Commodity raw material supplier evaluation method, system, device, medium, and program product

Abstract

The application discloses a kind of bulk raw material supplier evaluation method, system, equipment, medium and program product, the method includes: based on the original relevant data of multiple sample suppliers is collected to multiple source heterogeneous data source, and the original relevant data is preprocessed, obtain sample data, based on business requirement information to sample supplier carries out static analysis, generates static evaluation parameter, and based on static evaluation parameter constructs initial evaluation model, based on sample data to sample supplier carries out dynamic analysis, and based on dynamic analysis result to initial evaluation model is adjusted, obtains target evaluation model, the data to be analyzed of the supplier to be analyzed is input to target evaluation model and is evaluated;Since the application is adjusted by dynamic analysis optimization static model, realize static and dynamic combination evaluation, to accurately evaluate the real ability and potential risk of supplier, realize quickly, efficiently screen out the most suitable supplier, improve the decision-making ability of bulk raw material procurement.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data processing, in particular to a bulk raw material supplier evaluation method, system, device, medium and program product. BACKGROUND

[0002] Under the industrial environment of global commodity trade and domestic industrial production, enterprises face a complex and dynamic market pattern when carrying out bulk raw material (covering imported mines, domestic mines, domestic coal, imported coal and other categories) procurement. As a key link at the front end of the production chain, bulk raw material procurement needs to support the group-based multi-base production layout. However, there are problems such as less information source, scattered collection channel, poor timeliness, and difficult adjustment of index system in the supplier evaluation of procurement plan and procurement contract formulation. At the same time, the evaluation method mainly depends on manual experience and simple data statistics, and lacks comprehensive, systematic and accurate evaluation means, which leads to the fact that the evaluation results cannot accurately reflect the real ability and potential risks of the suppliers. When facing complex market environment and diverse procurement demands, it is difficult to quickly and efficiently select the most suitable suppliers, affecting the scientificity of procurement decision. SUMMARY

[0003] The main purpose of the present application is to provide a bulk raw material supplier evaluation method, system, device, medium and program product, which aims to solve the technical problem that the prior art cannot accurately reflect the real ability and potential risks of the suppliers when facing complex market environment and diverse procurement demands, leading to the fact that it is difficult to quickly and efficiently select suitable suppliers, affecting the scientificity of procurement decision.

[0004] To achieve the above-mentioned purpose, the present application provides a bulk raw material supplier evaluation method, which comprises the following steps:

[0005] Based on a plurality of heterogeneous data sources, raw related data of a plurality of sample suppliers is collected, and the raw related data is preprocessed to obtain sample data, the sample data comprising business change data, delivery time sequence data and historical behavior data;

[0006] Based on business demand information, the sample suppliers are statically analyzed to generate static evaluation parameters, and an initial evaluation model is constructed based on the static evaluation parameters, the static evaluation parameters comprising a plurality of static evaluation dimensions and a weight vector corresponding to each static evaluation dimension, and the business demand information comprising evaluation conditions of bulk raw material procurement;

[0007] Based on the sample data, the sample suppliers are dynamically analyzed, and the initial evaluation model is adjusted based on the dynamic analysis result to obtain a target evaluation model, the dynamic analysis comprising business change analysis, delivery time sequence analysis and historical behavior analysis;

[0008] Collecting original relevant data of the supplier to be analyzed, and preprocessing the original relevant data to obtain data to be analyzed;

[0009] The data to be analyzed is input into the target evaluation model for evaluation to obtain an evaluation result of the supplier to be analyzed.

[0010] Optionally, performing static analysis on the sample suppliers based on the business demand information to generate static evaluation parameters, and constructing an initial evaluation model based on the static evaluation parameters includes:

[0011] Decompose business demand information into multiple static evaluation dimensions based on hierarchical analysis strategy;

[0012] Constructing a judgment matrix according to the multiple static evaluation dimensions, and calculating the maximum eigenvalue of the judgment matrix and the eigenvector corresponding to the maximum eigenvalue;

[0013] Normalizing each dimensional component of the feature vector to obtain a weight vector for each static evaluation dimension;

[0014] The consistency parameters of the judgment matrix are calculated based on the maximum eigenvalue, and the consistency parameters include a consistency index, a random consistency index, and a consistency ratio:

[0015] ;

[0016] ;

[0017] in, represents the consistency index, represents the random consistency index, represents the consistency ratio, Represents the dimension of the judgment matrix;

[0018] Determining whether the consistency of the judgment matrix meets a preset condition according to the consistency parameter;

[0019] If the consistency of the judgment matrix meets the preset conditions, it is determined that the weight vector configuration is reasonable, and an initial evaluation model is constructed based on the multiple static evaluation dimensions and the weight vector.

[0020] Optionally, the business demand information is decomposed into multiple static evaluation dimensions based on the hierarchical analysis strategy, including:

[0021] Decompose business demand information into multiple candidate evaluation dimensions based on hierarchical analysis strategy;

[0022] performing historical behavior analysis on the historical behavior data to obtain historical behavior characteristics of the sample supplier, the historical behavior characteristics including price fluctuation characteristics, delivery cycle characteristics, quality characteristics and service characteristics, the price fluctuation characteristics being obtained based on a formula as follows:

[0023] ;

[0024] wherein, denotes a standard deviation of price fluctuation, used to measure the price fluctuation characteristics of the sample supplier, denotes a sample number, denotes a th price data point, denotes an average value of all the price data points;

[0025] the delivery cycle characteristics being obtained based on a formula as follows:

[0026] ;

[0027] wherein, denotes a standard deviation of delivery time, used to measure the delivery cycle characteristics of the sample supplier, denotes a th delivery time data point, denotes an average value of all the delivery time data points;

[0028] the quality characteristics being obtained based on a formula as follows:

[0029] ;

[0030] wherein, denotes a product qualified batch ratio of the sample supplier, used to measure the quality characteristics of the sample supplier, denotes a qualified batch number passed through quality inspection, denotes a total delivery batch number;

[0031] the service characteristics being obtained based on a formula as follows:

[0032] ;

[0033] wherein, denotes an average time required for processing complaints, used to measure the service characteristics of the sample supplier, denotes a time for processing a th complaint, denotes a total complaint number;

[0034] screening a plurality of static evaluation dimensions from the candidate evaluation dimensions based on the historical behavior characteristics.

[0035] Optionally, the dynamic analysis is performed on the sample supplier based on the sample data, and the initial evaluation model is adjusted based on a dynamic analysis result to obtain a target evaluation model, including:

[0036] Change information monitoring is performed on the sample supplier based on the business change data to obtain a dynamic change monitoring result, and the change information includes business change information and market change information;

[0037] A dynamic change portrait of the sample supplier is constructed according to the dynamic change monitoring result;

[0038] The weight vectors of each static evaluation dimension in the initial evaluation model are redistributed based on the dynamic change portrait to obtain a target weight vector;

[0039] The initial evaluation model is adjusted based on the target weight vector to obtain a target evaluation model.

[0040] Optionally, the initial evaluation model is adjusted based on the target weight vector to obtain a target evaluation model, including:

[0041] Data cleaning is performed on the delivery time series data to obtain an initial delivery time series;

[0042] Delivery feature analysis is performed on the initial delivery time series, and the initial delivery time series is decomposed into a transaction price time series and a material quotation time series based on a delivery feature analysis result;

[0043] Price difference analysis is performed based on the transaction price time series and the material quotation time series to generate a price difference sub-time series;

[0044] The number of windows is determined according to the step length of the price difference sub-time series, and the price difference sub-time series is divided into a plurality of sub-sequences based on the number of windows;

[0045] The sub-sequences are respectively input into an ARIMA model for time series prediction to obtain a plurality of sub-prediction features;

[0046] The plurality of sub-prediction features are fused to obtain a target delivery prediction result;

[0047] The weight vector of the market price dimension in the target weight vector is adjusted based on the target delivery prediction result to obtain a market price weight vector;

[0048] The initial evaluation model is adjusted according to the market price weight vector and the target weight vector to obtain a target evaluation model.

[0049] Optionally, the adjusting the initial evaluation model according to the market price weight vector and the target weight vector to obtain a target evaluation model comprises:

[0050] performing historical behavior analysis on the sample supplier based on historical behavior data to generate historical behavior labels of the sample supplier and time labels corresponding to each historical behavior label;

[0051] determining associated users having correlations with the sample supplier according to the historical behavior labels;

[0052] determining association weights of each associated user based on the time labels;

[0053] generating an associated entity path according to the association weights and the associated users;

[0054] generating an associated entity knowledge graph of the sample supplier based on the associated entity path and the associated users;

[0055] performing risk analysis according to the associated entity knowledge graph to obtain a risk infection path of the sample supplier;

[0056] constructing a risk portrait of the sample supplier based on the risk infection path;

[0057] adjusting the initial evaluation model according to the risk portrait, the market price weight vector and the target weight vector.

[0058] In addition, to achieve the above object, the present application further provides a bulk raw material supplier evaluation system, which comprises:

[0059] In addition, to achieve the above object, the present application further provides a bulk raw material supplier evaluation device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the computer program is configured to implement the steps of the bulk raw material supplier evaluation method as described above.

[0060] In addition, to achieve the above object, the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps of the bulk raw material supplier evaluation method as described above.

[0061] In addition, to achieve the above object, the present application further provides a computer program product, which comprises a computer program, and the computer program is executed by a processor to implement the steps of the bulk raw material supplier evaluation method as described above.

[0062] The application is based on collecting original related data of a plurality of sample suppliers from a plurality of heterogeneous data sources, and pre-processing the original related data to obtain sample data, the sample data including business change data, delivery time sequence data and historical behavior data; based on business demand information, performing static analysis on the sample suppliers to generate static evaluation parameters, and based on the static evaluation parameters, constructing an initial evaluation model, the static evaluation parameters including a plurality of static evaluation dimensions and a weight vector corresponding to each static evaluation dimension, and the business demand information including evaluation conditions of bulk raw material procurement; based on the sample data, performing dynamic analysis on the sample suppliers, and based on the dynamic analysis result, adjusting the initial evaluation model to obtain a target evaluation model, the dynamic analysis including business change analysis, delivery time sequence analysis and historical behavior analysis; collecting original related data of a to-be-analyzed supplier, and pre-processing the original related data to obtain to-be-analyzed data; inputting the to-be-analyzed data into the target evaluation model for evaluation to obtain an evaluation result of the to-be-analyzed supplier; since the application collects original related data of a plurality of sample suppliers from a plurality of heterogeneous data sources and pre-processes the original related data, the comprehensiveness and quality reliability of evaluation data dimensions are ensured, the initial evaluation model is constructed based on a plurality of static evaluation dimensions through static analysis, so that the evaluation is in line with actual business demand, meets the diversified procurement demand of a complex market environment, and the initial evaluation model is optimized and adjusted through dynamic analysis, so that the evaluation model can accurately reflect the real ability and potential risk of the supplier, through the evaluation mode combining static analysis and dynamic analysis, the most suitable supplier is quickly and efficiently screened out, and the decision-making ability of bulk raw material procurement is improved. BRIEF DESCRIPTION OF DRAWINGS

[0063] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, for those skilled in the art, other drawings can also be obtained without creative labor.

[0064] Figure 1 is a structural schematic diagram of a bulk raw material supplier evaluation device of a hardware running environment related to the embodiment scheme of the present application;

[0065] Figure 2 is a flowchart of a first embodiment of a bulk raw material supplier evaluation method of the present application;

[0066] Figure 3 is a flowchart of a second embodiment of a bulk raw material supplier evaluation method of the present application;

[0067] Figure 4This is a structural block diagram of the first embodiment of the bulk raw material supplier evaluation system of the present invention.

[0068] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0069] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0070] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of a bulk raw material supplier evaluation device in the hardware operating environment involved in the embodiment of the present invention.

[0071] like Figure 1 As shown, the bulk raw material supplier assessment device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display and an input unit, such as a keyboard. Optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a wireless fidelity (WI-FI) interface). The memory 1005 may be a high-speed random access memory (RAM) or a stable non-volatile memory (NVM), such as a disk drive. The memory 1005 may also be a storage system independent of the processor 1001.

[0072] Those skilled in the art will understand that Figure 1 The structure shown in the figure does not constitute a limitation on the bulk raw material supplier evaluation device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0073] like Figure 1 As shown, the memory 1005 as a computer-readable storage medium may include an operating system, a network communication module, a user interface module, and a bulk raw material supplier evaluation program.

[0074] exist Figure 1In the bulk raw material supplier evaluation device shown, the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 in the bulk raw material supplier evaluation device can be arranged in the bulk raw material supplier evaluation device, and the bulk raw material supplier evaluation device calls a bulk raw material supplier evaluation program stored in the memory 1005 through the processor 1001, and executes the bulk raw material supplier evaluation method provided in the embodiment of the application.

[0075] The embodiment of the application provides a bulk raw material supplier evaluation method, which refers to Figure 2 , Figure 2 FIG. 1 is a flowchart of a bulk raw material supplier evaluation method according to a first embodiment of the application.

[0076] In the embodiment, the bulk raw material supplier evaluation method comprises the following steps:

[0077] Step S10: Collecting original relevant data of a plurality of sample suppliers based on a plurality of source heterogeneous data sources, and pre-processing the original relevant data to obtain sample data.

[0078] It should be noted that the embodiment is applied to the capability and risk evaluation of a bulk raw material supplier, and the evaluation method combining static analysis and dynamic analysis ensures that the evaluation result accurately reflects the real capability and potential risk of the supplier, so that the most suitable supplier can be quickly and efficiently screened in the face of complex market environment and various procurement demands, and the scientificity of procurement decision is improved.

[0079] It should be understood that the execution subject of the embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone and the like, or a terminal electronic device capable of realizing the above functions. The following takes a bulk raw material supplier evaluation device (referred to as an evaluation device) as an example to describe the embodiment and each of the following embodiments.

[0080] It should be noted that the sample data comprises business change data, delivery time sequence data and historical behavior data.

[0081] It should be noted that the plurality of source heterogeneous data sources can comprise a plurality of data source channels such as a procurement center, a fund payment table, a settlement table, a contract table and a quality inspection table.

[0082] It should be noted that the original relevant data can be multi-dimensional heterogeneous data, for example, the original relevant data can include delivery time series data of a supplier, historical behavior data, business change time series data, identity data, credit data, default data, etc. The data types of the original relevant data can include numerical values, texts, images, audios, and videos, etc.

[0083] In some embodiments, the evaluation device can perform data cleaning, data standardization processing, feature extraction, and data structuring processing on the original relevant data. For example, the evaluation device can map different types of original relevant data to the same dimension for analysis, such as mapping supplier-related image data and text data to the same dimension of feature vectors through image feature extraction (such as target detection and edge detection, etc.) and NLP processing.

[0084] For example, for the missing actual delivery date in the delivery time series data, if the missing proportion is less than 10%, the mean value interpolation method is used. Taking supplier B as an example, if the actual delivery date of an order is missing, the average delivery cycle of other orders in the same month is 15 days, then the actual delivery date of the order is set to the expected delivery date + 15 days. If the missing proportion is greater than 10%, the data record is directly deleted.

[0085] Step S20: performing static analysis on the sample supplier based on the business demand information, generating a static evaluation parameter, and constructing an initial evaluation model based on the static evaluation parameter.

[0086] It should be noted that the static evaluation parameter includes a plurality of static evaluation dimensions and a weight vector corresponding to each static evaluation dimension, and the business demand information includes an evaluation condition of bulk raw material procurement.

[0087] In some embodiments, the static evaluation dimensions can include key dimensions such as basic qualification and compliance, supply capacity, financial and credit risk, overall cooperation situation, historical material supply situation, market price, etc. Each dimension can be further divided into a plurality of specific evaluation indexes, such as enterprise attribute, registered scale, mine owner / trader, region, subject legality, business scope of the enterprise, etc. under basic qualification and compliance.

[0088] It can be understood that the present embodiment sets a weight for each evaluation dimension and specific index, and establishes an initial evaluation model for bulk raw material suppliers. Since the scoring weights of different raw material types (such as imported ore, domestic ore, domestic coal, and imported coal) can be different, the present embodiment decomposes the business demand information into a plurality of static evaluation dimensions, and reflects the respective unique evaluation requirements through the evaluation model of multi-dimensional evaluation.

[0089] Further, in order to effectively decompose the complex multi-objective bulk raw material procurement demand and evaluate the suppliers in multiple dimensions, the step S20 can include:

[0090] Step S201: decompose the business demand information into multiple static evaluation dimensions based on the analytic hierarchy strategy;

[0091] Step S202: construct a judgment matrix according to the multiple static evaluation dimensions, and calculate the maximum eigenvalue of the judgment matrix and the eigenvector corresponding to the maximum eigenvalue;

[0092] Step S203: normalize each dimension component of the eigenvector to obtain a weight vector of each static evaluation dimension;

[0093] Step S204: calculate the consistency parameter of the judgment matrix based on the maximum eigenvalue;

[0094] Step S205: determine whether the consistency degree of the judgment matrix meets the preset condition according to the consistency parameter;

[0095] Step S206: if the consistency degree of the judgment matrix meets the preset condition, it is determined that the weight vector is reasonable, and an initial evaluation model is constructed based on the multiple static evaluation dimensions and the weight vector.

[0096] It should be noted that the analytic hierarchy strategy can be the analytic hierarchy process (AHP), which is a decision-making method that decomposes complex multi-objective decision-making problems into target, criterion, scheme, etc. levels, and combines qualitative and quantitative analysis.

[0097] For example, the evaluation device can use the analytic hierarchy process (AHP) to determine the weight distribution of each dimension and establish a scorecard system containing five first-level indicators (qualification review 20%, supply capacity 25%, financial risk 20%, cooperation performance 15%, and historical performance 20%).

[0098] In some embodiments, the evaluation device can process the quantitative problem of qualitative indicators by fuzzy comprehensive evaluation method, such as converting "cooperation degree" into a 1-5 Likert scale.

[0099] In a specific implementation, the evaluation device can divide the decision-making target, consideration factor (criterion), and decision-making object of the bulk raw material procurement business demand into the highest layer (decision-making purpose), the middle layer (consideration factor, criterion), and the lowest layer (alternative scheme) according to their relationships, draw a hierarchical structure diagram, and clearly define the relationships between each layer, thereby decomposing the business demand into multiple static evaluation dimensions.

[0100] It should be noted that the evaluation device can compare the importance of the same level factors with each other according to the criteria of the last level by using the Saaty 1-9 scale method (1 represents equal importance, 3 represents slightly important, …, 9 represents extremely important, and 2 and 4 are intermediate values between adjacent judgments), to form a judgment matrix, and to reflect the relative importance between the factors.

[0101] It should be noted that the evaluation device calculates the eigenvector corresponding to the maximum eigenvalue of the judgment matrix, and obtains the ordering weight vector of the same level factors to the upper level factors after normalization, and calculates a consistency parameter to determine the rationality of the weight vector. The consistency parameter includes a consistency index, a random consistency index, and a consistency ratio. The consistency parameter is calculated according to the following formula:

[0102] ;

[0103] ;

[0104] wherein, the consistency index is represented by CI, the random consistency index is represented by CR, the consistency ratio is represented by RI, and the dimension of the judgment matrix is represented by n.

[0105] Further, in order to improve the evaluation efficiency, reduce the complexity of decision-making, and ensure the rationality of decision-making, the above step S201 can include:

[0106] Step S2011: decomposing the business requirement information into a plurality of candidate evaluation dimensions based on the analytic hierarchy strategy;

[0107] Step S2012: performing historical behavior analysis on the historical behavior data to obtain historical behavior characteristics of the sample supplier;

[0108] Step S2013: screening a plurality of static evaluation dimensions from the candidate evaluation dimensions based on the historical behavior characteristics.

[0109] It should be noted that the evaluation device decomposes the business requirements (such as cost control and supply chain stability) into a plurality of candidate evaluation dimensions based on the analytic hierarchy strategy, and performs historical behavior analysis combined with the historical behavior data of the sample supplier, so as to screen out key and important static evaluation dimensions.

[0110] In some embodiments, the evaluation device learns a model by establishing a self-learning mechanism, analyzes the historical behavior of the user by using a machine learning algorithm, automatically labels the related data, and preferentially recommends the labeled dimensions (i.e., the static evaluation dimensions) when evaluating the same type of business.

[0111] It can be understood that the historical behavior characteristics include price fluctuation characteristics, delivery cycle characteristics, quality characteristics and service characteristics, the price fluctuation characteristics are obtained based on the following formula:

[0112] ;

[0113] wherein, represents the standard deviation of price fluctuation, used to measure the price fluctuation characteristics of the sample supplier, represents the sample quantity, represents the th price data point, represents the average value of all price data points;

[0114] The delivery cycle characteristics are obtained based on the following formula:

[0115] ;

[0116] wherein, represents the standard deviation of delivery time, used to measure the delivery cycle characteristics of the sample supplier, represents the th delivery time data point, represents the average value of all delivery time data points;

[0117] The quality characteristics are obtained based on the following formula:

[0118] ;

[0119] wherein, represents the product qualified batch ratio of the sample supplier, used to measure the quality characteristics of the sample supplier, represents the qualified batch quantity passed the quality inspection, represents the total delivery batch quantity;

[0120] The service characteristics are obtained based on the following formula:

[0121] ;

[0122] wherein, represents the average time required for handling complaints, used to measure the service characteristics of the sample supplier, represents the time for handling the th complaint, represents the total complaint quantity.

[0123] Step S30: dynamically analyzing the sample supplier based on the sample data, and adjusting the initial evaluation model based on the dynamic analysis result to obtain a target evaluation model.

[0124] It should be noted that the dynamic analysis includes business change analysis, delivery time series analysis and historical behavior analysis.

[0125] In some embodiments, the evaluation device can analyze the business change information of the sample supplier by monitoring the business data and market behavior data of the sample supplier, so as to dynamically adjust the weight of each static evaluation dimension.

[0126] Listen to the supplier portrait, business data table structure and market major changes, automatically trigger the change of static evaluation dimensions. At the same time, using the data obtained through the real-time data acquisition interface, combined with the dynamic weight distribution algorithm, automatically adjust the weight vector of each static evaluation dimension, so as to realize the adjustment and optimization of the initial evaluation model.

[0127] In some embodiments, the evaluation device can dynamically adjust the initial evaluation model by performing time series analysis and prediction on the delivery time series data of the supplier based on the delivery time series prediction results, for example, according to the comparison between the recent transaction price and the material quotation, and timely adjust the weight score of the market price dimension.

[0128] In some embodiments, the evaluation device can establish a model self-learning mechanism, analyze user historical behavior using machine learning algorithms, automatically label related data, and preferentially recommend labeled dimensions during the evaluation of the same type of business.

[0129] In some embodiments, the evaluation device can apply a reinforcement learning framework to dynamically adjust the weight of each static evaluation dimension according to the historical decision effect (such as increasing the timeliness of delivery weight from 30% to 45% when a certain type of raw material is out of stock).

[0130] Step S40: Collecting original relevant data of the supplier to be analyzed, and pre-processing the original relevant data to obtain the data to be analyzed.

[0131] It should be noted that the preprocessing can include missing value filling, abnormal data processing, numerical standardization processing, feature extraction, data integration and the like.

[0132] For example, for capacity data, registered capital and other numerical data, Z-score standardization processing is adopted:

[0133] ;

[0134] Wherein, is the capacity data, registered capital and other numerical data in the original relevant data, is the data mean, is the standard deviation, is the numerical data after standardization processing.

[0135] In some embodiments, the evaluation device uses one-hot encoding to process non-numeric data such as credit ratings (e.g., AAA, AA, A, etc.), litigation results (win, loss), etc. For example, the credit rating "AAA" is encoded as [1, 0, 0], "AA" is encoded as [0, 1, 0], and "A" is encoded as [0, 0, 1].

[0136] In some embodiments, the evaluation device can perform data cleaning, standardization, and structuring on the collected raw data of the supplier to be analyzed, and organize the data into a model-wide table through data governance to ensure the accuracy and consistency of the data.

[0137] Step S50: inputting the data to be analyzed into the target evaluation model for evaluation to obtain the evaluation result of the supplier to be analyzed.

[0138] In some embodiments, the evaluation device can input the data to be analyzed into the target evaluation model for evaluation to score and rank multiple suppliers to be analyzed.

[0139] In some embodiments, the evaluation device can adaptively adjust the static evaluation dimensions and weight vectors in the target evaluation model based on the relevant characteristics of the supplier to be analyzed and the actual procurement requirements, thereby adaptively adjusting the target evaluation model. The adaptively adjusted target evaluation model can analyze the business data according to the new evaluation rules to generate a new evaluation result.

[0140] The embodiment is based on a plurality of sample suppliers, and original related data of the plurality of sample suppliers is collected from a plurality of source heterogeneous data sources, and the original related data is preprocessed to obtain sample data, the sample data including business change data, delivery time sequence data and historical behavior data; the sample suppliers are statically analyzed based on business demand information to generate static evaluation parameters, and an initial evaluation model is constructed based on the static evaluation parameters, the static evaluation parameters including a plurality of static evaluation dimensions and a weight vector corresponding to each static evaluation dimension, and the business demand information including evaluation conditions for bulk raw material procurement; the sample suppliers are dynamically analyzed based on the sample data, and the initial evaluation model is adjusted based on a dynamic analysis result to obtain a target evaluation model, the dynamic analysis including business change analysis, delivery time sequence analysis and historical behavior analysis; original related data of a to-be-analyzed supplier is collected, and the original related data is preprocessed to obtain to-be-analyzed data; the to-be-analyzed data is input into the target evaluation model for evaluation to obtain an evaluation result of the to-be-analyzed supplier; since the original related data of the plurality of sample suppliers is collected from the plurality of source heterogeneous data sources and preprocessed, the evaluation data dimension is ensured to be comprehensive and reliable in quality, the initial evaluation model is constructed based on a plurality of static evaluation dimensions through static analysis, so that the evaluation is ensured to be in line with actual business demand and meet diversified procurement demand in a complex market environment, the initial evaluation model is optimized and adjusted through dynamic analysis, so that the evaluation model is ensured to accurately reflect the real ability and potential risks of the supplier, and the most suitable supplier is quickly and efficiently screened out through the evaluation mode combining static analysis and dynamic analysis, and the decision-making ability for bulk raw material procurement is improved.

[0141] Reference Figure 3 , Figure 3 FIG. 2 is a flowchart of a bulk raw material supplier evaluation method according to a second embodiment of the present application.

[0142] Based on the first embodiment, in the present embodiment, the step S30 further includes:

[0143] Step S301: Based on the business change data, change information monitoring of the sample suppliers is performed to obtain a dynamic change monitoring result.

[0144] It should be noted that the change information includes business change information and market change information.

[0145] In some embodiments, the evaluation device can automatically trigger change of the evaluation dimension by monitoring supplier portraits, business data table structures and market major changes. Meanwhile, data obtained by using a real-time data acquisition interface is combined with a dynamic weight distribution algorithm to automatically adjust the weight of each dimension.

[0146] Step S302: constructing a dynamic change portrait of the sample supplier according to the dynamic change monitoring result.

[0147] In some embodiments, the evaluation device can construct a static supplier portrait based on business data, market data, behavioral data and entity relationship data of the sample supplier, adjust the static supplier portrait based on the change information of the supplier by monitoring the change information of the supplier, and generate a dynamic change portrait of the sample supplier.

[0148] In some embodiments, the evaluation device can integrate and understand various data source information, including news, financial reports, supplier information on Qichacha, and give the final result to the supplier portrait model by combining reinforcement learning and analysis prediction model.

[0149] In some embodiments, the evaluation device can use knowledge graph technology to construct the relationship between the supplier and the related entities, and form the supplier portrait; through multi-dimensional visualization presentation, the network structure and risk transmission path of the supplier are displayed.

[0150] Step S303: re-distribute the weight vector of each static evaluation dimension in the initial evaluation model based on the dynamic change portrait, and obtain a target weight vector.

[0151] In some embodiments, the evaluation device can re-distribute the weight vector based on the following formula:

[0152] ;

[0153] wherein, represents the weight vector before the change, represents the target weight vector, represents the weight adjustment coefficient, represents the change value of the key indicator in the dynamic change portrait (such as an increase of 10% in market sensitivity).

[0154] Step S304: adjusting the initial evaluation model based on the target weight vector to obtain a target evaluation model.

[0155] In specific implementation, the evaluation device replaces the weight vector of each static evaluation dimension in the initial evaluation model with the target weight vector, thereby realizing dynamic adjustment and optimization of the initial evaluation model, and obtaining the target evaluation model.

[0156] Further, in order to dynamically analyze the supplier from the time dimension and thereby improve the evaluation accuracy of the evaluation model, the above step S304 can include:

[0157] Step S3041: data cleaning on the delivery time series data to obtain an initial delivery time series;

[0158] Step S3042: performing delivery feature analysis on the initial delivery time sequence, and decomposing the initial delivery time sequence into a transaction price time sequence and a material offer time sequence based on the delivery feature analysis result;

[0159] Step S3043: performing price difference analysis based on the transaction price time sequence and the material offer time sequence, and generating a price difference sub-time sequence;

[0160] Step S3044: determining a total number of windows according to a step of the price difference sub-time sequence, and cutting the price difference sub-time sequence into a plurality of sub-sequences based on the total number of windows;

[0161] Step S3045: inputting the sub-sequences into an ARIMA model respectively to perform time sequence prediction, and obtaining a plurality of sub-prediction features;

[0162] Step S3046: performing feature fusion on the plurality of sub-prediction features to obtain a target delivery prediction result;

[0163] Step S3047: performing weight adjustment on a weight vector of a market price dimension in the target weight vector based on the target delivery prediction result to obtain a market price weight vector;

[0164] Step S3048: adjusting the initial evaluation model according to the market price weight vector and the target weight vector to obtain a target evaluation model.

[0165] It should be noted that the evaluation device can perform outlier rejection on the delivery time sequence data, for example, identifying and correcting delivery times that are outside a reasonable range (e.g., data that is more than 3 times the contractually agreed period); filling missing values in the delivery time sequence data, for example, using linear interpolation or mean filling for missing delivery times; and performing standardization processing on the time stamp.

[0166] In some embodiments, the evaluation device can separate the initial delivery time sequence into a transaction price time sequence and a material offer time sequence through a regression model, the transaction price time sequence can reflect the change of actual transaction price over time, and the material offer time sequence can reflect the adjustment of supplier offer strategy, referring to the following formula:

[0167] ;

[0168] wherein, denotes the initial delivery time sequence, and denote regression coefficients, denotes the transaction price time sequence, denotes the material offer time sequence, denotes a residual term.

[0169] The evaluation device can perform price difference analysis based on the transaction price time series and the material quotation time series, and generate a price difference time series , which can reflect the deviation degree of the actual transaction price and the quotation, and refer to the following formula:

[0170] ;

[0171] In some embodiments, the evaluation device can set a window step (such as 30 days) according to the fluctuation period (such as quarterly fluctuation) of the price difference sequence, determine the total number of windows based on the window step, and divide the price difference time series into N sub-sequences based on the total number of windows, each sub-sequence has a length of S.

[0172] It can be understood that the evaluation device can introduce a time series analysis algorithm to dynamically adjust the scoring model of the supplier according to the real-time data of the business system. For example, according to the comparison between the recent transaction price and the material quotation, the weight score of the market price dimension is adjusted in time.

[0173] In some embodiments, the evaluation device can apply an ARIMA model to predict the supplier delivery punctuality rate trend, and automatically trigger weight adjustment when the prediction deviation exceeds ± 15%, for example. Exponential smoothing method is used for dynamic weighting processing of sudden events (such as raw material price fluctuations).

[0174] Further, in order to accurately identify potential risks existing in the supplier and improve the evaluation accuracy, the above step S3048 can include:

[0175] Step S30481: performing historical behavior analysis on the sample supplier based on historical behavior data, and generating historical behavior labels of the sample supplier and time labels corresponding to each historical behavior label;

[0176] Step S30482: determining associated users associated with the sample supplier according to the historical behavior labels;

[0177] Step S30483: determining the association weight of each associated user based on the time label;

[0178] Step S30484: generating an associated entity path according to the association weight and the associated user;

[0179] Step S30485: generating an associated entity knowledge graph of the sample supplier based on the associated entity path and the associated user;

[0180] Step S30486: performing risk analysis according to the associated entity knowledge graph to obtain a risk transmission path of the sample supplier;

[0181] Step S30487: constructing a risk profile of the sample supplier based on the risk infection path;

[0182] Step S30488: adjusting the initial evaluation model according to the risk profile, the market price weight vector and the target weight vector.

[0183] It should be noted that the data source of the historical behavior label can include: quality behavior: batch qualification rate, defect type (such as “excessive impurities” “dimensional deviation”); delivery behavior: on-time delivery rate, delay days (such as “delayed for 3 days” “advanced for 2 days”); financial behavior: order amount fluctuation, account period change (such as “account period from 30 days to 60 days”).

[0184] In some embodiments, the evaluation device can cluster orders with similar behavior characteristics through cluster analysis and generate labels (such as “high-frequency delayed delivery” “low-quality batch concentration”). The evaluation device can also assign a timestamp or time range to each behavior event, thereby outputting a time label corresponding to the historical behavior label.

[0185] It should be noted that the associated user can include a direct associated user and an indirect associated user, and the direct associated user can be a customer, a partner (such as a purchaser, a logistics company) having a direct transaction relationship with the supplier, and the indirect associated user can be a user associated through a third party (such as a supplier C of a customer B of a supplier A).

[0186] In some embodiments, the association weight can be the association strength between the associated user and the supplier, for example, the evaluation device can determine the association weight based on transaction frequency, relative geographic location, industry attribute, and time interval from the last transaction to the current time.

[0187] In some embodiments, the evaluation device can determine the shortest path from the sample supplier to the associated user (such as “supplier A - customer X - logistics company Y”).

[0188] In some embodiments, the evaluation device can take the supplier, the customer, the logistics company, the financial institution, etc. as entity nodes, take the historical behavior label (such as “high-risk delivery”) as a behavior node, determine an association edge based on the relationship between entities, determine a behavior edge based on the association between the entity and the behavior label, and construct an associated entity knowledge graph based on the entity node, the behavior node, the association edge and the behavior edge.

[0189] In some embodiments, the evaluation device can use the XGBoost ensemble learning algorithm to construct a supplier risk early warning model, input the associated entity knowledge graph into the supplier risk early warning model for risk analysis, obtain the risk infection path of the sample supplier, and thereby predict risk behaviors such as supply interruption and low-quality delivery in advance.

[0190] In some embodiments, the evaluation device uses Neo4j to build a supplier relationship network, for example, to build an associated entity knowledge graph of sample suppliers containing more than 200 entity nodes (enterprises, products, qualifications, etc.) and more than 50 relationship types (holding, cooperation, litigation, etc.).

[0191] In some embodiments, the evaluation device can generate a supplier feature vector through a Graph Embedding algorithm to support similarity retrieval (such as finding alternative suppliers similar to the technical features of strategic supplier A).

[0192] The embodiment performs change information monitoring on the sample suppliers based on the business change data, obtains a dynamic change monitoring result, the change information includes business change information and market change information, constructs a dynamic change portrait of the sample suppliers according to the dynamic change monitoring result, re-distributes a weight vector of each static evaluation dimension in the initial evaluation model based on the dynamic change portrait, obtains a target weight vector, adjusts the initial evaluation model based on the target weight vector, obtains a target evaluation model, accurately configures reasonable evaluation weights for each static evaluation dimension, thereby optimizing and adjusting the initial evaluation model, ensuring that the evaluation result is more in line with the business needs of the user, and providing the user with the optimal supplier of bulk raw materials.

[0193] In addition, the embodiment of the present application also proposes a computer readable storage medium, the computer readable storage medium stores a bulk raw material supplier evaluation program, the bulk raw material supplier evaluation program is executed by the processor to realize the steps of the bulk raw material supplier evaluation method as described above.

[0194] The computer readable storage medium provided in the application may be, for example, a U disk, but is not limited to an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, system, or device, or any combination of the above. More specific examples of the computer readable storage medium may include, but are not limited to, an electrical connection with one or more conductive wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the embodiment, the computer readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system, or device. The program code contained on the computer readable storage medium can be transmitted by any suitable medium, including but not limited to an electrical wire, an optical cable, an RF (Radio Frequency), and the like, or any suitable combination of the above.

[0195] The above computer readable storage medium may be contained in the bulk raw material supplier evaluation device, or may exist separately without being assembled into the bulk raw material supplier evaluation device.

[0196] In addition, the embodiment of the application further provides a computer program product comprising a bulk raw material supplier evaluation program, which realizes the steps of the bulk raw material supplier evaluation method as described above when executed by a processor.

[0197] The computer program product embodiment of the application is basically the same as the above-mentioned bulk raw material supplier evaluation method embodiments, and will not be repeated here.

[0198] Reference Figure 4 , Figure 4 is a structural block diagram of the first embodiment of the bulk raw material supplier evaluation system of the application.

[0199] As Figure 4 indicated, the bulk raw material supplier evaluation system provided in the embodiment of the application comprises:

[0200] The data processing module 10 is configured to collect original relevant data of a plurality of sample suppliers based on a plurality of source heterogeneous data sources, and pre-process the original relevant data to obtain sample data, wherein the sample data comprises business change data, delivery time sequence data, and historical behavior data.

[0201] The static analysis module 20 is configured to perform static analysis on the sample supplier based on business requirement information, generate static evaluation parameters, and construct an initial evaluation model based on the static evaluation parameters, wherein the static evaluation parameters include a plurality of static evaluation dimensions and a weight vector corresponding to each static evaluation dimension, and the business requirement information includes evaluation conditions for bulk raw material procurement;

[0202] The dynamic analysis module 30 is configured to perform dynamic analysis on the sample supplier based on the sample data, and adjust the initial evaluation model based on the dynamic analysis result to obtain a target evaluation model, wherein the dynamic analysis includes business change analysis, delivery time sequence analysis, and historical behavior analysis;

[0203] The data processing module 10 is further configured to collect original relevant data of a supplier to be analyzed, and pre-process the original relevant data to obtain to-be-analyzed data.

[0204] The supplier evaluation module 40 is configured to input the to-be-analyzed data into the target evaluation model for evaluation to obtain an evaluation result of the to-be-analyzed supplier.

[0205] The embodiment is based on a plurality of sample suppliers, and collects original related data of the plurality of sample suppliers from a plurality of heterogeneous data sources, and pre-processes the original related data to obtain sample data, the sample data including business change data, delivery time sequence data and historical behavior data; based on business demand information, the sample suppliers are statically analyzed to generate static evaluation parameters, and based on the static evaluation parameters, an initial evaluation model is constructed, the static evaluation parameters including a plurality of static evaluation dimensions and a weight vector corresponding to each static evaluation dimension, and the business demand information including evaluation conditions of bulk raw material procurement; based on the sample data, the sample suppliers are dynamically analyzed, and based on a dynamic analysis result, the initial evaluation model is adjusted to obtain a target evaluation model, the dynamic analysis including business change analysis, delivery time sequence analysis and historical behavior analysis; original related data of a to-be-analyzed supplier is collected, and the original related data is pre-processed to obtain to-be-analyzed data; the to-be-analyzed data is input into the target evaluation model for evaluation to obtain an evaluation result of the to-be-analyzed supplier; since the embodiment collects original related data of a plurality of sample suppliers from a plurality of heterogeneous data sources and pre-processes the original related data, the comprehensiveness and quality reliability of evaluation data dimensions are ensured, the initial evaluation model is constructed based on a plurality of static evaluation dimensions through static analysis, so that the evaluation is ensured to be in line with actual business demand, to meet diversified procurement demand in a complex market environment, and the initial evaluation model is optimized and adjusted through dynamic analysis, so that the evaluation model can accurately reflect the real ability and potential risks of the supplier, through the evaluation mode combining static analysis and dynamic analysis, the most suitable supplier is quickly and efficiently screened out, and the decision-making ability of bulk raw material procurement is improved.

[0206] The bulk raw material supplier evaluation system provided in the application adopts the bulk raw material supplier evaluation method in the above embodiment, and can solve the technical problem of bulk raw material supplier evaluation. Compared with the prior art, the bulk raw material supplier evaluation system provided in the application has the same beneficial effects as the bulk raw material supplier evaluation method provided in the above embodiment, and other technical features in the bulk raw material supplier evaluation system are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0207] It should be understood that the above is only an example, and does not constitute any limitation on the technical solutions of the application. In specific applications, those skilled in the art can set it up as needed, and the application does not limit this.

[0208] It should be noted that the above-described workflow is only illustrative and does not limit the scope of protection of the application. In actual application, those skilled in the art can select part or all of them to achieve the purpose of the embodiment scheme according to actual needs, which is not limited here.

[0209] In addition, technical details not described in detail in the present embodiment can be found in the bulk raw material supplier evaluation method provided by any embodiment of the present application, which will not be described here.

[0210] It should be noted that in this document, the terms "comprising", "containing", or any other variant thereof are intended to cover non-exclusive inclusions, so that a process, method, article or system that includes a series of elements not only includes those elements, but also includes other elements not explicitly listed, or further includes elements inherent to such a process, method, article or system. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of other identical elements in the process, method, article or system that includes the element.

[0211] The above-mentioned embodiment numbers of the present application are only for description, not representing the advantages and disadvantages of the embodiments.

[0212] From the above description of the embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be realized by means of software and necessary general hardware platforms, of course, they can also be realized by hardware, but in many cases the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory / random access memory, a magnetic disk, an optical disk), and includes a number of instructions to make a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) execute the methods described in various embodiments of the present application.

[0213] The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application, and any equivalent structure or equivalent process transformation, or direct or indirect application in other related technical fields based on the content of the present application specification and drawings, are also included in the patent protection scope of the present application.

Claims

1. A method for evaluating bulk raw material suppliers, characterized in that: The bulk raw material supplier evaluation method includes: Collecting original relevant data of multiple sample suppliers based on multi-source heterogeneous data sources, and preprocessing the original relevant data to obtain sample data, wherein the sample data includes business change data, delivery time series data, and historical behavior data; Performing a static analysis on the sample suppliers based on the business demand information to generate static evaluation parameters, and constructing an initial evaluation model based on the static evaluation parameters. The static evaluation parameters include multiple static evaluation dimensions and weight vectors corresponding to each static evaluation dimension. The business demand information includes evaluation conditions for bulk raw material procurement. Performing a dynamic analysis on the sample supplier based on the sample data, and adjusting the initial evaluation model based on the dynamic analysis results to obtain a target evaluation model, wherein the dynamic analysis includes business change analysis, delivery time series analysis, and historical behavior analysis; Collecting original relevant data of the supplier to be analyzed, and preprocessing the original relevant data to obtain data to be analyzed; The data to be analyzed is input into the target evaluation model for evaluation to obtain an evaluation result of the supplier to be analyzed.

2. The method for evaluating bulk raw material suppliers according to claim 1, wherein: The static analysis of the sample suppliers based on the business demand information to generate static evaluation parameters, and constructing an initial evaluation model based on the static evaluation parameters includes: Decompose business demand information into multiple static evaluation dimensions based on hierarchical analysis strategy; Constructing a judgment matrix according to the multiple static evaluation dimensions, and calculating the maximum eigenvalue of the judgment matrix and the eigenvector corresponding to the maximum eigenvalue; Normalizing each dimensional component of the feature vector to obtain a weight vector for each static evaluation dimension; The consistency parameters of the judgment matrix are calculated based on the maximum eigenvalue, and the consistency parameters include a consistency index, a random consistency index, and a consistency ratio: ； ； in, represents the consistency index, represents the random consistency index, represents the consistency ratio, Represents the dimension of the judgment matrix; Determining whether the consistency of the judgment matrix meets a preset condition according to the consistency parameter; If the consistency of the judgment matrix meets a preset condition, it is determined that the weight vector configuration is reasonable, and an initial evaluation model is constructed based on the multiple static evaluation dimensions and the weight vector.

3. The method for evaluating bulk raw material suppliers according to claim 2, wherein: The hierarchical analysis strategy decomposes business demand information into multiple static evaluation dimensions, including: Decompose business demand information into multiple candidate evaluation dimensions based on hierarchical analysis strategy; Perform historical behavior analysis on the historical behavior data to obtain historical behavior characteristics of the sample supplier. The historical behavior characteristics include price fluctuation characteristics, delivery cycle characteristics, quality characteristics, and service characteristics. The price fluctuation characteristics are obtained based on the following formula: ； in, Represents the standard deviation of price fluctuations, which is used to measure the price fluctuation characteristics of sample suppliers. represents the number of samples, Indicates the Price data points, Represents the average of all price data points; The lead time characteristics are obtained based on the following formula: ； in, It represents the standard deviation of delivery time, which is used to measure the delivery cycle characteristics of sample suppliers. Indicates the delivery time data points, represents the average of all delivery time data points; The quality characteristics are obtained based on the following formula: ； in, It indicates the qualified batch ratio of the sample supplier's products, which is used to measure the quality characteristics of the sample supplier. Indicates the number of qualified batches that have passed quality inspection. Indicates the total number of delivered batches; The service characteristics are obtained based on the following formula: ； in, It represents the average time required to handle complaints and is used to measure the service characteristics of sample suppliers. Indicates processing Time for a complaint, Indicates the total number of complaints; A plurality of static evaluation dimensions are screened out from the candidate evaluation dimensions based on the historical behavior characteristics.

4. The method for evaluating bulk raw material suppliers according to any one of claims 1 to 3, characterized in that: The dynamically analyzing the sample suppliers based on the sample data and adjusting the initial evaluation model based on the dynamic analysis results to obtain a target evaluation model includes: Monitor change information of the sample suppliers based on the business change data to obtain dynamic change monitoring results, where the change information includes business change information and market change information; Constructing a dynamic change profile of the sample supplier based on the dynamic change monitoring results; Redistributing the weight vectors of the static evaluation dimensions in the initial evaluation model based on the dynamic change portrait to obtain a target weight vector; The initial evaluation model is adjusted based on the target weight vector to obtain a target evaluation model.

5. The method for evaluating bulk raw material suppliers according to claim 4, wherein: The adjusting the initial evaluation model based on the target weight vector to obtain a target evaluation model includes: Performing data cleaning on the delivery time series data to obtain an initial delivery time series; Performing a delivery feature analysis on the initial delivery time series, and decomposing the initial delivery time series into a transaction price time series and a material quotation time series based on the delivery feature analysis results; Perform price difference analysis based on the transaction price time series and the material quotation time series to generate a price difference time series; determining a total number of windows according to a step size of the price difference time series, and dividing the price difference time series into a plurality of subsequences based on the total number of windows; Inputting the subsequences into the ARIMA model respectively for time series prediction to obtain multiple sub-prediction features; Fusing the multiple sub-prediction features to obtain a target delivery prediction result; Adjusting the weight vector of the market price dimension in the target weight vector based on the target delivery prediction result to obtain a market price weight vector; The initial evaluation model is adjusted according to the market price weight vector and the target weight vector to obtain a target evaluation model.

6. The method for evaluating bulk raw material suppliers according to claim 5, wherein: The adjusting the initial evaluation model according to the market price weight vector and the target weight vector to obtain a target evaluation model includes: Performing a historical behavior analysis on the sample supplier based on the historical behavior data to generate historical behavior tags for the sample supplier and time tags corresponding to each historical behavior tag; Determining associated users associated with the sample supplier based on the historical behavior tags; determining an association weight for each associated user based on the time tag; generating an associated entity path according to the associated weight and the associated user; Generate a related entity knowledge graph of the sample supplier based on the related entity path and the related user; Perform risk analysis based on the associated entity knowledge graph to obtain the risk contagion path of the sample supplier; Constructing a risk profile of the sample supplier based on the risk contagion path; The initial assessment model is adjusted according to the risk profile, the market price weight vector and the target weight vector.

7. A bulk raw material supplier evaluation system, characterized by: The bulk raw material supplier evaluation system includes: A data processing module is used to collect original relevant data of multiple sample suppliers based on multi-source heterogeneous data sources, and pre-process the original relevant data to obtain sample data, wherein the sample data includes business change data, delivery time series data, and historical behavior data; a static analysis module for performing a static analysis on the sample suppliers based on business demand information, generating static evaluation parameters, and constructing an initial evaluation model based on the static evaluation parameters, wherein the static evaluation parameters include multiple static evaluation dimensions and weight vectors corresponding to each static evaluation dimension, and the business demand information includes evaluation conditions for bulk raw material procurement; A dynamic analysis module, configured to perform a dynamic analysis on the sample supplier based on the sample data, and adjust the initial evaluation model based on the dynamic analysis results to obtain a target evaluation model, wherein the dynamic analysis includes business change analysis, delivery time series analysis, and historical behavior analysis; The data processing module is further used to collect original relevant data of the supplier to be analyzed, and pre-process the original relevant data to obtain the data to be analyzed; The supplier evaluation module is used to input the data to be analyzed into the target evaluation model for evaluation, and obtain the evaluation result of the supplier to be analyzed.

8. A bulk raw material supplier evaluation device, characterized in that: The bulk raw material supplier evaluation device includes: a memory, a processor, and a bulk raw material supplier evaluation program stored in the memory and executable on the processor, wherein the bulk raw material supplier evaluation program is configured to implement the bulk raw material supplier evaluation method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a bulk raw material supplier evaluation program, and when the bulk raw material supplier evaluation program is executed by a processor, the bulk raw material supplier evaluation method according to any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that The computer program product includes a bulk raw material supplier evaluation program, which implements the steps of the bulk raw material supplier evaluation method according to any one of claims 1 to 6 when executed by a processor.

Images