Purchase project auditing method and device based on large model, medium and equipment

Abstract

The invention provides a purchase project auditing method and device based on a large model, a medium and equipment. The method comprises the following steps: performing information extraction on bidding and tendering related files of a current purchase project by adopting an information extraction model, and storing the extracted information in a database; a similarity model is adopted to retrieve similar historical purchase items of the current purchase item from a file library, and the current purchase item and the similar historical purchase items are compared on at least one element; according to each item of information extracted for the current purchase project, analyzing whether the bid opening process of the current purchase project has a mixed bid problem and an accompanying bid problem; according to the information extracted from the bid winning notice in the bidding and tendering related file, the scoring condition of an expert judge in the bidding process is analyzed; and generating a full life cycle analysis report of the current purchase project. According to the invention, the processing efficiency can be improved, and comprehensive deep auditing of the purchase project is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of procurement project auditing, and in particular to a procurement project auditing method, device, medium and equipment based on a large model. Background Art

[0002] Procurement project audit is an important part of ensuring the efficiency of fund use and the fairness of public resource allocation. However, traditional audit methods often have problems such as low processing efficiency and insufficient analysis depth when faced with massive data, complex business logic and changing market environment. Summary of the invention

[0003] In response to at least one of the above technical problems, an embodiment of the present invention provides a procurement project audit method, device, medium, and equipment based on a large model.

[0004] According to the first aspect, the procurement project audit method based on the big model provided by the embodiment of the present invention includes:

[0005] Use the information extraction model to extract information from the bidding documents of the current procurement project, and store the extracted information in the database;

[0006] According to the information extracted from the current procurement project, similar historical procurement projects of the current procurement project are retrieved from the document library using a similarity model, and the current procurement project is compared with the similar historical procurement projects in at least one element;

[0007] Analyze whether there are problems of collusion and bid spoofing in the bidding process of the current procurement project based on the information extracted from the current procurement project;

[0008] Analyze the scores of expert judges during the bid evaluation process based on the information extracted from the winning notice in the bidding documents;

[0009] Analyze whether there are any abnormalities in the execution of the contract based on the agreed information in the contract documents in the bidding documents and the actual progress of the current procurement project;

[0010] A full life cycle analysis report for the current procurement project is generated based on the information extracted from the current procurement project, the comparison results between the current procurement project and similar historical procurement projects in at least one element, the analysis results of the bid opening process, the analysis results of the bid evaluation process and the analysis results of the contract execution.

[0011] In one embodiment, the extracting information from the bidding documents of the current procurement project using the information extraction model includes:

[0012] Different information extraction models are used to extract information from different bidding-related documents; different information extraction models need to extract different information items;

[0013] The bidding-related documents include bidding documents, tender documents, notice of winning, notice of change, contract notice and contract documents;

[0014] The information items to be extracted by the information extraction model for extracting information from the bidding document include at least one of the basic information of the project, the construction content, the purchaser information and the bidding time information; the basic information of the project includes at least one of the project number, the project name, the procurement method, the budget amount and the maximum price limit;

[0015] The information extraction model for extracting information from the bidding document needs to extract at least one of the following information items: project number, quotation, supplier information, budget, construction period and technical standards;

[0016] The information extraction model for extracting information from the bid winning notice needs to extract at least one of the following information items: project number, bid winning supplier name, bid winning supplier address, bid winning price, review expert information, and bid winning supplier score;

[0017] The information extraction model for extracting information from the change notice needs to extract information items including procurement project change information and project number;

[0018] The information extraction model for extracting information from the contract announcement needs to extract at least one of the following information items: project number, contract number, name, procurement item, signing time and announcement time;

[0019] The information extraction model used to extract information from the contract document needs to extract at least one of the names, addresses, contact information, contract signing date and contract term, contract subject, price and payment terms, performance period, place and method, and liability for breach of contract of the two parties to the contract.

[0020] In one embodiment, the extracting information from the bidding documents of the current procurement project using the information extraction model includes:

[0021] Determine whether the file type of each bidding-related document is a scanned PDF;

[0022] If yes, then use optical character recognition technology to perform text recognition on the bidding related document to obtain a text file corresponding to the bidding related document;

[0023] Determine whether the number of text words in the text file is greater than the maximum limit of the number of input words for the information extraction model of the bidding related document;

[0024] If it is greater than the maximum limit, the text file is segmented using a text paragraph segmentation model to obtain multiple paragraphs, and each paragraph is input into the information extraction model of the bidding-related file to obtain extracted information;

[0025] If it is less than or equal to the maximum limit, the text file is input into the information extraction model of the bidding related documents to obtain the extracted information.

[0026] In one embodiment, comparing the current procurement item with the similar historical procurement item on at least one element includes:

[0027] The current procurement project is compared with the similar historical procurement projects in at least one element of procurement content, procurement method and budget amount.

[0028] In one embodiment, analyzing whether there are problems of collusion and bid spoofing in the bidding process of the current procurement project includes:

[0029] Determine whether the supplier's bid price meets the preset linear rule, and determine whether there is a problem of collusion based on the judgment result; and, based on the supplier's historical bidding data and corporate equity penetration information, determine whether the supplier has a problem of accompanying bidding.

[0030] In one embodiment, analyzing the scores of the expert judges during the evaluation process includes calculating at least one of the range, variance, and interquartile range of the scores of the expert judges during the evaluation process.

[0031] In one embodiment, the method further comprises at least one of the following:

[0032] Check whether the bidding agency for the current procurement project has legal qualifications;

[0033] Testing the legality of enterprises during the bidding process;

[0034] Determine the information extracted from historical procurement projects to determine whether there are any abnormal suppliers.

[0035] According to the second aspect, the procurement project audit device based on the big model provided by the embodiment of the present invention includes:

[0036] The information extraction module is used to extract information from the bidding documents related to the current procurement project using the information extraction model, and store the extracted information in the database;

[0037] A project comparison module, configured to retrieve similar historical procurement projects of the current procurement project from a document library using a similarity model according to various information extracted from the current procurement project, and compare the current procurement project with the similar historical procurement project in at least one element;

[0038] A bid opening analysis module, used to analyze whether there are problems of collusion and bid spoofing in the bid opening process of the current procurement project according to various information extracted from the current procurement project;

[0039] The bid evaluation analysis module is used to analyze the scores of the expert judges in the bid evaluation process based on the information extracted from the winning notice in the bidding related documents;

[0040] An execution analysis module is used to analyze whether there are abnormalities in the execution of the contract based on the agreed information in the contract documents in the bidding related documents and the actual promotion of the current procurement project;

[0041] The report generation module is used to generate a full life cycle analysis report for the current procurement project based on the information extracted from the current procurement project, the comparison results between the current procurement project and similar historical procurement projects in at least one element, the analysis results of the bid opening process, the analysis results of the bid evaluation process and the analysis results of the contract execution.

[0042] In one embodiment, the information extraction module is specifically used to: use different information extraction models to extract information from different bidding related documents; wherein different information extraction models need to extract different information items; wherein the bidding related documents include bidding documents, tender documents, notice of winning, change notice, contract notice and contract documents; the information extraction model used to extract information from the bidding documents needs to extract at least one of the following: basic information of the project, construction content, purchaser information and bidding time information; the basic information of the project includes at least one of the following: project number, project name, procurement method, budget amount and maximum price; the information extraction model used to extract information from the tender documents needs to extract at least one of the following: project number, quotation, supplier information, budget, construction period and technical standards Items; the information extraction model used to extract information from the notice of successful bid includes at least one of the project number, the name of the successful bidder, the address of the successful bidder, the successful bid price, the evaluation expert information and the successful bidder score; the information extraction model used to extract information from the change announcement includes procurement project change information and the project number; the information extraction model used to extract information from the contract announcement includes at least one of the project number, contract number, name, procurement project, signing time and announcement time; the information extraction model used to extract information from the contract document includes at least one of the name, address, contact information, contract signing date and contract term, contract subject, price and payment terms, performance period, place and method and liability for breach of contract.

[0043] In one embodiment, the information extraction module is specifically used to: determine whether the file type of each bidding related document is a scanned version of PDF; if so, use optical character recognition technology to perform text recognition on the bidding related document to obtain a text file corresponding to the bidding related document; determine whether the number of text words in the text file is greater than the maximum limit on the number of input words of the information extraction model of the bidding related document; if it is greater than the maximum limit, use a text paragraph segmentation model to segment the text file into paragraphs to obtain multiple paragraphs, and input each paragraph into the information extraction model of the bidding related document to obtain extracted information; if it is less than or equal to the maximum limit, input the text file into the information extraction model of the bidding related document to obtain extracted information.

[0044] In one embodiment, the project comparison module is specifically used to compare the current procurement project with the similar historical procurement projects in at least one element of procurement content, procurement method and budget amount.

[0045] In one embodiment, the bid opening analysis module is specifically used to: determine whether the supplier's bid price meets the preset linear law, and determine whether there is a problem of collusion in bidding based on the judgment result; and determine whether the supplier has a problem of accompanying bidding based on the supplier's historical bidding data and corporate equity penetration information.

[0046] In one embodiment, the evaluation criteria analysis module is specifically used to calculate at least one of the range, variance and interquartile range of scores given by expert judges during the evaluation criteria process.

[0047] In one embodiment, the device may further include:

[0048] The first detection module is used to perform at least one of the following: check whether the bidding agency of the current procurement project has legal qualifications; check the legality of the enterprise during the bid opening process; determine the information extracted from historical procurement projects to determine whether there are abnormal suppliers.

[0049] According to a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, enables the computer to execute the method provided in the first aspect.

[0050] According to a fourth aspect, an embodiment of the present invention provides a computing device, including a memory and a processor, wherein the memory stores executable code, and when the processor executes the executable code, the method provided in the first aspect is implemented.

[0051] The procurement project audit method, device, medium, and equipment based on a large model provided by the embodiment of the present invention use an information extraction model to extract information from the bidding documents related to the current procurement project, and store the extracted information in a database, and then use a similarity model to retrieve similar historical procurement projects of the current procurement project from the file library, and compare the current procurement project with the similar historical procurement project. Then, based on the extracted information, analyze whether there are collusion and accompanying bidding problems in the bidding process of the current procurement project, the scoring of expert judges in the evaluation process, and whether there are abnormal contract execution situations, and then form a full life cycle analysis report of the current procurement project with the extracted information and the analysis of each stage. It can be seen that the embodiment of the present invention realizes procurement audit analysis based on a large model. Since the large model can process massive amounts of data and automatically perform many tedious data analysis and comparison tasks, it can significantly reduce the manual workload and time consumption of auditors, and can quickly identify anomalies and potential risks in procurement data, helping auditors to quickly lock in audit priorities. Compared with traditional auditing methods, the embodiments of the present invention can improve processing efficiency and achieve comprehensive and in-depth auditing of procurement projects when faced with massive data, complex business logic and changing market environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flowchart of a procurement project audit method based on a large model in one embodiment of the present invention;

[0053] Figure 2 The present invention is a structural block diagram of a procurement project audit device based on a large model in one embodiment of the present invention. DETAILED DESCRIPTION

[0054] In a first aspect, an embodiment of the present invention provides a procurement project audit method based on a large model, see Figure 1 The method comprises the following steps S110 to S160:

[0055] S110, extracting information from the bidding documents related to the current procurement project using an information extraction model, and storing the extracted information in a database;

[0056] In one embodiment, extracting information from the bidding documents related to the current procurement project using the information extraction model in S110 may include:

[0057] Different information extraction models are used to extract information from different bidding-related documents; different information extraction models need to extract different information items;

[0058] The bidding related documents may include bidding documents, tender documents, notice of winning the bid, notice of change, contract notice and contract documents;

[0059] The information items to be extracted by the information extraction model for extracting information from the bidding document include at least one of the basic information of the project, the construction content, the purchaser information and the bidding time information; the basic information of the project includes at least one of the project number, the project name, the procurement method, the budget amount and the maximum price limit;

[0060] The information extraction model for extracting information from the bidding document needs to extract at least one of the following information items: project number, quotation, supplier information, budget, construction period and technical standards;

[0061] The information extraction model for extracting information from the bid winning notice needs to extract at least one of the following information items: project number, bid winning supplier name, bid winning supplier address, bid winning price, review expert information, and bid winning supplier score;

[0062] The information extraction model for extracting information from the change notice needs to extract information items including procurement project change information and project number;

[0063] The information extraction model for extracting information from the contract announcement needs to extract at least one of the following information items: project number, contract number, name, procurement item, signing time and announcement time;

[0064] The information extraction model used to extract information from the contract document needs to extract at least one of the names, addresses, contact information, contract signing date and contract term, contract subject, price and payment terms, performance period, place and method, and liability for breach of contract of the two parties to the contract.

[0065] Among them, the purchaser information may include name, address, contact person, contact information, etc.

[0066] Among them, the bidding time information may include the bidding announcement time, the bidding document acquisition time, the deadline for submitting bidding documents, the bid opening time, etc.

[0067] In one embodiment, the information extraction model used in S110 to extract information from the bidding documents related to the current procurement project may include the following five steps:

[0068] 1. Determine whether the file type of each bidding-related document is a scanned PDF;

[0069] 2. If yes, then use optical character recognition technology to perform text recognition on the bidding related document to obtain a text file corresponding to the bidding related document;

[0070] 3. Determine whether the number of words in the text file is greater than the maximum limit on the number of input words of the information extraction model of the bidding-related document;

[0071] 4. If it is greater than the maximum limit, the text file is segmented using a text paragraph segmentation model to obtain multiple paragraphs, and each paragraph is input into the information extraction model of the bidding-related document to obtain the extracted information;

[0072] 5. If it is less than or equal to the maximum limit, the text file is input into the information extraction model of the bidding related documents to obtain the extracted information.

[0073] It can be seen that if the file type of a bidding-related document is a scanned PDF, the OCR technology is used to perform text recognition on the bidding-related document to obtain the text file corresponding to the bidding-related document. Then, it is determined whether the number of text words in the text file exceeds the maximum limit of the number of input words of the information extraction model of the bidding-related document. If it exceeds, the text paragraph segmentation model is used to segment the text file to obtain multiple paragraphs, and each paragraph is input into the information extraction model of the bidding-related document to obtain the extracted information. If it does not exceed, the text file is directly input into the information extraction model of the bidding-related document to obtain the extracted information.

[0074] Among them, the text word is the Token.

[0075] Among them, the text paragraph segmentation model can use SeqModel, BERT, etc. BERT can encode multiple sentences at the same time, model the dependency relationship between longer contexts, calculate the sentence vector, and finally predict whether text segmentation is performed after each sentence.

[0076] Among them, the information extraction model can adopt the prompt model. Use the information extraction model with large parameter quantity as much as possible, so that more tokens can be input at one time, ensuring the extraction effect and converting unstructured text into structured data.

[0077] S120, based on the information extracted from the current procurement project, using a similarity model to retrieve similar historical procurement projects of the current procurement project from a document library, and comparing the current procurement project with the similar historical procurement projects in at least one element;

[0078] It can be seen that S120 searches the document library for historical procurement projects whose similarity with the current procurement project is higher than a certain value based on the extracted information and similarity model, and then compares the current procurement project with similar historical procurement projects in at least one factor, thereby obtaining a multi-dimensional comparison result.

[0079] That is, by using the large model text similarity calculation capability, the extracted construction content is compared with the construction content of historical procurement projects, and historical procurement projects with similar construction content are output for auditors’ reference.

[0080] In one embodiment, comparing the current procurement item with the similar historical procurement item on at least one element in S120 may include:

[0081] The current procurement project is compared with the similar historical procurement projects in at least one element of procurement content, procurement method and budget amount.

[0082] It can be seen that a comparison is made on at least one element among the procurement content, procurement method and budget amount, thereby obtaining a comparison result on at least one element among the procurement content, procurement method and budget amount.

[0083] S130, analyzing whether there are problems of collusion and bid spoofing in the bidding process of the current procurement project according to various information extracted from the current procurement project;

[0084] In one embodiment, the analyzing in S130 whether there is a problem of collusion and bid spoofing in the bidding process of the current procurement project may include:

[0085] Determine whether the supplier's bid price meets the preset linear rule, and determine whether there is a problem of collusion based on the judgment result; and, based on the supplier's historical bidding data and corporate equity penetration information, determine whether the supplier has a problem of accompanying bidding.

[0086] In other words, we can determine whether there is a bidding collusion phenomenon based on whether the supplier's bid price meets a certain linear rule. Combining the supplier's historical bidding data and the company's equity penetration information, we can determine whether the supplier has a bidding collusion phenomenon.

[0087] It is understandable that S130 is an analysis of the bid opening process.

[0088] S140, analyzing the scores of the expert judges in the bid evaluation process according to the information extracted from the winning notice in the bidding related documents;

[0089] In one embodiment, analyzing the expert judges' scores during the evaluation process in S140 may include calculating at least one of the range, variance, and interquartile range of the expert judges' scores during the evaluation process.

[0090] It can be understood that S140 is an analysis of the evaluation process.

[0091] S150, analyzing whether there are any abnormalities in the execution of the contract based on the agreed information in the contract documents in the bidding related documents and the actual progress of the current procurement project;

[0092] Specifically, based on the key information extracted from the contract-related documents and the actual project promotion data, a timeline can be drawn to determine whether there are any abnormalities in the contract execution. For example, whether the supplier carries out project construction and delivery in accordance with the time and requirements stipulated in the contract. By comparing the time of project payment, bid opening date, and contract signing date, it can be determined whether there are problems such as suppliers entering the site early or paying before the tender is accepted.

[0093] It can be seen that S150 is an analysis of the contract execution status.

[0094] S160. Generate a full life cycle analysis report for the current procurement project based on various information extracted from the current procurement project, the comparison results between the current procurement project and similar historical procurement projects in at least one element, the analysis results of the bid opening process, the analysis results of the bid evaluation process, and the analysis results of the contract execution.

[0095] In other words, the key information obtained from the above steps and the analysis data from each stage are formed into a full life cycle analysis report for the current procurement project.

[0096] In one embodiment, the method may further include at least one of the following:

[0097] (1) Check whether the bidding agency for the current procurement project has legal qualifications;

[0098] (2) Testing the legality of enterprises during the bidding process;

[0099] (3) Determine the information extracted from historical procurement projects and determine whether there are any abnormal suppliers.

[0100] Among them, in (2), the legality of the enterprise can be tested by checking whether the enterprise has a major illegal record in the past three years, whether it is included in the list of serious illegal and dishonest behavior records in government procurement, the list of dishonest persons subject to execution, the list of parties involved in major tax law violations, etc.

[0101] Among them, in (3), by analyzing the relationship network between suppliers, auditors can also discover potential collusion, frequent cooperation, binding bidding and other abnormal phenomena.

[0102] In actual scenarios, the big model can also be used to extract and list the main information of the bid winning notice to determine whether the name of the winning bidder is consistent with the winning bidder information. In the bid evaluation process, the composition of the bid evaluation experts can also be listed, such as the expert's name, work unit, and number of people on the bid evaluation committee.

[0103] In actual scenarios, it is also possible to build a knowledge graph through the big model's capabilities of entity object extraction, entity attribute extraction, entity relationship extraction, knowledge fusion and disambiguation, to assist in analyzing the correlation between bidding company personnel, projects, etc.

[0104] In the embodiment of the present invention, a large model analysis technology is introduced. The large model analysis technology is a method based on big data and machine learning algorithms to deeply mine and analyze massive data by building a large-scale model. It uses advanced data processing technology and algorithms to extract valuable information and patterns from complex data, providing a scientific basis for decision-making.

[0105] It can be seen that the embodiment of the present invention implements procurement audit analysis based on the big model. Since the big model can process massive amounts of data and automatically perform many tedious data analysis and comparison tasks, it significantly reduces the manual workload and time consumption of auditors, and can quickly identify anomalies and potential risks in procurement data, helping auditors quickly lock in audit focus. Compared with traditional audit methods, the embodiment of the present invention can improve processing efficiency when facing massive amounts of data, complex business logic and changing market environments, and achieve a comprehensive and in-depth audit of procurement projects.

[0106] In a second aspect, an embodiment of the present invention provides a procurement project audit device based on a large model, see Figure 2 , the device 100 comprises:

[0107] The information extraction module 110 is used to extract information from the bidding documents related to the current procurement project using the information extraction model, and store the extracted information in the database;

[0108] A project comparison module 120 is used to retrieve similar historical procurement projects of the current procurement project from a document library using a similarity model based on various information extracted from the current procurement project, and compare the current procurement project with the similar historical procurement project in at least one element;

[0109] The bid opening analysis module 130 is used to analyze whether there are problems of collusion and bid spoofing in the bid opening process of the current procurement project according to various information extracted from the current procurement project;

[0110] The bid evaluation analysis module 140 is used to analyze the scores of the expert judges in the bid evaluation process according to the information extracted from the bid winning notice in the bidding related documents;

[0111] An execution analysis module 150 is used to analyze whether there is any abnormality in the execution of the contract based on the agreed information in the contract document in the bidding related documents and the actual promotion of the current procurement project;

[0112] The report generation module 160 is used to generate a full life cycle analysis report for the current procurement project based on the information extracted from the current procurement project, the comparison results between the current procurement project and the similar historical procurement projects in at least one element, the analysis results of the bid opening process, the analysis results of the bid evaluation process and the analysis results of the contract execution.

[0113] In one embodiment, the information extraction module is specifically used to: use different information extraction models to extract information from different bidding related documents; wherein different information extraction models need to extract different information items; wherein the bidding related documents include bidding documents, tender documents, notice of winning, change notice, contract notice and contract documents; the information extraction model used to extract information from the bidding documents needs to extract at least one of the following: basic information of the project, construction content, purchaser information and bidding time information; the basic information of the project includes at least one of the following: project number, project name, procurement method, budget amount and maximum price; the information extraction model used to extract information from the tender documents needs to extract at least one of the following: project number, quotation, supplier information, budget, construction period and technical standards Items; the information extraction model used to extract information from the notice of successful bid includes at least one of the project number, the name of the successful bidder, the address of the successful bidder, the successful bid price, the evaluation expert information and the successful bidder score; the information extraction model used to extract information from the change announcement includes procurement project change information and the project number; the information extraction model used to extract information from the contract announcement includes at least one of the project number, contract number, name, procurement project, signing time and announcement time; the information extraction model used to extract information from the contract document includes at least one of the name, address, contact information, contract signing date and contract term, contract subject, price and payment terms, performance period, place and method and liability for breach of contract.

[0114] In one embodiment, the information extraction module is specifically used to: determine whether the file type of each bidding related document is a scanned version of PDF; if so, use optical character recognition technology to perform text recognition on the bidding related document to obtain a text file corresponding to the bidding related document; determine whether the number of text words in the text file is greater than the maximum limit on the number of input words of the information extraction model of the bidding related document; if it is greater than the maximum limit, use a text paragraph segmentation model to segment the text file into paragraphs to obtain multiple paragraphs, and input each paragraph into the information extraction model of the bidding related document to obtain extracted information; if it is less than or equal to the maximum limit, input the text file into the information extraction model of the bidding related document to obtain extracted information.

[0115] In one embodiment, the project comparison module is specifically used to compare the current procurement project with the similar historical procurement projects in at least one element of procurement content, procurement method and budget amount.

[0116] In one embodiment, the bid opening analysis module is specifically used to: determine whether the supplier's bid price meets the preset linear law, and determine whether there is a problem of collusion in bidding based on the judgment result; and determine whether the supplier has a problem of accompanying bidding based on the supplier's historical bidding data and corporate equity penetration information.

[0117] In one embodiment, the evaluation criteria analysis module is specifically used to calculate at least one of the range, variance and interquartile range of scores given by expert judges during the evaluation criteria process.

[0118] In one embodiment, the device may further include:

[0119] The first detection module is used to perform at least one of the following: check whether the bidding agency of the current procurement project has legal qualifications; check the legality of the enterprise during the bid opening process; determine the information extracted from historical procurement projects to determine whether there are abnormal suppliers.

[0120] It is understandable that the explanation, specific implementation, beneficial effects, examples, etc. of the relevant contents in the device provided in the embodiment of the present invention can be found in the corresponding parts of the method provided in the first aspect, and will not be repeated here.

[0121] In a third aspect, an embodiment of the present invention provides a computer-readable medium, on which computer instructions are stored. When the computer instructions are executed by a processor, the processor executes the method provided in the first aspect.

[0122] Specifically, a system or device equipped with a storage medium can be provided, on which software program code that implements the functions of any of the above-mentioned embodiments is stored, and a computer (or CPU or MPU) of the system or device can be enabled to read and execute the program code stored in the storage medium.

[0123] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0124] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0125] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0126] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or to a memory provided in an expansion module connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or expansion module is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0127] It is understandable that the explanation, specific implementation methods, beneficial effects, examples, etc. of the relevant contents in the computer-readable medium provided in the embodiment of the present invention can be found in the corresponding parts of the method provided in the first aspect, and will not be repeated here.

[0128] In a fourth aspect, an embodiment of the present specification provides a computing device, including a memory and a processor, wherein the memory stores executable code, and when the processor executes the executable code, it implements the method in any embodiment of the specification.

[0129] It is understandable that the explanation, specific implementation, beneficial effects, examples, etc. of the relevant contents in the computing device provided in the embodiment of the present invention can be found in the corresponding parts of the method provided in the first aspect, and will not be repeated here.

[0130] 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 device 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.

[0131] Those skilled in the art should be aware that in one or more of the above examples, the functions described in the present invention can be implemented by hardware, software, widgets, or any combination thereof. When implemented by software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium.

[0132] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. A procurement project audit method based on a large model, characterized in that: include: Use the information extraction model to extract information from the bidding documents of the current procurement project, and store the extracted information in the database; According to the information extracted from the current procurement project, similar historical procurement projects of the current procurement project are retrieved from the document library using a similarity model, and the current procurement project is compared with the similar historical procurement projects in at least one element; Analyze whether there are problems of collusion and bid spoofing in the bidding process of the current procurement project based on the information extracted from the current procurement project; Analyze the scores of expert judges during the bid evaluation process based on the information extracted from the winning notice in the bidding documents; Analyze whether there are any abnormalities in the execution of the contract based on the agreed information in the contract documents in the bidding documents and the actual progress of the current procurement project; A full life cycle analysis report for the current procurement project is generated based on the information extracted from the current procurement project, the comparison results between the current procurement project and similar historical procurement projects in at least one element, the analysis results of the bid opening process, the analysis results of the bid evaluation process and the analysis results of the contract execution.

2. The method according to claim 1, characterized in that The information extraction model is used to extract information from the bidding documents related to the current procurement project, including: Different information extraction models are used to extract information from different bidding related documents; different information extraction models need to extract different information items; The bidding-related documents include bidding documents, tender documents, notice of winning, notice of change, contract notice and contract documents; The information items to be extracted by the information extraction model for extracting information from the bidding document include at least one of the basic information of the project, the construction content, the purchaser information and the bidding time information; the basic information of the project includes at least one of the project number, the project name, the procurement method, the budget amount and the maximum price limit; The information extraction model for extracting information from the bidding document needs to extract at least one of the following information items: project number, quotation, supplier information, budget, construction period and technical standards; The information extraction model for extracting information from the bid winning notice needs to extract at least one of the following information items: project number, bid winning supplier name, bid winning supplier address, bid winning price, review expert information, and bid winning supplier score; The information extraction model for extracting information from the change notice needs to extract information items including procurement project change information and project number; The information extraction model for extracting information from the contract announcement needs to extract at least one of the following information items: project number, contract number, name, procurement item, signing time and announcement time; The information extraction model used to extract information from the contract document needs to extract at least one of the names, addresses, contact information, contract signing date and contract term, contract subject, price and payment terms, performance period, place and method, and liability for breach of contract of the two parties to the contract.

3. The method according to claim 1, characterized in that The information extraction model is used to extract information from the bidding documents related to the current procurement project, including: Determine whether the file type of each bidding-related document is a scanned PDF; If yes, then use optical character recognition technology to perform text recognition on the bidding related document to obtain a text file corresponding to the bidding related document; Determine whether the number of text words in the text file is greater than the maximum limit of the number of input words for the information extraction model of the bidding related document; If it is greater than the maximum limit, the text file is segmented using a text paragraph segmentation model to obtain multiple paragraphs, and each paragraph is input into the information extraction model of the bidding-related file to obtain extracted information; If it is less than or equal to the maximum limit, the text file is input into the information extraction model of the bidding related documents to obtain the extracted information.

4. The method according to claim 1, characterized in that: The comparing the current procurement project with the similar historical procurement project on at least one factor includes: The current procurement project is compared with the similar historical procurement projects in at least one element of procurement content, procurement method and budget amount.

5. The method according to claim 1, characterized in that The analysis of whether there are problems of collusion and bidding with others in the bidding process of the current procurement project includes: Determine whether the supplier's bid price meets the preset linear rule, and determine whether there is a problem of collusion based on the judgment result; and, based on the supplier's historical bidding data and corporate equity penetration information, determine whether the supplier has a problem of accompanying bidding.

6. The method according to claim 1, characterized in that The analysis of the expert judges' scoring during the evaluation process includes: calculating at least one of the range, variance and interquartile range of the expert judges' scoring during the evaluation process.

7. The method according to claim 1, characterized in that Also includes at least one of the following: Check whether the bidding agency for the current procurement project has legal qualifications; Testing the legality of enterprises during the bidding process; Determine the information extracted from historical procurement projects to determine whether there are any abnormal suppliers.

8. A procurement project audit device based on a large model, characterized in that: include: The information extraction module is used to extract information from the bidding documents related to the current procurement project using the information extraction model, and store the extracted information in the database; A project comparison module, configured to retrieve similar historical procurement projects of the current procurement project from a document library using a similarity model according to various information extracted from the current procurement project, and compare the current procurement project with the similar historical procurement project in at least one element; A bid opening analysis module, used to analyze whether there are problems of collusion and bid spoofing in the bid opening process of the current procurement project according to various information extracted from the current procurement project; The bid evaluation analysis module is used to analyze the scores of the expert judges in the bid evaluation process based on the information extracted from the winning notice in the bidding related documents; An execution analysis module is used to analyze whether there are abnormalities in the execution of the contract based on the agreed information in the contract documents in the bidding related documents and the actual promotion of the current procurement project; The report generation module is used to generate a full life cycle analysis report for the current procurement project based on the information extracted from the current procurement project, the comparison results between the current procurement project and similar historical procurement projects in at least one element, the analysis results of the bid opening process, the analysis results of the bid evaluation process and the analysis results of the contract execution.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed in a computer, the computer is caused to execute the method according to any one of claims 1 to 7.

10. A computing device, characterized in that: The method comprises a memory and a processor, wherein the memory stores executable codes, and when the processor executes the executable codes, the method according to any one of claims 1 to 7 is implemented.

Images