Project review method, device and equipment based on large model and storage medium

Abstract

The invention discloses a project review method and device based on a large model, equipment and a storage medium, and relates to the technical field of artificial intelligence, and the method comprises the steps: carrying out the data preprocessing of a target investment construction project of a preset administrative department based on a preset large model, and obtaining a target scheme summary, target construction content and a target budget amount; determining a target construction content analysis report of the target construction content based on the historical investment construction project in the target knowledge base; analyzing the target scheme summary to obtain a target policy file analysis result and target investment budget reference information; determining a target review result based on a preset large model and the target evaluation list, and determining a target type of to-be-purchased equipment of the target investment construction project; and generating a project review result based on the target budget amount, the target construction content analysis report, the target policy file analysis result, the target investment budget reference information, the target review result and the target type. The accuracy of the project review result can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence, and particularly to a project review method, device, equipment and storage medium based on a large model. Background Art

[0002] With the rapid development of information technology, the investment of administrative departments in informatization projects has been increasing day by day, and these projects cover all aspects from infrastructure construction to public service optimization. However, the traditional project review process faces multiple challenges. First of all, the traditional review method highly relies on manual work, which is not only time-consuming and laborious, but also extremely vulnerable to human factors, resulting in subjectivity and uncertainty in the review results. Secondly, due to the complexity and diversity of informatization projects, information asymmetry and lack of flexibility further reduce work efficiency. In addition, the efficient and reasonable allocation of financial funds is crucial to ensure the successful implementation of informatization projects, but the existing review mechanisms are difficult to meet this requirement. At present, it has become an inevitable trend to use modern information technology to improve the intelligent level of the review process. However, most of the existing technology applications based on large models are concentrated in the commercial field, and the exploration in project review is still limited.

[0003] In summary, how to use large model technology to improve the review efficiency of investment construction projects and the objectivity and accuracy of review results is a technical problem that needs to be solved urgently at present. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a project review method, device, equipment and storage medium based on a large model, which can use large model technology to improve the review efficiency of investment construction projects and the objectivity and accuracy of review results. The specific solutions are as follows:

[0005] In the first aspect, the present application provides a project review method based on a large model, including:

[0006] Performing data preprocessing on the target investment construction project of a preset administrative department based on a preset large model to obtain a target plan overview, target construction content and target budget amount corresponding to the target investment construction project;

[0007] Using the preset large model to determine an analysis report of the target construction content corresponding to the target construction content based on historical investment construction projects in a target knowledge base;

[0008] Using the preset large model to analyze the target plan overview based on the target knowledge base to obtain a corresponding target policy document parsing result and target investment budget reference information;

[0009] Determine the target review result corresponding to the target investment and construction project based on the preset large model and the target evaluation list, and determine the target type of the equipment to be purchased corresponding to the target investment and construction project based on the preset large model;

[0010] Generate the project review result corresponding to the target investment and construction project based on the target budget amount, the target construction content analysis report, the target policy document analysis result, the target investment budget reference information, the target review result, and the target type.

[0011] Optionally, perform data preprocessing on the target investment and construction project of the preset administrative department based on the preset large model to obtain the target plan overview and target construction content corresponding to the target investment and construction project, including:

[0012] Use the preset large model to find the target location corresponding to the target keyword in the target investment and construction project, and extract the target paragraph content in the target investment and construction project based on the target location;

[0013] Generate the target plan overview corresponding to the target investment and construction project based on the target paragraph content, and extract the target construction content corresponding to the target investment and construction project based on the preset paragraph structure.

[0014] Optionally, perform data preprocessing on the target investment and construction project of the preset administrative department based on the preset large model to obtain the target budget amount corresponding to the target investment and construction project, including:

[0015] Use the preset large model to determine the content type corresponding to the target investment and construction project;

[0016] If the content type corresponding to the target investment and construction project is text, determine the target budget amount paragraph in the target investment and construction project, and determine the target budget amount corresponding to the target investment and construction project based on the target budget amount paragraph;

[0017] If the content type corresponding to the target investment and construction project is a table and the table types corresponding to the target investment and construction project are the same, extract the target budget amount corresponding to the target investment and construction project based on the preset hierarchical extraction strategy;

[0018] If the content type corresponding to the target investment and construction project is a table and the table types corresponding to the target investment and construction project are different, extract the target budget amount corresponding to the target investment and construction project based on the preset classification extraction strategy.

[0019] Optionally, using the preset large model to determine an analysis report of the target construction content corresponding to the target construction content based on historical investment and construction projects in the target knowledge base, including:

[0020] Using the preset large model to determine the first similarity between each paragraph in each of the historical investment and construction projects in the target knowledge base and the target construction content, and determining a first target similarity exceeding a first preset similarity threshold from the first similarities based on a preset knowledge retrieval service;

[0021] Determining a first target paragraph corresponding to the first target similarity, and determining a target text from the historical investment and construction projects based on the first target paragraph;

[0022] Using the preset large model to perform semantic understanding on the target text and the target construction content to obtain a target similarity analysis result corresponding to the target construction content, and determining the analysis report of the target construction content corresponding to the target construction content based on the target similarity analysis result.

[0023] Optionally, determining the analysis report of the target construction content corresponding to the target construction content based on the target similarity analysis result includes:

[0024] Using the preset large model to extract target function points corresponding to the target construction content, and determining each initial paragraph from each of the historical investment and construction projects in the target knowledge base based on the target function points;

[0025] Determining target administrative content and a target construction unit corresponding to the target construction content, determining a second target paragraph in each of the initial paragraphs based on the target administrative content, and determining a third target paragraph in each of the initial paragraphs based on the target construction unit;

[0026] Determining a target historical investment and construction project in the historical investment and construction projects based on the second target paragraph and the third target paragraph, and determining the analysis report of the target construction content corresponding to the target construction content based on the target similarity analysis result and the target historical investment and construction project.

[0027] Optionally, using the preset large model to analyze the target scheme overview based on the target knowledge base to obtain a corresponding target policy document parsing result, including:

[0028] Using the preset large model to extract keywords corresponding to the target scheme overview, and retrieving the target knowledge base based on the keywords to determine a second similarity between each policy document paragraph in the target knowledge base and the target scheme overview;

[0029] Determine a second target similarity that exceeds the second preset similarity threshold from the second similarity, and determine the target policy document paragraph corresponding to the second target similarity and the target policy document content corresponding to the target policy document paragraph;

[0030] Determine the target policy document parsing result corresponding to the target solution overview based on the target policy document content.

[0031] Optionally, use the preset large model to analyze the target solution overview based on the target knowledge base to obtain corresponding target investment budget reference information, including:

[0032] Use the preset large model to extract the target construction plan summary corresponding to the target solution overview, and retrieve the target knowledge base based on the target construction plan summary to obtain the target policy name corresponding to the target construction plan summary;

[0033] Determine the target investment time and target investment amount corresponding to the target policy name based on the target knowledge base, and determine the target investment budget reference information corresponding to the target solution overview based on the target investment time and the target investment amount.

[0034] In a second aspect, the present application provides a project review device based on a large model, including:

[0035] A data preprocessing module, configured to perform data preprocessing on the target investment construction project of a preset administrative department based on a preset large model to obtain a target solution overview, target construction content, and target budget amount corresponding to the target investment construction project;

[0036] A target construction content analysis report determination module, configured to use the preset large model to determine a target construction content analysis report corresponding to the target construction content based on historical investment construction projects in the target knowledge base;

[0037] A target solution overview analysis module, configured to use the preset large model to analyze the target solution overview based on the target knowledge base to obtain corresponding target policy document parsing results and target investment budget reference information;

[0038] A target type determination module, configured to determine a target review result corresponding to the target investment construction project based on the preset large model and a target evaluation list, and determine the target type of the equipment to be purchased corresponding to the target investment construction project based on the preset large model;

[0039] A project review result generation module, configured to generate a project review result corresponding to the target investment and construction project based on the target budget amount, the target construction content analysis report, the target policy document parsing result, the target investment budget reference information, the target review result, and the target type.

[0040] In a third aspect, the present application provides an electronic device, including:

[0041] A memory, configured to store a computer program;

[0042] A processor, configured to execute the computer program to implement the foregoing project review method based on a large model.

[0043] In a fourth aspect, the present application provides a computer-readable storage medium, configured to store a computer program; wherein, when the computer program is executed by a processor, the foregoing project review method based on a large model is implemented.

[0044] In this application, first, data preprocessing is performed on the target investment and construction project of a preset administrative department based on a preset large model to obtain the target plan overview, target construction content, and target budget amount corresponding to the target investment and construction project. Then, the preset large model is used to determine the target construction content analysis report corresponding to the target construction content based on the historical investment and construction projects in the target knowledge base. Next, the preset large model is used to analyze the target plan overview based on the target knowledge base to obtain the corresponding target policy document parsing result and target investment budget reference information. After that, the target review result corresponding to the target investment and construction project is determined based on the preset large model and the target evaluation list, and the target type of the equipment to be purchased corresponding to the target investment and construction project is determined based on the preset large model. Finally, the project review result corresponding to the target investment and construction project is generated based on the target budget amount, the target construction content analysis report, the target policy document parsing result, the target investment budget reference information, the target review result, and the target type. As can be seen from the above, in this application, first, the preset large model is used to determine the target plan overview, target construction content, and target budget amount corresponding to the target investment and construction project. Then, based on the historical investment and construction projects in the target knowledge base, the target construction content analysis report corresponding to the target construction content is determined. Next, the target policy document parsing result and target investment budget reference information are obtained by analyzing the target plan overview. After that, the target review result corresponding to the target investment and construction project, as well as the target type of the equipment to be purchased corresponding to the target investment and construction project, are determined. Finally, the project review result corresponding to the target investment and construction project is generated based on the target budget amount, the target construction content analysis report, the target policy document parsing result, the target investment budget reference information, the target review result, and the target type. In this way, this application realizes the intelligent analysis of investment and construction projects through a preset large model, automatically generates a report containing the review result, greatly reduces the workload of manual review, and improves the review efficiency. At the same time, due to the adoption of deep learning technology, the model can learn a large amount of historical data and continuously optimize its evaluation ability, thereby improving the accuracy and objectivity of the review result. In this way, this application can improve the efficiency and quality of project review work, provide a scientific basis for project decision-making, help optimize resource allocation, and promote the effective implementation of investment and construction projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only the embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on the provided drawings.

[0046] Figure 1A flowchart of a project review method based on a large model provided by this application;

[0047] Figure 2 A specific data extraction flowchart provided by this application;

[0048] Figure 3 A specific flowchart for analyzing the scheme overview provided by this application;

[0049] Figure 4 A specific project review flowchart provided by this application;

[0050] Figure 5 A specific flowchart of a project review method based on a large model provided by this application;

[0051] Figure 6 A schematic structural diagram of a project review device based on a large model provided by this application;

[0052] Figure 7 A structural diagram of an electronic device provided by this application. Specific implementation manners

[0053] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0054] With the rapid development of information technology, the investment of administrative departments in informatization projects has been increasing day by day, and these projects cover all aspects from infrastructure construction to public service optimization. However, the traditional project review process faces multiple challenges. First, the traditional review method highly relies on manual work, which is not only time-consuming and laborious, but also extremely vulnerable to human factors, resulting in subjectivity and uncertainty in the review results. Second, due to the complexity and diversity of informatization projects, information asymmetry and lack of flexibility further reduce work efficiency. In addition, the efficient and reasonable allocation of financial funds is crucial to ensure the successful implementation of informatization projects, but the existing review mechanisms are difficult to meet this requirement. At present, it has become an inevitable trend to use modern information technology to improve the intelligent level of the review process. However, most of the existing technology applications based on large models are concentrated in the commercial field, and the exploration in project review is still limited. Therefore, this application provides a project review solution based on a large model, which can use large model technology to improve the review efficiency of investment construction projects and the objectivity and accuracy of review results.

[0055] See Figure 1As shown in the figure, an embodiment of the present invention discloses a project review method based on a large model, which may include:

[0056] Step S11: Perform data preprocessing on the target investment and construction project of a preset administrative department based on a preset large model to obtain a target plan overview, target construction content, and target budget amount corresponding to the target investment and construction project.

[0057] In this embodiment, first, a large model dedicated to investment and construction project review is constructed through deep learning and training by integrating multi-source data such as policy documents, historical investment and construction projects, and bidding documents based on the Hairo Government Affairs Large Model. Refer to Figure 2 As shown in the figure, in order to achieve efficient data processing and analysis, the system first performs text preprocessing on the incoming target investment and construction project, extracts the core content, and converts it into structured data. Specifically, the above-mentioned data preprocessing of the target investment and construction project of a preset administrative department based on a preset large model to obtain the target plan overview and target construction content corresponding to the target investment and construction project may include: first, using the preset large model to find the target location corresponding to the target keyword in the target investment and construction project based on the target keyword, and extracting the target paragraph content in the target investment and construction project based on the target location; then generating the target plan overview corresponding to the target investment and construction project based on the target paragraph content, and extracting the target construction content corresponding to the target investment and construction project based on a preset paragraph structure. Specifically, first, search for keywords such as "construction objectives" and "overall objectives" to obtain the target location corresponding to the target keyword, and extract the relevant paragraph content in the target investment and construction project based on the target location to obtain the target paragraph content. Then, generate a concise and clear target plan overview based on the target paragraph content. At the same time, in order to ensure the integrity of the plan content and provide accurate basic data for subsequent analysis, this embodiment can extract specific construction content from the target investment and construction project and maintain the original cascading paragraph structure.

[0058] It should be noted that in order to improve the efficiency and accuracy of processing large-scale budget data, refer to Figure 2As shown, data preprocessing is performed on the target investment and construction project of the preset administrative department based on the preset large model to obtain the target budget amount corresponding to the target investment and construction project, which may include: First, use the preset large model to determine the content type corresponding to the target investment and construction project. In the first specific implementation manner, if the content type corresponding to the target investment and construction project is text, determine the target budget amount paragraph in the target investment and construction project, and determine the target budget amount corresponding to the target investment and construction project based on the target budget amount paragraph. That is, when the content type corresponding to the target investment and construction project is text, relevant budget paragraphs in the target investment and construction project can be directly extracted and summarized to obtain the target budget amount. In the second specific implementation manner, if the content type corresponding to the target investment and construction project is a table and the table types corresponding to the target investment and construction project are the same, extract the target budget amount corresponding to the target investment and construction project based on the preset hierarchical extraction strategy. That is, when the content type corresponding to the target investment and construction project is a table and the table types are the same, hierarchical extraction can be performed according to the size of the table. For small tables, the budget amount can be directly extracted, and for large tables, the budget amount can be extracted in modules, ensuring that the extracted content meets the actual budget requirements. In the third specific implementation manner, if the content type corresponding to the target investment and construction project is a table and the table types corresponding to the target investment and construction project are different, extract the target budget amount corresponding to the target investment and construction project based on the preset classification extraction strategy. That is, when the content type corresponding to the target investment and construction project is a table and the table types are different. For example, if the tables are the total investment budget table and the hardware equipment purchase list, the total investment budget amount can be directly extracted from the total investment budget table, and the hardware purchase budget amount can be extracted from the hardware equipment purchase list.

[0059] Step S12: Use the preset large model to determine the target construction content analysis report corresponding to the target construction content based on the historical investment and construction projects in the target knowledge base.

[0060] In this embodiment, the method for determining the analysis report of the target construction content corresponding to the target construction content based on the historical investment and construction projects in the target knowledge base by using the preset large model may include: First, use the preset large model to determine the first similarity between each paragraph in each of the historical investment and construction projects in the target knowledge base and the target construction content, and determine the first target similarity exceeding the first preset similarity threshold from the first similarities based on the preset knowledge retrieval service; then determine the first target paragraph corresponding to the first target similarity, and determine the target text from the historical investment and construction projects based on the first target paragraph; finally, use the preset large model to perform semantic understanding on the target text and the target construction content to obtain the target similarity analysis result corresponding to the target construction content, and determine the analysis report of the target construction content corresponding to the target construction content based on the target similarity analysis result. Specifically, through a knowledge retrieval service, such as Elasticsearch, each paragraph in the target knowledge base can be retrieved based on the first similarity between each paragraph in each historical investment and construction project in the target knowledge base and the target construction content, so as to obtain the first target paragraph exceeding the first preset similarity threshold. Then, the target text can be determined according to the first target paragraph, and the preset large model can be used to perform semantic understanding on the target text and the target construction content, and output the duplicate analysis result of the target construction content and the target text, that is, the target similarity analysis result.

[0061] It should be noted that the above-mentioned analysis report of the target construction content corresponding to the target construction content determined based on the target similarity analysis result may include: First, use the preset large model to extract the target function points corresponding to the target construction content, and determine each initial paragraph from each of the historical investment and construction projects in the target knowledge base based on the target function points; then determine the target administrative content and the target construction unit corresponding to the target construction content, and determine the second target paragraph in each of the initial paragraphs based on the target administrative content, and determine the third target paragraph in each of the initial paragraphs based on the target construction unit; finally, determine the target historical investment and construction project in the historical investment and construction projects based on the second target paragraph and the third target paragraph, and determine the analysis report of the target construction content corresponding to the target construction content based on the target similarity analysis result and the target historical investment and construction project. Specifically, first, the target function points corresponding to the target construction content can be extracted, and each initial paragraph matching the target function points can be determined from each of the historical investment and construction projects in the target knowledge base through the knowledge retrieval interface. Then, determine the target administrative content and the target construction unit corresponding to the target construction content, and determine the second target paragraph from each of the initial paragraphs based on the target administrative content, and determine the third target paragraph from each of the initial paragraphs based on the target construction unit. Then, the historical investment and construction project corresponding to the second target paragraph and the third target paragraph can be determined to obtain the target historical investment and construction project. In this way, by determining the target administrative content and the target construction unit corresponding to the target construction content, the target historical investment and construction project matching the target administrative content and the target construction unit can be obtained, providing a reference for subsequent decision-making. At the same time, in this embodiment, the function points of the target construction content can be checked for duplication to obtain the duplication check result corresponding to the target construction content. Specifically, first, the key elements of the target construction content are extracted, and then the similarity between different key elements is judged. If the similarity is higher than the similarity threshold, the similar content is returned; if the similarity is lower than the similarity threshold, it is marked as unique content, where the similarity threshold can be set to 0.75. Finally, the analysis report of the target construction content and the corresponding suggestions can be generated based on the target similarity analysis result, the target historical investment and construction project, and the duplication check result.

[0062] Step S13: Use the preset large model to analyze the target program overview based on the target knowledge base to obtain the corresponding target policy document parsing result and target investment budget reference information.

[0063] In this embodiment, similar policy document paragraphs can be retrieved from the knowledge base according to the target program overview, and highly relevant content can be returned. See Figure 3As shown in the figure, the above analysis of the target program overview based on the target knowledge base using the preset large model to obtain the corresponding target policy document parsing result may include: First, use the preset large model to extract the keywords corresponding to the target program overview, and retrieve the target knowledge base based on the keywords to determine the second similarity between each policy document paragraph in the target knowledge base and the target program overview; then determine the second target similarity that exceeds the second preset similarity threshold from the second similarities, and determine the target policy document paragraph corresponding to the second target similarity and the target policy document content corresponding to the target policy document paragraph; finally, determine the target policy document parsing result corresponding to the target program overview based on the target policy document content. Specifically, first perform content analysis on the target program overview to extract the keywords of the target program overview, then retrieve the target knowledge base based on the keywords, and based on the second similarity between each policy document paragraph in the target knowledge base and the target program overview, determine the target policy document paragraph that exceeds the second preset similarity threshold, and return the target policy document content corresponding to the target policy document paragraph. Then, perform policy docking and project compliance review on the target program overview based on the target policy document content, and output the target policy document parsing result corresponding to the target program overview. In this way, it can be ensured that the investment and construction project can be adjusted according to the latest policy requirements.

[0064] See Figure 3 As shown in the figure, in order to quickly obtain historical investment data related to the target investment and construction project, the above analysis of the target program overview based on the target knowledge base using the preset large model to obtain the corresponding target investment budget reference information may include: First, use the preset large model to extract the target construction plan outline corresponding to the target program overview, and retrieve the target knowledge base based on the target construction plan outline to obtain the target policy name corresponding to the target construction plan outline; then determine the target investment time and target investment amount corresponding to the target policy name based on the target knowledge base, and determine the target investment budget reference information corresponding to the target program overview based on the target investment time and the target investment amount. Specifically, first perform content analysis on the target program overview to extract the target construction plan outline corresponding to the target program overview, and retrieve the target knowledge base to obtain the target policy name corresponding to the target construction plan outline, then obtain the target investment time and target investment amount, so as to determine the corresponding target investment budget reference information based on the target investment time and target investment amount. In this way, relevant policy names, investment time, and investment amounts can be retrieved from the target knowledge base, providing an important basis for the investment decision-making of the project.

[0065] Step S14: Determine the target review result corresponding to the target investment and construction project based on the preset large model and the target evaluation list, and determine the target type of the equipment to be purchased corresponding to the target investment and construction project based on the preset large model.

[0066] In this embodiment, to ensure the compliance of the equipment, as shown in Figure 4 , first, the equipment information in the target investment and construction project can be checked based on the preset large model, and it can be determined whether the model of the equipment is in the preset XC evaluation list. If the model of the equipment is not in the preset XC evaluation list, it is marked as "does not meet the XC evaluation list"; if the model of the equipment is in the preset XC evaluation list, it is marked as "meets"; if the model of the equipment is missing, a prompt is given to supplement the model. Finally, the target review result corresponding to the target investment and construction project and corresponding suggestions can be output.

[0067] It should be noted that, to accurately identify the requirements of software and hardware equipment and provide data support for the implementation of the project, in this embodiment, it can be identified whether each piece of equipment in the equipment purchase list corresponding to the target investment and construction project is software or hardware, and the corresponding identification results can be obtained by matching with the Xinchuang national measurement list and the component library. In this way, in this embodiment, not only the accuracy of the equipment list is ensured, but also the efficient progress of subsequent procurement, management, and review is facilitated.

[0068] Step S15: Generate the project review result corresponding to the target investment and construction project based on the target budget amount, the target construction content analysis report, the target policy document analysis result, the target investment budget reference information, the target review result, and the target type.

[0069] In this embodiment, the project review result can be generated based on the target budget amount, the target construction content analysis report, the target policy document analysis result, the target investment budget reference information, the target review result, and the target type obtained by analyzing the target investment and construction project. It can be seen that in this embodiment, an efficient retrieval mechanism is established in the knowledge base, and based on the preset large model combined with the efficient knowledge retrieval technology, the input investment and construction content can be quickly identified to obtain the corresponding analysis results. It can be understood that the establishment and maintenance of the knowledge base require continuous updating of historical data and ensuring the accuracy and timeliness of the data.

[0070] It should be noted that in this embodiment, a three - layer architecture system is designed, including a large - model layer, a background service layer, and a front - end page layer, providing an intuitive and easy - to - use operation interface for users. The function modules cover the whole process from project upload, project review, to result confirmation, and support the automatic generation and export of review reports. Among them, the large - model layer can be used to receive the input content of investment construction projects and perform intelligent analysis on the content; the background service layer is responsible for data processing, storage, and data interaction with the front - end page layer; the front - end page layer provides an intuitive and easy - to - use operation interface for users, supports custom retrieval conditions, and displays review results. Specifically, a large model is trained using deep - learning algorithms, which can automatically identify similar items, unreasonable items, and incorrect items. The large - model layer can provide a knowledge - retrieval service for quickly retrieving historical project paragraphs similar to the input content in the target knowledge base; and the data - preprocessing unit of the large - model layer can convert the incoming target investment construction project into structured data. The background service layer includes an efficient database management system and supports high - concurrency queries to ensure fast response under a large amount of data; a data - integration module for integrating data from multiple sources, such as policy documents, historical investment construction projects, and bidding documents, etc.; a report - generation module for automatically generating review reports based on detailed data analysis and evaluation models. The front - end page layer includes an operation interface for allowing users to upload target investment construction projects and set retrieval conditions; a result - display area for displaying review results and relevant suggestions in the form of lists or cards; and a special - review function for ensuring that all investment construction projects comply with the current relevant national standards and regulations.

[0071] As can be seen from the above, referring to Figure 5As shown in the figure, in this embodiment, first, data preprocessing is performed on the target investment and construction project of the preset administrative department based on the preset large model to obtain the target plan overview, target construction content, and target budget amount corresponding to the target investment and construction project; then, the preset large model is used to determine the target construction content analysis report corresponding to the target construction content based on the historical investment and construction projects in the target knowledge base; then, the preset large model is used to analyze the target plan overview based on the target knowledge base to obtain the corresponding target policy document analysis result and target investment budget reference information; then, based on the preset large model and the target evaluation list, the target review result corresponding to the target investment and construction project is determined, and the target type of the equipment to be purchased corresponding to the target investment and construction project is determined based on the preset large model; finally, the project review result corresponding to the target investment and construction project is generated based on the target budget amount, the target construction content analysis report, the target policy document analysis result, the target investment budget reference information, the target review result, and the target type. As can be seen from the above, in this embodiment, first, the preset large model is used to determine the target plan overview, target construction content, and target budget amount corresponding to the target investment and construction project, then, based on the historical investment and construction projects in the target knowledge base, the target construction content analysis report corresponding to the target construction content is determined, then, the target policy document analysis result and target investment budget reference information are obtained by analyzing the target plan overview, then, the target review result corresponding to the target investment and construction project, and the target type of the equipment to be purchased corresponding to the target investment and construction project are determined, and finally, the project review result corresponding to the target investment and construction project is generated based on the target budget amount, the target construction content analysis report, the target policy document analysis result, the target investment budget reference information, the target review result, and the target type. In this way, in this embodiment, the intelligent analysis of the investment and construction project is realized through the preset large model, and a report including the review result is automatically generated, greatly reducing the workload of manual review and improving the review efficiency. At the same time, due to the adoption of deep learning technology, the model can learn a large amount of historical data and continuously optimize its evaluation ability, thereby improving the accuracy and objectivity of the review result. In this way, this embodiment can improve the efficiency and quality of the project review work, provide a scientific basis for project decision-making, help optimize resource allocation, and promote the effective implementation of investment and construction projects.

[0072] Correspondingly, referring to Figure 6 As shown in the figure, an embodiment of the present application further provides a project review device based on a large model, which may include:

[0073] A data preprocessing module 11, configured to perform data preprocessing on the target investment and construction project of the preset administrative department based on the preset large model to obtain the target plan overview, target construction content, and target budget amount corresponding to the target investment and construction project;

[0074] The target construction content analysis report determination module 12 is used to determine the target construction content analysis report corresponding to the target construction content by using the preset large model based on the historical investment construction projects in the target knowledge base;

[0075] The target plan overview analysis module 13 is used to analyze the target plan overview based on the target knowledge base by using the preset large model to obtain the corresponding target policy document analysis result and target investment budget reference information;

[0076] The target type determination module 14 is used to determine the target review result corresponding to the target investment construction project based on the preset large model and the target evaluation list, and determine the target type of the equipment to be purchased corresponding to the target investment construction project based on the preset large model;

[0077] The project review result generation module 15 is used to generate the project review result corresponding to the target investment construction project based on the target budget amount, the target construction content analysis report, the target policy document analysis result, the target investment budget reference information, the target review result and the target type.

[0078] As can be seen from the above, in this application, first, data preprocessing is performed on the target investment and construction project of the preset administrative department based on the preset large model to obtain the target plan overview, target construction content, and target budget amount corresponding to the target investment and construction project; then, the preset large model is used to determine the target construction content analysis report corresponding to the target construction content based on the historical investment and construction projects in the target knowledge base; then, the preset large model is used to analyze the target plan overview based on the target knowledge base to obtain the corresponding target policy document parsing result and target investment budget reference information; then, based on the preset large model and the target evaluation list, the target review result corresponding to the target investment and construction project is determined, and the target type of the equipment to be purchased corresponding to the target investment and construction project is determined based on the preset large model; finally, the project review result corresponding to the target investment and construction project is generated based on the target budget amount, the target construction content analysis report, the target policy document parsing result, the target investment budget reference information, the target review result, and the target type. As can be seen from the above, in this application, first, the preset large model is used to determine the target plan overview, target construction content, and target budget amount corresponding to the target investment and construction project, then, based on the historical investment and construction projects in the target knowledge base, the target construction content analysis report corresponding to the target construction content is determined, then, the target policy document parsing result and target investment budget reference information are obtained by analyzing the target plan overview, then, the target review result corresponding to the target investment and construction project, and the target type of the equipment to be purchased corresponding to the target investment and construction project are determined, and finally, the project review result corresponding to the target investment and construction project is generated based on the target budget amount, the target construction content analysis report, the target policy document parsing result, the target investment budget reference information, the target review result, and the target type. In this way, this application realizes the intelligent analysis of investment and construction projects through the preset large model, automatically generates a report including the review result, greatly reduces the workload of manual review, and improves the review efficiency. At the same time, due to the adoption of deep learning technology, the model can learn a large amount of historical data and continuously optimize its evaluation ability, thereby improving the accuracy and objectivity of the review result. In this way, this application can improve the efficiency and quality of project review work, provide a scientific basis for project decision-making, help optimize resource allocation, and promote the effective implementation of investment and construction projects.

[0079] In some specific embodiments, the data preprocessing module 11 may include:

[0080] A target paragraph content extraction unit, configured to use the preset large model to find the target position corresponding to the target keyword in the target investment and construction project based on the target keyword, and extract the target paragraph content in the target investment and construction project based on the target position;

[0081] The target construction content extraction unit is used to generate the target plan overview corresponding to the target investment and construction project based on the target paragraph content, and extract the target construction content corresponding to the target investment and construction project based on the preset paragraph structure.

[0082] In some specific embodiments, the data preprocessing module 11 may include:

[0083] The content type determination unit is used to determine the content type corresponding to the target investment and construction project by using the preset large model;

[0084] The first target budget amount determination unit is used to, if the content type corresponding to the target investment and construction project is text, determine the target budget amount paragraph in the target investment and construction project, and determine the target budget amount corresponding to the target investment and construction project based on the target budget amount paragraph;

[0085] The second target budget amount determination unit is used to, if the content type corresponding to the target investment and construction project is a table and the table types corresponding to the target investment and construction project are the same, extract the target budget amount corresponding to the target investment and construction project based on the preset hierarchical extraction strategy;

[0086] The third target budget amount determination unit is used to, if the content type corresponding to the target investment and construction project is a table and the table types corresponding to the target investment and construction project are different, extract the target budget amount corresponding to the target investment and construction project based on the preset classification extraction strategy.

[0087] In some specific embodiments, the target construction content analysis report determination module 12 may include:

[0088] The first target similarity determination sub-module is used to determine the first similarity between each paragraph in the target knowledge base of the historical investment and construction projects and the target construction content by using the preset large model, and determine the first target similarity exceeding the first preset similarity threshold from the first similarities based on the preset knowledge retrieval service;

[0089] The target text determination sub-module is used to determine the first target paragraph corresponding to the first target similarity, and determine the target text from the historical investment and construction projects based on the first target paragraph;

[0090] The target construction content analysis report determination sub-module is used to perform semantic understanding on the target text and the target construction content by using the preset large model to obtain the target similarity analysis result corresponding to the target construction content, and determine the target construction content analysis report corresponding to the target construction content based on the target similarity analysis result.

[0091] In some specific embodiments, the sub-module for determining the analysis report of the target construction content may include:

[0092] An initial paragraph determination unit, configured to use the preset large model to extract target function points corresponding to the target construction content, and determine each initial paragraph from each of the historical investment and construction projects in the target knowledge base based on the target function points;

[0093] A third target paragraph determination unit, configured to determine the target administrative content and the target construction unit corresponding to the target construction content, and determine a second target paragraph in each of the initial paragraphs based on the target administrative content, and determine a third target paragraph in each of the initial paragraphs based on the target construction unit;

[0094] A target construction content analysis report determination unit, configured to determine a target historical investment and construction project in the historical investment and construction projects based on the second target paragraph and the third target paragraph, and determine the analysis report of the target construction content corresponding to the target construction content based on the target similarity analysis result and the target historical investment and construction project.

[0095] In some specific embodiments, the target solution overview analysis module 13 may include:

[0096] A second similarity determination unit, configured to use the preset large model to extract keywords corresponding to the target solution overview, and retrieve the target knowledge base based on the keywords to determine the second similarity between each policy document paragraph in the target knowledge base and the target solution overview;

[0097] A target policy document content determination unit, configured to determine a second target similarity exceeding a second preset similarity threshold from the second similarities, and determine the target policy document paragraph corresponding to the second target similarity and the target policy document content corresponding to the target policy document paragraph;

[0098] A target policy document parsing result determination unit, configured to determine the target policy document parsing result corresponding to the target solution overview based on the target policy document content.

[0099] In some specific embodiments, the target solution overview analysis module 13 may include:

[0100] A target policy name determination unit, configured to use the preset large model to extract the target construction plan outline corresponding to the target solution overview, and retrieve the target knowledge base based on the target construction plan outline to obtain the target policy name corresponding to the target construction plan outline;

[0101] A target investment budget reference information determination unit, configured to determine a target investment time and a target investment amount corresponding to the target policy name based on the target knowledge base, and determine target investment budget reference information corresponding to the target plan overview based on the target investment time and the target investment amount.

[0102] Furthermore, an embodiment of the present application also discloses an electronic device. Figure 7 It is a structural diagram of an electronic device 20 shown according to an exemplary embodiment, and the content in the figure should not be considered as any limitation to the scope of use of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. Among them, the memory 22 is used to store a computer program, and the computer program is loaded and executed by the processor 21 to implement the relevant steps in the project review method based on a large model disclosed in any of the foregoing embodiments. In addition, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0103] In this embodiment, the power supply 23 is used to provide operating voltages for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol it follows is any communication protocol applicable to the technical solution of the present application, and no specific limitation is imposed on it here; the input / output interface 25 is used to obtain external input data or output data to the outside, and its specific interface type can be selected according to specific application needs, and no specific limitation is made here.

[0104] In addition, as a carrier for resource storage, the memory 22 may be a read-only memory, a random access memory, a disk, or an optical disc, etc., and the resources stored thereon may include an operating system 221, a computer program 222, etc., and the storage method may be short-term storage or permanent storage.

[0105] Among them, the operating system 221 is used to manage and control each hardware device on the electronic device 20 and the computer program 222, and it may be Windows Server, Netware, Unix, Linux, etc. In addition to the computer program that can be used to complete the project review method based on a large model executed by the electronic device 20 disclosed in any of the foregoing embodiments, the computer program 222 may further include computer programs that can be used to complete other specific tasks.

[0106] Further, the present application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned project review method based on a large model. For the specific steps of this method, reference can be made to the corresponding content disclosed in the foregoing embodiments, and details will not be elaborated herein.

[0107] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the various embodiments, reference can be made to each other. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and reference can be made to the description in the method part for related parts.

[0108] Those skilled in the art can further realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the components and steps of the examples have been generally described according to their functions in the above description. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0109] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein can be directly implemented by hardware, a software module executed by a processor, or a combination of the two. The software module can be placed in a random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium well-known in the technical field.

[0110] Finally, it should also be noted that in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising a..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0111] The above has introduced the technical solution provided by the present application in detail. Specific examples are used herein to elaborate on the principle and implementation manner of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those of ordinary skill in the art, according to the idea of the present application, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A project review method based on a large model, characterized in that Including: Performing data preprocessing on the target investment and construction project of a preset administrative department based on a preset large model to obtain a target plan overview, target construction content, and target budget amount corresponding to the target investment and construction project; Using the preset large model to determine a target construction content analysis report corresponding to the target construction content based on historical investment and construction projects in the target knowledge base; Using the preset large model to analyze the target plan overview based on the target knowledge base to obtain corresponding target policy document analysis results and target investment budget reference information; Determining a target review result corresponding to the target investment and construction project based on the preset large model and a target evaluation list, and determining a target type of equipment to be purchased corresponding to the target investment and construction project based on the preset large model; Generating a project review result corresponding to the target investment and construction project based on the target budget amount, the target construction content analysis report, the target policy document analysis results, the target investment budget reference information, the target review result, and the target type.

2. The project review method based on a large model according to claim 1, wherein The performing data preprocessing on the target investment and construction project of a preset administrative department based on a preset large model to obtain a target plan overview and target construction content corresponding to the target investment and construction project includes: Using the preset large model to search for the target investment and construction project based on target keywords to obtain a target position corresponding to the target keywords, and extracting target paragraph content in the target investment and construction project based on the target position; Generating the target plan overview corresponding to the target investment and construction project based on the target paragraph content, and extracting the target construction content corresponding to the target investment and construction project based on a preset paragraph structure.

3. The project review method based on the large model according to claim 1, wherein The performing data preprocessing on the target investment and construction project of a preset administrative department based on a preset large model to obtain a target budget amount corresponding to the target investment and construction project includes: Using the preset large model to determine a content type corresponding to the target investment and construction project; If the content type corresponding to the target investment and construction project is text, determining a target budget amount paragraph in the target investment and construction project, and determining the target budget amount corresponding to the target investment and construction project based on the target budget amount paragraph; If the content type corresponding to the target investment and construction project is a table and the table types corresponding to the target investment and construction project are the same, extracting the target budget amount corresponding to the target investment and construction project based on a preset hierarchical extraction strategy; If the content type corresponding to the target investment and construction project is a table and the table types corresponding to the target investment and construction project are different, extracting the target budget amount corresponding to the target investment and construction project based on a preset classification extraction strategy.

4. The project review method based on a large model according to claim 1, wherein The using the preset large model to determine a target construction content analysis report corresponding to the target construction content based on historical investment and construction projects in the target knowledge base includes: Using the preset large model to determine the first similarity between each paragraph in each of the historical investment and construction projects in the target knowledge base and the target construction content, and determining, based on a preset knowledge retrieval service, a first target similarity that exceeds a first preset similarity threshold from the first similarities; Determining a first target paragraph corresponding to the first target similarity, and determining target text from the historical investment and construction projects based on the first target paragraph; Using the preset large model to perform semantic understanding on the target text and the target construction content to obtain a target similarity analysis result corresponding to the target construction content, and determining a target construction content analysis report corresponding to the target construction content based on the target similarity analysis result.

5. The project review method based on a large model according to claim 4, wherein The determining the target construction content analysis report corresponding to the target construction content based on the target similarity analysis result includes: Using the preset large model to extract target function points corresponding to the target construction content, and determining each initial paragraph from each of the historical investment and construction projects in the target knowledge base based on the target function points; Determining target administrative content and a target construction unit corresponding to the target construction content, determining a second target paragraph in each of the initial paragraphs based on the target administrative content, and determining a third target paragraph in each of the initial paragraphs based on the target construction unit; Determining a target historical investment and construction project in the historical investment and construction projects based on the second target paragraph and the third target paragraph, and determining the target construction content analysis report corresponding to the target construction content based on the target similarity analysis result and the target historical investment and construction project.

6. The project review method based on a large model according to claim 1, wherein The using the preset large model to analyze the target program overview based on the target knowledge base to obtain a corresponding target policy document analysis result includes: Using the preset large model to extract keywords corresponding to the target program overview, and retrieving the target knowledge base based on the keywords to determine the second similarity between each policy document paragraph in the target knowledge base and the target program overview; Determining a second target similarity that exceeds a second preset similarity threshold from the second similarities, and determining a target policy document paragraph corresponding to the second target similarity and target policy document content corresponding to the target policy document paragraph; Determining the target policy document analysis result corresponding to the target program overview based on the target policy document content.

7. The project review method based on a large model according to any one of claims 1 to 6, characterized in that Using the preset large model to analyze the target program overview based on the target knowledge base to obtain corresponding target investment budget reference information, including: Using the preset large model to extract a target construction plan summary corresponding to the target program overview, and retrieving the target knowledge base based on the target construction plan summary to obtain a target policy name corresponding to the target construction plan summary; Determining a target investment time and a target investment amount corresponding to the target policy name based on the target knowledge base, and determining the target investment budget reference information corresponding to the target program overview based on the target investment time and the target investment amount.

8. A project review device based on a large model, characterized in that, Including: A data preprocessing module for preprocessing data of a target investment and construction project of a preset administrative department based on a preset large model to obtain a target plan overview, target construction content, and target budget amount corresponding to the target investment and construction project; A target construction content analysis report determination module for determining a target construction content analysis report corresponding to the target construction content by using the preset large model based on historical investment and construction projects in a target knowledge base; A target plan overview analysis module for analyzing the target plan overview based on the target knowledge base by using the preset large model to obtain corresponding target policy document analysis results and target investment budget reference information; A target type determination module for determining a target review result corresponding to the target investment and construction project based on the preset large model and a target evaluation list, and determining a target type of equipment to be purchased corresponding to the target investment and construction project based on the preset large model; A project review result generation module for generating a project review result corresponding to the target investment and construction project based on the target budget amount, the target construction content analysis report, the target policy document analysis results, the target investment budget reference information, the target review result, and the target type.

9. An electronic device, characterized in that, The electronic device includes a processor and a memory; wherein, the memory is used for storing a computer program, and the computer program is loaded and executed by the processor to implement the large model-based project review method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, For saving a computer program, the computer program, when executed by a processor, implements the large model-based project review method according to any one of claims 1 to 7.

Images