Bidder qualification judgment method and device, equipment, medium and product

By cleaning and labeling engineering bidding data, training and adjusting the qualification judgment model, the problem of unreliable manual judgment is solved, the automatic and accurate judgment of bidder qualifications is realized, and the efficiency and fairness of the bidding process is improved.

CN119027137APending Publication Date: 2024-11-26SHANGHAI OUYE FINANCIAL INFORMATION SERVICE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411110964.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-14
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

During the bidding process in the existing construction industry, the bidder qualification review results rely on manual judgment, which leads to time consumption and unreliable results.

Method used

By obtaining and cleaning the relevant data of the project bidding, annotating and model training based on the characteristics of the project, dividing the training set, testing set and verification set, adjusting and evaluating the bidder's qualification judgment model, generating qualification categories and levels, and realizing automated judgment.

Benefits of technology

It improves the accuracy and professionalism of bidder qualification judgments, avoids manual subjective influence, and ensures the reliability of bidder qualification review results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119027137A_ABST
    Figure CN119027137A_ABST
Patent Text Reader

Abstract

The invention discloses a bidder qualification judgment method and device, equipment, a medium and a product, and relates to the field of bidder qualification judgment, and the method comprises the steps: obtaining original corpus data, cleaning the original corpus data, obtaining target corpus data, labeling the target corpus data, and obtaining labeled corpus data; dividing the tagged corpus data into a training set, a test set and a verification set, inputting the training set into a preset model for training to obtain a trained bidder qualification judgment model, and adjusting the bidder qualification judgment model based on the test set to obtain an adjusted bidder qualification judgment model, and optimizing the adjusted bidder qualification judgment model based on the verification set, inputting the general situation of the current project into the optimized bidder qualification judgment model, outputting a qualification category and a qualification grade required by the current project, and judging the bidder qualification according to the qualification category and the qualification grade. According to the invention, the unreliability of the bidder qualification examination result is avoided.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of bidder qualification determination, and in particular to a bidder qualification determination method, device, equipment, medium and product. Background Art

[0002] In the current bidding process in the construction industry, the tenderer must carefully compare the numerous "Enterprise Qualification Standards" issued by the Ministry of Housing and Urban-Rural Development based on the specific needs of the project to determine which bidding companies' qualifications and the qualifications of the project managers meet the requirements.

[0003] In the existing bidding process, the qualifications of bidders are often judged and reviewed manually. The manual review process is not only time-consuming, but the subjective experience and judgment of the reviewers often affect the results, resulting in unreliable bidder qualification review results. Summary of the invention

[0004] The purpose of this application is to provide a bidder qualification determination method, device, equipment, medium and product that can avoid the unreliability of the bidder qualification review results.

[0005] To achieve the above objectives, this application provides the following solutions:

[0006] In the first aspect, the present application provides a method for determining the qualifications of bidders, including:

[0007] Get the original corpus data;

[0008] Cleaning the original corpus data to obtain target corpus data, wherein the original corpus data includes project bidding related data;

[0009] Annotating the target corpus data based on the project characteristics to obtain annotated corpus data;

[0010] Dividing the annotated corpus data into a training set, a test set and a validation set;

[0011] Inputting the training set into a preset model for training to obtain a trained bidder qualification judgment model;

[0012] Adjusting the bidder qualification judgment model based on the test set to obtain the adjusted bidder qualification judgment model;

[0013] Evaluating the adjusted bidder qualification judgment model based on the verification set;

[0014] The current project profile is input into the evaluated bidder qualification judgment model, and the qualification category and qualification level required for the current project are output. The bidder's qualifications are judged based on the qualification category and qualification level, wherein the qualification category includes the general contracting sequence category and / or the professional contracting sequence category.

[0015] In a second aspect, the present application provides a bidder qualification judgment device, comprising:

[0016] The acquisition module is used to obtain the original corpus data;

[0017] A cleaning module, used for cleaning the original corpus data to obtain target corpus data, wherein the original corpus data includes project bidding related data;

[0018] A labeling module, used to label the target corpus data based on the characteristics of the engineering project to obtain labeled corpus data;

[0019] A partitioning module, used to partition the annotated corpus data into a training set, a test set and a validation set;

[0020] A training module, used for inputting the training set into a preset model for training to obtain a trained bidder qualification judgment model;

[0021] An adjustment module, used to adjust the bidder qualification judgment model based on the test set to obtain the adjusted bidder qualification judgment model;

[0022] An evaluation module, used for evaluating the adjusted bidder qualification judgment model based on the verification set;

[0023] The judgment module is used to input the current project profile into the evaluated bidder qualification judgment model, output the qualification category and qualification level required for the current project, and judge the bidder's qualifications based on the qualification category and qualification level, wherein the qualification category includes the general contracting sequence category and / or the professional contracting sequence category.

[0024] In a third aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of any of the above-mentioned methods for determining the qualifications of bidders.

[0025] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the above-mentioned methods for determining the qualifications of bidders.

[0026] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of any of the above-mentioned methods for determining the qualifications of bidders.

[0027] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0028] The present application provides a bidder qualification judgment method, device, equipment, medium and product, which obtains original corpus data and cleans the original corpus data to ensure the accuracy and consistency of the data, and obtains target corpus data, wherein the original corpus data includes engineering bidding related data; the target corpus data is annotated based on the characteristics of the engineering project to obtain annotated corpus data, the annotated corpus data is divided into a training set, a test set and a verification set, and the training set is input into a preset model for training to obtain a trained bidder qualification judgment model, and the professionalism and accuracy of the bidder qualification judgment in engineering bidding are improved by training the preset model. The bidder qualification judgment model is adjusted based on the test set to obtain an adjusted bidder qualification judgment model. The model is enabled to accurately generate results consistent with the expected output. The adjusted bidder qualification judgment model is evaluated based on the verification set. The current project profile is input into the evaluated bidder qualification judgment model, and the qualification category and qualification level required for the current project are output, and the bidder qualification is judged according to the qualification category and qualification level, wherein the qualification category includes the general contracting sequence category and / or the professional contracting sequence category. The bidder qualification judgment method avoids judging the bidder's qualifications by qualification categories and qualification levels that do not match the project, thereby ensuring the accuracy of the bidder qualification review results. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0030] Figure 1 This is an application environment diagram of a bidder qualification judgment method in an embodiment of the present application;

[0031] Figure 2 A flowchart of a method for determining the qualifications of a bidder provided in one embodiment of the present application;

[0032] Figure 3 A flowchart of a method for determining the qualifications of a bidder provided in another embodiment of the present application;

[0033] Figure 4A flowchart of a method for determining the qualifications of a bidder provided in another embodiment of the present application;

[0034] Figure 5 A flowchart of a method for determining the qualifications of a bidder provided in another embodiment of the present application;

[0035] Figure 6 A flowchart of a method for determining the qualifications of a bidder provided in another embodiment of the present application;

[0036] Figure 7 A schematic diagram of the functional modules of a bidder qualification determination device provided in one embodiment of the present application;

[0037] Figure 8 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0039] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0040] The bidder qualification determination method provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the original corpus data to be processed to the server 104. After the server 104 receives the original corpus data to be processed, the server 104 cleans the original corpus data to be processed to obtain the target corpus data, annotates the target corpus data based on the characteristics of the engineering project, obtains the annotated corpus data, divides the annotated corpus data into a training set and a test set, and inputs the training set into a preset model for training to obtain a trained bidder qualification judgment model. The server 104 can feedback the obtained bidder qualification judgment model to the terminal 102. In addition, in some embodiments, the bidder qualification judgment method can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly process the original corpus data to be processed, or the server 104 can obtain the original corpus data from the data storage system and process the original corpus data to be processed.

[0041] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0042] In an exemplary embodiment, Figure 2 As shown, a method for determining the qualifications of bidders is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or can be executed by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate the method, which includes the following steps 201 to 206. Among them:

[0043] Step 201, obtaining original corpus data.

[0044] Specifically, a corpus related to engineering bidding is constructed, which includes but is not limited to the "Construction Enterprise Qualification Standards", "Engineering Project Bidding Documents", "Engineering Project Cases and Review Reports", and "Related Laws, Regulations and Technical Specifications". The "Construction Enterprise Qualification Standards" is a set of grading standards for the qualifications of construction enterprises formulated by the Ministry of Housing and Urban-Rural Development of the People's Republic of China (hereinafter referred to as "MOHURD") for regulating the qualifications of construction enterprises. This set of standards has important guiding significance for construction enterprises engaged in new construction, expansion, and reconstruction of civil engineering, construction engineering, line pipeline equipment installation engineering, and decoration engineering within the territory of the People's Republic of China. When extracting information from the project profile in previous engineering project documents, the extracted content is not limited to the project initiator, construction scale, construction content, investment amount, project location, and project type.

[0045] Step 202, clean the original corpus data to obtain target corpus data, wherein the original corpus data includes project bidding related data.

[0046] Specifically, in the process of processing and analyzing a large amount of data, the original corpus data needs to be cleaned first to ensure the consistency and accuracy of the data. This process involves the screening, sorting and optimization of the original data, with the aim of removing errors, duplications or irrelevant information, so as to obtain purer and more valuable target corpus data. In this case, the original corpus data mainly covers various types of data related to engineering bidding, including bidding notices, bidding documents, contract terms, project progress reports, project qualification standards, etc. By cleaning these data, the accuracy of subsequent analysis and processing can be ensured, providing reliable data support for decision-making in engineering bidding.

[0047] Step 203: annotate the target corpus data based on the project characteristics to obtain annotated corpus data.

[0048] Specifically, project characteristics refer to the key attributes and features that describe a project, and through project characteristics, we can have a comprehensive understanding of the project, such as project type, project scale, project cost, project cycle, etc. By annotating the target corpus data, the accuracy of model training is improved.

[0049] Step 204: Divide the annotated corpus data into a training set, a test set, and a validation set

[0050] Specifically, in the process of processing and preparing annotated corpus data, it is first necessary to reasonably distribute these data into different subsets to facilitate subsequent model training and evaluation. Generally, the annotated corpus data needs to be divided into three main parts: training set, test set, and validation set. Training Set: used to train the model, accounting for about 60%-80%. Test Set: used to adjust model parameters and prevent overfitting, accounting for about 10%-20%. Validation Set: used to evaluate the final performance of the model, accounting for about 10%-20%.

[0051] For example, you can use the sklearn.model_selection.train_test_split function in Python to randomly split the data.

[0052] As you can understand, the training set is used to train the model. Through the training set, the model can learn the mapping relationship from input to output, so as to make predictions for new input data. The training set contains a large amount of labeled sample data, which will be used to guide the model to learn and identify different features and patterns. Through repeated iterations and optimization on the training set, the model can gradually grasp the regularity in the data and establish preliminary prediction capabilities. The test set is used to evaluate the generalization ability of the model, that is, the model's prediction ability for new data. After the model training is completed, the model is tested on the test set to verify the generalization ability of the model on unseen data. The data in the test set is the data that the model has never encountered during the training process, so the model can be adjusted through the test set. The role of the validation set is to monitor performance and adjust parameters during the model training process. The data in the validation set also comes from the original annotated corpus, but they are neither involved in the training of the model nor used for the final performance test, but serve as a reference standard for model tuning. The evaluation of the validation set can more objectively reflect the true performance of the model.

[0053] Step 205: input the training set into the preset model for training to obtain a trained bidder qualification judgment model.

[0054] Specifically, select a locally deployed pre-trained large language model with strong general language understanding capabilities, such as the Qwen2-72B open source large model. This type of model has been preliminarily trained on a wide range of general corpora and has basic language understanding and text generation capabilities. The basic model is trained with the training set, and the model learns to extract keywords from the project profile, and makes judgments based on the qualification category and qualification level to generate reasonable judgment results.

[0055] Step 206, adjusting the bidder qualification judgment model based on the test set to obtain an adjusted bidder qualification judgment model.

[0056] Specifically, after the bidder qualification judgment model is obtained through training, in order to make the model output more accurate, the model needs to be further adjusted and optimized. By fine-tuning the model parameters, it can be ensured that the model can more accurately generate the qualification category and qualification level required for the project based on the project profile, thereby providing more reliable decision-making support for the tendering party.

[0057] Step 207, evaluating the adjusted bidder qualification judgment model based on the validation set.

[0058] Specifically, the deficiencies of the model identified during the evaluation process may involve careful review and necessary adjustments to the training data set, or manual optimization and adjustment of the model's parameter settings. In addition, in order to ensure that the performance of the model in actual applications can continue to improve, the validation set needs to be updated regularly to capture the latest data trends and changes. By continuously repeating the above evaluation process, we can ensure that the performance of the model in the engineering bidding qualification judgment task is continuously improved and optimized. Through continuous optimization, we ensure that the model is always in the best working state and provide technical support for engineering bidding qualification judgment.

[0059] Step 208, input the current project profile into the evaluated bidder qualification judgment model, output the qualification category and qualification level required for the current project, and judge the bidder's qualifications based on the qualification category and qualification level, where the qualification category includes the general contracting sequence category and / or the professional contracting sequence category.

[0060] Specifically, in the construction industry, engineering contracting is usually divided into two main models: general contracting and specialized contracting. The main differences between the two models lie in the contract structure, scope of responsibility, and service content. General Contracting (GC): The general contractor is responsible for the construction management of the entire project, including but not limited to design coordination, material procurement, site management, etc. Specialty Contracting: Specialty contractors focus on specific construction services or processes, such as electrical installation, pipeline installation, air conditioning system installation, etc. The general contractor is responsible for the management and coordination of the entire project, while the specialized contractor focuses on specific professional areas. In actual operation, a large engineering project may include both general contracting and specialized contracting models.

[0061] The general contracting sequence includes 12 categories: general contracting for construction engineering, general contracting for highway engineering, general contracting for railway engineering, general contracting for port and waterway engineering, general contracting for water conservancy and hydropower engineering, general contracting for power engineering, general contracting for mining engineering, general contracting for metallurgical engineering, general contracting for petrochemical engineering, general contracting for municipal public works, general contracting for communication engineering, and general contracting for electromechanical engineering. The qualification levels under each category are generally divided into special, first, second, and third grades. Determine which level of qualification is required based on the project scale, complexity, and technical requirements.

[0062] The professional contracting sequence includes 36 categories: professional contracting for foundation engineering, professional contracting for lifting equipment installation engineering, professional contracting for ready-mixed concrete, professional contracting for electronic and intelligent engineering, professional contracting for fire protection facilities engineering, professional contracting for waterproofing, anti-corrosion and insulation engineering, professional contracting for bridge engineering, professional contracting for tunnel engineering, professional contracting for steel structure engineering, professional contracting for formwork and scaffolding, professional contracting for building decoration and decoration engineering, professional contracting for building mechanical and electrical installation engineering, professional contracting for building curtain wall engineering, professional contracting for ancient building engineering, professional contracting for urban and road lighting engineering, professional contracting for highway pavement engineering, professional contracting for highway roadbed engineering, professional contracting for highway traffic engineering, professional contracting for railway electrical engineering, and professional contracting for railway track laying. Professional contracting for beam engineering, professional contracting for railway electrification engineering, professional contracting for airport runway engineering, professional contracting for civil aviation air traffic control engineering and airport weak current system engineering, professional contracting for airport visual aids engineering, professional contracting for port and coastal engineering, professional contracting for waterway engineering, professional contracting for navigation building engineering, professional contracting for port and shipping equipment installation and water traffic control engineering, professional contracting for hydraulic metal structure fabrication and installation engineering, professional contracting for water conservancy and hydroelectric power installation engineering, professional contracting for river and lake regulation engineering, professional contracting for power transmission and transformation engineering, professional contracting for nuclear engineering, professional contracting for offshore oil engineering, professional contracting for environmental protection engineering, professional contracting for special engineering. The qualification levels under each professional contracting sequence are generally divided into level one, level two, and level three.

[0063] For example, the following project profile information is input into the bidding management system:

[0064] 1. Name of the bidding project: Sound insulation and noise reduction project in the ore dressing workshop of a mining company.

[0065] 2. Construction scale: about 12 million yuan (tax included).

[0066] 3. Planned construction period: 120 calendar days.

[0067] The large language model first extracts keywords from the project profile information:

[0068] Project initiator: a mining company.

[0069] Construction content: Ore dressing workshop, sound insulation and noise reduction project.

[0070] Construction scale: about 12 million yuan (tax included).

[0071] Investment amount: 12 million yuan.

[0072] Based on the above keywords and in comparison with the "Construction Enterprise Qualification Standards", the large language model uses its own reasoning ability to determine that this project requires the "Environmental Protection Engineering Professional Contracting" qualification based on the keyword "sound insulation and noise reduction engineering".

[0073] Determination of general contracting qualification level or professional contracting qualification level:

[0074] Through the large language model, the classification standards for the general contracting qualification or professional contracting qualification related to the project are extracted from the "Construction Enterprise Qualification Standards", and the minimum qualification level matching the project is determined based on the project construction scale or investment amount.

[0075] In this embodiment, after the large language model screens out the environmental protection engineering professional contracting qualifications, the qualification level of the environmental protection engineering professional contracting is judged according to the construction scale, including the following steps:

[0076] 1. Determine the category of environmental protection projects. Through the semantic understanding and reasoning ability of the large language model, it is determined that "sound insulation and noise reduction projects" belong to the "noise and vibration control" category of environmental protection projects.

[0077] 2. Using the prompt words, extract the scale classification standard of "noise and vibration control" from the "Construction Enterprise Qualification Standards":

[0078] Large-scale projects: investment amount ≥ 1.5 million yuan, corresponding to first-level qualification;

[0079] Medium-sized projects: investment of 500,000-1.5 million yuan, corresponding to the second-level qualification;

[0080] Small project: investment amount less than 500,000 yuan, corresponding to level 3 qualification.

[0081] In this embodiment, the project investment amount is 12 million yuan, which is a large project. Therefore, the large language model determines that the bidder for this project needs to have the first-level qualification of environmental protection engineering professional contractor.

[0082] It should be noted that if the project construction content is large and multiple qualifications are required to be completed at the same time, or multiple qualifications can meet the project construction requirements, the large language model needs to clarify the logical relationship between multiple qualifications. The project content can be covered by any one of the qualifications, or it can be covered by both the general contracting qualification and the professional contracting qualification.

[0083] It can be understood that the large language model automatically outputs the sequence categories and minimum qualification levels required for the project based on the above judgment results, integrates them according to the logical relationship, and generates the final qualification requirement format. When the project content can be covered by any one of the qualifications, use "or" to connect; when it needs to be covered by both the general contracting and professional contracting qualifications, use "and" to connect.

[0084] For example: [General contracting·Construction engineering·Level 1] (inclusive) or above or [General contracting·Metallurgical engineering·Level 1] (inclusive) or above and [Professional contracting·Building decoration and renovation engineering·Level 2] (inclusive) or above.

[0085] In another exemplary embodiment of the present application, Figure 3 As shown, the above step 202 is replaced by the following steps 2021 to 2024:

[0086] Step 2021: Remove HTML tags, special characters and spaces from the original corpus data.

[0087] Step 2022: Check the duplicates in the original corpus data and delete the duplicates to ensure the uniqueness and accuracy of the target corpus data.

[0088] Step 2023: unify the format of the original corpus data.

[0089] Step 2024: normalize the original corpus data to obtain target corpus data.

[0090] Specifically, in the process of processing the original corpus data, first of all, it is necessary to remove all HTML tags contained therein, which may interfere with subsequent data processing. In addition, it is also necessary to remove special characters in the data, which may affect the purity and readability of the data. Finally, delete all blank characters, including spaces, tabs, and line breaks, to ensure the neatness and consistency of the data. After ensuring data cleanliness, check the original corpus data to identify and delete duplicates. Duplicate data items may cause deviations in the analysis results, so these duplicates must be identified through appropriate algorithms or tools and removed from the data set. The purpose of this process is to ensure the uniqueness and accuracy of the target corpus data, thereby providing a basis for subsequent data analysis and processing. After processing the duplicates, the format of the original corpus data is unified. For example, if the data comes from different sources, they need to be converted to the same encoding format, or the order and length of the fields need to be adjusted to ensure the consistency and comparability of all data.

[0091] It is understandable that after completing the data format unification, the original corpus data needs to be normalized to convert the data into a standard range for comparison and analysis. For example, scaling the data to a range between 0 and 1, or converting the data into a form that conforms to the standard normal distribution, can eliminate the impact of different dimensions and orders of magnitude on data analysis and ensure that the data is compared on the same scale. Through normalization, the target corpus data can be obtained for further analysis and model training.

[0092] In another exemplary embodiment of the present application, Figure 4 As shown, the above step 203 is replaced by the following steps 2031 to 2033:

[0093] Step 2031, determine the label content, the label content includes project overview, qualification category and qualification level.

[0094] Step 2032: extract key information from the target corpus data, where the key information corresponds to the tag content.

[0095] Step 2033: annotate the key information according to the corresponding label content through the preset annotation tool to obtain project overview data, qualification category data and qualification level data.

[0096] Specifically, before labeling, it is necessary to first clarify and determine the label content. The label content should cover the general information of the project, including key elements such as project name, project scale, and project location. At the same time, it should also include the qualification category, specifically involving the type of qualification, scope of application, etc.; in addition, the clarification of the qualification level is also essential, which usually involves the classification of qualifications. The combination of this information constitutes the basic framework of the labeling work. After clarifying the label content, extract the key information corresponding to the label content from the target corpus data. The extraction of key information not only includes text descriptions, but may also involve information in various forms such as data, charts, and images. Finally, the key information extracted previously is annotated through the preset annotation tools. In this process, by annotating the key information, the model training is made more accurate and efficient.

[0097] In another exemplary embodiment of the present application, Figure 5 As shown, the above step 206 is replaced by the following steps 2061 to 2065:

[0098] Step 2061: input the project profile data in the test set into the bidder qualification judgment model to obtain predicted output data, the predicted output data including predicted qualification category data and predicted qualification grade data, wherein the test set includes multiple test samples, the test samples include project profile data and expected output data, and the expected output data includes expected qualification category data and expected qualification grade data;

[0099] Step 2062: Calculate a loss function based on the predicted output data and the expected output data;

[0100] Step 2063: Update the model parameters based on the loss function;

[0101] Step 2064: iteratively train the bidder qualification judgment model based on multiple test samples and updated model parameters to obtain an adjusted bidder qualification judgment model.

[0102] Specifically, for the training sample (x i ,y i ), x i Indicates the first project overview, y i Represents the expected output, which includes the expected qualification category and the expected qualification level. i Input the preset model to get the predicted output, which includes the predicted qualification category data and the predicted qualification level data. The predicted output is expressed as: Where f represents the model function and θ represents the model parameters. The loss function is calculated based on the expected output data and the predicted output data. The loss function is expressed as:

[0103] It can be understood that the loss function is calculated by cross-entropy loss or mean squared error. In the field of machine learning and deep learning, the loss function is an important tool to measure the difference between the model's predicted value and the true value. The choice of loss function has a direct impact on the training effect of the model. In practical applications, commonly used loss functions include cross-entropy loss (Cross-Entropy Loss) and mean squared error (Mean Squared Error, MSE). The cross-entropy loss function is mainly used for classification problems. It measures the difference between two probability distributions.

[0104] For example, when we have a model's predicted probability distribution p and a true probability distribution q, the cross entropy loss function can be expressed as:

[0105] \mathcal{L}{CE}=-\sum{i}q_i\log(p_i)

[0106] Among them, (q_i) is the probability distribution corresponding to the true label, and (p_i) is the probability distribution predicted by the model. In the binary classification problem, the cross entropy loss function can be simplified as:

[0107] \mathcal{L}_{CE}=-[y\log(p)+(1-y)\log(1-p)]

[0108] Where (y) is the true label (0 or 1) and (p) is the probability of the model predicting the positive class.

[0109] The mean square error is a commonly used loss function in regression problems. It calculates the average of the squares of the differences between the predicted values ​​and the true values. For a set of samples, the mean square error can be expressed as:

[0110] \mathcal{L}{MSE}=\frac{1}{n}\sum{i=1}^{n}(y_i-\hat{y}_i)^2

[0111] Among them, (y_i) is the true value of the (i)th sample, (\hat{y}_i) is the value predicted by the model, and (n) is the total number of samples.

[0112] During model training, we usually use optimization algorithms, such as gradient descent, to minimize the loss function and adjust the model parameters so that the model's predicted value is as close to the true value as possible. In this way, the model can learn the rules in the data and make accurate predictions on new data.

[0113] Update the model parameters based on the loss function to minimize the loss function The update rule is expressed as: in, represents the updated model parameters, η represents the learning rate, Represents the loss function The gradient of the parameter θ. The preset model is iteratively trained based on multiple training samples and updated model parameters to obtain an adjusted bidder qualification judgment model.

[0114] Specifically, the gradient descent method is used to update the model parameters θ so that the loss function Minimize. The goal of the gradient descent method is to minimize the loss function by gradually adjusting the parameter θ in an iterative manner so that the loss function Reach the minimum value. For the current parameter θ, calculate the loss function The gradient of θ is calculated, and the parameter θ is updated according to the gradient and learning rate until the maximum number of iterations is reached or the gradient is close to 0, then the updating is stopped. The model at this time is the adjusted bidder qualification judgment model.

[0115] In another exemplary embodiment of the present application, Figure 6 As shown, the above step 207 is replaced by the following steps 2071 to 2074:

[0116] Step 2071: input the project profile data in the verification set into the adjusted bidder qualification judgment model to obtain the model judgment result, wherein the model judgment result includes the verification qualification category data and the verification qualification level data;

[0117] Step 2072: compare the model judgment result with the expert-given result. If the model judgment result is completely consistent with the expert-given result, it is recorded as the model judgment is correct.

[0118] Step 2073, counting the total number of correct model judgments and the total number of model judgments, and calculating the accuracy rate of the bidder qualification judgment based on the total number of correct model judgments and the total number of model judgments;

[0119] Step 2074: Evaluate the adjusted bidder qualification judgment model based on the bidder qualification judgment accuracy rate.

[0120] Specifically, the project profile data in the validation set is first input into the adjusted bidder qualification judgment model, which is designed to automatically evaluate the qualifications of bidders. After processing and analysis by the model, we obtain a judgment result containing verification qualification category data and verification qualification level data. This result is the judgment of the bidder's qualifications given by the model based on the input project profile data, combined with its internal logic and rules. In order to evaluate the accuracy of the model's judgment results, we compare the judgment results given by the model with the given results given by the expert team. The expert team is composed of experienced industry experts who independently evaluate the qualifications of the bidders based on their professional knowledge and experience. If the judgment results of the model are completely consistent with the results given by the expert team in all aspects, then this judgment result is marked as the correct model judgment.

[0121] Optionally, in order to evaluate the overall performance of the model, choose to perform statistical analysis on the correctness of the model's judgment. Record the total number of correct judgments made by the model and compare it with the total number of judgments made by the model. Through these two data, we can calculate the accuracy of a bidder's qualification judgment, qualification judgment accuracy = number of correct judgments / total number of judgments. This accuracy reflects the proportion of correct judgments made by the model in all judgments. It is an important indicator for measuring model performance and can intuitively reflect the reliability of the model in practical applications. Based on the calculated accuracy of the bidder's qualification judgment, the model can be further optimized.

[0122] It is understandable that if the accuracy has reached a satisfactory level, the model can be considered for application in the actual bidding qualification assessment work. If there is still room for improvement in the accuracy, the model will analyze the cases where the model made incorrect judgments, find out the shortcomings of the model, and adjust and improve the model's algorithm, parameters or training data. Through such an optimization process, the model's judgment accuracy is continuously improved, thereby improving the efficiency and fairness of the entire bidding process.

[0123] In another exemplary embodiment of the present application, the bidder qualification determination method further includes:

[0124] Integrate the bidder qualification judgment model into the bidding management system through the API interface to realize automated bidder qualification judgment.

[0125] Specifically, in order to effectively integrate the bidding qualification judgment model into the bidding management system, it is necessary to ensure that the model has the ability to interact through the API interface, including the following steps:

[0126] S1. Encapsulate the bidding qualification judgment model so that it can run as an independent service and deploy the model on a server that can receive external requests and return processing results. In this process, the model needs to be properly adjusted to ensure that it can process input data from the API interface and output results in a standardized format.

[0127] S2. Design an API interface, which serves as a bridge between the bidding qualification judgment model and the bidding management system. The design of the API interface should follow the RESTful principle to ensure the flexibility and scalability of the interface. When implementing the API interface, it is necessary to consider the security of data transmission, ensure that all data transmission is encrypted, and only authorized users can access the interface.

[0128] S3. After the API interface is designed, the integration work begins. Code is added to the bidding management system so that the bidding management system can call the API interface. In order to ensure the accuracy and reliability of the bidding qualification judgment model, it needs to be tested and maintained regularly. The model is updated to adapt to new regulatory requirements or optimized according to actual usage.

[0129] S4. In order to improve the user experience, some auxiliary functions are added to the bidding management system, such as real-time display of bidding qualification judgment results, or providing a user-friendly interface so that users can easily view and manage relevant information on bidding qualifications. The addition of these functions makes the entire bidding process more efficient and transparent.

[0130] Through the above steps, the bidding qualification judgment model can be integrated into the bidding management system through the API interface, thereby improving the automation and intelligence level of the entire bidding process and improving the efficiency and fairness of the bidding.

[0131] Based on the same inventive concept, the embodiment of the present application also provides a device for implementing the above-mentioned bidder qualification judgment device. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in one or more embodiments of the bidder qualification judgment device provided below can refer to the limitations of the bidder qualification judgment method above, and will not be repeated here.

[0132] In an exemplary embodiment, Figure 7 As shown, a bidder qualification judgment device is provided, comprising:

[0133] An acquisition module 710 is used to acquire original corpus data;

[0134] A cleaning module 720 is used to clean the original corpus data to obtain target corpus data, wherein the original corpus data includes project bidding related data;

[0135] The annotation module 730 is used to annotate the target corpus data based on the characteristics of the engineering project to obtain annotated corpus data;

[0136] A partitioning module 740 is used to partition the annotated corpus data into a training set, a test set, and a validation set;

[0137] A training module 750 is used to input the training set into a preset model for training to obtain a trained bidder qualification judgment model;

[0138] An adjustment module 760, configured to adjust the bidder qualification judgment model based on the test set to obtain an adjusted bidder qualification judgment model;

[0139] An evaluation module 770, for evaluating the adjusted bidder qualification judgment model based on the validation set;

[0140] The judgment module 780 is used to input the current project profile into the evaluated bidder qualification judgment model, output the qualification category and qualification level required for the current project, and judge the bidder's qualifications based on the qualification category and qualification level, where the qualification category includes the general contracting sequence category and / or the professional contracting sequence category.

[0141] As an optional implementation, the cleaning module 720 is specifically used for:

[0142] Remove HTML tags, special characters and spaces from the original corpus data;

[0143] Check the duplicates in the original corpus data and delete them to ensure the uniqueness and accuracy of the target corpus data;

[0144] Unify the format of the original corpus data;

[0145] The original corpus data is normalized to obtain the target corpus data.

[0146] As an optional implementation, the marking module 730 is specifically used to:

[0147] Determine the label content, which includes project overview, qualification category and qualification level;

[0148] Extract key information from the target corpus data, where the key information corresponds to the label content;

[0149] Through the preset annotation tools, key information is annotated according to the corresponding label content to obtain project overview data, qualification category data and qualification level data.

[0150] As an optional implementation, the adjustment module 760 is specifically configured to:

[0151] Input the project profile data in the test set into the bidder qualification judgment model to obtain predicted output data, the predicted output data including predicted qualification category data and predicted qualification level data, wherein the test set includes multiple test samples, the test samples include project profile data and expected output data, and the expected output data includes expected qualification category data and expected qualification level data;

[0152] Calculate the loss function based on the predicted output data and the expected output data;

[0153] Update the model parameters based on the loss function;

[0154] The bidder qualification judgment model is iteratively trained based on multiple test samples and updated model parameters to obtain an adjusted bidder qualification judgment model.

[0155] As an optional implementation, the evaluation module 770 is specifically configured to:

[0156] Input the project profile data in the verification set into the adjusted bidder qualification judgment model to obtain the model judgment result, wherein the model judgment result includes verification qualification category data and verification qualification level data;

[0157] Compare the model judgment result with the expert's given result. If the model judgment result is completely consistent with the expert's given result, it is recorded as the model judgment is correct.

[0158] Count the total number of correct model judgments and the total number of model judgments, and calculate the accuracy of the bidder qualification judgment based on the total number of correct model judgments and the total number of model judgments;

[0159] The adjusted bidder qualification judgment model is evaluated based on the accuracy of bidder qualification judgment.

[0160] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 8 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store original corpus data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for determining the qualifications of a bidder is implemented.

[0161] Those skilled in the art will understand that Figure 8 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0162] In an exemplary embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0163] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0164] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0165] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0166] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0167] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0168] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0169] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for determining the qualifications of a bidder, characterized in that: The method for determining the qualifications of bidders includes: Get the original corpus data; Cleaning the original corpus data to obtain target corpus data, wherein the original corpus data includes project bidding related data; Annotating the target corpus data based on the project characteristics to obtain annotated corpus data; Dividing the annotated corpus data into a training set, a test set and a validation set; Inputting the training set into a preset model for training to obtain a trained bidder qualification judgment model; Adjusting the bidder qualification judgment model based on the test set to obtain the adjusted bidder qualification judgment model; Evaluating the adjusted bidder qualification judgment model based on the verification set; The current project profile is input into the evaluated bidder qualification judgment model, and the qualification category and qualification level required for the current project are output. The bidder's qualifications are judged based on the qualification category and the qualification level, wherein the qualification category includes the general contracting sequence category and / or the professional contracting sequence category.

2. The method for determining the qualifications of bidders according to claim 1, characterized in that: The cleaning of the original corpus data to obtain the target corpus data specifically includes: Removing HTML tags, special characters and blanks from the original corpus data; Checking duplicate items in the original corpus data and deleting the duplicate items; Unifying the format of the original corpus data; The original corpus data is normalized to obtain the target corpus data.

3. The method for determining the qualifications of bidders according to claim 1, characterized in that: The labeling of the target corpus data based on the project characteristics to obtain the labeled corpus data specifically includes: Determine the label content, which includes project overview, qualification category and qualification level; Extracting key information from the target corpus data, wherein the key information corresponds to the tag content; The key information is annotated according to the corresponding label content through a preset annotation tool to obtain project overview data, qualification category data and qualification level data.

4. The method for determining the qualifications of bidders according to claim 3, characterized in that: The step of adjusting the bidder qualification judgment model based on the test set to obtain the adjusted bidder qualification judgment model specifically includes: Inputting the project profile data in the test set into the bidder qualification judgment model to obtain predicted output data, the predicted output data including predicted qualification category data and predicted qualification grade data, wherein the test set includes a plurality of test samples, the test samples include the project profile data and expected output data, and the expected output data includes expected qualification category data and expected qualification grade data; Calculating a loss function based on the predicted output data and the expected output data; Updating model parameters based on the loss function; The bidder qualification judgment model is iteratively trained based on the multiple test samples and the updated model parameters to obtain the adjusted bidder qualification judgment model.

5. The method for determining the qualifications of bidders according to claim 3, characterized in that: The evaluating the adjusted bidder qualification judgment model based on the verification set specifically includes: Inputting the project profile data in the verification set into the adjusted bidder qualification judgment model to obtain a model judgment result, wherein the model judgment result includes verification qualification category data and verification qualification level data; Compare the model judgment result with the expert-given result. If the model judgment result is completely consistent with the expert-given result, it is recorded as the model judgment is correct. Count the total number of correct model judgments and the total number of model judgments, and calculate the accuracy rate of the bidder qualification judgment based on the total number of correct model judgments and the total number of model judgments; The adjusted bidder qualification judgment model is evaluated according to the bidder qualification judgment accuracy rate.

6. The method for determining the qualifications of bidders according to claim 1, characterized in that: The method for determining the qualifications of bidders also includes: The bidder qualification judgment model is integrated into the bidding management system through the API interface to realize automated bidder qualification judgment.

7. A bidder qualification judgment device, characterized in that: The bidder qualification judgment device comprises: The acquisition module is used to obtain the original corpus data; A cleaning module, used for cleaning the original corpus data to obtain target corpus data, wherein the original corpus data includes project bidding related data; A labeling module, used to label the target corpus data based on the characteristics of the engineering project to obtain labeled corpus data; A partitioning module, used to partition the annotated corpus data into a training set, a test set and a validation set; A training module, used for inputting the training set into a preset model for training to obtain a trained bidder qualification judgment model; An adjustment module, used to adjust the bidder qualification judgment model based on the test set to obtain the adjusted bidder qualification judgment model; An evaluation module, used for evaluating the adjusted bidder qualification judgment model based on the verification set; The judgment module is used to input the current project profile into the evaluated bidder qualification judgment model, output the qualification category and qualification level required for the current project, and judge the bidder's qualifications based on the qualification category and qualification level, wherein the qualification category includes the general contracting sequence category and / or the professional contracting sequence category.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the bidder qualification determination method described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the bidder qualification determination method described in any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the bidder qualification determination method described in any one of claims 1 to 6 are implemented.