Intelligent bidder equity analysis system and method based on cloud collaboration

Abstract

The invention belongs to the field of bidding engineering technology research and development, and discloses an intelligent bidder equity analysis system and method based on cloud collaboration. Bidder unit industry and commerce information data are collected through a cloud end, and bidding risk features and bidding document similar features are obtained through acquisition and processing; and constructing a bidder stock equity intelligent risk analysis model by using the unit industry and commerce information data, the bidding risk features, the bidding document similar features and corresponding bidder risk assessment coefficients, and generating a new bidder risk assessment coefficient by using the model to perform risk analysis. A data generation model based on unit industry and commerce information data correction is used to carry out training data set expansion of the model so as to obtain a more accurate bidder risk assessment coefficient; according to the bidder risk analysis method and system, data accurate generation is carried out on related characteristics used for analyzing bidder risks by using the multi-fusion model, and a more generalized risk analysis model is obtained based on the generated data, so that the intelligence and accuracy of bidder risk analysis are improved.

Description

Technical Field

[0001] This invention belongs to the field of bidding data processing engineering technology, and in particular relates to an intelligent analysis system and method for bidder equity based on cloud collaboration. Background Technology

[0002] Early methods of bidder risk analysis relied primarily on expert review of individual bidders' qualification documents and bids. This approach was not only inefficient but also lacked cross-sectional comparison, making it difficult to detect collusion or bid-rigging, such as different bids being prepared by the same person, or identical IP addresses or machine codes. Background research has led to systems that collect documents from all bidding companies, using multi-document comparative analysis and feature extraction to evaluate each company within a competitive environment, thereby more accurately predicting winning risks and identifying document anomalies.

[0003] Therefore, existing technologies have the following problems: how to collect data on the impact of bidders on bidding risks through existing data storage clients, and what type of data to use to construct an intelligent risk analysis model for bidder equity suitable for risk analysis in bidding scenarios; in addition, since the amount and set of relevant data generated by similar bids are relatively small, how to use the model to generate more sets of relevant data on bidder risk assessment coefficients for risk analysis; how to use a data generation model based on corrected business information data to expand the training dataset of the model to obtain more accurate bidder risk assessment coefficients; and how to use a multi-fusion model to accurately generate data on relevant features used to analyze bidder risks, and use this data to obtain a more generalized risk analysis model, thereby improving the intelligence and accuracy of bidder risk analysis. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes a cloud-based collaborative intelligent analysis system and method for bidder equity.

[0005] In a first aspect of the present invention, a cloud-based collaborative intelligent analysis method for bidder equity is provided, the method comprising:

[0006] D1. Collect the business information data of the first unit of the bidder through the cloud, and obtain the risk characteristics of the first bidder and the similarity characteristics of the first bidding document through collection and processing, and obtain the risk assessment coefficient of the first bidder set by the bidding review experts according to the classification.

[0007] D2. The risk assessment coefficient of the bidder is expanded by using a generative adversarial network model based on the first unit's business information data to obtain a second bidder risk assessment coefficient. The second bidder risk assessment coefficient and the first bidder risk assessment coefficient together constitute the third bidder risk assessment coefficient for training the intelligent risk analysis model of bidder equity. The third bidder risk assessment coefficient has corresponding third unit business information data, third bid risk characteristics and third bidding document similarity characteristics.

[0008] D3. Receive and process the business registration information data of the third entity, the risk characteristics of the third bidder, and the similarity characteristics of the third tender document to obtain the bidder's bid risk analysis characteristics, and construct an intelligent risk analysis model for the bidder's equity using the bidder's bid risk analysis characteristics and the risk assessment coefficient of the third bidder;

[0009] D4. Using the intelligent risk analysis model for bidder equity, conduct bidding risk analysis for different bidder entities to obtain the risk assessment coefficient of the fourth bidder, and use the risk assessment coefficient of the fourth bidder to conduct bidding risk analysis for the bidder.

[0010] Furthermore, the third entity's business registration information data includes the first entity's business registration information data and the second entity's business registration information data; the third bidding risk characteristics include the first bidding risk characteristics and the second bidding risk characteristics; the third bidding document similarity characteristics include the first bidding document similarity characteristics and the second bidding document similarity characteristics; the third bidder risk assessment coefficient includes the first bidder risk assessment coefficient and the second bidder risk assessment coefficient; the first bidder risk assessment coefficient and its corresponding characteristics are collection coefficients and characteristics; and the second bidder risk assessment coefficient and its corresponding characteristics are generation coefficients and characteristics.

[0011] Furthermore, the business registration information data of the first entity, the second entity, or the third entity includes business registration location information, equity structure hierarchy data, and actual controller data.

[0012] Furthermore, the first, second, or third bidding risk features are visualized using the IP address and MAC address entered by the bidder when logging into the system before uploading the bid documents, obtained from the cloud. Then, image similarity calculation is performed on the images of the IP address and MAC address entered by other bidders when logging into the system before uploading their bid documents, and the maximum similarity value is taken.

[0013] Furthermore, the first, second, or third similarity features of the tender documents are obtained by using natural language processing technology to perform similarity detection between the tender documents uploaded by the bidder and any other tender documents uploaded by the bidder, and then calculating the average value.

[0014] Furthermore, the generative adversarial network model based on the first unit's business information data employs an improved generative activation function.

[0015] Furthermore, the third bidder's bidding risk analysis features are obtained by horizontally concatenating the third entity's business information data, the third bid risk features, and the similarity features of the third bidding document to obtain the corresponding feature vector as the bidder's bidding risk analysis features.

[0016] Furthermore, the intelligent risk analysis model for the bidder's equity is a neural network model.

[0017] It also provides a cloud-based collaborative intelligent analysis system for bidder equity, which includes a cloud-based bid data collection module, a bidder risk assessment feature generation module, a bidder bid risk analysis feature fusion module, a bidder equity intelligent risk analysis model construction module, and a bidder equity intelligent analysis module.

[0018] The cloud-based bidding data collection module collects the business information data of the first entity of the bidder through the cloud, and obtains the first bid risk characteristics and the first bidding document similarity characteristics through collection and processing, and obtains the first bidder risk assessment coefficient set by the bidding review experts according to the classification.

[0019] The bidder risk assessment feature generation module: uses a generative adversarial network model based on the first entity's business information data to expand the bidder risk assessment coefficient to obtain a second bidder risk assessment coefficient. The second bidder risk assessment coefficient and the first bidder risk assessment coefficient together constitute a third bidder risk assessment coefficient for training the intelligent risk analysis model of bidder equity. The third bidder risk assessment coefficient has corresponding third entity business information data, third bid risk features, and third tender document similarity features.

[0020] The bidder's bid risk analysis feature fusion module receives and processes the third entity's business information data, the third bid risk features, and the third tender document similarity features to obtain the bidder's bid risk analysis features;

[0021] The intelligent risk analysis model construction module for bidder equity: processes the bidder's bidding risk analysis features and the third bidder's risk assessment coefficient to construct an intelligent risk analysis model for bidder equity. The third entity's business registration information data includes the first entity's business registration information data and the second entity's business registration information data. The third bidding risk features include the first bidding risk features and the second bidding risk features. The third bidding document similarity features include the first bidding document similarity features and the second bidding document similarity features. The third bidder's risk assessment coefficient includes the first bidder's risk assessment coefficient and the second bidder's risk assessment coefficient. The first bidder's risk assessment coefficient and its corresponding features are collection coefficients and features, and the second bidder's risk assessment coefficient and its corresponding features are generation coefficients and features.

[0022] The intelligent analysis module for bidder equity: uses the intelligent risk analysis model for bidder equity to conduct bidding risk analysis on different subsequent bidders to obtain the risk assessment coefficient of the fourth bidder, and uses the risk assessment coefficient of the fourth bidder to conduct bidding risk analysis on the bidder.

[0023] This invention collects business registration information data of bidders through the cloud, and obtains bidding risk characteristics and bidding document similarity characteristics through data collection and processing. It then constructs an intelligent risk analysis model for bidder equity using the business registration information data, the bidding risk characteristics, the bidding document similarity characteristics, and the corresponding bidder risk assessment coefficients. The model generates new bidder risk assessment coefficients for risk analysis. Furthermore, it expands the training dataset of the model using a data generation model based on corrected business registration information data to obtain more accurate bidder risk assessment coefficients. This invention utilizes a multi-fusion model to accurately generate data on relevant characteristics used for bidder risk analysis, and uses this data to obtain a more generalized risk analysis model, thereby improving the intelligence and accuracy of bidder risk analysis. Attached Figure Description

[0024] Figure 1 This is a flowchart of the intelligent analysis method for bidder equity based on cloud collaboration according to the present invention;

[0025] Figure 2 This is a schematic diagram of the cloud-based collaborative intelligent analysis system for bidder equity of the present invention;

[0026] Figure 3 This is a schematic diagram of the generative adversarial network model in this invention;

[0027] Figure 4 This is a schematic diagram of the equity layering structure of the bidders in an embodiment of the present invention;

[0028] Figure 5This is a schematic diagram of an electronic device structure for implementing the method of the present invention in an embodiment of the present invention. Detailed Implementation

[0029] The invention will now be further described in conjunction with the accompanying drawings and specific embodiments. The generative adversarial network model used in this invention is an improved generative adversarial network model based on the unit business information data used for training, and the activation function is a corresponding improved processing method for the application scenario of bidder risk assessment coefficient data augmentation.

[0030] like Figure 2 As shown, this invention belongs to the field of bidding engineering technology research and development in the evaluation of bidding projects, and is a research and development of engineering technology for extended analysis and processing of bidding data. Therefore, it belongs to other engineering technology research and development besides marine engineering equipment, new materials, biotechnology, new energy, new energy vehicles, energy conservation, and environmental protection.

[0031] In a first aspect of the present invention, a cloud-based collaborative intelligent analysis method for bidder equity is provided, the method comprising:

[0032] D1. Collect the business information data of the first unit of the bidder through the cloud, and obtain the risk characteristics of the first bidder and the similarity characteristics of the first bidding document through collection and processing, and obtain the risk assessment coefficient of the first bidder set by the bidding review experts according to the classification.

[0033] D2. The risk assessment coefficient of the bidder is expanded by using a generative adversarial network model based on the first unit's business information data to obtain a second bidder risk assessment coefficient. The second bidder risk assessment coefficient and the first bidder risk assessment coefficient together constitute the third bidder risk assessment coefficient for training the intelligent risk analysis model of bidder equity. The third bidder risk assessment coefficient has corresponding third unit business information data, third bid risk characteristics and third bidding document similarity characteristics.

[0034] D3. Receive and process the business registration information data of the third entity, the risk characteristics of the third bidder, and the similarity characteristics of the third tender document to obtain the bidder's bid risk analysis characteristics, and construct an intelligent risk analysis model for the bidder's equity using the bidder's bid risk analysis characteristics and the risk assessment coefficient of the third bidder;

[0035] D4. Using the intelligent risk analysis model for bidder equity, conduct bidding risk analysis for different bidder entities to obtain the risk assessment coefficient of the fourth bidder, and use the risk assessment coefficient of the fourth bidder to conduct bidding risk analysis for the bidder.

[0036] Furthermore, the third entity's business registration information data includes the first entity's business registration information data and the second entity's business registration information data; the third bidding risk characteristics include the first bidding risk characteristics and the second bidding risk characteristics; the third bidding document similarity characteristics include the first bidding document similarity characteristics and the second bidding document similarity characteristics; the third bidder risk assessment coefficient includes the first bidder risk assessment coefficient and the second bidder risk assessment coefficient; the first bidder risk assessment coefficient and its corresponding characteristics are collection coefficients and characteristics; and the second bidder risk assessment coefficient and its corresponding characteristics are generation coefficients and characteristics.

[0037] In this embodiment, the second entity's business registration information data, the second bid risk characteristics, and the similarity characteristics of the second bidding document are the corresponding generated risk assessment coefficients of the second bidder. Here, the second entity's business registration information data, the second bid risk characteristics, the second bidding document similarity characteristics, and the corresponding risk assessment coefficients of the second bidder are generated data generated using the generative adversarial network model of this invention. The first entity's business registration information data, the first bid risk characteristics, the first bidding document similarity characteristics, and the corresponding risk assessment coefficients of the first bidder are collected data collected from the cloud. The above-mentioned first relevant data and second relevant data combined constitute the training data set for training the intelligent risk analysis model of bidder equity.

[0038] Furthermore, the business registration information data of the first entity, the second entity, or the third entity includes business registration location information, equity structure hierarchy data, and actual controller data.

[0039] When conducting risk analysis on bidders, some companies registered in certain regions may pose risks such as bid rigging and collusion. Therefore, the registration information of bidders is taken into account when constructing the model.

[0040] The complexity of equity structure hierarchy typically impacts the risk analysis of bidders. For example, a single-level equity structure lacks hierarchical management, thus presenting bidding risks. Therefore, this invention considers equity structure hierarchy data in constructing an intelligent risk analysis model for bidder equity. (See attached...) Figure 4 The equity structure shown has two levels.

[0041] In addition, there are risks associated with the actions of the actual controllers of some entities. This invention takes into account the risks of the actual controllers' actions and incorporates multi-level data considerations into the construction of the intelligent risk analysis model for bidder equity, thereby improving the model's intelligent identification of bidder risks.

[0042] In this embodiment, the business registration location information is represented by numerical values ​​to represent the corresponding vector features. Similarly, the equity structure hierarchy data is set by the upper and lower equity levels of the bidder. If the bidder has a superior company and a subordinate company, its equity structure hierarchy data is set to 3. The actual controller data is set by the legal risk of the company's ultimate superior actual controller. If the actual controller faces litigation risk, it is set to 1; if there is no risk, it is set to 0.

[0043] Furthermore, the first, second, or third bidding risk feature is visualized using the IP address and MAC address entered by the bidder when logging into the system before uploading their bid documents, obtained from the cloud. Then, it is compared with the visualized IP addresses and MAC addresses entered by other bidders when logging into the system before uploading their bid documents. The maximum similarity value is then calculated, and the similarity calculation formula is:

[0044]

[0045] In the formula, S represents the similarity between the bidder's IP address and MAC address after being graphically displayed and any other bidder. B is the image feature vector of the bidder's IP address and MAC address, and A is the image feature vector of any other bidder's IP address and MAC address. and The norm of an image feature vector. Represents the dot product of vectors.

[0046] In this invention, the cosine similarity algorithm is used to calculate the similarity of identical images, and the maximum similarity value is taken as the bidding risk feature. If the maximum similarity value is 1, it indicates that the IP addresses and MAC addresses of the two bidders are similar, and the bidding risk feature is 1. This feature vector is used to construct the subsequent intelligent risk analysis model for bidder equity, indicating that the risk of bidder analysis is extremely high. This is a bidding risk feature value in this embodiment, and no actual limitation is made in this invention. The similarity is between -1 and 1, including -1 and 1.

[0047] Furthermore, the first, second, or third similarity features of the bidding documents are obtained by using natural language processing technology to perform similarity detection between the bidder's uploaded bid documents and any other bidder's uploaded bid documents, and the average value is calculated. The calculation formula is as follows:

[0048]

[0049] In the formula, The average similarity between the bidder's uploaded bid documents and those uploaded by all other bidders, where n is the number of bid documents uploaded by all bidders. The Word2Vec model is used to convert the bid document text into text vectors for bidders to upload their bid documents. The text vector generated by converting the bid document uploaded by the i-th other bidder using the Word2Vec model. and This represents the norm of the transformed text feature vector. Represents the dot product of vectors.

[0050] In this embodiment, the bid text vector is used as one of the influencing factors for risk analysis of bid texts from different bidders, thereby enabling the subsequent construction of an intelligent risk analysis model for bidder equity.

[0051] Furthermore, the generative adversarial network model based on the first unit's business information data adopts an improved generative activation function as follows:

[0052]

[0053] In the formula, To generate activation function values, m represents the number of first-unit business information data points used in training the intelligent risk analysis model for bidder equity. The input bidder risk analysis characteristics are weighted and biased. The risk coefficient of the first unit of business information is transformed from the first unit of business information data of the i-th training data.

[0054] In this invention, m represents the number of feature vectors collected for training. During the expansion and generation of bidder risk assessment coefficients, it was found that the unit business registration information data has a significant impact on the generated data when assessing bidder risk. To ensure the calculation of the activation function, the first unit business registration information data, which contains three feature vector values—business registration location information, equity structure hierarchy data, and actual controller data—is artificially transformed into a risk coefficient. This transformation is based on its impact on the final result, ensuring the reasonable generation of bidder risk assessment coefficients. This allows the generated data to be more closely aligned with a reasonable range, while also ensuring that most of the generated bidder risk assessment coefficients fall within the set range of 0.5 to 6. Therefore, the first unit business registration information data from the actual data collected in the training dataset is used to improve the generative adversarial network model. Specifically, the most important aspect, the generation activation function, is improved, making the generative adversarial network model more generalizable to the generated data and improving its applicability in the scenario of bidder risk analysis.

[0055] As attached Figure 3 As shown, Generative Adversarial Networks (GANs) are deep learning models deeply inspired by the concept of "two-player zero-sum games" in game theory. Their core architecture comprises two collaborating neural networks: a generator and a discriminator. Through a unique adversarial training mechanism, this model aims to learn the latent distribution of real-world datasets, thereby generating new and realistic data samples.

[0056] Core components: Generator and Discriminator. The Generator functions like a "forger." It receives a random noise vector z sampled from a prior distribution (such as a Gaussian or uniform distribution) and attempts to map this noise to the data space, outputting a fake data sample G(z) (e.g., an image). The generator's goal is to make the distribution of G(z) as consistent as possible with the distribution of the real data.

[0057] Discriminator: Its role is similar to that of an "identifier." It receives a data sample (which may come from a real dataset or be a sample generated by a generator) and outputs a scalar probability value to determine how likely the input sample is to come from the real data distribution. The goal of the discriminator is to distinguish between real samples and generated samples as accurately as possible.

[0058] Furthermore, the third bidder's bidding risk analysis features are obtained by horizontally concatenating the third entity's business information data, the third bid risk features, and the similarity features of the third bidding document to obtain the corresponding feature vector as the bidder's bidding risk analysis features.

[0059] In this embodiment, the horizontal concatenation of vectors is a feature processing method before the feature vectors are input into the model. Here, the feature vectors are represented by horizontal concatenation of the feature vector values.

[0060] Furthermore, the intelligent risk analysis model for the bidder's equity is a neural network model.

[0061] Furthermore, the activation function calculation formula for the intelligent risk analysis model for bidder equity is as follows:

[0062]

[0063] In the formula, For the activation function value, The input bidder risk analysis features are biased and weighted, and the bidder risk analysis features include the bidder risk analysis features collected and processed, as well as the bidder risk analysis features corresponding to the second bidder risk assessment coefficient generated by the generative adversarial network model based on the first unit's business information data.

[0064] The risk assessment coefficient for bidders is output based on the final result of the neural network, calculated using an activation function. For example, if the final output value of the neural network is 5.86, the risk assessment coefficient for the bidder is determined to be 4. This is how risk monitoring personnel determine that the bidder's risk is high. More specific progress values ​​and neural network output values ​​will not be elaborated here.

[0065] It also provides a cloud-based collaborative intelligent analysis system for bidder equity, which includes a cloud-based bid data collection module, a bidder risk assessment feature generation module, a bidder bid risk analysis feature fusion module, a bidder equity intelligent risk analysis model construction module, and a bidder equity intelligent analysis module.

[0066] The cloud-based bidding data collection module collects the business information data of the first entity of the bidder through the cloud, and obtains the first bid risk characteristics and the first bidding document similarity characteristics through collection and processing, and obtains the first bidder risk assessment coefficient set by the bidding review experts according to the classification.

[0067] The bidder risk assessment feature generation module: uses a generative adversarial network model based on the first entity's business information data to expand the bidder risk assessment coefficient to obtain a second bidder risk assessment coefficient. The second bidder risk assessment coefficient and the first bidder risk assessment coefficient together constitute a third bidder risk assessment coefficient for training the intelligent risk analysis model of bidder equity. The third bidder risk assessment coefficient has corresponding third entity business information data, third bid risk features, and third tender document similarity features.

[0068] The bidder's bid risk analysis feature fusion module receives and processes the third-party business information data, third-party bid risk features, and third-party bidding document similarity features to obtain the bidder's bid risk analysis features;

[0069] The intelligent risk analysis model construction module for bidder equity: processes the bidder's bidding risk analysis features and the third bidder's risk assessment coefficient to construct an intelligent risk analysis model for bidder equity. The third entity's business registration information data includes the first entity's business registration information data and the second entity's business registration information data. The third bidding risk features include the first bidding risk features and the second bidding risk features. The third bidding document similarity features include the first bidding document similarity features and the second bidding document similarity features. The third bidder's risk assessment coefficient includes the first bidder's risk assessment coefficient and the second bidder's risk assessment coefficient. The first bidder's risk assessment coefficient and its corresponding features are collection coefficients and features, and the second bidder's risk assessment coefficient and its corresponding features are generation coefficients and features.

[0070] The intelligent analysis module for bidder equity: uses the intelligent risk analysis model for bidder equity to conduct bidding risk analysis on different subsequent bidders to obtain the risk assessment coefficient of the fourth bidder, and uses the risk assessment coefficient of the fourth bidder to conduct bidding risk analysis on the bidder.

[0071] Therefore, the beneficial effects of this invention are as follows: by collecting business registration information data of bidders in the cloud, and processing the data to obtain bidding risk characteristics and similarity characteristics of bidding documents, an intelligent risk analysis model for bidder equity is constructed using the business registration information data, the bidding risk characteristics, the similarity characteristics of bidding documents, and the corresponding bidder risk assessment coefficients. The model generates new bidder risk assessment coefficients for risk analysis. Furthermore, a data generation model based on corrected business registration information data is used to expand the training dataset of the model to obtain more accurate bidder risk assessment coefficients. This invention utilizes a multi-fusion model to accurately generate data on relevant characteristics used for analyzing bidder risk, and uses this data to obtain a more generalized risk analysis model, thereby improving the intelligence and accuracy of bidder risk analysis.

[0072] The combination of multiple embodiments of the present invention can achieve all the above-mentioned effects, but it is not required that each embodiment of the present invention achieve all the above-mentioned advantages and effects, because each embodiment of the present invention can constitute a separate technical solution and make one or more contributions to the prior art.

[0073] For any module structures not specifically defined in this invention, the existing technical specifications shall prevail. The existing technical specifications mentioned in the foregoing background and specific embodiments sections are considered part of this invention and are used to understand the meaning of certain technical features or parameters. The scope of protection of this invention is determined by the actual contents of the claims.

Claims

1. A cloud-based collaborative intelligent analysis method for bidder equity, characterized in that, The method includes: D1. Collect the business information data of the first unit of the bidder through the cloud, and obtain the risk characteristics of the first bidder and the similarity characteristics of the first bidding document through collection and processing, and obtain the risk assessment coefficient of the first bidder set by the bidding review experts according to the classification. D2. The risk assessment coefficient of the bidder is expanded by using a generative adversarial network model based on the first unit's business information data to obtain a second bidder risk assessment coefficient. The second bidder risk assessment coefficient and the first bidder risk assessment coefficient together constitute the third bidder risk assessment coefficient for training the intelligent risk analysis model of bidder equity. The third bidder risk assessment coefficient has corresponding third unit business information data, third bid risk characteristics and third bidding document similarity characteristics. D3. Receive and process the business registration information data of the third entity, the risk characteristics of the third bidder, and the similarity characteristics of the third bidding document to obtain the bidder's bid risk analysis characteristics, and use the bidder's bid risk analysis characteristics and the risk assessment coefficient of the third bidder to construct an intelligent risk analysis model for the bidder's equity; D4. Using the intelligent risk analysis model for bidder equity, conduct bidding risk analysis for different bidder entities to obtain the risk assessment coefficient of the fourth bidder, and use the risk assessment coefficient of the fourth bidder to conduct bidding risk analysis for the bidder.

2. The intelligent analysis method for bidder equity based on cloud collaboration as described in claim 1, characterized in that: The third entity's business registration information data includes the first entity's business registration information data and the second entity's business registration information data. The third bidding risk characteristics include the first bidding risk characteristics and the second bidding risk characteristics. The third bidding document similarity characteristics include the first bidding document similarity characteristics and the second bidding document similarity characteristics. The third bidder risk assessment coefficient includes the first bidder risk assessment coefficient and the second bidder risk assessment coefficient. The first bidder risk assessment coefficient and its corresponding characteristics are collection coefficients and characteristics. The second bidder risk assessment coefficient and its corresponding characteristics are generation coefficients and characteristics.

3. The intelligent analysis method for bidder equity based on cloud collaboration as described in claim 2, characterized in that: The business registration information of the first entity, the second entity, or the third entity includes information on the place of business registration, equity structure hierarchy, and actual controller.

4. The intelligent analysis method for bidder equity based on cloud collaboration as described in claim 2 or 3, characterized in that: The first, second, or third bidding risk features are visualized using the IP address and MAC address entered by the bidder when logging into the system before uploading the bid documents, obtained from the cloud. Then, the similarity between these images is calculated and displayed using the IP address and MAC address entered by other bidders when logging into the system before uploading their bid documents. The maximum similarity value is then taken.

5. The intelligent analysis method for bidder equity based on cloud collaboration as described in claim 4, characterized in that: The first, second, or third similarity features of the tender documents are obtained by using natural language processing technology to perform similarity detection between the tender documents uploaded by the bidder and any other tender documents uploaded by the bidder, and then calculating the average value.

6. The intelligent analysis method for bidder equity based on cloud collaboration as described in claim 1, characterized in that: The generative adversarial network model based on the first unit's business information data employs an improved generative activation function.

7. The intelligent analysis method for bidder equity based on cloud collaboration as described in claim 6, characterized in that: The bidder's bid risk analysis features are obtained by horizontally splicing the third entity's business information data, the third bid risk features, and the third tender document similarity features to obtain the corresponding feature vectors as the bidder's bid risk analysis features.

8. A cloud-based collaborative intelligent analysis method for bidder equity as described in claim 3 or 7, characterized in that: The intelligent risk analysis model for bidder equity is a neural network model.

9. A cloud-based collaborative intelligent analysis system for bidder equity, characterized in that... The system includes: Cloud-based bidding data collection module: Collects the business information data of the first bidder's entity through the cloud, and obtains the risk characteristics of the first bidder and the similarity characteristics of the first bidding document through collection and processing, and obtains the risk assessment coefficient of the first bidder set by the bidding review experts according to the classification. Bidder Risk Assessment Feature Generation Module: The bidder risk assessment coefficient is expanded using a generative adversarial network model based on the first entity's business information data to obtain a second bidder risk assessment coefficient. The second bidder risk assessment coefficient and the first bidder risk assessment coefficient together constitute the third bidder risk assessment coefficient for training the intelligent risk analysis model of bidder equity. The third bidder risk assessment coefficient has corresponding third entity business information data, third bid risk features, and third tender document similarity features. Bidder's Bid Risk Analysis Feature Fusion Module: Receives and processes third-party business information data, third-party bid risk features, and third-party bidding document similarity features to obtain bidder's bid risk analysis features; The intelligent risk analysis model construction module for bidder equity: This module processes the bidder's bidding risk analysis features and the third bidder's risk assessment coefficient to construct an intelligent risk analysis model for bidder equity. The third entity's business registration information data includes the first entity's business registration information data and the second entity's business registration information data. The third bidding risk features include the first bidding risk features and the second bidding risk features. The third bidding document similarity features include the first bidding document similarity features and the second bidding document similarity features. The third bidder risk assessment coefficient includes the first bidder risk assessment coefficient and the second bidder risk assessment coefficient. The first bidder risk assessment coefficient and its corresponding features are collection coefficients and features, while the second bidder risk assessment coefficient and its corresponding features are generation coefficients and features. Intelligent analysis module for bidder equity: The intelligent risk analysis model for bidder equity is used to analyze the bidding risks of different bidders and obtain the risk assessment coefficient of the fourth bidder. The risk assessment coefficient of the fourth bidder is then used to analyze the bidding risks of the bidder.

Images