Item risk detection method and device, storage medium and electronic equipment

Abstract

The invention provides a project risk detection method and apparatus, a storage medium and an electronic device. The method comprises the steps of obtaining first feature data of each bidder participating in a current bidding project in each first information dimension and second feature data of each bidder in each second information dimension; based on each piece of first feature data, determining each bidder with a dominant relationship, and determining dominant relationship information of each bidder with the dominant relationship; based on each piece of second feature data, determining a recessive association degree between the bidders; and generating bidding risk information of the bidding project based on the recessive association degree between the bidders and the dominant relationship information of each bidder with the dominant relationship. The dominant association relationship and the implicit association relationship among the bidders are mined, and the bidding risk information is generated based on the association relationship among the bidders, so that an enterprise issuing a bidding document takes corresponding measures to avoid the bidding risk according to the bidding risk information, and the interests of the enterprise are maintained.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a project risk detection method and device, a storage medium and an electronic device. Background Art

[0002] With the development of the economy, many enterprises or government departments usually publish tender documents when conducting transactions or constructing engineering projects, and then determine the final bidder from among many bidders for transaction.

[0003] At present, there is a risk of bid rigging among bidders in the bidding process. Bid rigging refers to the behavior in which multiple related bidders privately select a winning company among the bidders participating in the bidding, and then other companies jointly submit bids for the same project in the form of accomplices. This behavior seriously endangers the fairness of the bidding process, and joint bid rigging will monopolize the bidding project and raise the bid price. At present, it is impossible to determine the risk of bid rigging in bidding projects, making the interests of enterprises that issue bids vulnerable to bid rigging in bidding projects. Summary of the invention

[0004] In view of this, the present invention provides a project risk detection method and device, storage medium and electronic device, which analyze the relationship between each bidder and generate bid-rigging risk information based on the relationship between each bidder, so that the enterprise that publishes the bidding project can avoid risks based on the bid-rigging risk information and avoid damage to its own interests.

[0005] To achieve the above objectives, the embodiments of the present invention provide the following technical solutions:

[0006] A risk detection method for a project, comprising:

[0007] Identify the various bidders participating in the current bidding project;

[0008] Acquire first characteristic data of each of the bidders in each preset first information dimension, and second characteristic data of each of the bidders in each preset second information dimension;

[0009] Based on each of the first characteristic data, determine each bidder with an explicit relationship, and determine explicit relationship information of each of the bidders with an explicit relationship, wherein the explicit relationship information includes each explicit association relationship of the bidders;

[0010] Determining the implicit correlation between the bidders based on the second characteristic data of the bidders;

[0011] Based on the implicit correlation between each of the bidders and the explicit relationship information of each of the bidders with an explicit relationship, the bid-rigging risk information of the bidding project is generated.

[0012] A project risk detection device, comprising:

[0013] A first determining unit, used to determine each bidder participating in the current bidding project;

[0014] An acquisition unit, used for acquiring first characteristic data of each of the bidders in each preset first information dimension, and second characteristic data of each of the bidders in each preset second information dimension;

[0015] A second determining unit, configured to determine each bidder having an explicit relationship based on each of the first characteristic data, and determine explicit relationship information of each of the bidders having an explicit relationship, wherein the explicit relationship information includes each explicit association relationship of the bidders;

[0016] A third determining unit, configured to determine the implicit correlation between the bidders based on the second characteristic data of the bidders;

[0017] A generating unit is used to generate the bid-rigging risk information of the bidding project based on the implicit correlation degree between each of the bidders and the explicit relationship information of each of the bidders with an explicit relationship.

[0018] A storage medium includes stored instructions, wherein when the instructions are executed, the device where the storage medium is located is controlled to execute the risk detection method for the project as described above.

[0019] An electronic device includes a memory and one or more instructions, wherein the one or more instructions are stored in the memory and are configured to be executed by one or more processors to perform the risk detection method of the project as described above.

[0020] Compared with the prior art, the present invention has the following advantages:

[0021] The present invention provides a project risk detection method and device, storage medium and electronic device, the method comprising: determining each bidder participating in the current bidding project; obtaining first feature data of each bidder in each preset first information dimension, and second feature data in each preset second information dimension; determining each bidder with an explicit relationship based on each first feature data, and determining explicit relationship information of each bidder with an explicit relationship, wherein the explicit relationship information includes each explicit association relationship of the bidder with an explicit relationship; determining the implicit association degree between each bidder based on each second feature data of each bidder; generating bid-rigging risk information of the bidding project based on the implicit association degree between each bidder and the explicit relationship information of each bidder with an explicit relationship. The present invention deeply mines the explicit association relationship and implicit association relationship between each bidder, and generates bid-rigging risk information based on the association relationship between each bidder, so that the enterprise that issues the bid document can determine the bid-rigging risk of the current bidding project according to the bid-rigging risk information, so as to take corresponding measures to avoid the bid-rigging risk and safeguard its own interests. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0023] Figure 1 A method flow chart of a project risk detection method provided by an embodiment of the present invention;

[0024] Figure 2 A flow chart of a method for determining the implicit correlation between bidders based on the second characteristic data of each bidder provided in an embodiment of the present invention;

[0025] Figure 3 A flow chart of a method for obtaining a three-dimensional vector of each bidder provided in an embodiment of the present invention;

[0026] Figure 4 A flow chart of a method for obtaining the implicit correlation between each bidder provided in an embodiment of the present invention;

[0027] Figure 5 A system flow chart of risk detection of a project provided by an embodiment of the present invention;

[0028] Figure 6 A schematic diagram of the structure of a risk detection device for a project provided by an embodiment of the present invention;

[0029] Figure 7 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] In this application, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element.

[0032] Terminology explanation:

[0033] Newton's law of cooling: the law followed by an object with a temperature higher than the surrounding environment when it transfers heat to the surrounding medium and gradually cools down. This is one of the basic laws of heat transfer and is used to calculate the amount of convection heat. The specific formula is:

[0034]

[0035] Wherein, T is the current object temperature, H is the current ambient temperature, T0 is the initial temperature of the object, t-t0 is the time difference between the initial time and the current time, and α is the thermal conductivity coefficient.

[0036] TF-IDF: The full name is Term Frequency-Inverse Document Frequency, which is often used in natural language processing to evaluate the importance of a paragraph to the text. The formula is:

[0037] TF-IDF = TF (term frequency) × IDF (inverse document frequency);

[0038] Among them, word frequency refers to the frequency of a specific word appearing in a text, and reverse file frequency refers to the reverse processing of the frequency of a certain term appearing in other texts, that is, the smaller the frequency of a certain term appearing in other texts, the more important this term is to the text.

[0039] Graph embedding algorithm: It is a process of mapping graph data (usually high-dimensional dense matrices) into low-dimensional dense vectors.

[0040] Euclidean distance: It is a measure of the straight-line distance between two points in Euclidean space. In Euclidean space, point x = (x1,…,x n ) and y=(y1,…,y n ) is:

[0041]

[0042] Node2Vec algorithm: A graph embedding algorithm model that can comprehensively consider breadth-first traversal (BFS) and depth-first traversal (DFS) when learning graph features.

[0043] As can be seen from the background technology, timely identification of bid-rigging risks in the bidding process is very important for corporate risk control. At present, logistics companies have two ways to identify such risks. One is to manually compare the information submitted by bidders before and after bidding to see whether there is a connection between companies to determine whether there is a bid-rigging risk; the other is to identify such risky behavior after bidders have submitted their bids or after violations occur during operations.

[0044] The existing schemes for identifying the risk of bid rigging in projects have the following problems:

[0045] 1. The information provided by the bidders during the bidding process is used to determine whether there is any relationship between the bidders, giving the bidders room to operate. The information reviewed by the company comes from the bidders, and the information submitted by the bidders who really have the tendency to engage in bid rigging will deliberately avoid obvious or identifiable related information.

[0046] 2. The identification of whether there is a relationship between bidders through the information provided by bidders during the bidding process is limited and lacks comprehensiveness. The information provided by bidders is mostly used to prove that the bidders have the qualifications for the three-party transportation capacity required by the enterprise. It is easy to capture the relationship between bidders through these information incompletely.

[0047] 3. It is not objective to judge whether there is a relationship between bidders based on the information provided by bidders during the bidding process. Human judgment is mostly based on personal experience, and different people have different judgment standards. At the same time, there is a risk of collusion between tendering personnel and bidding companies through human judgment.

[0048] 4. There is a lag and uncertainty in judging whether a company has engaged in bid rigging based on violations committed by the company after bidding or during operation, making it impossible to stop losses in a timely manner.

[0049] The present invention analyzes the implicit and explicit associations between bidders, thereby determining the risk of bid rigging in the bid according to the associations between bidders, thereby avoiding the risk.

[0050] The present invention can be used in many general or special computing device environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multi-processor devices, distributed computing environments including any of the above devices or devices, etc. The present invention can be used in a bidding system, which can be composed of distributed computing devices or other types of data processing devices.

[0051] Reference Figure 1 , which is a method flow chart of a project risk detection method provided by an embodiment of the present invention, and the specific description is as follows.

[0052] S101. Determine the bidders participating in the current bidding project.

[0053] Get the bidder list of the current bidding project, parse the bidder list, and get each bidder.

[0054] It should be noted that the bidder is a participant in the bidding project and can be a carrier or other company that wants to undertake the bidding project.

[0055] Preferably, the bidding project of the present invention may be a project on transportation service, and the bidder in this case may be a carrier.

[0056] S102: Obtain first characteristic data of each bidder in each preset first information dimension and second characteristic data in each preset second information dimension.

[0057] The process of obtaining the first characteristic data of each bidder in each preset first information dimension and the second characteristic data in each preset second information dimension is as follows:

[0058] Obtain the corporate identity information and corporate operation information of each bidder;

[0059] Processing the corporate identity information of each bidder to obtain first characteristic data of each bidder in each preset first information dimension;

[0060] The enterprise operation information of each bidder is processed to obtain second characteristic data of each bidder in each preset second information dimension.

[0061] The bidder's corporate identity information includes but is not limited to the bidder's industrial and commercial registration information, operating qualification information, corporate shareholder information, corporate operating financial information and corporate guarantee information. Among them, the industrial and commercial registration information includes the bidder's legal person information and the same social credit code, the operating qualification information includes the bidder's road license code, the corporate shareholder information includes the bidder's corporate shareholders and natural person shareholders information, the corporate operating financial information includes the company's bank account number and account name, and the corporate guarantee information includes the company's guarantee object.

[0062] The bidder's business operation information includes but is not limited to the bidder's historical bidding information, historical operation information, information on affiliated drivers and vehicles, and contact information used during business registration and operation.

[0063] It should be noted that the identity information of each enterprise is requested for processing, and the unstructured data is converted into structured data, so as to obtain the first characteristic information of each bidder in each preset first information dimension.

[0064] The various first information dimensions include but are not limited to the unified social credit code dimension, the road permit code dimension, the enterprise bank account dimension, the corporate legal person identity information dimension, the shareholder identity information dimension, the head office identity information dimension, the branch identity information dimension, and the guarantor and guaranteed company or individual identity information dimension.

[0065] The first characteristic information includes but is not limited to the unified social credit code, road license code, corporate bank account number, corporate legal person identity information, shareholder identity information, head office identity information, branch identity information, and guaranteeing and guaranteed company or individual identity information.

[0066] The system cleans and processes the bidder's business operation information, converting the unstructured data in the business operation information into structured data.

[0067] The second information dimensions include but are not limited to the operating route dimension, the tender submission dimension, the affiliation information dimension, and the contact information dimension; the second characteristic data include but are not limited to historical bidding projects, historical operating routes and transportation conditions on the operating routes, the driver's identity information and vehicle license plate information, the company's contact number, contact person, and contact email address.

[0068] Based on the corporate identity information and corporate operation information of the bidders, the present invention can deeply mine the first feature data of the bidders in each first information dimension and the second feature data in each second information dimension; by mining the feature data of the bidders in multiple dimensions, it is convenient to subsequently determine the relationship between the bidders based on the mined feature data, and then determine the risk of the bidding project.

[0069] S103. Based on each first characteristic data, determine each bidder with an explicit relationship, and determine explicit relationship information of each bidder with an explicit relationship, wherein the explicit relationship information includes each explicit association relationship of the bidder with an explicit relationship.

[0070] For each bidder, data correlation analysis is performed on each first characteristic data of the bidder and each first characteristic data of the remaining bidders to obtain the data analysis result of the bidder. When the data analysis result indicates that the bidder is associated with other bidders, the bidder is determined as a bidder with an explicit relationship.

[0071] Preferably, for each bidder, the bidder is determined as a target bidder, and when any of the following scenarios occurs between each first characteristic data of the target bidder and each first characteristic data of the remaining bidders, the target bidder can be determined as a bidder with an explicit relationship.

[0072] Scenario 1: Among the remaining bidders, there is a bidder whose unified social information code is consistent with the unified social credit code of the target bidder.

[0073] Scenario 2: Among the remaining bidders, there is a bidder whose road license code is the same as the road license code of the target bidder.

[0074] Scenario 3: Among the remaining bidders, there is a bidder whose company account number is the same as that of the target bidder.

[0075] Scenario 4: Among the remaining bidders, there is a bidder whose company registration address is the same as that of the target bidder.

[0076] Scenario 5: Among the remaining bidders, there is a bidder that is a subsidiary of the target bidder.

[0077] Scenario 6: Among the remaining bidders, there is a bidder whose parent company is the same as that of the target bidder.

[0078] Scenario 7: Among the remaining bidders, there is a bidder that has a controlling stake in the target bidder.

[0079] Scenario 8: Among the remaining bidders, there is a bidder that owns the same controlling shareholder as the target bidder.

[0080] Scenario 9: Among the remaining bidders, there is a bidder whose legal entity is the same as the target bidder.

[0081] Scenario 10: Among the remaining bidders, there is a bidder that has guaranteed the target bidder.

[0082] Scenario 11: Among the remaining bidders, there is a bidder that has been guaranteed by the target bidder.

[0083] It should be noted that the above-mentioned scenarios 1 to 10 are scenarios where there is a relationship between bidders exemplified in the present invention. The scenarios where there is a relationship between bidders covered by the present invention include but are not limited to the above-mentioned scenarios 1 to 10, which will not be repeated here.

[0084] When the target bidder meets any of the above scenarios, it is determined that there is an explicit relationship between the bidder that makes the target bidder meet the scenario and the target bidder. It should be noted that the target bidder has multiple explicit relationships, each displayed relationship may be of a different type, and there may be multiple explicit relationships of the same type; explicit relationships include but are not limited to head office and branch relationships, holding relationships, belonging to the same legal person relationship, guarantee relationships, and suspected relationships with the same company; among them, different scenarios correspond to different explicit relationships, the explicit relationships corresponding to scenarios 1 to 4 are suspected relationships with the same company, the explicit relationships corresponding to scenarios 5 and 6 are head office and branch relationships, the explicit relationships corresponding to scenarios 7 and 8 are holding relationships, the explicit relationships corresponding to scenario 9 are belonging to the same legal person relationship, and the explicit relationships corresponding to scenarios 10 and 11 are guarantee relationships.

[0085] Exemplarily, for the target bidder and bidders 1, 2, 3, and 4, there is an explicit association relationship between the target bidder and bidder 1, and the explicit association relationship is a head office-branch relationship; there is an explicit association relationship between the target bidder and bidder 2, and the explicit association relationship is a head office-branch relationship; there is an explicit association relationship between the target bidder and bidder 3, and the explicit association relationship is a controlling relationship; there is an explicit association relationship between the target bidder and bidder 4, and the explicit association relationship is a guarantee relationship.

[0086] Furthermore, each first characteristic data can be divided into two types of data, one type is natural person type data, and the other type is enterprise type data. Each first characteristic data is the unique identification information of the bidder. If the first characteristic data of the bidder can be associated with other bidders, it can be confirmed that the two bidders are associated, and the explicit association relationship between the two bidders can be judged based on the specific association data.

[0087] S104: Determine the implicit correlation between the bidders based on the second characteristic data of the bidders.

[0088] The second characteristic data of each bidder can be analyzed to explore the implicit correlation between the bidders; Figure 2, which is a flow chart of a method for determining the implicit correlation between bidders based on the second characteristic data of each bidder provided in an embodiment of the present invention, and is specifically described as follows.

[0089] S201. Generate each relationship circle based on the historical bidding items and historical operation routes in each second characteristic data of each bidder.

[0090] It should be noted that each relationship circle is a bid document relationship circle or a line relationship circle. The bid document relationship circle corresponds to a bidding project, and the bidding project exists in the historical bidding projects of each bidder in the bid document relationship circle. The line relationship circle corresponds to an operating route, and the operating route exists in the historical operating routes of each bidder in the line relationship circle. Furthermore, the bid document relationship circle can also be a bidding relationship circle.

[0091] Each bidder can be divided into different relationship circles according to their historical bidding projects and historical operating routes. Specifically, bidders who have bid for a bidding project at the same time in the past are considered to be one relationship circle, and bidders who have operated the same operating route in the past are considered to be one relationship circle. Preferably, the operating route can be from Beijing to Shanghai.

[0092] S202: Determine a plurality of relationship pairs based on the vehicle operation data in each second characteristic data of each bidder, wherein each relationship pair has two bidders.

[0093] Vehicle operation data includes information about the bidder’s drivers and vehicles as well as contact information.

[0094] Preferably, the vehicle operation data are compared one by one. When some data of two vehicle operation data are the same, the two bidders to which the two vehicle operation data belong are determined as a relationship pair. For example, if the two bidders have vehicles with the same driver or license plate, or the two bidders have used the same contact number or contact person, in this case the two bidders can be determined as a relationship pair.

[0095] Furthermore, in the present invention, C represents a bidder, E represents the connection between bidders, and U represents a relationship circle. Let the line relationship circle be U L , the bid relationship circle is U B If there are K bidders who have bid for the same bidding project, let the bidder be C i (i=1,2,3,……,K), then if bidder C i and C j Because the relationship associates the features, bidder C i and C j There is a correlation between ij , if bidder Ci and C j Participated in the operation of the line m If C i ∈U lm , C j ∈U lm .

[0096] S203: construct an association relationship graph based on each relationship circle and each relationship pair.

[0097] According to the relationship circle characteristics and relationship pair characteristics of each bidder, the bidders are divided into relationship circles and the bidders that can be associated are connected, that is, an association relationship map corresponding to each bidder is constructed. Furthermore, the points in the association relationship map represent bidders, the lines represent the relationship pair characteristics of two connected bidders forming a relationship pair, and the circles represent the relationship circles to which the bidders belong.

[0098] S204. Based on the correlation relationship map, generate a bidding project correlation degree measurement matrix, an operation route correlation degree measurement matrix, and a capacity overlap degree measurement matrix.

[0099] Based on the association relationship map, combined with the relationship circle features and relationship pair features in the association relationship map, the association strength between bidders belonging to the same relationship circle or that can be associated by the relationship pair features is preliminarily calculated. The association strength includes three parts: the overlap rate of the bid relationship circle, the overlap rate of the route relationship circle and the relationship pair association strength. Among them, the overlap rate of the bid relationship circle and the bid, after obtaining the association strength, after feature extraction and calculation, the association matrix between the bidders is obtained. The association matrix contains the measurement information of the degree of association between bidders under different dimensions. Independent matrices under each dimension are extracted from the association matrix, namely the bidding project association degree measurement matrix EW1, the operating route association degree measurement matrix EW2 and the capacity overlap degree measurement matrix EW3.

[0100] It should be noted that the TF-IDF algorithm, Newton's cooling law, and set operation rules are used in the process of calculating the overlap rate of bid relationship circles, the overlap rate of line relationship circles, and the strength of relationship pair association.

[0101] The present invention uses a specific example to illustrate the overlap rate of the bidding relationship circles between bidders. Specifically, let bidder C i and C j The overlap rate of the bidding relationship circle is CIB(C i ,C j ), Bidder C i The bidding relationship circle is CUB i , similarly, bidder C j The bidding document relationship circle is CUBj , then bidder C i and C j The overlap rate of the bidding relationship circle CIB (C i ,C j ) are as follows:

[0102]

[0103] As for the overlap rate of the line relationship circles between bidders, since the characteristics of the bidder's operating line include the two dimensions of the transport task volume and the transport task time of this operating line, the calculation of the overlap rate of the line relationship circles cannot be a simple set operation similar to the calculation of the bid overlap rate. The specific transport tasks of the bidders on this operating line and the time to complete the transport tasks must be considered. Since the transport tasks are time-sensitive, the time factor needs to be considered when calculating the overlap rate of the operations of two bidders on the same operating line. Therefore, Newton's cooling law is introduced here. Let bidder C i and C j The line relationship circle overlap rate is CIL(C i ,C j ), Bidder C i The operating line collection is LC i , bidder C j The operating line collection is LC j . Bidder C i On line l m The time set corresponding to the transportation task is JTC i (l m ), let the current time be T0, k be the cooling parameter in Newton's cooling law, which can be evaluated by measuring the effectiveness of the bidder's historical operation status for the current operation status. Then the bidder C i and C j The line relationship circle overlap rate CIL(C i ,C j ) is as follows:

[0104]

[0105] The strength of association between bidders is calculated. The bidders can be associated with each other through the identity information of the driver they rely on, the license plate information of the vehicle, and the contact information provided during the registration and operation of the enterprise, including the phone number, email address, and contact identity information. However, compared with rule recognition, these features do not have unique discrimination. Therefore, such features are used as features for calculating implicit association relationships. Let these feature sets be F = {f k}, for example, f1 refers to two bidders using the same driver, f2 refers to two bidders using the same license plate, etc. Then Nij (f1) is bidder C i and C j The number of drivers who are jointly registered, N i (f1) is bidder C i The number of drivers registered. i and C j The relationship strength is CI (C i ,C j ).

[0106] First, calculate the target bidder C i and C j The same relationship characteristics occupy bidder C i The proportion of the number of this relationship feature, let the target bidder C i and C j The same relationship characteristics occupy bidder C i In the feature f k The proportion of TF ij , then the proportion of TF ij Specifically expressed as:

[0107]

[0108] At the same time, let the target bidder C i In the feature f k The total number of associated bidders is CF ki , let the target bidder C i In the feature f k The inverse document frequency is IDF i (f k ), the inverse document frequency IDF i (f k ) is expressed as:

[0109]

[0110] The target bidder C i and C j The strength of the relationship is:

[0111] CI(C i ,C j )=∑ k TF ij (f k )×IDF i (f k ).

[0112] Furthermore, after calculating the overlap rate of the bid document relationship circles, the overlap rate of the operating line relationship circles and the strength of the relationship pairs among the bidders, the system obtains the correlation matrix EW of the related bidders, where the correlation matrix EW (C i ,C j )=[CIB(C i ,C j ),CIL(C i ,C j ),CI(C i ,C j )].

[0113] After feature extraction and calculation, the system obtains the correlation matrix between bidders and separates this correlation matrix into three independent matrices, EW1, EW2, and EW3, where EW1 is the correlation degree measurement matrix of bidding items between bidders, EW2 is the correlation degree measurement matrix of operating routes between bidders, and EW3 is the capacity overlap degree measurement matrix between bidders.

[0114] Among them, EW1(C i ,C j )=CIB(C i ,C j ); EW2(C i ,C j )=CIL(C i ,C j ); EW3(C i ,C j )=CI(C i ,C j ).

[0115] S205, processing the bidding project correlation degree measurement matrix, the operation route correlation degree measurement matrix and the transport capacity overlap degree measurement matrix to obtain a three-dimensional vector of each bidder.

[0116] Reference Figure 3 , which is a flow chart of a method for obtaining a three-dimensional vector of each bidder provided in an embodiment of the present invention, and is specifically described as follows.

[0117] S301. Based on each bidder, obtain a bidder matrix and a bidder association matrix.

[0118] The bidder matrix C and the bidder correlation matrix E corresponding to each bidder are acquired.

[0119] S302, combining the bidder matrix, the bidder association matrix, the bidding project association degree measurement matrix, the operation route association degree measurement matrix and the capacity overlap degree measurement matrix to obtain a first matrix combination, a second matrix combination and a third matrix combination.

[0120] The first matrix combination includes the bidder matrix, the bidder correlation matrix and the bidding project correlation degree measurement matrix, the second matrix combination includes the bidder matrix, the bidder correlation matrix and the operating route correlation degree measurement matrix, and the third matrix combination includes the bidder matrix, the bidder correlation matrix and the capacity overlap degree measurement matrix.

[0121] Preferably, the first matrix combination can be expressed as G1=(C, E, EW1), the second matrix combination can be expressed as G2=(C, E, EW2), and the third matrix combination can be expressed as G3=(C, E, EW3).

[0122] S303, inputting the first matrix combination, the second matrix combination and the third matrix combination into a preset algorithm model respectively, so that the algorithm model outputs a first dimensional vector, a second dimensional vector and a third dimensional vector of each bidder.

[0123] The three combinations in S302 are used as initial data to input into the Node2Vec algorithm model, and the parameters p and q are introduced to control whether the Node2Vec output results are biased towards homogeneity or isomorphism. In this scenario, Node2Vec should be biased towards homogeneity. At the same time, AliasSampling is selected for sampling in the calculation. This sampling method is selected because it is a non-uniform random sampling method, so the sampling result depends on the weight, which is in line with the setting of this scenario; secondly, this solution has a low time complexity, so it can improve the computing efficiency when used online without affecting system interaction.

[0124] S304. Obtain a three-dimensional vector of each bidder based on the first dimensional vector, the second dimensional vector and the third dimensional vector of each bidder.

[0125] The algorithm in the above S303 will output the vector corresponding to the input bidder. Let the output result be V. Then for G1=(C,E,EW1), G2=(C,E,EW2) and G3=(C,E,EW3), the output results are V1, V2 and V3, V=[V1, V2, V3]. At this time, the three-dimensional vectors of all bidders are obtained.

[0126] S206: Calculate the three-dimensional vector of each bidder to obtain the implicit correlation between each bidder.

[0127] Reference Figure 4 , which is a flow chart of a method for obtaining the implicit correlation between each bidder provided in an embodiment of the present invention, and is specifically described as follows.

[0128] S401. Perform Euclidean operation on the three-dimensional vector of each bidder to obtain the Euclidean distance between each bidder.

[0129] S402: Determine the implicit correlation degree of each bidder based on the Euclidean distance between each bidder.

[0130] The system calculates the distance between each bidder through the three-dimensional vector V of the bidder and the Euclidean distance; by calculating the distance between each bidder, let bidder C i The vector v i =[v 1i ,v 2i ,v 3i ], Bidder C j The vector v j =[v 1j ,v 2j ,v 3j ], then bidder C i and Bidder C j The distance between is:

[0131]

[0132] At this point, the system obtains the relationship strength matrix L between bidders, and then obtains the probability of association between each bidder through normalization. i and Bidder C j The probability of association between them is:

[0133]

[0134] Furthermore, the correlation probability can be used as an implicit correlation. Thus, the system obtains the implicit correlation between the bidders of the current bidding project.

[0135] It should be noted that the present invention may also use other distance calculation methods to calculate the distance between bidders, such as a cosine similarity algorithm.

[0136] S105. Generate bid-rigging risk information for the bidding project based on the implicit correlation between each bidder and the explicit relationship information of each bidder with an explicit relationship.

[0137] After obtaining the implicit correlation between each bidder and the explicit relationship information of each bidder with an explicit relationship, the bid-rigging risk information of the bidding project can be generated. The bid-rigging risk information includes the relationship between each bidder participating in the bidding. The relationship between each bidder participating in the bidding can be directly determined through the bid-rigging risk information, thereby avoiding the bid-rigging risk of the bidding project and avoiding damage to the interests of the enterprise that issued the bid.

[0138] In the method provided by the embodiment of the present invention, each bidder participating in the current bidding project is determined; the first characteristic data of each bidder in each preset first information dimension and the second characteristic data in each preset second information dimension are obtained; based on each first characteristic data, each bidder with an explicit relationship is determined, and the explicit relationship information of each bidder with an explicit relationship is determined, and the explicit relationship information includes each explicit association relationship of the bidder with an explicit relationship; based on each second characteristic data of each bidder, the implicit correlation between each bidder is determined; based on the implicit correlation between each bidder and the explicit relationship information of each bidder with an explicit relationship, the bid-rigging risk information of the bidding project is generated. The present invention deeply mines the explicit association relationship and implicit association relationship between each bidder, and generates bid-rigging risk information based on the association relationship between each bidder, so that the enterprise that issues the bid document can determine the bid-rigging risk of the current bidding project according to the bid-rigging risk information, so as to take corresponding measures to avoid the bid-rigging risk and safeguard its own interests.

[0139] Reference Figure 5 , which is a system flow chart of risk detection of a project provided by an embodiment of the present invention, and is specifically described as follows:

[0140] The bidders participating in the bidding are input into the system, and the system cleans and processes the bidder's industrial and commercial registration information, bidder's operating qualification information, bidder's corporate shareholder information, bidder's operating financial information, bidder's corporate guarantee record and other data of each bidder, and uses information matching rules to process the cleaned and processed data, thereby identifying the explicit association relationship between each bidder. When it can be identified by the information matching rules, the explicit association relationship between bidders is identified; when it cannot be identified by the information matching rules, the bidder's historical bidding information, bidder's historical operating information, drivers and vehicles registered under the bidder's name, bidder's contact method and other data of each bidder are cleaned and processed, thereby obtaining the bidder's characteristic data, based on the bidder's characteristic data, determine the bidder's relationship circle and the bidder's relationship pair, use the relationship circle and relationship pair to construct an association relationship map, calculate the association strength of the association relationship map, obtain the implicit association degree between bidders, and then output the implicit association degree between bidders.

[0141] The system locates and identifies the risk of bid rigging among bidders participating in the same project by identifying the associations between bidders and calculating the strength of associations. The system explores the associations between bidders through information matching rules and association algorithm recognition models, and outputs explicit and implicit associations between bidders. Among them, explicit associations refer to the system being able to obtain the identity information of bidders, such as qualification information, financial information, or foreign investment information, and match information to directly associate with bidders other than the target bidder. At this time, the system believes that there is an explicit association between the two, so the system will directly output the pair of bidders with an explicit association, as well as the explicit association between this pair of bidders, such as a controlling relationship, a guarantee relationship, a total score relationship, etc. Implicit associations refer to when any two bidders cannot determine the relationship between them through the information already obtained, and the system needs to infer and analyze the relationship between them through the information already obtained, or calculate the strength of the association between them. At this time, the system will obtain the associated feature information of the target bidder for feature engineering, such as operation information, historical bidding information, etc., divide the bidders into relationship circles and match relationship pairs, and use the relationship circle and relationship pair feature representation to calculate the association strength between the bidders that can be associated, and input it into the graph embedding algorithm for feature embedding. The feature embedding value obtained can then be used to calculate the association strength value between the bidder pairs through a distance algorithm or an unsupervised clustering algorithm, and output the association strength value of the two.

[0142] The present invention aims to invent a system that can identify the correlation between bidders to locate the risk of bid rigging based on the bidders' business registration data, historical bidding and historical operation data, combined with information matching rules and graph embedding algorithm models. The system obtains information on bidders in various dimensions, including corporate business information, historical bidding and historical operation information, and uses information matching rules and graph embedding algorithm models to build a model that explores the correlation strength and correlation between bidders. The information sources are diversified and cover various dimensions at the same time. This system automatically processes input information throughout the process and outputs the correlation relationship and correlation strength between bidders. The output results depend on information matching rules and the use of algorithm models, do not include manual intervention, avoid the risk of manual intervention, and enhance the objectivity of the system.

[0143] The present invention can be applied to the bidding process in the bidding system. When the system receives the list of bidders, it obtains information to make rule judgments and calculates the statistical algorithm to obtain the strength of any association between all bidders and outputs it. The whole process is online and carried out in the bidding process, which plays a role in timely preventing risks and preventing human intervention. By using a combination strategy, combining rules and algorithms to establish an association model for use in the bidding system, by exploring the explicit association between bidders participating in the bidding, and combining statistical and algorithmic models, the implicit association between bidders is calculated to locate and identify the risk of bid rigging in bidding.

[0144] When selecting and processing the features required to establish an implicit association relationship model between bidders, the present invention relies on the bidding risk scenario and the characteristics of the bidders, selects the corresponding algorithm, and combines the algorithms according to the data performance. According to the bidding scenario, the present invention introduces the characteristics of the bidder's historical bidding behavior, establishes associations with the bidders through this characteristic, and measures the association degree of the bidders under this characteristic through set operation rules; according to the characteristics of the bidders, the characteristics of the transport route and the transport task are introduced, and the bidders are associated through this characteristic, and the unique time characteristics of the transport route and the transport task are considered, that is, the influence of the historical data will slowly decay over time, and the physical law of Newton's cooling law is introduced to measure the effectiveness of the historical operation task on the current data, and the association degree of the bidders under this characteristic is measured in combination with the set operation rules; according to the characteristics of the bidders, the characteristics of the attached transport resources (drivers and vehicles) and contact information are introduced, and considering that in the real situation, the driver and the vehicle may not belong to a single enterprise, and the contact person may flow between various enterprises according to the change of personnel, so the TF-IDF algorithm is selected. This algorithm for measuring the importance of text for unstructured data is used to measure the association strength of the bidders for structured data, which not only considers the information overlap, but also converges the hot dimension information; preferably, the present invention can also use other algorithms with convergence to replace the TF-IDF algorithm for data calculation.

[0145] The system is divided into relationship circle characteristics and relationship pair characteristics according to the characteristics of the bidders. The characteristics of the bidders include both group characteristics and their own independent characteristics. The relationship circle characteristics take into account the relationship circle structure existing in the network, while the relationship pair characteristics focus on the exploration of the relationship between the bidder and another bidder. The system combines the two, taking into account both the relationship between users in the group and the interaction between target individuals. There are generally two or more bidders for bidding projects and operating routes, so they belong to the relationship circle attributes in this model, and the overlap of the transport resources and contact methods proves the interaction between the target individuals, so they are divided into relationship pair characteristics.

[0146] When establishing an implicit association relationship model between bidders and performing feature engineering on the processed features, the present invention selects corresponding algorithms based on bidding risk scenarios and bidder characteristics, and combines the algorithms based on data performance.

[0147] (1) This scenario is based on the identification of bid-rigging risks among bidders in the corporate bidding process. Confirmation of the risk identification results requires time to accumulate negative samples, so the sample labeling cost is too high to obtain. Therefore, an algorithm based on unlabeled samples for feature engineering and result prediction is selected, namely, the graph embedding algorithm and Euclidean distance calculation to obtain the results.

[0148] (2) This scenario is based on the identification of bid-rigging risks among bidders in the corporate bidding process. Some of the bidders may have no historical bidding information or historical operation information. Therefore, when a graph algorithm is used and there is no specific contact information between two bidders, the relationship between the two bidders can be inferred through the graph structure information.

[0149] (3) This system uses graph embedding algorithms to learn features when engineering the implicit association model between bidders. Traditional clustering algorithms only measure the strength of associations between bidders based on the performance of features between bidders when exploring the strength of associations. Graph algorithms take into account the structural characteristics of the group to which the bidder belongs, as well as the characteristics of all bidders associated with the bidder. Therefore, this algorithm can locate and identify the relationship between bidders in the group. At the same time, the Node2Vec algorithm can be selected to perform biased exploration of the homogeneity and isomorphism of the model, satisfying the exploration of associations and the exploration of the network structure of bidders in this scenario.

[0150] (4) This system uses a multi-dimensional graph embedding algorithm when using the Node2Vec algorithm to perform a graph embedding algorithm for implicit associations between bidders. Since features contain multiple dimensions, when performing a graph embedding algorithm, if the input needs to reduce multiple dimensions, however, reducing dimensions may lose information in the absence of positive and negative samples. Therefore, a separate graph embedding algorithm is selected for the features, and a multi-dimensional vector of bidders is finally obtained. This multi-dimensional vector can be used to calculate distances in a variety of ways, and then the association probability between bidders is output through normalization.

[0151] and Figure 1 Correspondingly, a project risk detection device provided by an embodiment of the present invention can be applied in a bidding system to support Figure 1 The method shown is a specific implementation.

[0152] Reference Figure 6, is a schematic diagram of the structure of a project risk detection device provided by an embodiment of the present invention, and is specifically described as follows:

[0153] The first determining unit 601 is used to determine each bidder participating in the current bidding project;

[0154] An acquisition unit 602 is used to acquire first characteristic data of each of the bidders in each preset first information dimension and second characteristic data in each preset second information dimension;

[0155] A second determining unit 603 is used to determine each bidder with an explicit relationship based on each of the first characteristic data, and determine explicit relationship information of each of the bidders with an explicit relationship, wherein the explicit relationship information includes each explicit association relationship of the bidders;

[0156] A third determining unit 604 is used to determine the implicit correlation between the bidders based on the second characteristic data of the bidders;

[0157] The generating unit 605 is used to generate the bid-rigging risk information of the bidding project based on the implicit correlation degree between each of the bidders and the explicit relationship information of each of the bidders with an explicit relationship.

[0158] In the device provided by the embodiment of the present invention, each bidder participating in the current bidding project is determined; the first characteristic data of each bidder in each preset first information dimension and the second characteristic data in each preset second information dimension are obtained; based on each first characteristic data, each bidder with an explicit relationship is determined, and the explicit relationship information of each bidder with an explicit relationship is determined, and the explicit relationship information includes each explicit association relationship of the bidder with an explicit relationship; based on each second characteristic data of each bidder, the implicit correlation between each bidder is determined; based on the implicit correlation between each bidder and the explicit relationship information of each bidder with an explicit relationship, the bid-rigging risk information of the bidding project is generated. The present invention deeply mines the explicit association relationship and implicit association relationship between each bidder, and generates the bid-rigging risk information based on the association relationship between each bidder, so that the enterprise that publishes the bid document can determine the bid-rigging risk of the current bidding project according to the bid-rigging risk information, so as to take corresponding measures to avoid the bid-rigging risk and safeguard its own interests.

[0159] In another embodiment provided by the present invention, the acquisition unit 602 of the device includes:

[0160] An acquisition subunit, used to acquire the enterprise identity information and enterprise operation information of each of the bidders;

[0161] A first processing subunit is used to process the corporate identity information of each of the bidders to obtain first characteristic data of each of the bidders in each preset first information dimension;

[0162] The second processing subunit is used to process the enterprise operation information of each of the bidders to obtain second characteristic data of each of the bidders in each preset second information dimension.

[0163] In another embodiment provided by the present invention, the second determining unit 603 of the device includes:

[0164] The analysis subunit is used to perform data correlation analysis on each of the first characteristic data of the bidder and the first characteristic data of the remaining bidders for each of the bidders to obtain the data analysis result of the bidder. When the data analysis result indicates that the bidder is associated with other bidders, the bidder is determined as a bidder with an explicit relationship.

[0165] In another embodiment provided by the present invention, the third determining unit 604 of the device includes:

[0166] A first generating subunit is configured to generate each relationship circle based on the historical bidding items and historical operation routes in each second characteristic data of each bidder, each relationship circle being a tender document relationship circle or a line relationship circle, the tender document relationship circle corresponding to a bidding item, the bidding item being present in the historical bidding items of each bidder in the tender document relationship circle, the line relationship circle corresponding to an operation route, the operation route being present in the historical operation routes of each bidder in the line relationship circle;

[0167] A first determining subunit, configured to determine a plurality of relationship pairs based on vehicle operation data in each second characteristic data of each of the bidders, wherein each of the relationship pairs has two bidders;

[0168] A construction subunit, used to construct an association relationship graph based on each of the relationship circles and each of the relationship pairs;

[0169] A second generating subunit is used to generate a bidding project correlation degree measurement matrix, an operation route correlation degree measurement matrix and a transport capacity overlap degree measurement matrix based on the correlation relationship map;

[0170] A third processing subunit is used to process the bidding project correlation degree measurement matrix, the operation route correlation degree measurement matrix and the transport capacity overlap degree measurement matrix to obtain a three-dimensional vector of each bidder;

[0171] The operation subunit is used to operate on the three-dimensional vector of each bidder to obtain the implicit correlation between each bidder.

[0172] In another embodiment provided by the present invention, the computing subunit of the device includes:

[0173] A calculation module, used for performing Euclidean operation on the three-dimensional vector of each bidder to obtain the Euclidean distance between each bidder;

[0174] The determination module is used to determine the implicit correlation degree of each of the bidders based on the Euclidean distance between each of the bidders.

[0175] In another embodiment provided by the present invention, the third processing subunit of the device includes:

[0176] A first acquisition module is used to acquire a bidder matrix and a bidder association matrix based on each bidder;

[0177] a combination module, for combining the bidder matrix, the bidder association matrix, the bidding project association degree measurement matrix, the operating route association degree measurement matrix and the transportation capacity overlap degree measurement matrix to obtain a first matrix combination, a second matrix combination and a third matrix combination, wherein the first matrix combination includes the bidder matrix, the bidder association matrix and the bidding project association degree measurement matrix, the second matrix combination includes the bidder matrix, the bidder association matrix and the operating route association degree measurement matrix, and the third matrix combination includes the bidder matrix, the bidder association matrix and the transportation capacity overlap degree measurement matrix;

[0178] An input module, used for inputting the first matrix combination, the second matrix combination and the third matrix combination into a preset algorithm model respectively, so that the algorithm model outputs a first dimension vector, a second dimension vector and a third dimension vector of each bidder;

[0179] The second acquisition module is used to obtain the three-dimensional vector of each bidder based on the first dimensional vector, the second dimensional vector and the third dimensional vector of each bidder.

[0180] The implementation contents of each unit, sub-unit and module in the device provided by the present invention can refer to the method example provided by the present invention, and the implementation contents of each unit, sub-unit and module in the device provided by the present invention will not be repeated here.

[0181] An embodiment of the present invention further provides a storage medium, which includes stored instructions, wherein when the instructions are executed, the device where the storage medium is located is controlled to execute the risk detection method of the above-mentioned project.

[0182] The embodiment of the present invention further provides an electronic device, the structural diagram of which is shown in FIG. Figure 7 As shown, it specifically includes a memory 701 and one or more instructions 702, wherein the one or more instructions 702 are stored in the memory 701 and are configured to be executed by one or more processors 703 to execute the one or more instructions 702 to execute the risk detection method of the above-mentioned project.

[0183] The specific implementation processes and derivative methods of the above-mentioned embodiments are all within the protection scope of the present invention.

[0184] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The system and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0185] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0186] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A project risk detection method, characterized in that: include: Identify the various bidders participating in the current bidding project; Acquire first characteristic data of each of the bidders in each preset first information dimension, and second characteristic data of each of the bidders in each preset second information dimension; Based on each of the first characteristic data, determine each bidder with an explicit relationship, and determine explicit relationship information of each of the bidders with an explicit relationship, wherein the explicit relationship information includes each explicit association relationship of the bidders; Determining the implicit correlation between the bidders based on the second characteristic data of the bidders; Based on the implicit correlation between each bidder and the explicit relationship information of each bidder with an explicit relationship, the bid-rigging risk information of the bidding project is generated.

2. The method according to claim 1, characterized in that The obtaining of the first characteristic data of each of the bidders in each preset first information dimension and the second characteristic data in each preset second information dimension includes: Obtaining corporate identity information and corporate operation information of each of the bidders; Processing the corporate identity information of each of the bidders to obtain first characteristic data of each of the bidders in each preset first information dimension; The enterprise operation information of each of the bidders is processed to obtain second characteristic data of each of the bidders in each preset second information dimension.

3. The method according to claim 1, characterized in that The step of determining each bidder having an explicit relationship based on each of the first characteristic data comprises: For each of the bidders, data correlation analysis is performed on each of the first characteristic data of the bidder and each of the first characteristic data of the remaining bidders to obtain the data analysis result of the bidder. When the data analysis result indicates that the bidder is associated with other bidders, the bidder is determined as a bidder with an explicit relationship.

4. The method according to claim 1, characterized in that: The determining the implicit correlation between the bidders based on the second characteristic data of the bidders includes: Based on the historical bidding items and historical operating routes in each second characteristic data of each bidder, each relationship circle is generated, each relationship circle is a tender document relationship circle or a line relationship circle, the tender document relationship circle corresponds to a bidding item, and the bidding item exists in the historical bidding items of each bidder in the tender document relationship circle, and the line relationship circle corresponds to an operating route, and the operating route exists in the historical operating routes of each bidder in the line relationship circle; Based on the vehicle operation data in each second characteristic data of each of the bidders, a plurality of relationship pairs are determined, each of the relationship pairs having two bidders; Based on each of the relationship circles and each of the relationship pairs, construct an association relationship graph; Based on the association relationship map, a bidding project association degree measurement matrix, an operation route association degree measurement matrix and a capacity overlap degree measurement matrix are generated; Processing the bidding project correlation degree measurement matrix, the operation route correlation degree measurement matrix and the transport capacity overlap degree measurement matrix to obtain a three-dimensional vector of each bidder; The three-dimensional vector of each of the bidders is operated to obtain the implicit correlation between each of the bidders.

5. The method according to claim 4, characterized in that The operation on the three-dimensional vector of each of the bidders to obtain the implicit correlation between each of the bidders includes: Performing Euclidean operation on the three-dimensional vector of each of the bidders to obtain the Euclidean distance between each of the bidders; Based on the Euclidean distance between each of the bidders, the implicit correlation degree of each of the bidders is determined.

6. The method according to claim 4, characterized in that The three-dimensional vector of each bidder is obtained by processing the bidding project correlation degree measurement matrix, the operation route correlation degree measurement matrix and the transport capacity overlap degree measurement matrix, including: Based on each bidder, obtain a bidder matrix and a bidder correlation matrix; The bidder matrix, the bidder association matrix, the bidding project association degree measurement matrix, the operating route association degree measurement matrix and the transportation capacity overlap degree measurement matrix are combined to obtain a first matrix combination, a second matrix combination and a third matrix combination, wherein the first matrix combination includes the bidder matrix, the bidder association matrix and the bidding project association degree measurement matrix, the second matrix combination includes the bidder matrix, the bidder association matrix and the operating route association degree measurement matrix, and the third matrix combination includes the bidder matrix, the bidder association matrix and the transportation capacity overlap degree measurement matrix; Inputting the first matrix combination, the second matrix combination and the third matrix combination into a preset algorithm model respectively, so that the algorithm model outputs a first dimension vector, a second dimension vector and a third dimension vector of each bidder; Based on the first dimension vector, the second dimension vector and the third dimension vector of each bidder, a three-dimensional vector of each bidder is obtained.

7. A project risk detection device, characterized in that: include: A first determining unit, used to determine each bidder participating in the current bidding project; An acquisition unit, used for acquiring first characteristic data of each of the bidders in each preset first information dimension, and second characteristic data of each of the bidders in each preset second information dimension; A second determining unit, configured to determine each bidder having an explicit relationship based on each of the first characteristic data, and determine explicit relationship information of each of the bidders having an explicit relationship, wherein the explicit relationship information includes each explicit association relationship of the bidders; A third determining unit, configured to determine the implicit correlation between the bidders based on the second characteristic data of the bidders; A generating unit is used to generate the bid-rigging risk information of the bidding project based on the implicit correlation degree between each of the bidders and the explicit relationship information of each of the bidders with an explicit relationship.

8. The device according to claim 7, characterized in that The second determining unit includes: The analysis subunit is used to perform data correlation analysis on each of the first characteristic data of the bidder and the first characteristic data of the remaining bidders for each of the bidders to obtain the data analysis result of the bidder. When the data analysis result indicates that the bidder is associated with other bidders, the bidder is determined as a bidder with an explicit relationship.

9. A storage medium, characterized in that: The storage medium includes stored instructions, wherein when the instructions are executed, the device where the storage medium is located is controlled to execute the risk detection method for the project according to any one of claims 1 to 6.

10. An electronic device, characterized in that: It comprises a memory and one or more instructions, wherein the one or more instructions are stored in the memory and are configured to be executed by one or more processors to perform the risk detection method for the project as described in any one of claims 1 to 6.

Images