Recommendation method and device for bid invitation project suppliers, electronic equipment and storage medium

Abstract

The invention provides a bid invitation project supplier recommendation method and device, electronic equipment and a computer readable storage medium, and relates to the technical field of supplier recommendation. The recommendation method comprises the steps that bidding portraits of a plurality of suppliers are constructed, and the bidding portraits are used for representing the multi-dimensional matching relation between the suppliers and bidding items; according to the bidding portrait, evaluating a first bidding score and a first bidding response probability of each supplier; and determining a target supplier of the bid invitation project according to the first bid invitation score and the first bid invitation response probability, and outputting a recommendation result. The problems of low intelligent degree, insufficient data fusion and limited prediction and recommendation ability in related technologies are at least solved. The method is suitable for bid invitation recommendation and management scenes.

Description

Technical Field

[0001] This invention relates to the field of supplier recommendation technology, and in particular to a method, apparatus, electronic device, and computer-readable storage medium for recommending suppliers for bidding projects. Background Technology

[0002] With the popularization of digital procurement and electronic bidding, more and more enterprises and institutions are using information technology to conduct bidding and procurement activities. Traditional bidding preparation and supplier response prediction mainly rely on manual experience and static data analysis, which has pain points such as difficulty in predicting the number of responding suppliers, lack of scientific basis in the design of evaluation methods, and inaccurate supplier recommendations, which directly affect the smooth progress of bidding projects and the fairness of the winning bid results.

[0003] In recent years, extensive research and product development have been conducted in the fields of intelligent bidding assistance, supplier management, and response prediction. Existing technologies mainly include: 1. Segmenting and extracting structured information from bidding documents using keyword matching and regular expressions based on rule bases and templates. 2. Integrating publicly available data on business registration, qualifications, and performance to establish a basic supplier information database and perform simple tagging management. 3. Using historical response rates and industry average parameters, and performing rough predictions through linear regression or weighted average methods. 4. Screening suppliers based on qualification criteria and performing simple ranking based on historical winning rates and scoring methods. 5. Utilizing BERT (Bidirectional Encoder Representations from Transformers) and GPT (Generative Pretrained Transformer) models for entity recognition and relation extraction from bidding texts, improving structured processing capabilities.

[0004] However, existing technologies have limited ability to analyze complex, non-standardized bidding documents, making it difficult to automatically identify implicit constraints and complex scoring methods; supplier profiles have few dimensions and lack in-depth modeling; AI (Artificial Intelligence) / NLP (Natural Language Processing) technologies are mostly single-point applications, failing to achieve end-to-end intelligent decision support. In summary, existing technologies still suffer from low levels of intelligence, insufficient data fusion, and limited predictive and recommendation capabilities. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to address the above-mentioned shortcomings of the prior art by providing a method, apparatus, electronic device and computer-readable storage medium for recommending suppliers for bidding projects. This method can realize intelligent and effective recommendation of suppliers for bidding projects.

[0006] In a first aspect, the present invention provides a method for recommending suppliers for bidding projects, comprising: constructing bidding profiles of several suppliers, wherein the bidding profiles are used to characterize the multi-dimensional matching relationship between suppliers and bidding projects; evaluating the first bidding score and the first bidding response probability of each supplier according to the bidding profiles; determining the target supplier for the bidding project according to the first bidding score and the first bidding response probability, and outputting the recommendation result.

[0007] Preferably, constructing bidding profiles for several suppliers specifically includes: obtaining bidding characteristics of several suppliers and the bidding documents of the bidding project, wherein the bidding characteristics include qualification characteristics, technical characteristics, experience characteristics, and historical response rate; inputting the bidding documents into a large language model and outputting the first bidding rule characteristics of the bidding project, wherein the first bidding rule characteristics include the first qualification requirement, the first technical requirement, the first experience requirement, the first scoring weight, and the first attractiveness index; evaluating the first asset matching degree, the first technical matching degree, and the first experience matching degree between several suppliers and the bidding project, wherein the first asset matching degree, the first technical matching degree, and the first experience matching degree are used to characterize the similarity between qualification characteristics and the first qualification requirement, technical characteristics and the first technical requirement, and experience characteristics and the first experience requirement, respectively; constructing bidding profiles for several suppliers, wherein the bidding profile includes entities and first entity attributes, the entities include the bidding project entity, the supplier entity, and the matching identifier entity, the first entity attributes include historical response rate, first scoring weight, first attractiveness index, first asset matching degree, first technical matching degree, and first experience matching degree, and the matching identifier entity is used to characterize the matching relationship between the bidding project and the supplier.

[0008] Preferably, based on the bidding profile, the first bidding score and first bidding response probability of each supplier are evaluated respectively, specifically including: obtaining the scores corresponding to the first asset matching degree, the first technology matching degree, and the first experience matching degree according to the preset relationship between matching degree and score; calculating the first bidding score of each supplier according to the scores corresponding to the first asset matching degree, the first technology matching degree, and the first experience matching degree and the first scoring weight; inputting the historical response rate, the first attractiveness index, the first asset matching degree, the first technology matching degree, and the first experience matching degree into the AI ​​model to predict the first bidding response probability of each supplier respectively, wherein the AI ​​model includes at least one of the following: a logistic regression model and a neural network model.

[0009] Preferably, determining the target suppliers for the bidding project based on the first bidding score and the first bidding response probability specifically includes: performing sensitivity analysis on the bidding profiles of several suppliers to obtain sensitivity analysis profiles of several suppliers; evaluating the second bidding score and the second bidding response probability of each supplier based on the sensitivity analysis profiles; determining the first set of potential suppliers and the second set of potential suppliers for the bidding project, wherein the first potential supplier refers to the supplier whose first bidding score and first bidding response probability are greater than their corresponding thresholds, and the second potential supplier refers to the supplier whose second bidding score and second bidding response probability are greater than their corresponding thresholds; and identifying the suppliers common to both the first set of potential suppliers and the second set of potential suppliers as the target suppliers for the bidding project.

[0010] Preferably, sensitivity analysis is performed on the bidding profiles of several suppliers to obtain several supplier sensitivity analysis profiles. Specifically, this includes: adjusting one or more of the first qualification requirements, first technical requirements, first experience requirements, first scoring weights, and first attractiveness indicators to obtain the second bidding rule features of the bidding project. The second bidding rule features include second qualification requirements, second technical requirements, second experience requirements, second scoring weights, and second attractiveness indicators. The second asset matching degree, second technical matching degree, and second experience matching degree between several suppliers and the bidding project are evaluated respectively. The second asset matching degree, second technical matching degree, and second experience matching degree are used to characterize the similarity between qualification features and second qualification requirements, technical features and second technical requirements, and experience features and second experience requirements, respectively. Sensitivity analysis profiles of several suppliers are constructed. The sensitivity analysis profiles include entities and second entity attributes. The entities include the bidding project entity, the supplier entity, and the matching identifier entity. The second entity attributes include historical response rate, second scoring weights, second attractiveness indicators, second asset matching degree, second technical matching degree, and second experience matching degree.

[0011] Preferably, based on the sensitivity analysis profile, the second bidding score and second bidding response probability of each supplier are evaluated respectively, specifically including: obtaining the scores corresponding to the second asset matching degree, the second technology matching degree, and the second experience matching degree according to the preset relationship between matching degree and score; calculating the second bidding score of each supplier according to the scores corresponding to the second asset matching degree, the second technology matching degree, and the second experience matching degree and the second scoring weight; and inputting the historical response rate, the second attractiveness index, the second asset matching degree, the second technology matching degree, and the second experience matching degree into the AI ​​model to predict the second bidding response probability of each supplier respectively.

[0012] Preferably, after determining the first set of potential suppliers and the second set of potential suppliers for the bidding project, and before determining the suppliers common to the first set of potential suppliers and the second set of potential suppliers as the target suppliers for the bidding project, the method for recommending suppliers for the bidding project further includes: determining whether the number of suppliers in the first set of potential suppliers and the second set of potential suppliers is greater than a preset value; in response to the number of suppliers in the first set of potential suppliers or the second set of potential suppliers being less than the preset value, outputting a response insufficiency warning, and optimizing the bidding documents based on the bidding profile.

[0013] Secondly, the present invention also provides a supplier recommendation device for bidding projects, including a construction module, an evaluation module, and a determination module. The construction module is used to construct bidding profiles of several suppliers, wherein the bidding profiles are used to characterize the multi-dimensional matching relationship between suppliers and bidding projects. The evaluation module is connected to the construction module and is used to evaluate the first bidding score and the first bidding response probability of each supplier according to the bidding profiles. The determination module is connected to the evaluation module and is used to determine the target supplier of the bidding project according to the first bidding score and the first bidding response probability, and output the recommendation result.

[0014] Thirdly, the present invention also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to implement the method for recommending suppliers for bidding projects provided in the first aspect above.

[0015] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the method for recommending suppliers for bidding projects provided in the first aspect above.

[0016] This invention provides a method, apparatus, electronic device, and computer-readable storage medium for recommending suppliers for bidding projects. It dynamically captures multi-dimensional matching relationships between suppliers and bidding projects through bid profiles, and then evaluates the supplier's first bidding score and first bidding response probability based on these relationships to determine target suppliers. This makes the recommended results more aligned with the actual needs of the bidding project, reducing the risk of bid failure. Therefore, this invention enables intelligent and effective supplier recommendation for bidding projects. Attached Figure Description

[0017] Figure 1 This is a flowchart of a method for recommending suppliers for a bidding project according to Embodiment 1 of the present invention;

[0018] Figure 2 A flowchart for constructing bidding profiles of several suppliers in Embodiment 1 of the invention;

[0019] Figure 3This is a flowchart of the process of inputting the tender documents into a large language model and outputting the first tender rule feature of the tender project in Embodiment 1 of the invention;

[0020] Figure 4 This is an example image of a bidder's portrait in Embodiment 1 of the invention;

[0021] Figure 5 This is a flowchart of a sensitivity analysis in Embodiment 1 of the invention;

[0022] Figure 6 This is a flowchart of another sensitivity analysis in Embodiment 1 of the invention;

[0023] Figure 7 This is a flowchart illustrating the risk assessment and optimization of tender documents in Embodiment 1 of the invention;

[0024] Figure 8 A flowchart of a method for recommending suppliers for a bidding project, according to Embodiment 2 of the invention;

[0025] Figure 9 This is a schematic diagram of a supplier recommendation device for a bidding project according to Embodiment 3 of the present invention. Detailed Implementation

[0026] To enable those skilled in the art to better understand the technical solution of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0027] It is understood that the specific embodiments and accompanying drawings described herein are merely for explaining the invention and are not intended to limit the invention.

[0028] It is understood that, without conflict, the various embodiments and features in the embodiments of the present invention can be combined with each other.

[0029] It is understood that, for ease of description, only the parts related to the present invention are shown in the accompanying drawings, while the parts unrelated to the present invention are not shown in the drawings.

[0030] It is understood that each unit or module involved in the embodiments of the present invention may correspond to only one entity structure, or may be composed of multiple entity structures, or multiple units or modules may be integrated into one entity structure.

[0031] It is understood that, without conflict, the functions and steps marked in the flowcharts and block diagrams of this invention may occur in a different order than that marked in the accompanying drawings.

[0032] It is understood that the flowcharts and block diagrams of this invention illustrate the possible architecture, functions, and operations of systems, apparatuses, devices, and methods according to various embodiments of this invention. Each block in the flowchart or block diagram may represent a unit, module, program segment, or code, containing executable instructions for implementing the specified function. Furthermore, each block or combination of blocks in the block diagram and flowchart can be implemented using a hardware-based system to achieve the specified function, or using a combination of hardware and computer instructions.

[0033] It is understood that the units and modules involved in the embodiments of the present invention can be implemented by software or by hardware. For example, the units and modules can be located in a processor.

[0034] Example 1:

[0035] like Figure 1 As shown, this embodiment provides a method for recommending suppliers for bidding projects. The method for recommending suppliers for bidding projects includes:

[0036] S101, construct bidding profiles for several suppliers, whereby the bidding profiles are used to characterize the multi-dimensional matching relationship between suppliers and bidding projects.

[0037] Specifically, S101: Construct bidding profiles for several suppliers, including steps S1011-S1014:

[0038] S1011, obtain the bidding characteristics of several suppliers and the bidding documents of the bidding project. The bidding characteristics include qualification characteristics, technical characteristics, experience characteristics and historical response rate.

[0039] In this embodiment, bidding characteristics refer to a comprehensive set of indicators demonstrated by the supplier during the bidding process, reflecting its overall competitiveness. Qualification characteristics refer to the supplier's legal status and basic capability documentation, typically mandatory threshold conditions. Qualification characteristics include, but are not limited to: business license, industry license, audit report, bank credit certificate, tax records, social security payment certificate, no major litigation records, high-tech enterprise status, and patent certificates. Technical characteristics refer to the supplier's capabilities in providing technical solutions or products, reflecting technological advancement, feasibility, and adaptability. Experience characteristics refer to the supplier's industry-related experience accumulated from past projects, reflecting execution capabilities. Historical response rate reflects the supplier's participation enthusiasm and response efficiency in past bidding processes.

[0040] like Figure 2 As shown, several supplier bidding profiles are constructed, specifically including: obtaining basic information about several suppliers from business registration databases and industry qualification databases (i.e.,... Figure 2This involves integrating basic supplier information, historical bidding behavior, credit and performance data. Basic information includes, but is not limited to: business registration information (such as company name, registered capital, establishment date, legal representative), industry qualifications and certifications (such as ISO certification, industry licenses), and the scope of main products and services. From this basic information, historical bidding behavior, credit and performance data, we analyze qualification characteristics, technical characteristics, experience characteristics, and historical response rates (i.e.,...). Figure 2 (Modeling of competency and qualification characteristics in the context of [the text]).

[0041] It should be noted that this embodiment can also analyze the historical success rate from basic information, historical bidding behavior, credit and performance data. Analyzing the historical response rate and historical success rate from basic information, historical bidding behavior, credit and performance data specifically includes: calculating the number of historical bids N from basic information, historical bidding behavior, credit and performance data. bid Number of successful bids in history (N) win The total number of times N was invited to bid invite According to the formula: and Calculate the historical response rate R resp and historical success rate R win This embodiment can also employ visualization technology to statistically analyze the distribution of participating project types.

[0042] Taking a supplier who has been invited to bid 20 times in the past three years, actually bid 15 times, and won 5 bids as an example, the supplier's...

[0043] S1012, input the tender documents into the large language model, and output the first tender rule features of the tender project. The first tender rule features include the first qualification requirement, the first technical requirement, the first experience requirement, the first scoring weight, and the first attractiveness index.

[0044] In this embodiment, the first bidding rule feature refers to the set of core rules extracted from the bidding documents, used to quantitatively evaluate the supplier's entry standards and bidding priority. The first qualification requirement refers to the legal or industry-mandated entry conditions that suppliers must meet to participate in the bidding. The first technical requirement refers to the specific indicators of the project's technical capabilities required of the supplier, typically divided into hard and flexible indicators. The first experience requirement refers to the supplier's historical performance in similar projects. The first scoring weight refers to the allocation ratio of scores for each dimension of qualification, technology, and experience during the evaluation, determining the direction of bidding strategy optimization. The first attractiveness indicator is used to quantify the attractiveness of the bidding project to the supplier, such as budget and project scale, influencing the supplier's bidding priority decision.

[0045] like Figure 3As shown, the tender document is input into the large language model, and the first tender rule feature of the tender project is output. Specifically, this includes: preprocessing the tender document to obtain structured text blocks. This preprocessing includes, but is not limited to: file format parsing (PDF / Word / TXT to text), removing irrelevant formatting characters, images, header and footer noise, and retaining chapter titles, numbering, and clause structure information. The structured text blocks are then segmented into chapters to obtain the structured text blocks corresponding to each chapter. Let the text be T, the set of chapter titles be H, and the segmentation function be f. The segmentation result is: {S} i}=f(T,H), where S i Let S represent the structured text block corresponding to the i-th chapter. Based on an industry terminology database, contextual semantics and rule templates, an entity recognition model, and a relation extraction model, entity recognition, relation extraction, and rule normalization and standardization are performed on the structured text blocks corresponding to each chapter to obtain the first bidding rule feature of the bidding project. Let S be the structured text block, E be the set of entity categories in the industry terminology database, and M be the entity recognition model. NER Then the entity set is: ε = M NER (S,E), where ε represents all identified entities, and entity pairs (e) are defined. i ,e j ), the set of relation categories R in the industry terminology database, and the relation extraction model M. RE Then the set of relations is: This refers to all entity pairs and their relationships.

[0046] Taking the tender document containing sections such as "Chapter 3 Qualification Conditions: 1. Bidders must be independent legal entities with registered capital of no less than 10 million yuan; 2. Must possess ISO9001 quality management system certification. Chapter 4 Technical Requirements: 1. Equipment must meet GB / T12345 standards; 2. Response time no greater than 2 seconds. Chapter 5 Evaluation Method: 1. Technical score accounts for 40%, commercial score accounts for 30%, and price score accounts for 30%; 2. Bidders with ISO14001 certification receive 2 extra points; 3. Bids without an after-sales service plan will be rejected," as an example, regular expressions and NLP models are used to identify chapter titles (such as "Chapter 3 Qualification Conditions"). By segmenting the text according to chapter number and title, the structured text block S1 corresponding to "Chapter 3 Qualification Conditions" and the corresponding structured text block S1 corresponding to "Chapter 4 Technical Requirements" can be obtained. Structured text block S2 and structured text block S3 corresponding to "Chapter 5 Evaluation Methods" are used. Using a large language model and industry terminology database, entities such as "Registered Capital", "ISO9001 Certification", "GB / T12345 Standard", "Technical Score", "Business Score", "Price Score", "ISO14001 Certification", and "After-sales Service Plan" can be extracted from S1, S2, and S3. The "constraint" relationship between "Registered Capital" and "≥10 million yuan" is extracted; the "weight" relationship between "Technical Score" and "40%" is extracted; the "bonus" relationship between "ISO14001 Certification" and "add 2 points" is extracted; and the "veto" relationship between "After-sales Service Plan" and "veto bid" is extracted. After rule normalization and standardization, the first bidding rule features are obtained as shown in Table 1. This embodiment supports automatic parsing of various tender document formats; by performing multi-level entity and relationship recognition, it identifies implicit constraints and scoring details in complex text, significantly improving parsing accuracy; through rule normalization and standardization, the first tender rule features are made easier for subsequent AI simulation and optimization, and greatly improve the automated processing capability of tender data.

[0047] Table 1 Characteristics of the First Bidding Rule

[0048]

[0049] It should be noted that the data format for the first bidding rule features includes, but is not limited to, JSON, tables, or databases, and the entity recognition model and relation extraction model includes, but is not limited to, BERT and GPT.

[0050] S1013, evaluate the first asset matching degree, first technology matching degree, and first experience matching degree between several suppliers and the bidding project respectively. The first asset matching degree, first technology matching degree, and first experience matching degree are used to characterize the similarity between qualification characteristics and first qualification requirements, technical characteristics and first technical requirements, and experience characteristics and first experience requirements, respectively.

[0051] In this embodiment, as Figure 2 As shown, constructing bidding profiles for several suppliers also includes: according to the formula and Evaluate the first asset match, first technology match, and first experience match between several suppliers and the bidding project (i.e., Figure 2 (M) matching degree modeling, where M qual M represents the degree of qualification matching. tech M represents the degree of technical matching. exp N represents the empirical matching degree. qual_match N tech_match N exp_match N represents the number of qualification features identical to the first qualification requirement, the number of qualification features identical to the first technical requirement, and the number of qualification features identical to the first experience requirement, respectively. qual_req N tech_req N exp_req These represent the quantities of the first qualification requirement, the first technical requirement, and the first experience requirement, respectively.

[0052] It should be noted that this embodiment can also evaluate the matching score between the k-th technical feature and the first technical requirement. k_tech And set the weight w of the matching score. k_tech ; and then according to the formula Assess the primary technical match between several suppliers and the bidding project. Taking a bidding project that includes three primary qualification requirements and three primary technical requirements, where suppliers possess two identical qualification characteristics to the primary qualification requirements, and the matching scores for the three technical characteristics with the primary technical requirements are 1, 0.5, and 0 respectively, with weights of 0.5, 0.3, and 0.2 respectively.

[0053] This embodiment can also be based on the formula. Assess the first asset match between the i-th supplier and the bidding project. First technical compatibility First experience matching degree in, This represents the j-th qualification characteristic of the i-th supplier. This represents the j-th primary qualification requirement. This represents the c-th technical feature of the i-th supplier. This represents the c-th first technical requirement. This represents the d-th empirical feature of the i-th supplier. Let d represent the d-th first experience requirement, M represent the total number of first qualification requirements, C represent the total number of first technical requirements, and D represent the total number of first experience requirements. Indicates the indicator function, condition (i.e.) If true, the value is 1; otherwise, it is 0. According to the formula... Evaluate the overall matching degree between the i-th supplier and the bidding project.

[0054] S1014, construct several supplier bidding profiles, where each bidding profile includes an entity and a first entity attribute. The entity includes the bidding project entity, the supplier entity, and the matching identifier entity. The first entity attribute includes historical response rate, first score weight, first attractiveness index, first asset matching degree, first technology matching degree, and first experience matching degree. The matching identifier entity is used to characterize the matching relationship between the bidding project and the supplier.

[0055] In this embodiment, the matching identifier entities include asset matching identifier entities, technology matching identifier entities, and experience matching identifier entities. The historical response rate is the entity attribute of the supplier entity. The first scoring weight and the first attractiveness index are the entity attributes of the bidding project entity. The first asset matching degree, the first technology matching degree, and the first experience matching degree are the entity attributes of the asset matching identifier entity, the technology matching identifier entity, and the experience matching identifier entity, respectively. The bid profile is as follows: Figure 4 As shown, the supplier entity is... Figure 4 Supplier A in the tender project is... Figure 4 Item 'a' in the list matches the identified entity. Figure 4 In the matching of qualifications, skills, and experience, the first weighted score is... Figure 4 The rating weighting in the first attractiveness index is... Figure 4 The attractiveness of the bidding process. This embodiment integrates multi-source heterogeneous data from suppliers, combines historical behavior with current project characteristics to model the multi-dimensional connection between suppliers and bidding projects, thereby constructing a high-dimensional bidding profile of suppliers to support intelligent recommendations and adapt to different bidding needs.

[0056] It should be noted that, in this embodiment, the first qualification requirement, the first technical requirement, the first experience requirement, qualification characteristics, technical characteristics, and experience characteristics can also be associated with the bid profile.

[0057] S102, based on the bidding profile, evaluate the first bid score and first bid response probability of each supplier.

[0058] In this embodiment, the first bidding score refers to the supplier's overall evaluation in the current bidding project. It is calculated based on pre-set quantitative standards (such as technical solutions, prices, qualifications, and historical performance) and reflects its competitiveness ranking. The higher the first bidding score, the greater the likelihood of winning the bid. The first bidding response probability refers to the probability that the supplier will make a substantive response to the current bidding project (such as submitting a complete tender document), reflecting its willingness to participate and resource matching degree. The higher the first bidding response probability, the more likely the supplier is to participate in the competition.

[0059] Specifically, S102: Based on the bid profile, evaluate the first bid score and first bid response probability of each supplier, including steps S1021-S1023:

[0060] S1021, Based on the preset relationship between matching degree and score, obtain the scores corresponding to the first asset matching degree, the first technology matching degree, and the first experience matching degree respectively.

[0061] In this embodiment, the preset relationship between matching degree and score is shown in Table 2. Based on the preset relationship and formula between matching degree and score... You can obtain scores corresponding to the first asset matching degree, the first technical matching degree, and the first experience matching degree. f represents the score corresponding to the b-th match degree in the match degree set of the i-th supplier. b (·) represents the scoring function corresponding to the b-th matching degree. (i) Let represent the matching degree set of the i-th supplier. The matching degree set includes the first asset matching degree, the first technology matching degree, and the first experience matching degree. The scoring function includes, but is not limited to: linear, interval mapping, and machine learning model.

[0062] Table 2 Preset Relationship Between Matching Degree and Score

[0063] Matching range Score illustrate 90％～100％ 5 Almost perfect match 80％～89％ 4 Highly matched 70％～79％ 3 Medium match 60％～69％ 2 Partial Matching <60％ 1 Mismatch or requires manual review

[0064] S1022, Calculate the first bidding score for each supplier based on the scores corresponding to the first asset matching degree, the first technology matching degree, and the first experience matching degree, and the first scoring weight.

[0065] In this embodiment, as shown in Table 1, the first scoring weights are: "Technical score weight 40%; Business score weight 30%; Price score weight 30%; ISO14001 certification +2 points". These first scoring weights include the weights corresponding to the first asset matching degree, the first technical matching degree, and the first experience matching degree, as well as additional bonus points and additional deduction points, according to the formula... Calculate the first bid score for the i-th supplier. (i)B represents the total number of matching scores in the matching score set, w b This represents the weight corresponding to the b-th matching degree. V represents the a-th extra point for the i-th supplier. l (i) Let A represent the l-th additional deduction for the i-th supplier, A represent the total number of additional points added, and p represent the total number of additional deductions.

[0066] S1023, input the historical response rate, first attractiveness index, first asset matching degree, first technology matching degree, and first experience matching degree into the AI ​​model to predict the first tender response probability of each supplier. The AI ​​model includes at least one of the following: logistic regression model and neural network model.

[0067] In this embodiment, the historical response rate, the first attractiveness index, the first asset matching degree, the first technology matching degree, and the first experience matching degree are input into formula (1) to predict the first tender response probability of each supplier:

[0068]

[0069] in, Let represent the probability of the first bid response from the i-th supplier, σ(·) represent the Sigmoid function, and α0, α1, α2, α3, α4, and α5 represent, respectively. This represents the historical response rate of the i-th supplier. This represents the first qualification matching degree of the i-th supplier. This represents the first technical match degree of the i-th supplier. A represents the first empirical match degree of the i-th supplier. bid This indicates the primary attraction indicator.

[0070] Assuming α0=-1, α1=2, α2=1, α3=1, α4=1, α5=0.5, supplier R resp =0.75,M qual =0.67,M tech =0.65,M exp =0.8,A bid For example, =0.7 Therefore, the probability of a supplier responding to the first tender is 95%.

[0071] It should be noted that, in this embodiment, the probability of the first tender response can also be associated with the tender profile.

[0072] S103: Determine the target supplier for the bidding project based on the first bidding score and the first bidding response probability, and output the recommendation results.

[0073] Specifically, the target supplier for the bidding project is determined based on the first bidding score and the first bidding response probability, including steps S1031-S1034:

[0074] S1031, Sensitivity analysis is performed on the bidding profiles of several suppliers to obtain sensitivity analysis profiles of several suppliers.

[0075] Specifically, S1031: Sensitivity analysis is performed on the bidding profiles of several suppliers to obtain sensitivity analysis profiles of several suppliers, including: adjusting one or more of the first qualification requirements, first technical requirements, first experience requirements, first scoring weight, and first attractiveness index to obtain the second bidding rule features of the bidding project, wherein the second bidding rule features include second qualification requirements, second technical requirements, second experience requirements, second scoring weight, and second attractiveness index; the second asset matching degree, second technical matching degree, and second experience matching degree between several suppliers and the bidding project are evaluated respectively, wherein the second asset matching degree, second technical matching degree, and second experience matching degree are used to characterize the similarity between qualification features and second qualification requirements, technical features and second technical requirements, and experience features and second experience requirements, respectively; and a sensitivity analysis profile of several suppliers is constructed, wherein the sensitivity analysis profile includes entities and second entity attributes, the entities include the bidding project entity, the supplier entity, and the matching identifier entity, and the second entity attributes include historical response rate, second scoring weight, second attractiveness index, second asset matching degree, second technical matching degree, and second experience matching degree.

[0076] S1032, Based on the sensitivity analysis profile, evaluate the second tender score and second tender response probability of each supplier.

[0077] Specifically, S1032: Based on the sensitivity analysis profile, evaluate the second bidding score and second bidding response probability of each supplier, including: obtaining the scores corresponding to the second asset matching degree, second technology matching degree, and second experience matching degree according to the preset relationship between matching degree and score; calculating the second bidding score of each supplier according to the scores corresponding to the second asset matching degree, second technology matching degree, and second experience matching degree and the second scoring weight; and inputting the historical response rate, second attractiveness index, second asset matching degree, second technology matching degree, and second experience matching degree into the AI ​​model to predict the second bidding response probability of each supplier.

[0078] In this embodiment, as Figure 5 or Figure 6As shown, after evaluating the first bidding score and first bidding response probability of each supplier, this embodiment further includes: adjusting one or more of the first qualification requirements, first technical requirements, first experience requirements, first scoring weights, and first attractiveness indicators, such as relaxing the registered capital requirement or lowering the technical threshold, to obtain the second bidding rule feature of the bidding project. adj Similarly, by constructing the bid profile and evaluating the first bid score and the probability of response to the first bid, a sensitivity analysis profile is constructed to evaluate the second bid score and the probability of response to the second bid. This embodiment uses sensitivity analysis to monitor in real time the changes in the bid profile, bid score, and bid response probability after the dynamic adjustment of the first bid rule features, thereby assisting in the subsequent determination of target suppliers and improving the stability and accuracy of supplier recommendations.

[0079] S1033, determine the first set of potential suppliers and the second set of potential suppliers for the bidding project, wherein the first potential supplier refers to the supplier whose first bidding score and first bidding response probability are greater than their corresponding thresholds, and the second potential supplier refers to the supplier whose second bidding score and second bidding response probability are greater than their corresponding thresholds.

[0080] In this embodiment, the determination Among them, T resp The corresponding threshold representing the probability of responding to the first tender; if The i-th supplier is identified as the first potential supplier for the bidding project.

[0081] It should be noted that this embodiment can also be determined by judgment. Identify the first set of potential suppliers for the bidding project.

[0082] T resp The value range is [0.5, 0.7], and it can be dynamically adjusted based on historical data, project importance, and industry characteristics. resp The higher the value, the more stringent the supplier response, and the fewer the quantities; T resp The lower the threshold, the more supplier responses there will be, but the better the supplier's fit with the bidding project may be. The corresponding thresholds for the first and second bid response probabilities can be the same, denoted by T. resp For example, if the probability of responding to the first tender is 0.7, then only suppliers whose probability of responding to the first tender is greater than 0.7 are counted as first potential suppliers and second potential suppliers.

[0083] S1034, the suppliers common to the first set of potential suppliers and the second set of potential suppliers are identified as the target suppliers for the bidding project.

[0084] In this embodiment, after determining the target supplier for the bidding project, the embodiment further includes: according to the formula Generate and output the recommendation result Rank, where argsort(·) represents the descending order function, and Score... (h) tip Let H represent the first and second bid scores of the h-th target supplier, and let H represent the total number of target suppliers.

[0085] Optionally, after determining the first and second potential supplier sets for the bidding project in S1033, and before determining the suppliers common to the first and second potential supplier sets as target suppliers for the bidding project in S1034, the method for recommending suppliers for the bidding project further includes:

[0086] S1035, determine whether the number of suppliers in the first potential supplier set and the second potential supplier set is greater than a preset value.

[0087] S1036, in response to the fact that the number of suppliers in the first potential supplier set or the second potential supplier set is less than a preset value, output an insufficient response warning and optimize the bidding documents based on the bidding profile.

[0088] In this embodiment, to avoid a severe mismatch between bidding requirements and bidding suppliers, the embodiment also includes: risk assessment and optimization of the bidding documents. For example... Figure 7 As shown, risk assessment and optimization of the tender documents are carried out, specifically including: based on the first tender score and the first tender response probability (i.e. Figure 7 The response probability in the data and its corresponding threshold (i.e.) Figure 7 The response probability threshold in the data is used to obtain the first set of potential suppliers (i.e., Figure 7 (Batch response probability judgment in the formula); Count the number of suppliers in the first set of potential suppliers, N. resp Let N represent the number of suppliers in the first set of potential suppliers, and N represent the total number of suppliers. T resp For example, N = 0.7 resp =2 (the 1st and 3rd suppliers); according to the formula Conduct risk assessment and output insufficient response warnings, response probability distributions, response lists, and N. resp And optimize the tender documents, where N min This refers to a preset value, reflecting the minimum number of responding suppliers (e.g., 3). Risk_Alert indicates an insufficient response warning. This embodiment automatically prompts the bidding party to adjust the bidding documents through a risk warning mechanism, reducing the risk of bidding failure and increasing the success rate of bidding.

[0089] It should be noted that this embodiment can also be based on the formula Count the number of suppliers in the first set of potential suppliers, where N pot This represents the number of suppliers in the first set of potential suppliers.

[0090] Before outputting the response probability distribution, this embodiment also includes: according to the formula Calculate the concentration σ of the response probability distribution score ,in, Based on the concentration of the response probability distribution, an early warning is issued indicating an overly concentrated distribution and a shortage of potential successful bidders. The presentation of the response probability distribution includes, but is not limited to, histograms and quantiles.

[0091] The supplier recommendation method for bidding projects provided in this embodiment dynamically captures the multi-dimensional matching relationship between suppliers and bidding projects through bidding profiles. Then, it evaluates the supplier's first bidding score and first bidding response probability based on the multi-dimensional matching relationship to determine the target supplier. This makes the recommendation results more in line with the actual needs of the bidding project, reduces the risk of bidding failure, and realizes intelligent and effective recommendation of suppliers for bidding projects.

[0092] Example 2:

[0093] like Figure 8 As shown in this embodiment, a method for recommending suppliers for bidding projects is provided. The method for recommending suppliers for bidding projects includes:

[0094] S201, obtain the bidding characteristics of several suppliers and the bidding documents of the bidding project. The bidding characteristics include qualification characteristics, technical characteristics, experience characteristics and historical response rate.

[0095] S202: Input the tender documents into the large language model and output the first tender rule features of the tender project. The first tender rule features include the first qualification requirement, the first technical requirement, the first experience requirement, the first scoring weight, and the first attractiveness index.

[0096] In this embodiment, the large language model is... Figure 8 The rule analysis in the text, the first bidding rule feature is... Figure 8 Structured rules in the context of [the system / mechanism].

[0097] S203 assesses the first asset matching degree, first technology matching degree, and first experience matching degree between several suppliers and the bidding project, respectively. The first asset matching degree, first technology matching degree, and first experience matching degree are used to characterize the similarity between qualification characteristics and first qualification requirements, technical characteristics and first technical requirements, and experience characteristics and first experience requirements, respectively. Construct bidding profiles for several suppliers. The bidding profile includes entities and first entity attributes. Entities include bidding project entities, supplier entities, and matching identifier entities. First entity attributes include historical response rate, first scoring weight, first attractiveness index, first asset matching degree, first technology matching degree, and first experience matching degree. The matching identifier entity is used to characterize the matching relationship between the bidding project and the supplier.

[0098] In this embodiment, the bidder's portrait is... Figure 8 Supplier profiles in the database.

[0099] S204. Based on the bid profile, evaluate the first bid score and first bid response probability of each supplier.

[0100] S205, Perform sensitivity analysis on the bidding profiles of several suppliers to obtain sensitivity analysis profiles of several suppliers; Based on the sensitivity analysis profiles, evaluate the second bidding score and second bidding response probability of each supplier; Determine the first potential supplier set and the second potential supplier set for the bidding project, wherein the first potential supplier refers to the supplier whose first bidding score and first bidding response probability are greater than their corresponding thresholds, and the second potential supplier refers to the supplier whose second bidding score and second bidding response probability are greater than their corresponding thresholds.

[0101] In this embodiment, sensitivity analysis is... Figure 8 The simulation analysis in the text refers to the first set of potential suppliers and the second set of potential suppliers. Figure 8 Potential suppliers in the industry.

[0102] S206, determine whether the number of suppliers in the first potential supplier set and the second potential supplier set is greater than a preset value; in response to the number of suppliers in the first potential supplier set or the second potential supplier set being less than the preset value, output an insufficient response warning, and optimize the bidding documents based on the bidding profile.

[0103] In this embodiment, the output response insufficiency warning is... Figure 8 Risk warnings in the middle.

[0104] S207 identifies the suppliers common to both the first and second potential supplier sets as target suppliers for the bidding project and outputs the recommendation results.

[0105] In this embodiment, the target supplier is... Figure 8The list of potential suppliers in the database, and the recommended results are as follows: Figure 8 The recommended list.

[0106] The supplier recommendation method for bidding projects provided in this embodiment dynamically captures the multi-dimensional matching relationship between suppliers and bidding projects through bidding profiles. Then, it evaluates the supplier's first bidding score and first bidding response probability based on the multi-dimensional matching relationship to determine the target supplier. This makes the recommendation results more in line with the actual needs of the bidding project, reduces the risk of bidding failure, and realizes intelligent and effective recommendation of suppliers for bidding projects.

[0107] Example 3:

[0108] like Figure 9 As shown, this embodiment also provides a supplier recommendation device for bidding projects, including a construction module 31, an evaluation module 32, and a determination module 33. The construction module 31 is used to construct bidding profiles of several suppliers, wherein the bidding profiles are used to characterize the multi-dimensional matching relationship between suppliers and bidding projects. The evaluation module 32 is connected to the construction module 31 and is used to evaluate the first bidding score and the first bidding response probability of each supplier according to the bidding profiles. The determination module 33 is connected to the evaluation module 32 and is used to determine the target supplier of the bidding project according to the first bidding score and the first bidding response probability, and output the recommendation result.

[0109] Specifically, the construction module 31 includes: a first acquisition unit 311, an extraction unit 312, a first evaluation unit 313, and a construction unit 314. The first acquisition unit 311 is used to acquire the bidding characteristics of several suppliers and the bidding documents of the bidding project. The bidding characteristics include qualification characteristics, technical characteristics, experience characteristics, and historical response rate. The extraction unit 312 is used to input the bidding documents into the large language model and output the first bidding rule characteristics of the bidding project. The first bidding rule characteristics include the first qualification requirements, the first technical requirements, the first experience requirements, the first scoring weight, and the first attractiveness index. The first evaluation unit 313 is used to evaluate the first asset matching between several suppliers and the bidding project. The matching degree includes first asset matching degree, first technology matching degree, and first experience matching degree. The first asset matching degree, first technology matching degree, and first experience matching degree are used to characterize the similarity between qualification characteristics and first qualification requirements, technical characteristics and first technical requirements, and experience characteristics and first experience requirements, respectively. The construction unit 314 is used to construct bidding profiles of several suppliers. The bidding profile includes entities and first entity attributes. The entities include bidding project entities, supplier entities, and matching identifier entities. The first entity attributes include historical response rate, first scoring weight, first attractiveness index, first asset matching degree, first technology matching degree, and first experience matching degree. The matching identifier entity is used to characterize the matching relationship between the bidding project and the supplier.

[0110] Specifically, the evaluation module 32 includes: a second acquisition unit 321, a calculation unit 322, and a prediction unit 323. The second acquisition unit 321 is used to acquire the scores corresponding to the first asset matching degree, the first technology matching degree, and the first experience matching degree according to the preset relationship between the matching degree and the score. The calculation unit 322 is used to calculate the first bidding score of each supplier according to the scores corresponding to the first asset matching degree, the first technology matching degree, and the first experience matching degree and the first scoring weight. The prediction unit 323 is used to input the historical response rate, the first attractiveness index, the first asset matching degree, the first technology matching degree, and the first experience matching degree into the AI ​​model to predict the first bidding response probability of each supplier. The AI ​​model includes at least one of the following: a logistic regression model and a neural network model.

[0111] Specifically, the determination module 33 includes: an analysis unit 331, a second evaluation unit 332, a first determination unit 333, and a second determination unit 334. The analysis unit 331 is used to perform sensitivity analysis on the bidding profiles of several suppliers to obtain sensitivity analysis profiles of several suppliers. The second evaluation unit 332 is used to evaluate the second bidding score and the second bidding response probability of each supplier based on the sensitivity analysis profiles. The first determination unit 333 is used to determine the first potential supplier set and the second potential supplier set for the bidding project. The first potential supplier refers to the supplier whose first bidding score and the first bidding response probability are greater than their corresponding thresholds, and the second potential supplier refers to the supplier whose second bidding score and the second bidding response probability are greater than their corresponding thresholds. The second determination unit 334 is used to determine the suppliers common to the first potential supplier set and the second potential supplier set as the target suppliers for the bidding project.

[0112] Specifically, the analysis unit 331 includes: an adjustment subunit, an evaluation subunit, and a construction subunit. The adjustment subunit is used to adjust one or more of the first qualification requirements, first technical requirements, first experience requirements, first scoring weights, and first attractiveness indicators to obtain the second bidding rule features of the bidding project. The second bidding rule features include second qualification requirements, second technical requirements, second experience requirements, second scoring weights, and second attractiveness indicators. The evaluation subunit evaluates the second asset matching degree, second technical matching degree, and second experience matching degree between several suppliers and the bidding project. The second asset matching degree, second technical matching degree, and second experience matching degree are used to characterize the similarity between qualification features and second qualification requirements, technical features and second technical requirements, and experience features and second experience requirements, respectively. The construction subunit is used to construct sensitivity analysis profiles of several suppliers. The sensitivity analysis profiles include entities and second entity attributes. The entities include the bidding project entity, the supplier entity, and the matching identifier entity. The second entity attributes include historical response rate, second scoring weights, second attractiveness indicators, second asset matching degree, second technical matching degree, and second experience matching degree.

[0113] Specifically, the second evaluation unit 332 includes: an acquisition subunit, a calculation subunit, and a prediction subunit. The acquisition subunit is used to acquire the scores corresponding to the second asset matching degree, the second technology matching degree, and the second experience matching degree according to the preset relationship between the matching degree and the score. The calculation subunit is used to calculate the second bidding score of each supplier according to the scores corresponding to the second asset matching degree, the second technology matching degree, the second experience matching degree, and the second scoring weight. The prediction subunit is used to input the historical response rate, the second attractiveness index, the second asset matching degree, the second technology matching degree, and the second experience matching degree into the AI ​​model to predict the second bidding response probability of each supplier.

[0114] Optionally, the determining module 33 further includes: a judging unit 335 and an early warning unit 336. The judging unit 335 is used to judge whether the number of suppliers in the first potential supplier set and the second potential supplier set is greater than a preset value. The early warning unit 336 is used to output an insufficient response warning in response to the number of suppliers in the first potential supplier set or the second potential supplier set being less than the preset value, and to optimize the bidding documents according to the bidding profile.

[0115] Understandably, the above-described device for recommending suppliers for bidding projects executes the supplier recommendation method for bidding projects corresponding to Embodiment 1 provided above. Therefore, the beneficial effects it can achieve can be referred to the beneficial effects of the solution corresponding to the supplier recommendation method for bidding projects in Embodiment 1 above, which will not be repeated here.

[0116] Example 4:

[0117] This embodiment also provides an electronic device, including a memory and a processor. The memory stores a computer program, and the processor is configured to run the computer program to implement the method for recommending suppliers for bidding projects in Embodiment 1 or Embodiment 2 described above.

[0118] Example 5:

[0119] This embodiment also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, it implements the method for recommending suppliers for bidding projects in Embodiment 1 or Embodiment 2 above.

[0120] It is understood that the above embodiments are merely exemplary implementations used to illustrate the principles of the present invention, and the present invention is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A method of recommending suppliers for a tender project, characterized by The method comprises the following steps: constructing a bid image of a plurality of suppliers, wherein the bid image is used to represent a multi-dimensional matching relationship between the suppliers and the bidding project; evaluating a first bidding score and a first bidding response probability of each supplier according to the bid image; determining a target supplier of the bidding project according to the first bidding score and the first bidding response probability, and outputting a recommendation result.

2. The method of claim 1, wherein, The construction of the bid image of the plurality of suppliers specifically comprises: obtaining bid characteristics of a plurality of suppliers and a bidding document of a bidding project, wherein the bid characteristics include qualification characteristics, technical characteristics, experience characteristics and historical response rates; inputting the bidding document into a large language model to output first bidding rule characteristics of the bidding project, wherein the first bidding rule characteristics include first qualification requirements, first technical requirements, first experience requirements, first scoring weights and first attractiveness indicators; evaluating a first asset matching degree, a first technical matching degree and a first experience matching degree between the plurality of suppliers and the bidding project, respectively, wherein the first asset matching degree, the first technical matching degree and the first experience matching degree are used to represent the similarity between the qualification characteristics and the first qualification requirements, the technical characteristics and the first technical requirements, and the experience characteristics and the first experience requirements, respectively; constructing a bid image of a plurality of suppliers, wherein the bid image includes entities, first entity attributes, the entities include a bidding project entity, a supplier entity and a matching identification entity, the first entity attributes include historical response rates, first scoring weights, first attractiveness indicators, first asset matching degrees, first technical matching degrees and first experience matching degrees, and the matching identification entity is used to represent the matching relationship between the bidding project and the suppliers.

3. The method of claim 2, wherein, The evaluation of the first bidding score and the first bidding response probability of each supplier according to the bid image specifically comprises: obtaining scores corresponding to the first asset matching degree, the first technical matching degree and the first experience matching degree according to a preset relationship between the matching degrees and the scores; calculating the first bidding score of each supplier according to the scores corresponding to the first asset matching degree, the first technical matching degree and the first experience matching degree and the first scoring weights; inputting the historical response rates, the first attractiveness indicators, the first asset matching degrees, the first technical matching degrees and the first experience matching degrees into an AI model to predict the first bidding response probability of each supplier, wherein the AI model includes at least one of the following: a logistic regression model and a neural network model.

4. The method of claim 2, wherein the bid project supplier recommendation is based on a bid project supplier's past performance on bid projects. The determination of the target supplier of the bidding project according to the first bidding score and the first bidding response probability specifically comprises: performing sensitivity analysis on the bid images of the plurality of suppliers to obtain sensitivity analysis images of the plurality of suppliers; evaluating a second bidding score and a second bidding response probability of each supplier according to the sensitivity analysis images; determining a first potential supplier set and a second potential supplier set of the bidding project, wherein the first potential supplier refers to a supplier whose first bidding score and first bidding response probability are greater than the corresponding threshold values, and the second potential supplier refers to a supplier whose second bidding score and second bidding response probability are greater than the corresponding threshold values; The common suppliers in the first potential supplier set and the second potential supplier set are determined as target suppliers of the bidding project.

5. The method of claim 4, wherein, The sensitivity analysis portraits of the suppliers are obtained by performing sensitivity analysis on the bidding portraits of the suppliers, and the sensitivity analysis portraits of the suppliers are obtained. The second bidding rule characteristics of the bidding project are obtained by adjusting one or more of the first qualification requirement, the first technical requirement, the first experience requirement, the first score weight and the first attraction index, wherein the second bidding rule characteristics include a second qualification requirement, a second technical requirement, a second experience requirement, a second score weight and a second attraction index. The second asset matching degree, the second technical matching degree and the second experience matching degree between the suppliers and the bidding project are respectively evaluated, wherein the second asset matching degree, the second technical matching degree and the second experience matching degree are respectively used to represent the similarity between the qualification characteristics and the second qualification requirement, the technical characteristics and the second technical requirement, and the experience characteristics and the second experience requirement. The sensitivity analysis portraits of the suppliers are constructed, wherein the sensitivity analysis portraits include entities, second entity attributes, the entities include a bidding project entity, a supplier entity and a matching identifier entity, and the second entity attributes include a historical response rate, a second score weight, a second attraction index, a second asset matching degree, a second technical matching degree and a second experience matching degree.

6. The method of claim 4, wherein the bid project supplier recommendation is based on a bid project supplier's past performance on bid projects. The second bidding scores and the second bidding response probabilities of the suppliers are respectively evaluated according to the sensitivity analysis portraits, and the evaluation specifically includes: According to a preset relationship between the matching degree and the score, the scores corresponding to the second asset matching degree, the second technical matching degree and the second experience matching degree are respectively obtained. The second bidding scores of the suppliers are respectively calculated according to the scores corresponding to the second asset matching degree, the second technical matching degree and the second experience matching degree and the second score weight. The second bidding response probabilities of the suppliers are respectively predicted by inputting the historical response rate, the second attraction index, the second asset matching degree, the second technical matching degree and the second experience matching degree into an AI model.

7. The method of claim 4, wherein the bid project supplier recommendation is based on a bid project supplier's past performance on bid projects. After determining the first potential supplier set and the second potential supplier set of the bidding project, and before determining the common suppliers in the first potential supplier set and the second potential supplier set as the target suppliers of the bidding project, the method further includes: Judging whether the number of suppliers in the first potential supplier set and the second potential supplier set is greater than a preset value; In response to the number of suppliers in the first potential supplier set or the second potential supplier set being less than the preset value, outputting a response shortage warning, and optimizing the bidding document according to the bidding portrait.

8. A device for recommending a supplier of a bid project, characterized by The method includes a construction module, an evaluation module and a determination module, The construction module is configured to construct bidding portraits of the suppliers, wherein the bidding portraits are used to represent multi-dimensional matching relationships between the suppliers and the bidding project, The evaluation module is connected with the construction module and is configured to evaluate the first bidding scores and the first bidding response probabilities of the suppliers according to the bidding portraits, A determining module, connected with the evaluating module, is configured to determine a target supplier of the bidding project according to the first bidding score and the first bidding response probability, and output a recommendation result.

9. An electronic device, comprising: A computer program product, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the computer program to implement the method for recommending a supplier of a bidding project according to any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program product is executed by the processor to implement the method for recommending a supplier of a bidding project according to any one of claims 1 to 7.

Images