Electronic bidding correlation degree analysis method and system based on Pearson's correlation coefficient

Abstract

The invention belongs to the technical field of electronic bidding, and particularly relates to an electronic bidding correlation degree analysis method and system based on a Pearson's correlation coefficient. The method comprises the following steps: collecting electronic bidding related contents of suppliers; automatically extracting supplier information features and bidding information features used for anti-cross-bidding analysis from the collected data through a program; analyzing association degrees of different bidding companies based on the extracted features, judging whether the two different bidding companies are associated or not, and obtaining an association degree analysis result; the obtained correlation degree analysis result is displayed in the form of a chart for viewing and reference; the system comprises a data collection module, a feature extraction module / correlation analysis module and an analysis result display module. According to the method, the bidding information of the bidding company in the bidding process can be comprehensively and deeply analyzed, and the efficiency of identifying the potential bidding behavior can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of electronic bidding, and in particular relates to an electronic bidding correlation analysis method and system based on the Pearson correlation coefficient. Background Art

[0002] In the process of electronic bidding, bid collusion between bidding companies has always been an important issue that affects the fairness of bidding and market competitiveness. Bid collusion usually refers to the bidding companies reaching a tacit understanding on prices through private negotiations, thereby manipulating the bidding results and damaging the interests of other bidders and the tenderer. In most cases, the bid-collusive units often make multiple modifications based on a basic quotation list to generate quotation lists of other bid-collusive units, so the correlation analysis of the quotation lists of different bidding companies can detect bid collusion to a certain extent. Existing anti-bid collusion methods mainly rely on manual comparison of bidding companies' quotation information, but these methods are often time-consuming and may have omissions, and are difficult to meet current needs. Summary of the invention

[0003] The present invention solves the problem that existing anti-coordinated bidding methods mainly rely on manual comparison of bidding companies' quotation information, but these methods are often time-consuming and may cause omissions, making it difficult to meet current needs. An electronic bidding correlation analysis method and system based on the Pearson correlation coefficient is provided to identify whether there is any bid-collusive behavior between two bidding companies.

[0004] The technical solution claimed in the present invention is as follows:

[0005] A method for analyzing the correlation of electronic bidding based on the Pearson correlation coefficient comprises the following steps:

[0006] S1: Data collection: Collect the relevant content of the supplier's electronic bidding; the relevant content includes: the IP address used by the bidding company when bidding, the bidding qualification documents and quotation documents and the bidding company name;

[0007] S2: Feature extraction: The supplier information features and bidding information features used in the anti-bid rigging analysis are automatically extracted from the data collected in S1 through a program;

[0008] S3: Correlation analysis: Analyze the correlation of different bidding companies based on the features extracted in S2, determine whether there is a correlation between two different bidding companies, and obtain the correlation analysis results;

[0009] S4: Display the correlation analysis results obtained in S3 in the form of a chart for reference.

[0010] Preferably, the data collected in S1 is derived from an electronic bidding system; when a bidding company quotes, the electronic bidding system records the IP address used by the bidding company.

[0011] In the above method, S2 specifically includes: extracting the IP addresses used by all bidding companies when bidding; extracting the legal person, registered address, registered email address and contact number of each bidding company according to the company name of the bidding company; extracting the creation program and creator of each file from the electronic qualification file and electronic quotation file of the bidding company; extracting the unit price of each quotation sub-item from the quotation file of the bidding company.

[0012] In the above method, step S3 is specifically as follows: for the IP address, company legal person, registered address, registered email address, contact number, the creation program and creator of each file, and the unit price of each quotation sub-item extracted in S2, a correlation analysis is performed: IP address analysis refers to comparing the IP addresses used by different bidding companies extracted. If the same IP address exists, the corresponding bidding companies are associated; company legal person analysis refers to comparing the legal persons of different bidding companies. If the same legal person exists among different bidding companies, the corresponding bidding companies are associated; company registered address analysis refers to comparing the registered addresses of different bidding companies. If different bidding companies have the same registered address or a similar address, the corresponding bidding companies are associated; Telephone number analysis refers to comparing the contact numbers of different bidding companies. If different bidding companies have the same registered telephone number, there is a correlation between the corresponding bidding companies; registered email address analysis refers to comparing the registered email addresses of different bidding companies. If different bidding companies have the same registered email address, there is a correlation between the corresponding bidding companies; the creation program and creator analysis of each file is to compare the creation program and the name of the creator of the bidding documents of different bidding companies. If different bidding companies have the same creation program or creation person, and the creation program or creation person is not the factory default and has obvious characteristics, there is a correlation; the unit price analysis of each quotation sub-item is to analyze the sub-item unit prices provided by different bidding companies to determine whether there is a correlation.

[0013] In the above method, the creation procedure and creator analysis of each file can only be used as a reference and cannot be used as an accurate basis for judging collusion.

[0014] In the above method, the unit price analysis of each quotation sub-item includes the following steps:

[0015] Step 1: Collect the sub-items common in the quotations of all bidding companies, extract the name and unit price of each sub-item, and store them in groups according to the bidding companies, which are called unit price groups;

[0016] Step 2: finding the intersection of the unit price groups of all bidding companies in step 1 according to the sub-item names, so that the number of unit prices in each unit price group is the same and has the same sub-item name; preferably, the value of the correlation coefficient is 0-1;

[0017] Step 3: For all the unit price groups of bidding companies, use the Pearson correlation coefficient algorithm to perform correlation coefficient analysis between them to obtain the correlation coefficient; the correlation coefficient represents the strength of the correlation between the two groups of data;

[0018] Step 4: If the correlation between the two groups of sub-items is less than the preset correlation coefficient, remove the points with larger deviations, continue the correlation coefficient analysis, and record the number of removed points and the correlation coefficient after removal each time, until the correlation coefficient reaches the preset correlation coefficient; preferably, the preset correlation coefficient is 1;

[0019] Step 5: According to the percentage of deviation points eliminated when the correlation coefficient in step 4 reaches the preset correlation coefficient, combined with the correlation coefficient, the degree of correlation between the quotations of the two bidding companies is judged and displayed; preferably, the degree of correlation is specifically: 0.8-1.0 is extremely correlated; 0.6-0.8 is strongly correlated; 0.4-0.6 is moderately correlated; 0.2-0.4 is weakly correlated; 0.0-0.2 is very weakly correlated or no correlation.

[0020] The present invention also provides an electronic bidding correlation analysis system based on the Pearson correlation coefficient, comprising a data collection module connected in sequence for collecting supplier electronic bidding related content, a feature extraction module for automatically extracting supplier information features and bidding information features used for anti-coercion analysis from the data collected by the data collection module through a program, a correlation analysis module for analyzing the correlation of different bidding companies based on the features extracted in the feature extraction module to determine whether two different bidding companies are associated and obtain correlation analysis results, and an analysis result display module for displaying the correlation analysis results obtained by the correlation analysis module in the form of a chart for viewing and reference; the related content includes: the IP address used by the bidding company when bidding, the bidding qualification documents and quotation documents and the name of the bidding company.

[0021] Beneficial effects:

[0022] The present invention provides an electronic bidding correlation analysis method and system based on the Pearson correlation coefficient, which not only considers the bid amount of the bidding company, but also combines the IP address used in the quotation, the company registration information of the bidding company, the quotation file information and other multi-source data, thereby improving the comprehensiveness and accuracy of the analysis; automatically extracting features related to the collusion behavior through the program, avoiding the cumbersomeness and subjectivity of manually extracting features in the traditional method, and solving the problem that the existing anti-collusion method mainly relies on manual comparison of the bidding company's quotation information, but these methods are often time-consuming and may be missed, and it is difficult to meet the current needs; combined with the electronic bidding system, the analysis results are displayed in a visual form, which is convenient for the tenderer and the regulatory department to intuitively understand the abnormal situation in the bidding process. In summary, through the bidding correlation analysis method and system provided by the present invention, it is possible to achieve a comprehensive and in-depth analysis of the bidding information of the bidding company in the bidding process, which can improve the efficiency of identifying potential collusion behaviors and ensure the fairness and market competitiveness of bidding. At the same time, the method can also provide powerful decision-making support for the tenderer and the regulatory department, and improve the efficiency and transparency of bidding work. The present invention also supports real-time comparative monitoring during the bidding process, and can timely discover and deal with potential collusion behaviors. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 The present invention is a flowchart of an electronic bidding correlation analysis method based on the Pearson correlation coefficient according to an embodiment of the present invention.

[0024] Figure 2 Schematic diagram of correlation coefficient analysis of two groups of data in an embodiment of the present invention.

[0025] Figure 3 It is a display diagram of the price correlation coefficient of the embodiment of the present invention; wherein: the percentage represents the percentage of the eliminated deviation points in the total number of sub-items, and the numbers in the table represent the correlation coefficients of the bid quotations of the bidding companies.

[0026] Figure 4 The figure is a schematic diagram of an electronic bidding correlation analysis system based on the Pearson correlation coefficient according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0028] First group of embodiments: Electronic bidding correlation analysis method based on Pearson correlation coefficient

[0029] This group of embodiments provides an electronic bidding correlation analysis method based on the Pearson correlation coefficient. Figure 1 As shown, the following steps are included:

[0030] S1: Collect data: Collect relevant content of suppliers' electronic bidding; the relevant content includes: the IP address used by the bidding company when bidding, the qualification documents and quotation documents of the bidding and the name of the bidding company; the collected data comes from the electronic bidding system; when the bidding company quotes, the electronic bidding system records the IP address used by the bidding company; the quotation document contains the quotation list of the bidding company; after the bidding company quotes, the electronic bidding system stores the electronic bidding documents uploaded by the bidding company when quoting (including but not limited to qualification documents and quotation documents).

[0031] S2: Feature extraction: The program automatically extracts the supplier information features and bidding information features used in anti-bid-rigging analysis from the data collected in S1, specifically: extract the IP addresses used by all bidding companies when bidding; extract the legal person, registered address, registered email address and contact number of each bidding company based on the company name of the bidding company; extract the creation program and creator of each file from the bidding company's electronic qualification file and electronic quotation file; extract the unit price of each quotation sub-item from the bidding company's quotation file.

[0032] In a specific embodiment of the present invention, based on the name of the bidding company, the company's legal person, registered address, email address and contact number are obtained from the Tianyancha platform (including but not limited to the platform); from the bidding documents uploaded by the bidding company, the file creation program (such as a printer that scans and generates a PDF file, etc.) and the creator of the file (the user name of the Office software used, etc.) are extracted; from the quotation documents uploaded by the bidding company, the quotation amount is extracted and stored in groups.

[0033] S3: Correlation analysis: Based on the features extracted in S2, the correlation of different bidding companies is analyzed to determine whether there is any correlation between two different bidding companies, and the correlation analysis results are obtained. The bid collusion analysis mainly includes seven aspects: IP address, company legal person (legal person of the bidding company), registered address (registered address of the bidding company), registered email (registered email of the bidding company), contact number (contact number of the bidding company), the creation program and creator of each file, and the unit price of each quotation sub-item. Among them: IP address analysis refers to the comparison of the IP addresses used by different bidding companies extracted. If the same IP address exists, the corresponding bidding companies are associated; company legal person analysis refers to the comparison of legal persons of different bidding companies. If there are the same legal persons of different bidding companies, the corresponding bidding companies are associated; company registered address analysis refers to the comparison of registered addresses of different bidding companies. If there are different bidding companies with the same registered addresses or similar addresses, the corresponding bidding companies are associated; contact number analysis refers to the comparison of contact numbers of different bidding companies. If there are different bidding companies with the same registered telephone numbers, the corresponding bidding companies are associated; registered email analysis is Refers to comparing the registered email addresses of different bidding companies. If different bidding companies have the same registered email addresses, there is a correlation between the corresponding bidding companies. The creation program and creator analysis of each file is to compare the creation programs and creator names of the bidding documents of different bidding companies. If different bidding companies have the same creation program or creator, and the creation program or creator is not the factory default and has obvious characteristics, there is a correlation. This analysis can only be used as a reference and cannot be used as an accurate basis for judging collusion, because printers of the same model may have the same default name. The unit price analysis of each quotation sub-item is to analyze the sub-item unit prices provided by different bidding companies to determine whether there is a correlation.

[0034] The unit price analysis of each quotation sub-item includes the following steps:

[0035] Step 1: Collect the sub-items common in the quotations of all bidding companies, extract the name and unit price of each sub-item, and store them in groups according to the bidding companies, which are called unit price groups;

[0036] Step 2: Find the intersection of the unit price groups of all bidding companies in step 1 according to the sub-item name, so that the number of unit prices in each unit price group is the same and has the same sub-item name; the value of the correlation coefficient is 0-1;

[0037] Step 3: For all the unit price groups of bidding companies, use the Pearson correlation coefficient algorithm to perform correlation coefficient analysis between them to obtain the correlation coefficient; the correlation coefficient represents the strength of the correlation between the two groups of data;

[0038] Step 4: If the correlation between the two groups of sub-items is less than the preset correlation coefficient, remove the points with large deviations, continue the correlation coefficient analysis, and record the number of removed points and the correlation coefficient after removal each time, until the correlation coefficient is the preset correlation coefficient ( Figure 2 ); In some embodiments of the present invention, the preset correlation coefficient is 1; In other embodiments of the present invention, the preset correlation coefficient can be set according to requirements;

[0039] Step 5: According to the percentage of deviation points eliminated when the correlation coefficient in step 4 reaches the preset correlation coefficient, combined with the correlation coefficient, the degree of correlation between the quotations of the two bidding companies is judged and displayed; the specific degree of correlation is: 0.8-1.0 is extremely correlated; 0.6-0.8 is strongly correlated; 0.4-0.6 is moderately correlated; 0.2-0.4 is weakly correlated; 0.0-0.2 is very weakly correlated or no correlation.

[0040] S4: Display the correlation analysis results obtained in S3 in the form of a chart for reference; in a specific embodiment of the present invention, Figure 3 The results of the bid price analysis are displayed to facilitate the review and reference of tendering personnel or supervisory personnel.

[0041] Second group of embodiments: Electronic bidding correlation analysis system based on Pearson correlation coefficient

[0042] This group of embodiments provides an electronic bidding correlation analysis system based on the Pearson correlation coefficient. Figure 4 As shown, it includes a data collection module for collecting supplier electronic bidding related content, a feature extraction module for automatically extracting supplier information features and bidding information features used for anti-coercion analysis from the data collected by the data collection module through a program, a correlation analysis module for analyzing the correlation between different bidding companies based on the features extracted in the feature extraction module to determine whether there is a correlation between two different bidding companies and obtain the correlation analysis results, and an analysis result display module for displaying the correlation analysis results obtained by the correlation analysis module in the form of a chart for viewing and reference; the relevant content includes: the IP address used by the bidding company when bidding, the bidding qualification documents and quotation documents and the name of the bidding company.

[0043] It should be noted that the above-described specific implementations can enable those skilled in the art to more fully understand the invention, but do not limit the invention in any way. Therefore, although this specification has described the invention in detail with reference to the drawings and embodiments, those skilled in the art should understand that the invention can still be modified or replaced by equivalents. In short, all technical solutions and improvements that do not deviate from the spirit and scope of the invention should be included in the protection scope of the patent for the invention.

Claims

1. A method for analyzing the correlation between electronic bidding based on the Pearson correlation coefficient, characterized in that: The steps include: S1: Data collection: Collect the relevant content of the supplier's electronic bidding; the relevant content includes: the IP address used by the bidding company when bidding, the bidding qualification documents and quotation documents and the bidding company name; S2: Feature extraction: The supplier information features and bidding information features used in the anti-bid rigging analysis are automatically extracted from the data collected in S1 through a program; S3: Correlation analysis: Analyze the correlation of different bidding companies based on the features extracted in S2, determine whether there is a correlation between two different bidding companies, and obtain the correlation analysis results; S4: Display the correlation analysis results obtained in S3 in the form of a chart for reference.

2. The method for analyzing the correlation between electronic bidding and the Pearson correlation coefficient according to claim 1 is characterized in that: The data collected in S1 comes from the electronic bidding system; when bidding companies submit their bids, the electronic bidding system records the IP addresses used by the bidding companies.

3. The method for analyzing the correlation between electronic bidding and the Pearson correlation coefficient according to claim 1 is characterized in that: S2 specifically includes: extracting the IP addresses used by all bidding companies when bidding; extracting the legal person, registered address, registered email address and contact number of each bidding company based on the company name of the bidding company; extracting the creation program and creator of each file from the bidding company's electronic qualification files and electronic quotation files; extracting the unit price of each quotation sub-item from the bidding company's quotation file.

4. The method for analyzing the correlation between electronic bidding and the Pearson correlation coefficient according to claim 3 is characterized in that: Step S3 is specifically as follows: for the IP address, company legal person, registered address, registered email address, contact number, the creation program and creator of each file, and the unit price of each quotation sub-item extracted in S2, a correlation analysis is performed: IP address analysis refers to comparing the IP addresses used by different bidding companies extracted. If the same IP address exists, the corresponding bidding companies are associated; company legal person analysis refers to comparing the legal persons of different bidding companies. If the same legal person exists in different bidding companies, the corresponding bidding companies are associated; company registered address analysis refers to comparing the registered addresses of different bidding companies. If different bidding companies have the same registered address or a similar address, the corresponding bidding companies are associated; Contact phone number analysis refers to comparing the contact numbers of different bidding companies. If there are different bidding companies with the same registered phone number, then the corresponding bidding companies are related; Registration email analysis refers to comparing the registration emails of different bidding companies. If different bidding companies have the same registration emails, there is a correlation between the corresponding bidding companies. The creation program and creator analysis of each file refers to comparing the creation programs and creator names of the bidding documents of different bidding companies. If different bidding companies have the same creation programs or creators, and the creation programs or creators are not factory defaults and have obvious characteristics, there is a correlation. The unit price analysis of each quotation sub-item refers to analyzing the sub-item unit prices provided by different bidding companies to determine whether there is a correlation.

5. The method for analyzing the correlation between electronic bidding and the Pearson correlation coefficient according to claim 4 is characterized in that: The creation procedure and creator analysis of each file can only be used as a reference and cannot be used as an accurate basis for judging collusion.

6. The method for analyzing the correlation between electronic bidding and the Pearson correlation coefficient according to claim 5 is characterized in that: The unit price analysis of each quotation sub-item includes the following steps: Step 1: Collect the sub-items common in the quotations of all bidding companies, extract the name and unit price of each sub-item, and store them in groups by bidding company, which is called unit price group; Step 2: Find the intersection of the unit price groups of all bidding companies in step 1 by sub-item name, so that each unit price group has the same number of unit prices and the same sub-item name; Step 3: For all the unit price groups of bidding companies, use the Pearson correlation coefficient algorithm to perform correlation coefficient analysis between them to obtain the correlation coefficient; the correlation coefficient represents the strength of the correlation between the two groups of data; Step 4: If the correlation between the two groups of sub-items is less than the preset correlation coefficient, remove the points with large deviations, continue the correlation coefficient analysis, and record the number of removed points and the correlation coefficient after removal each time until the correlation coefficient reaches the preset correlation coefficient; Step 5: According to the percentage of deviation points eliminated when the correlation coefficient in step 4 reaches the preset correlation coefficient, combined with the correlation coefficient, the degree of correlation between the quotations of the two bidding companies is judged and displayed.

7. The method for analyzing the correlation between electronic bidding and the Pearson correlation coefficient according to claim 1 is characterized in that: The value of the correlation coefficient in step 2 is 0-1.

8. The method for analyzing the correlation between electronic bidding and the Pearson correlation coefficient according to claim 7 is characterized in that: The specific correlation degree in step 5 is: 0.8-1.0 is extremely correlated; 0.6-0.8 is strongly correlated; 0.4-0.6 is moderately correlated; 0.2-0.4 is weakly correlated; 0.0-0.2 indicates very weak correlation or no correlation.

9. The method for analyzing the correlation between electronic bidding and the Pearson correlation coefficient according to claim 7 is characterized in that: The preset correlation coefficient in step 4 is 1.

10. An electronic bidding correlation analysis system based on Pearson correlation coefficient, characterized in that: It includes a data collection module for collecting supplier electronic bidding related content, a feature extraction module for automatically extracting supplier information features and bidding information features used for anti-coercive bidding analysis from the data collected by the data collection module through a program, a correlation analysis module for analyzing the correlation between different bidding companies based on the features extracted in the feature extraction module to determine whether two different bidding companies are related and obtain correlation analysis results, and an analysis result display module for displaying the correlation analysis results obtained by the correlation analysis module in the form of a chart for viewing and reference; The relevant content includes: the IP address used by the bidding company when bidding, the bidding qualification documents and quotation documents and the name of the bidding company.

Images