the positional relationship of memory 611 relative to other devices. In practical applications, there can be one or more memories 611. When the processor 610 runs the computer program, it implements the method for generating the above-described process scheme.

[0063] The computing device may also include at least one network interface 612. The various components of the computing device are coupled together via a bus system 613. It is understood that the bus system 613 is used to implement communication between these components. In addition to a data bus, the bus system 613 also includes a power bus, a control bus, and a status signal bus. However, for clarity, in... Figure 4 The general designated all buses as Bus System 613.

[0064] Based on the same inventive concept as the foregoing embodiments, this embodiment also provides a computer-readable storage medium storing a computer program. The computer-readable storage medium can be a magnetic random access memory (FRAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM), etc.; it can also be various devices including one or any combination of the above-mentioned memories, such as mobile phones, computers, tablet devices, personal digital assistants, etc. When the computer program stored in the computer-readable storage medium is run by a processor, it implements the above-described process solution generation method. For the specific steps implemented when the computer program is executed by the processor, please refer to [link to relevant documentation]. Figure 1 The description of the illustrated embodiments will not be repeated here.

[0065] Through the above description of the embodiments, those skilled in the art can clearly understand that the embodiments of the present invention can be implemented in hardware or by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solutions of the embodiments of the present invention can be embodied in the form of a software product. This software product can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, mobile hard drive, etc.) and includes several instructions to cause a computer device (such as a personal computer, server, or network device, etc.) to execute the methods described in the various implementation scenarios of the embodiments of the present invention.

[0066] The preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the specific details of the above embodiments. The above embodiments and accompanying drawings are exemplary. The modules or processes in the accompanying drawings are not necessarily necessary for implementing the embodiments of the present invention and should not be construed as limiting the present invention. Within the scope of the technical concept of the present invention, various simple modifications and combinations can be made to the technical solutions of the present invention, and these simple modifications and combinations all fall within the protection scope of the present invention.

Claims

1. A supplier analysis and recommendation method, characterized in that, include: Obtain embedded supplier influencing factors and supplier performance trends; Construct an adjacency matrix of influencing factors based on the supplier influencing factors; Based on the supplier performance trend embedding and the adjacency matrix of influencing factors, a graph convolutional network is used to analyze the supplier performance score; The supplier performance scores are clustered using the LSTM algorithm, and the supplier analysis results and supplier recommendation schemes are determined based on the calculation results.

2. The supplier analysis and recommendation method as described in claim 1, characterized in that, The step of constructing the adjacency matrix of the influencing factors based on the supplier influencing factors includes: Based on each of the supplier influence factors, determine the correlation between the various supplier influence factors; Based on the aforementioned relationships, a similarity graph of influencing factors is generated; Based on the similarity graph of the influencing factors, the adjacency matrix of the influencing factors is determined.

3. The supplier analysis and recommendation method as described in claim 1, characterized in that, The steps to obtain embedded supplier performance trends include: Determine the historical supplier impact factors and the number of historical suppliers within the preset time period; The supplier performance trend embedding is determined based on the historical supplier influence factor and the number of historical suppliers.

4. The supplier analysis and recommendation method as described in claim 2, characterized in that, The steps for determining the adjacency matrix of the influencing factors based on the similarity graph of the influencing factors include: Based on the similarity diagram of the influencing factors, determine the correlation matrix among the supplier influencing factors; The identity matrix is ​​determined based on the number of influencing factors, and the adjacency matrix of the influencing factors is determined based on the identity matrix and the correlation matrix.

5. The supplier analysis and recommendation method as described in claim 1, characterized in that, The supplier analysis results include strategic, cooperative, and elimination types; The steps for determining the supplier analysis results and recommended supplier solutions based on the calculation results also include: The comprehensive score of each supplier is obtained based on the calculation results, and the supplier recommendation level is divided according to the comprehensive score. Suppliers whose comprehensive score exceeds the first preset value are identified as strategic type, suppliers whose comprehensive score is within the range of the second preset value to the first preset value are identified as cooperative type, and suppliers whose comprehensive score does not exceed the second preset value are identified as elimination type. The supplier recommendation scheme is generated based on the recommendation level.

6. The supplier analysis and recommendation method as described in claim 1, characterized in that, After performing cluster analysis on the supplier performance scores using the LSTM algorithm and determining the supplier analysis and recommendation scheme based on the calculation results, the process includes: The supplier analysis results and supplier recommendation schemes are output to the enterprise SRM system via API interface, and the enterprise SRM system supports multi-terminal visualization display.

7. The supplier analysis and recommendation method as described in claim 1, characterized in that, After performing cluster analysis on the supplier performance scores using the LSTM algorithm and determining the supplier analysis and recommendation scheme based on the calculation results, the method further includes: Based on the supplier recommendation scheme, Monte Carlo simulation was used to assess the risk of supply chain disruption. When the risk of supply chain disruption exceeds a threshold, an alert message is generated to remind the user to introduce a new supplier.

8. A supplier analysis and recommendation system, characterized in that, include: The data acquisition module obtains embedded supplier influencing factors and supplier performance trends; The data processing module constructs the adjacency matrix of the influencing factors based on the supplier influencing factors; The supplier analysis and recommendation module analyzes supplier performance scores using graph convolutional networks based on the supplier performance trend embedding and the adjacency matrix of influencing factors, performs cluster analysis on the supplier performance scores using the LSTM algorithm, and determines the supplier analysis results and supplier recommendation schemes based on the calculation results.

9. A computing device, characterized in that, The computing device includes a memory and a processor, wherein the memory stores a supplier analysis and recommendation program, and when the supplier analysis and recommendation program is executed by the processor, it implements the steps of the supplier analysis and recommendation method as described in any one of claims 1 to 7.

10. A readable storage medium, characterized in that, The readable storage medium stores a computer program that, when executed by a processor, implements the steps of the supplier analysis and recommendation method as described in any one of claims 1 to 7.

Images