Submitted material auditing method and device, electronic equipment, medium and program product

Abstract

The invention provides a submitted material auditing method and device, electronic equipment, a medium and a program product, and relates to the technical field of engineering project auditing. According to the method, through structured information extraction, automatic file analysis and information extraction of the submitted file are realized, and the material auditing efficiency is improved. Through the mutual exclusion relationship and the symbiotic relationship, intelligent adjustment of the quota quantity is realized, the accuracy of the quota quantity is improved, and the accuracy and efficiency of auditing the actually submitted materials are further improved. Meanwhile, intelligent auditing of the real submitted material list is achieved, manual intervention is reduced, the auditing error rate and auditing cost are reduced, and meanwhile the auditing efficiency is improved.

Description

Technical Field

[0001] This application relates to the field of engineering project auditing technology, and in particular to a method, apparatus, electronic equipment, medium and program product for auditing submitted materials. Background Technology

[0002] In mobile communication engineering projects, material auditing is a crucial step in ensuring cost control and the rational use of resources. By auditing materials, potential financial risks can be effectively identified, resources can be allocated appropriately, and the smooth progress of the project can be guaranteed.

[0003] After a telecommunications engineering project is completed, information on the materials used during the construction process is submitted to an auditing unit for review. The auditing unit verifies the authenticity and reasonableness of the quantity and price of the materials used. However, auditing is time-consuming and involves many complex audit points, including quota relationships, requisition and allocation documentation, site surveys, construction drawings, price comparisons, and offline price inquiries. The large amount and complexity of information inevitably leads to audit omissions.

[0004] To improve the quality and efficiency of engineering material audits, AI capabilities will be leveraged to intelligently identify the categories of submitted materials, theoretical workload, and reference information prices. This will assist in identifying audit risks and avoid quality loopholes caused by large amounts of information or insufficient audit time, thereby improving work efficiency and audit quality. We advocate for a management model based on AI technology, promoting AI+engineering audits to achieve high-quality development of smart engineering audit services.

[0005] However, existing AI-powered engineering audits cannot audit materials. Summary of the Invention

[0006] This application provides a method, apparatus, electronic device, medium, and program product for auditing submitted materials, in order to address the shortcomings of existing AI+engineering auditing that cannot audit materials, and to achieve automated auditing of materials in engineering projects.

[0007] Firstly, this application provides a method for reviewing submitted materials, including: Structured information extraction is performed on the settlement material information in the project submission documents to obtain the actual submitted material list and the material list to be submitted. The material list to be submitted includes the material to be submitted and the quota quantity of the material to be submitted. Based on the mutual exclusion and symbiotic relationships of the materials to be submitted for review, the quota quantity is adjusted to obtain the revised list of materials to be submitted for review. Based on the comparison between the actual submitted materials list and the revised materials list to be submitted, a risk warning is generated for the actual submitted materials list.

[0008] In one embodiment, the project includes multiple sub-projects, and the mutual exclusion relationship is determined based on the following method: Based on at least one material that should be submitted for review in each sub-project, construct a material information table for the sub-project; Based on at least one sub-item contained in each material to be submitted for review, construct a sub-item information table for the material; Extract any material to be submitted for review from the material sub-item information table to obtain the target material and a list of sub-items contained in the target material; Extract any sub-item from the list of sub-items included in the target material to obtain the target sub-item; Obtain the list of materials to be submitted for review for the target sub-project from the material information table of the sub-project; Based on the list of materials to be submitted for review for the target sub-project, obtain the set of related materials to be submitted for review for the target materials; Based on the complement of the associated set of materials to be submitted for review, obtain the set of mutually exclusive materials to be submitted for review for the target material. The materials to be submitted for review and the target material in the set of mutually exclusive materials to be submitted for review are mutually exclusive.

[0009] In one embodiment, the quota quantity of materials to be submitted for review is limited by multiple constraints, including pre-materials and post-materials that need to be constrained. The post-materials are derived from the pre-materials, and the symbiotic relationship is obtained based on the following method: Based on the coverage threshold and confidence threshold of the constraints, multiple constraints are filtered to obtain multiple target constraints; Based on the target constraint condition that includes the same materials to be submitted for review, obtain the equivalent constraint condition set, which includes multiple equivalent constraint conditions. In the set of equivalent constraints, based on the equivalent constraints with the largest coverage and the smallest number of previous materials, a set of filtered equivalent constraints is constructed. In the set of equivalent constraints that have been screened, the first constraint and the second constraint are obtained. If the set of materials to be submitted for review included by the first constraint is a subset of the set of materials to be submitted for review included by the second constraint, and the coverage of the first constraint is greater than the coverage of the second constraint, then it is determined that the first constraint and the second constraint have a symbiotic relationship.

[0010] In one embodiment, coverage is determined based on the following: For any constraint, based on the preceding and following materials included in the constraint, obtain the set of materials to be submitted for review under the constraint. Based on the material set with constraints, obtain the number of sub-items that satisfy the constraints. Sub-items that satisfy the constraints are sub-items that cover all materials in the material set to be submitted for review that meet the constraints. The coverage of the constraints is determined based on the number of sub-items that satisfy the constraints and the total number of sub-items.

[0011] In one embodiment, based on a comparison between the actual submitted materials list and the revised materials list to be submitted, a risk warning for the actual submitted materials list is generated, including: The revised list of materials to be submitted for review and the list of materials actually submitted for review are standardized to obtain a standardized list of materials to be submitted for review and a standardized list of materials actually submitted for review. Based on the comparison between the standardized list of materials to be submitted for review and the standardized list of materials actually submitted for review, risk warnings are generated for the list of materials actually submitted for review.

[0012] In one embodiment, the revised list of materials to be submitted for review includes non-standard material names, while the standardized list of materials to be submitted for review is obtained based on the following steps: Construct prompts based on the names of non-standard and standard materials; Input the prompts and the revised list of materials to be submitted for review into the translation model, and obtain the standardized list of materials to be submitted for review output by the translation model.

[0013] In one embodiment, the translation model is used for: Based on the prompt words, the non-standard material names in the revised list of materials to be submitted for review are converted into standard material names, resulting in a standardized list of materials to be submitted for review. The units of measurement in the list of materials to be submitted for review with standardized names are standardized to obtain a list of materials to be submitted for review with standardized units of measurement. When there is a reference price for the materials to be submitted for review, the reference price is added to the list of materials to be submitted for review with standardized units of measurement, thus obtaining a standardized list of materials to be submitted for review.

[0014] In one embodiment, the standardized list of materials to be submitted for review includes the materials to be submitted, the revised quota quantity, and the price to be submitted; the standardized list of materials actually submitted for review includes the materials actually submitted, the actual quantity submitted, and the actual price submitted. Based on the comparison between the standardized list of materials to be submitted for review and the standardized list of materials actually submitted for review, a risk warning for the list of materials actually submitted for review is generated, including: When the comparison result shows that the actual submitted materials and the materials to be submitted match, but the corrected quota quantity does not match the actual submitted quantity, a risk warning is generated indicating that the actual submitted materials have an incorrect submission quantity. When the comparison result shows that the submitted materials and the materials to be submitted match, but the price of the submitted materials does not match the price of the materials to be submitted, a risk warning is generated indicating that the price of the submitted materials is incorrect. When the comparison result shows that the submitted materials and the materials that should be submitted do not match, a risk warning is generated indicating that the submitted materials are not within the scope of submission.

[0015] Secondly, this application also provides a device for reviewing submitted materials, comprising: The information extraction module is used to extract structured information from the settlement material information in the submission documents of the project, and obtain the actual submission material list and the material list to be submitted. The material list to be submitted includes the material to be submitted and the quota quantity of the material to be submitted. The correction module is used to correct the quota quantity based on the mutual exclusion relationship and symbiotic relationship of the materials to be submitted for review, so as to obtain the corrected list of materials to be submitted for review. The comparison module is used to generate risk warnings for the actual submitted materials list based on the comparison results between the actual submitted materials list and the revised materials list to be submitted.

[0016] Thirdly, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the review method for any of the submitted materials described above.

[0017] Fourthly, this application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the review method for any of the submitted materials described above.

[0018] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the review method for any of the submitted materials described above.

[0019] The method, apparatus, electronic equipment, media, and program products for reviewing submitted materials provided in this application achieve automated document parsing and information extraction of submitted documents through structured information extraction, thereby improving the efficiency of material review. This application achieves intelligent adjustment of quota quantities through mutual exclusion and symbiotic relationships, improving the accuracy of quota quantities and thus enhancing the accuracy and efficiency of the review of actual submitted materials. Simultaneously, this application enables intelligent review of the actual submitted material list, reducing manual intervention, lowering the review error rate and review costs, and also improving review efficiency. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0021] Figure 1 This is one of the flowcharts illustrating the review method for the submitted materials provided in this application.

[0022] Figure 2 This is the second flowchart illustrating the review process for the submitted materials provided in this application.

[0023] Figure 3 This is a schematic diagram of the structure of the review device for the submitted materials provided in this application.

[0024] Figure 4 This is a schematic diagram of the structure of the electronic device provided in this application. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] After the project audit is submitted, the quotas in Table 3A of the settlement statement will be used to determine the authenticity and compliance of the usage scenarios, quantities, and unit prices of the submitted materials. The current methods for reducing materials mainly include the following steps.

[0027] 1. The auditing firm initiates the audit submission through the Project Management System (PMS).

[0028] 2. The construction unit transmits the audit settlement statement data to the Centralized Alert Management System (CAMS), and the auditing intermediary can download it themselves.

[0029] 3. Offline audit: The auditing intermediary finds the submitted settlement statement and checks the standard quota.

[0030] 4. Offline audit: The auditing intermediary reviews the material information associated with each quota in the 451 quota book based on the standard quota.

[0031] 5. Offline audit: Based on historical experience and the relationships specified in the 451 quota book, determine whether the materials should be submitted for audit.

[0032] 6. Offline audit: Calculate the material usage associated with each quota according to the calculation relationship specified in the 451 quota book, and perform summary calculation.

[0033] 7. Offline audit: Based on the city and time of submission, query the information price of the materials.

[0034] 8. Offline audit: Based on the above conditions, risk warnings are issued regarding whether materials should be submitted for audit, whether the usage is compliant, and whether the unit price is overreported.

[0035] 9. Offline Audit: Compiling and Approving the Results 10. The auditing agency uploads the audited settlement statement documents to CAMS to complete the materials audit.

[0036] The aforementioned audit process, lacking structured storage of submitted material data, failed to effectively utilize data assets for audit vulnerability verification and risk assessment. Furthermore, significant discrepancies in material names and specifications, along with inconsistent unit of measurement, necessitated time-consuming judgment and conversions of prices and workloads based on different units, further wasting time. Manual verification was slow, requiring meticulous searching and calculation of numerous quotas, consuming substantial auditor time. Moreover, reliance entirely on manual verification risked audit vulnerabilities arising from subjective choices and varying abilities of auditors. These issues resulted in reduced audit quality and efficiency, leading to economic losses and hindering the maximization of investment returns.

[0037] The following is combined Figures 1-4 Describe the examination methods, apparatus, and electronic equipment used for the materials submitted for review in this application.

[0038] Figure 1 This is one of the flowcharts illustrating the review method for the submitted materials provided in this application, such as... Figure 1 and Figure 2 As shown, the method includes steps S100 to S400, and the specific steps are as follows.

[0039] S100: Extract structured information from the settlement material information in the project's submission documents to obtain the actual submitted material list and the material list to be submitted.

[0040] The list of materials to be submitted for review includes the materials to be submitted for review and the quota quantity of the materials to be submitted for review.

[0041] The list of materials actually submitted for review includes the list of materials that were actually submitted for review. The list of materials that should be submitted for review includes the list of materials that should theoretically be submitted for review.

[0042] The engineering project covered by this application includes a telecommunications engineering project. The submitted documents include the submitted settlement statement and its attachments. Figure 2As shown, in the Project Management System (PMS), attachments to the submitted settlement statement are read from a specified path or data source. Appropriate parsing tools are used to perform structured parsing of the attachments. For example, for Excel files, Python code is written to use an Excel parsing library to obtain key field information. For PDF files, the pdf2txt tool is called to extract key information such as the project ID, city, and construction dates to obtain information about the project and its sub-projects. The parsed data is then stored in a structured format in an Excel or JSON file.

[0043] The submitted settlement statement includes Table 3A and Table 4. Table 3A includes the quota number and quota quantity of the actual submitted materials. Table 4 includes fields such as the name, constraints, unit, quantity (actual submitted quantity), and actual submitted price of the submitted materials.

[0044] The structured data in Tables 3A and 4 are cleaned to remove meaningless data (e.g., data with a quantity of 0 or empty data). During data cleaning, data validation is performed to ensure data integrity and accuracy. Tables 3A and 4 are then filtered. Settlement material information is extracted from the filtered Table 3A to obtain a list of materials to be submitted for review. Settlement material information is extracted from the filtered Table 4 to obtain a list of materials actually submitted for review. Settlement material information includes the quota number (representing the materials to be submitted for review) and quota quantity from Table 3A. Settlement material information also includes the name, constraints, unit, quantity (actual submission quantity), and actual submission price of the materials actually submitted for review from Table 4. Tables 3A and 4 are shown below.

[0045] Table 3A

[0046] Table 4

[0047] S200: Based on the mutual exclusion and symbiotic relationships of the materials to be submitted for review, the quota quantity is modified to obtain the modified list of materials to be submitted for review.

[0048] There are certain relationships between the materials used by users. Some materials are symbiotic, such as installing and debugging a "satellite system antenna," which necessarily requires installing a "lightning arrester." Symbiotic materials may have a certain quantitative relationship, such as the 1:1 relationship between the "satellite system antenna" and the "lightning arrester." However, some materials are mutually exclusive, such as a "power distribution rack (floor-mounted)" and a "power distribution rack (top-mounted)."

[0049] Based on the constraints obtained in S100, the mutual exclusion relationships and symbiotic relationships between the materials to be submitted for review are extracted, resulting in a list of symbiotic relationships and a list of mutual exclusion relationships. Partial lists of symbiotic relationships and mutual exclusion relationships are shown in Tables 1 and 2.

[0050] Table 1. Partial List of Symbiotic Relationships

[0051] Table 2. List of Partial Mutually Exclusive Relationships

[0052] Based on the mutual exclusion and symbiotic relationship, the quota quantity of the submitted materials is adjusted to obtain the revised quota quantity.

[0053] For example, by training a model to learn the symbiotic and mutually exclusive relationships, an AI symbiotic-mutually exclusive reasoning model can be obtained. Figure 2 As shown, the list of materials to be submitted for review is input into the AI ​​symbiotic mutual exclusion reasoning model. (1) The AI ​​symbiotic mutual exclusion reasoning model determines whether the quota number (material to be submitted for review) should exist. If it does not exist, the quota quantity is determined to be 0. (2) The AI ​​symbiotic mutual exclusion reasoning model also determines whether the current material to be submitted for review has a symbiotic relationship with other materials to be submitted for review. If there is a symbiotic relationship, it determines whether the symbiotic quota (symbiotic material to be submitted for review) of the current material to be submitted for review has been submitted for review. If it has not been submitted for review, the quota quantity of the current material to be submitted for review is adjusted to 0. If it has been submitted for review, it further determines whether there is a symbiotic quantity relationship. If there is a symbiotic quantity relationship, the quota quantity of the current material to be submitted for review is adjusted according to the symbiotic quantity relationship (e.g., reduced).

[0054] Furthermore, after determining the symbiotic relationship, the ratio between the materials to be submitted for review under the constraints corresponding to the symbiotic relationship is calculated. For example, two materials to be submitted for review are randomly selected from the constraints corresponding to the symbiotic relationship. For each sub-item, the ratio between the two materials to be submitted for review is calculated. The mean, minimum, maximum, and variance (std) of the ratio between the two materials to be submitted for review are calculated. Based on the mean and variance, the range interval [mean-std, mean+std] of the ratio between the two materials to be submitted for review is determined. If the minimum and maximum values ​​are within this range interval [mean-std, mean+std], the calculated ratio between the two materials to be submitted for review is considered valid; otherwise, the calculated ratio between the two materials to be submitted for review is considered invalid, and the revised quota quantity is further adjusted.

[0055] Furthermore, adjustments will be made to the list of materials to be submitted for review and the corresponding quota quantities as stipulated in the 451 quota library.

[0056] By revising the quota quantities, a large number of false reports can be found among the materials that should be submitted for review.

[0057] S300: Based on the comparison between the actual submitted materials list and the revised materials list to be submitted, generate a risk warning for the actual submitted materials list.

[0058] Compare the actual submitted materials list with the revised list of materials to be submitted to determine if the actual submitted materials list falls within the scope of the revised list of materials to be submitted. If the actual submitted materials list falls within the scope of the revised list of materials to be submitted, then the actual submitted materials list is deemed risk-free. If the actual submitted materials list does not fall within the scope of the revised list of materials to be submitted, then the actual submitted materials list is deemed risky.

[0059] In one embodiment, a risk warning for the actual submitted materials list is generated based on a comparison between the actual submitted materials list and the revised materials list to be submitted, including the following steps: The revised list of materials to be submitted for review and the list of materials actually submitted for review are standardized to obtain a standardized list of materials to be submitted for review and a standardized list of materials actually submitted for review. Based on the comparison between the standardized list of materials to be submitted for review and the standardized list of materials actually submitted for review, risk warnings are generated for the list of materials actually submitted for review.

[0060] The revised lists of materials to be submitted for review and the actual lists of materials submitted for review are standardized to obtain standardized lists of materials to be submitted for review and standardized lists of materials submitted for review. For example, non-standard terms in the revised lists of materials to be submitted for review are converted into standardized terms based on a large language model, resulting in a standardized list of materials to be submitted for review. Similarly, non-standard terms in the actual lists of materials submitted for review are converted into standardized terms based on a large language model, resulting in a standardized list of materials submitted for review. The large language model includes the Qianxun (e.g., Qwen-72B) large model. Non-standardized terms include non-standardized terms for material names, prices, and units of measurement.

[0061] Compare the standardized list of materials to be submitted for review with the standardized list of materials actually submitted for review. If the contents of the standardized list of materials actually submitted for review fall within the scope of the standardized list of materials to be submitted for review, the list of materials actually submitted for review is deemed risk-free. If the contents of the standardized list of materials actually submitted for review do not fall within the scope of the standardized list of materials to be submitted for review, the list of materials actually submitted for review is deemed risky.

[0062] The method for reviewing submitted materials provided in this application, through structured information extraction, achieves automated document parsing and information extraction of submitted documents, improving the efficiency of material review. This application utilizes mutual exclusion and symbiotic relationships to achieve intelligent adjustment of quota quantities, improving the accuracy of quota quantities and thus enhancing the accuracy and efficiency of reviewing actual submitted materials. Simultaneously, this application enables intelligent review of the actual submitted material list, reducing manual intervention, lowering the error rate and cost of review, while also improving review efficiency.

[0063] Based on the above embodiments, the engineering project includes multiple sub-projects, and the mutual exclusion relationship is determined in the following way: Based on at least one material that should be submitted for review in each sub-project, construct a material information table for the sub-project; Based on at least one sub-item contained in each material to be submitted for review, construct a sub-item information table for the material; Extract any material to be submitted for review from the material sub-item information table to obtain the target material and a list of sub-items contained in the target material; Extract any sub-item from the list of sub-items included in the target material to obtain the target sub-item; Obtain the list of materials to be submitted for review for the target sub-project from the material information table of the sub-project; Based on the list of materials to be submitted for review for the target sub-project, obtain the set of related materials to be submitted for review for the target materials; Based on the complement of the associated set of materials to be submitted for review, obtain the set of mutually exclusive materials to be submitted for review for the target material. The materials to be submitted for review and the target material in the set of mutually exclusive materials to be submitted for review are mutually exclusive.

[0064] Determining mutual exclusion involves the following steps.

[0065] (1) Construct a material information table for each sub-project based on at least one material to be submitted for review. Construct a sub-project information table for each material to be submitted for review based on at least one sub-project. The results of the material information table (D) for sub-projects and the sub-project information table (T) for materials are as follows.

[0066] Sub-project Material Information Sheet D={ "project_1":{"T5G2-046x":1.0,"T5G2-065y":1.0,"T5G2-119z":3.6,...}, "project_2":{"T5G2-046":1.0,"T5G1-048":2.0,"T5G2-121":2.2,...}, ... } # Sub-item information table of materials T={ "T5G2-046x":{"project_1":1.0,"project_2":1.0,...}, "T5G2-065y":{"project_1":1.0,...}, "T5G2-119z":{"project_1":3.6,...} } Among them, "project" refers to a sub-project, and "T5G2-046x", "T5G2-065y", and "T5G2-119z" are the quota numbers of the materials to be submitted for review.

[0067] (2) Delete materials that should be submitted for review if the number of sub-items is less than 30 in the material sub-item information table to reduce noise.

[0068] (3) Traverse the materials to be submitted for review in the sub-item information table T to obtain the target material (e.g., T5G2-046x). Extract the list of sub-items contained in the target material (T5G2-046x) from the sub-item information table T. For example, the list of sub-items contained in the target material is {"project_1":1.0,"project_2":1.0,...}.

[0069] (3) Extract any sub-item from the list of sub-items contained in the target material to obtain the target sub-item. Iterate through the sub-items in the list of sub-items contained in the target material to obtain the target sub-item (e.g., "project_1":1.0).

[0070] (4) Obtain the list of materials to be submitted for review for the target sub-project ("project_1":1.0) from the material information table D of the sub-project. For example, the list of materials to be submitted for review for the target sub-project is {"T5G2-046x":1.0,"T5G2-065y":1.0,"T5G2-119z":3.6,...}.

[0071] (5) Based on the list of materials to be submitted for review for the target sub-project, obtain the set of related materials to be submitted for review for the target material. For example, if the target material is T5G2-046x, the set of related materials to be submitted for review for the target material is {T5G2-046x, T5G2-065y, T5G2-119z}.

[0072] (6) Based on the complement of the set of materials to be submitted for review, obtain the set of mutually exclusive materials to be submitted for review for the target material. The materials to be submitted for review in the set of mutually exclusive materials to be submitted for review and the target material are mutually exclusive. According to the sub-item information table T of the material, obtain the set of all materials to be submitted for review. For example, the set of all materials to be submitted for review is {T5G2-046x,T5G2-065y,T5G2-119z}. From the set of all materials to be submitted for review, obtain the complement of the set of materials to be submitted for review related to the target material. For example, if the set of materials to be submitted for review related to the target material is {T5G2-046x,T5G2-065y,T5G2-119z}, then the complement is empty, and the set of mutually exclusive materials to be submitted for review for the target material is empty. According to the set of mutually exclusive materials to be submitted for review for the target material, obtain the materials that are mutually exclusive with the target material.

[0073] (7) Repeat step (3) above to traverse each material to be submitted for review in the sub-item information table T of the materials, and then obtain the materials that form a mutually exclusive relationship with each material to be submitted for review.

[0074] This application achieves automatic identification of the mutual exclusion relationships between materials to be submitted for review by constructing a material information table for sub-projects and a material sub-project information table, which helps to improve the accuracy of the quota quantity of materials to be submitted for review.

[0075] Based on the above embodiments, the quota quantity of materials to be submitted for review is limited by multiple constraints. These constraints include the preceding and subsequent materials that need to be constrained. The subsequent materials are derived from the preceding materials, and the symbiotic relationship is obtained in the following way: Based on the coverage threshold and confidence threshold of the constraints, multiple constraints are filtered to obtain multiple target constraints; Based on the target constraint condition that includes the same materials to be submitted for review, obtain the equivalent constraint condition set, which includes multiple equivalent constraint conditions. In the set of equivalent constraints, based on the equivalent constraints with the largest coverage and the smallest number of previous materials, a set of filtered equivalent constraints is constructed. In the set of equivalent constraints that have been screened, the first constraint and the second constraint are obtained. If the set of materials to be submitted for review included by the first constraint is a subset of the set of materials to be submitted for review included by the second constraint, and the coverage of the first constraint is greater than the coverage of the second constraint, it is determined that the first constraint and the second constraint have a symbiotic relationship.

[0076] Constraints are used to define the materials to be submitted for review. For example, constraint 1 requires materials A and B. It can be deduced that the requirement for materials A and B presupposes the requirement for material C. Therefore, the materials to be submitted under constraint 1 can be simplified as (A, B) => (C). Here, materials A and B are the preceding materials, and material C is the following material.

[0077] Determining mutual exclusion involves the following steps.

[0078] (1) Obtain the constraint set of the project. In actual projects, the recall rate of the constraint set is 53%, and the precision is 97.4%. The low recall rate is because some materials with symbiotic relationships do not have a quantitative relationship, such as electricity meters and cables. If the user declares more cables, the constraint cannot be found, which requires combining drawings and site surveys.

[0079] (2) Calculate the coverage of each constraint in the constraint set. The constraint set includes multiple constraints. Based on the coverage threshold and confidence threshold of the constraints, the multiple constraints are filtered to obtain multiple target constraints. For example, multiple constraints with a coverage greater than the coverage threshold (e.g., 5%) and a confidence level of 100% (confidence threshold) are filtered from the constraint set to obtain multiple target constraints, thus obtaining the target constraint set.

[0080] (3) Take the target constraints that contain the same materials to be submitted for review as equivalent constraints, and obtain the equivalent constraint set. The equivalent constraint set includes multiple equivalent constraints. For example, for constraint 1, which includes materials B=>C (where material B is the preceding material and material C is the following material), the materials to be submitted for review are material B and material C. For constraint 2, which includes materials BC (containing only the two preceding materials B and C), the materials to be submitted for review are also material B and material C. Then, constraint 1 and constraint 2 are determined to be equivalent constraints. Obtain the equivalent constraint set from the target constraint set.

[0081] (4) In the set of equivalent constraints, construct a set of filtered equivalent constraints based on the equivalent constraints with the largest coverage and the smallest quantity of preceding materials. In the set of equivalent constraints, obtain the equivalent constraint with the largest coverage. Further, extract the equivalent constraint with the smallest quantity of preceding materials from the equivalent constraint with the largest coverage to obtain the set of filtered equivalent constraints.

[0082] (5) From the selected set of equivalent constraints, obtain the first constraint R1 and the second constraint R2. Determine that the set of materials to be submitted for review included in the first constraint (e.g., a set consisting of materials A and B) is a subset of the set of materials to be submitted for review included in the second constraint (e.g., a set consisting of materials A, B, and C). Further, if the coverage of the first constraint is greater than the coverage of the second constraint, determine that the first and second constraints have a symbiotic relationship. If the coverage of the first constraint is the same as the coverage of the second constraint, delete the first constraint.

[0083] This application achieves automatic identification of symbiotic relationships between materials to be submitted for review by adjusting the coverage and confidence of constraints, which helps to improve the accuracy of the quota quantity of materials to be submitted for review.

[0084] Coverage is determined based on the following method: For any constraint, based on the preceding and following materials included in the constraint, obtain the set of materials to be submitted for review under the constraint. Based on the material set with constraints, obtain the number of sub-items that satisfy the constraints. Sub-items that satisfy the constraints are sub-items that cover all materials in the material set to be submitted for review that meet the constraints. The coverage of the constraints is determined based on the number of sub-items that satisfy the constraints and the total number of sub-items.

[0085] For example, the set of materials to be submitted for review under constraint 1 is {material A, material B, material C}. If there exists a sub-item 1 that covers materials {material A, material B, material C, material D}, then sub-item 1 is determined to be a sub-item satisfying constraint 1. All sub-items satisfying constraint 1 are obtained to determine the number of sub-items satisfying the constraint. The coverage of the constraint is obtained based on the ratio of the number of sub-items satisfying the constraint to the total number of sub-items.

[0086] This application determines the set of materials to be submitted for review based on the preceding and following materials included in the constraints, and determines the number of sub-items that satisfy the constraints, thereby achieving accurate calculation of the coverage of the constraints and thus improving the accuracy of determining symbiotic relationships.

[0087] Based on the above embodiments, the revised list of materials to be submitted for review includes non-standard material names, while the standardized list of materials to be submitted for review is obtained based on the following steps: Construct prompts based on the names of non-standard and standard materials; Input the prompts and the revised list of materials to be submitted for review into the translation model, and obtain the standardized list of materials to be submitted for review output by the translation model.

[0088] In one embodiment, the translation model is used for: Based on the prompt words, the non-standard material names in the revised list of materials to be submitted for review are converted into standard material names, resulting in a standardized list of materials to be submitted for review. The units of measurement in the list of materials to be submitted for review with standardized names are standardized to obtain a list of materials to be submitted for review with standardized units of measurement. When there is a reference price for the materials to be submitted for review, the reference price is added to the list of materials to be submitted for review with standardized units of measurement, thus obtaining a standardized list of materials to be submitted for review.

[0089] Due to the inconsistency in the names and specifications of submitted materials, information prices, and quota books, which makes identification difficult, a translation model was developed. This model translates the non-standard material names and specifications input by various provinces into standard material names and specifications.

[0090] The names and specifications of materials input by different provinces are not uniform, and the quantity is enormous, reaching millions. Standardizing and unifying them by audit experts would be too costly in terms of manpower. However, the standard library is unique and definitive, with a data volume of approximately 90,000. Therefore, a large model like Qwen-72B, which has high understanding and generation capabilities, can be used to convert the material names and specifications in the standard library into non-standard material names and specifications that may be input by various provinces during the review process.

[0091] Collect data on the names of all materials submitted for review from all provinces, merge the data, remove duplicates, and prepare a list of the names of materials submitted for review.

[0092] Prompts are constructed based on non-standard and standard material names. The prompts and the revised list of materials to be submitted for review are input into the translation model, which outputs a standardized list of materials to be submitted for review. [ { "instruction": "You are a construction project cost engineer expert and need to map the input non-standard material names and specifications to their corresponding standard material names and specifications." "input":"The name of the non-standard material entered is: 8-pin modular information socket, and the specification is: 11A". "output": "The standard material name is: 8-position modular information socket, and the specification is: 10A" }, { "instruction": "You are a construction project cost engineer expert, and you need to map the input non-standard material names and specifications to their corresponding standard material names and specifications." "input":"The name of the non-standard material entered is: 8-hole information socket, and the specification is: 10A". "output": "The standard material name is: 8-position modular information socket, and the specification is: 10A" }, { "instruction": "You are a construction project cost engineer expert, and you need to map the input non-standard material names and specifications to their corresponding standard material names and specifications." "input":"The non-standard material name entered is: 8-hole modular network socket, and the specification is: 12A". "output": "The standard material name is: 8-position modular information socket, and the specification is: 10A" }, ... ].

[0094] like Figure 2 As shown, based on the previously obtained information on engineering projects and their sub-projects, a material information price database is generated. This database includes the city and year of submission.

[0095] Input the prompts and the revised list of materials to be submitted for review into the translation model. Based on the prompts, the translation model converts the non-standard names of the materials to be submitted in the revised list into standard names, resulting in a standardized list of materials to be submitted. Then, the translation model standardizes the units of measurement in the standardized list of materials to be submitted, resulting in a list of materials to be submitted with standardized units of measurement. For example, units of measurement including "kilometer" and "hundred-meter" are uniformly converted to "meter" in the backend.

[0096] The translation model is also used to compare the standardized list of materials actually submitted for review with the standardized list of materials to be submitted for review, in order to determine whether there are reference prices for the materials to be submitted for review in the standardized list of materials to be submitted for review. When it is determined that there are reference prices for the materials to be submitted for review, the corresponding reference prices are extracted from the standardized material information price database and added to the standardized list of materials to be submitted for review in terms of unit of measurement, thus obtaining the standardized list of materials to be submitted for review.

[0097] Optionally, the translation model can also be used to generate a standardized list of materials actually submitted for review. For example, prompt words and the list of materials actually submitted for review are input into the translation model. Based on the prompt words, the translation model converts the non-standard names of the materials actually submitted for review in the list into standard material names, resulting in a list of materials actually submitted for review with standardized names. Then, the translation model standardizes the units of measurement in the list of materials actually submitted for review with standardized names, resulting in a standardized list of materials actually submitted for review.

[0098] Optionally, the translation module can also be used to generate a standardized material information price database. For example, prompt words and the material information price database are input into the translation model. Based on the prompt words, the translation model converts the non-standard names of materials in the material information price database into standard material names, resulting in a material information price database with standardized names. Then, the translation model standardizes the units of measurement in the material information price database with standardized names, resulting in a standardized material information price database.

[0099] This application constructs prompt words using non-standard and standard material names, enabling automatic conversion of non-standard material names in the corresponding submitted material list and the actual submitted material list, reducing the need for manual annotation. Through a translation model, this application achieves automatic standardization of the corresponding submitted material list, the actual submitted material list, and the material information price database, ensuring the accuracy and consistency of information, and realizing standardized processing of submissions, facilitating cross-regional and cross-project material management and auditing.

[0100] Based on the above embodiments, the standardized list of materials to be submitted for review includes the materials to be submitted, the revised quota quantity, and the price to be submitted. The standardized list of materials actually submitted for review includes the materials actually submitted, the actual quantity submitted, and the actual price submitted. Based on the comparison results between the standardized list of materials to be submitted for review and the standardized list of materials actually submitted for review, a risk warning for the list of materials actually submitted for review is generated, including the following steps: When the comparison result shows that the actual submitted materials and the materials to be submitted match, but the corrected quota quantity does not match the actual submitted quantity, a risk warning is generated indicating that the actual submitted materials have an incorrect submission quantity. When the comparison result shows that the submitted materials and the materials to be submitted match, but the price of the submitted materials does not match the price of the materials to be submitted, a risk warning is generated indicating that the price of the submitted materials is incorrect. When the comparison result shows that the submitted materials and the materials that should be submitted do not match, a risk warning is generated indicating that the submitted materials are not within the scope of submission.

[0101] The standardized list of materials to be submitted for review and the standardized list of materials actually submitted for review are compared. First, it is determined whether the materials actually submitted for review should be submitted. For example, if the materials actually submitted for review match the materials to be submitted for review, then the materials actually submitted for review should be submitted. If the materials actually submitted for review and the materials to be submitted for review do not match, then the materials actually submitted for review should not be submitted, and a risk warning is generated indicating that the materials actually submitted for review are not within the scope of review.

[0102] If it is determined that the submitted materials and the materials to be submitted match, then it is further determined whether the actual quantity of the submitted materials matches the revised quota quantity. If the revised quota quantity does not match the actual quantity submitted, a risk warning for an incorrect quantity of submitted materials is generated. If the revised quota quantity does not match the actual quantity submitted, no risk is determined.

[0103] Simultaneously, if it is determined that the submitted materials and the materials to be submitted match, the system further determines whether the actual submitted price matches the required price. If the actual submitted price and the required price do not match, a risk warning is generated indicating an error in the submitted price. If the actual submitted price and the required price match, no risk is identified.

[0104] Furthermore, the intermediary confirms the risks through PMS and submits the review results of the actual submitted materials list.

[0105] This application achieves automatic identification of risk alerts in the actual submitted materials list by comparing the submitted materials list with the required materials list, providing decision support for auditors and helping to promptly identify and correct potential errors and risks.

[0106] This application resolves the issue of material parsing in settlement statements, enabling data asset accumulation. It also addresses the problem of inconsistent material names and specifications in submitted materials, information prices, and quota books, leading to difficulty in identification. Furthermore, it resolves the conversion issue caused by inconsistent units in submitted materials, information prices, and quota books. Finally, it addresses the time-consuming issue of querying quota materials and information prices in the 451 quota book, improving audit efficiency.

[0107] The following describes the examination device for the submitted materials provided in this application. The examination device for the submitted materials described below can be referred to in correspondence with the examination method for the submitted materials described above.

[0108] like Figure 3 As shown, an auditing device for submitted materials includes an information extraction module 301, a correction module 302, and a comparison module 303.

[0109] The information extraction module 301 is used to extract structured information from the settlement material information in the submission documents of the project, and obtain the actual submission material list and the material list to be submitted. The material list to be submitted includes the material to be submitted and the quota quantity of the material to be submitted.

[0110] The correction module 302 is used to correct the quota quantity based on the mutual exclusion relationship and symbiotic relationship of the materials to be submitted for review, so as to obtain the corrected list of materials to be submitted for review.

[0111] The comparison module 303 is used to generate risk warnings for the actual submitted materials list based on the comparison results between the actual submitted materials list and the revised materials list to be submitted.

[0112] The material review device provided in this application, through structured information extraction, achieves automated document parsing and information extraction of submitted documents, improving the efficiency of material review. This application utilizes mutual exclusion and symbiotic relationships to achieve intelligent adjustment of quota quantities, improving the accuracy of quota quantities and thus enhancing the accuracy and efficiency of actual material review. Simultaneously, this application enables intelligent review of the actual material list, reducing manual intervention, lowering the error rate and cost of review, while also improving review efficiency.

[0113] In one embodiment, the engineering project includes multiple sub-projects, and the correction module 302 is further configured to: construct a material information table for each sub-project based on at least one material to be submitted for review contained in each sub-project; construct a sub-project information table for each material based on at least one sub-project contained in each material to be submitted for review; extract any material to be submitted for review from the sub-project information table of the material to obtain a target material and a list of sub-projects contained in the target material; extract any sub-project from the list of sub-projects contained in the target material to obtain a target sub-project; obtain a list of materials to be submitted for review for the target sub-project from the material information table of the sub-project; obtain a set of associated materials to be submitted for review for the target material based on the list of materials to be submitted for review for the target sub-project; and obtain a set of mutually exclusive materials to be submitted for review for the target material based on the complement of the set of associated materials to be submitted for review, wherein the materials to be submitted for review in the set of mutually exclusive materials to be submitted for review and the target material form a mutually exclusive relationship.

[0114] In one embodiment, the quota quantity of materials to be submitted for review is limited by multiple constraints, including pre-materials and post-materials that need to be constrained. The post-materials are derived from the pre-materials. The correction module 302 is further configured to: filter multiple constraints according to the coverage threshold and confidence threshold of the constraints to obtain multiple target constraints; obtain an equivalent constraint set based on the target constraints that contain the same materials to be submitted for review, the equivalent constraint set including multiple equivalent constraints; construct a filtered equivalent constraint set based on the equivalent constraint with the largest coverage and the smallest quantity of pre-materials in the equivalent constraint set; obtain a first constraint and a second constraint in the filtered equivalent constraint set; if the set of materials to be submitted for review contained in the first constraint is a subset of the set of materials to be submitted for review contained in the second constraint, and the coverage of the first constraint is greater than the coverage of the second constraint, it is determined that the first constraint and the second constraint have a symbiotic relationship.

[0115] In one embodiment, the correction module 302 is further configured to: for any constraint, obtain the set of materials to be submitted for review based on the preceding and following materials included in the constraint; obtain the number of sub-items that satisfy the constraint based on the set of materials for review, wherein the sub-items that satisfy the constraint are sub-items that cover all materials in the set of materials to be submitted for review; and determine the coverage of the constraint based on the number of sub-items that satisfy the constraint and the total number of sub-items.

[0116] In one embodiment, the comparison module 303 is used to: standardize the revised list of materials to be submitted for review and the list of materials actually submitted for review to obtain a standardized list of materials to be submitted for review and a standardized list of materials actually submitted for review; and generate a risk warning for the list of materials actually submitted for review based on the comparison result of the standardized list of materials to be submitted for review and the standardized list of materials actually submitted for review.

[0117] In one embodiment, the revised list of materials to be submitted for review includes non-standard material names. The comparison module 303 is used to: construct prompt words based on the non-standard material names and standard material names; input the prompt words and the revised list of materials to be submitted for review into the translation model, and obtain the standardized list of materials to be submitted for review output by the translation model.

[0118] In one embodiment, the translation model is used to: convert non-standard material names in the revised list of materials to be submitted for review into standard material names based on prompt words, to obtain a list of materials to be submitted for review with standardized names; standardize the units of measurement in the list of materials to be submitted for review with standardized units of measurement, to obtain a list of materials to be submitted for review with standardized units of measurement; and when there is a reference price for the materials to be submitted for review, add the reference price to the list of materials to be submitted for review with standardized units of measurement, to obtain a standardized list of materials to be submitted for review.

[0119] In one embodiment, the standardized list of materials to be submitted for review includes the materials to be submitted, the revised quota quantity, and the price to be submitted. The standardized list of materials actually submitted for review includes the materials actually submitted, the actual quantity submitted, and the actual price submitted. The comparison module 304 is used to: generate a risk warning that the quantity of materials actually submitted for review is incorrect when the comparison result shows that the materials actually submitted for review match the materials to be submitted for review, but the revised quota quantity does not match the actual quantity submitted for review; generate a risk warning that the price of materials actually submitted for review is incorrect when the comparison result shows that the materials actually submitted for review match the materials to be submitted for review, but the actual price submitted for review does not match the price submitted for review; and generate a risk warning that the materials actually submitted for review are not within the scope of submission when the comparison result shows that the materials actually submitted for review match the materials to be submitted for review.

[0120] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device may include: a processor 410, a communications interface 420, a memory 430, and a communication bus 440. The processor 410, communications interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute a method for auditing submitted materials. This method includes: extracting structured information from the settlement material information in the project's submission documents to obtain a list of submitted materials and a list of materials to be submitted, the list of materials to be submitted including the materials to be submitted and their quota quantities; correcting the quota quantities based on the mutual exclusion and symbiotic relationships of the materials to be submitted, obtaining a corrected list of materials to be submitted; and generating a risk warning for the actual submitted materials list based on the comparison between the actual submitted materials list and the corrected list of materials to be submitted.

[0121] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0122] On the other hand, this application also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the audit method for submitted materials provided by the above methods. The method includes: extracting structured information from the settlement material information in the submission documents of the engineering project to obtain a list of submitted materials and a list of materials to be submitted, wherein the list of materials to be submitted includes the materials to be submitted and their quota quantities; modifying the quota quantities based on the mutual exclusion and symbiotic relationships of the materials to be submitted to obtain a modified list of materials to be submitted; and generating a risk warning for the actual submitted materials list based on the comparison results between the actual submitted materials list and the modified list of materials to be submitted.

[0123] On another front, this application also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the auditing method for submitted materials provided by the above-mentioned methods. The method includes: extracting structured information from the settlement material information in the submission documents of the engineering project to obtain a list of actually submitted materials and a list of materials to be submitted, wherein the list of materials to be submitted includes the materials to be submitted and their quota quantities; modifying the quota quantities based on the mutual exclusion and symbiotic relationships of the materials to be submitted to obtain a modified list of materials to be submitted; and generating a risk warning for the actual submitted materials list based on the comparison results between the actual submitted materials list and the modified list of materials to be submitted.

[0124] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0125] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for reviewing submitted materials, characterized in that, include: Structured information extraction is performed on the settlement material information in the project submission documents to obtain the actual submitted material list and the material list to be submitted. The material list to be submitted includes the material to be submitted and the quota quantity of the material to be submitted. Based on the mutual exclusion relationship and the symbiotic relationship of the materials to be submitted for review, the quota quantity is modified to obtain a modified list of materials to be submitted for review. Based on the comparison between the actual submitted materials list and the revised materials list to be submitted, a risk warning for the actual submitted materials list is generated.

2. The method for reviewing submitted materials according to claim 1, characterized in that, The project comprises multiple sub-projects, and the mutual exclusion relationships are determined based on the following method: Based on at least one of the materials to be submitted for review contained in each of the sub-projects, construct a material information table for the sub-projects; Based on at least one of the sub-items contained in each of the materials to be submitted for review, construct a sub-item information table for the materials; Extract any of the materials to be submitted for review from the sub-item information table of the materials to obtain the target materials and a list of sub-items contained in the target materials; Extract any sub-item from the list of sub-items contained in the target material to obtain the target sub-item; Obtain the list of materials to be submitted for review for the target sub-project from the material information table of the sub-project; Based on the list of materials to be submitted for review for the target sub-project, obtain the set of associated materials to be submitted for review for the target materials; Based on the complement of the associated set of materials to be submitted for review, a set of mutually exclusive materials to be submitted for review for the target material is obtained, and the materials to be submitted for review in the set of mutually exclusive materials to be submitted for review and the target material form the mutual exclusion relationship.

3. The method for reviewing submitted materials according to claim 1, characterized in that, The quota quantity of the materials to be submitted for review is subject to multiple constraints, including pre-materials and post-materials that need to be constrained. The post-materials are derived from the pre-materials. The symbiotic relationship is obtained based on the following method: Based on the coverage threshold and confidence threshold of the constraint conditions, multiple constraint conditions are filtered to obtain multiple target constraint conditions; Based on the target constraint condition that includes the same materials to be submitted for review, obtain the equivalent constraint condition set, which includes multiple equivalent constraint conditions. In the set of equivalent constraints, a filtered set of equivalent constraints is constructed based on the equivalent constraints with the maximum coverage and the minimum quantity of previous materials. In the selected set of equivalent constraints, a first constraint and a second constraint are obtained. If the set of materials to be submitted for review included by the first constraint is a subset of the set of materials to be submitted for review included by the second constraint, and the coverage of the first constraint is greater than the coverage of the second constraint, then it is determined that the first constraint and the second constraint have the symbiotic relationship.

4. The method for reviewing submitted materials according to claim 3, characterized in that, The coverage is determined based on the following method: For any of the constraints, based on the preceding and following materials included in the constraint, obtain the set of materials to be submitted for review under the constraint. Based on the material set with the constraints, obtain the number of sub-items that satisfy the constraints. The sub-items that satisfy the constraints are sub-items that cover all materials in the material set to be submitted for review under the constraints. The coverage of the constraints is determined based on the number of sub-items that satisfy the constraints and the total number of sub-items.

5. The method for reviewing submitted materials according to claim 1, characterized in that, The risk warning generated for the actual submitted materials list based on the comparison results between the revised list of materials to be submitted includes: The revised list of materials to be submitted for review and the list of materials actually submitted for review are standardized to obtain a standardized list of materials to be submitted for review and a standardized list of materials actually submitted for review. Based on the comparison between the standardized list of materials to be submitted for review and the standardized list of materials actually submitted for review, a risk warning for the list of materials actually submitted for review is generated.

6. The method for reviewing submitted materials according to claim 5, characterized in that, The revised list of materials to be submitted for review includes non-standard material names, and the standardized list of materials to be submitted for review is obtained based on the following steps: Construct prompt words based on the non-standard material names and standard material names; Input the prompt words and the revised list of materials to be submitted for review into the translation model, and obtain the standardized list of materials to be submitted for review output by the translation model.

7. The method for reviewing submitted materials according to claim 6, characterized in that, The translation model is used for: Based on the prompt words, the non-standard material names in the revised list of materials to be submitted for review are converted into the standard material names to obtain a standardized list of materials to be submitted for review. The units of measurement in the standardized list of materials to be submitted for review are standardized to obtain a standardized list of materials to be submitted for review. When a reference price exists for the materials to be submitted for review, the reference price is added to the standardized list of materials to be submitted for review, thus obtaining the standardized list of materials to be submitted for review.

8. The method for reviewing submitted materials according to claim 5, characterized in that, The standardized list of materials to be submitted for review includes the materials to be submitted, the revised quota quantity, and the price to be submitted. The standardized list of materials actually submitted for review includes the materials actually submitted, the actual quantity submitted, and the actual price submitted. The risk warning for the actual submitted materials list is generated based on the comparison between the standardized list of materials to be submitted for review and the standardized list of materials actually submitted for review, including: When the comparison result shows that the actual submitted materials and the materials to be submitted match, but the corrected quota quantity does not match the actual submitted quantity, a risk warning is generated indicating that the actual submitted materials have an incorrect submission quantity. When the comparison result shows that the submitted materials and the materials to be submitted match, but the submitted price does not match the submitted price, a risk warning is generated indicating that the submitted price of the submitted materials is incorrect. When the comparison result shows that the submitted materials and the materials to be submitted do not match, a risk warning is generated indicating that the submitted materials are not within the scope of submission.

9. A device for reviewing submitted materials, characterized in that, include: The information extraction module is used to extract structured information from the settlement material information in the submission documents of the project, and obtain the actual submission material list and the material list to be submitted. The material list to be submitted includes the material to be submitted and the quota quantity of the material to be submitted. The correction module is used to correct the quota quantity based on the mutual exclusion relationship and the symbiotic relationship of the materials to be submitted for review, so as to obtain a corrected list of materials to be submitted for review. The comparison module is used to generate a risk warning for the actual submitted materials list based on the comparison results between the actual submitted materials list and the revised materials list to be submitted.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the review method for submitted materials as described in any one of claims 1 to 8.

11. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the review method for submitted materials as described in any one of claims 1 to 8.

12. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the review method for submitted materials as described in any one of claims 1 to 8.

Images