A feasibility study report matching analysis system and method based on artificial intelligence

Abstract

The present invention relates to the field of data analysis, and specifically to a feasibility study report matching analysis system and method based on artificial intelligence, comprising a data processing module, a project data matching analysis unit, and a threshold judgment module; the data processing module is used to collect approved project feasibility study reports and project application materials from historical projects, and modularly classify the data therein according to functions and content sections and perform structured processing; the project data matching analysis unit is used to analyze the matching degree between project application material sub-data and current project requirements, the matching degree between project application material data module predicted data and actual data, and the comprehensive matching degree between project application material automatically generated data obtained according to the project feasibility study report and actual data; the threshold judgment module is used to perform a first threshold judgment on the project application material sub-data, a second threshold judgment on the project application material data module, and a third threshold judgment on the matching degree of the project feasibility study report.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a feasibility study report matching analysis system and method based on artificial intelligence. Background Art

[0002] With the continuous adjustment and improvement of my country's industrial structure, the number of project applications in various industries is increasing. In project approval, the review of the matching degree between the feasibility study report and the project application is of great significance to the feasibility of the project.

[0003] In the current project review, most of them adopt the manual review method. Before the project application is submitted, the project feasibility study report and the application form are manually checked and verified. There is a lot of room for improvement in the efficiency of project review. In addition, in the matching degree verification between the feasibility study report and the project application form, the matching degree calculation and analysis method is single, the comparison standard of the data indicators in the materials is vague, and the review of the data of each module in the feasibility study report lacks a more detailed judgment.

[0004] Therefore, there is a need for an artificial intelligence-based feasibility study report matching analysis system and method to realize the automatic review and screening function of projects based on the matching degree of project feasibility study reports. Summary of the invention

[0005] The purpose of the present invention is to provide a feasibility study report matching analysis system and method based on artificial intelligence to solve the problems raised in the above background technology.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] A feasibility study report matching analysis method based on artificial intelligence includes the following analysis steps:

[0008] Step S100: Collect feasibility study reports and project application materials from historical project approval materials, perform modular classification and data structuring processing;

[0009] Step S200: Structuring the feasibility study report and project application materials of the current project, setting a first threshold for the structured data according to the requirements of the current project, and performing a first threshold judgment on the feasibility study report and project application materials of the current project;

[0010] Step S300: Analyze the relevance of sub-data of each module in the structured data of the feasibility study report and the project application materials, and then use multivariate linear fitting to train the project application materials data generation model, and use the structured data of the feasibility study report to generate corresponding project application materials data;

[0011] Step S400: Compare the automatically generated data of the project application materials with the actual data to calculate the matching degree of each data module between the feasibility study report and the project application materials, and perform a second threshold judgment on the current project application materials according to the matching results of each data module;

[0012] Step S500: Perform comprehensive matching calculation on the feasibility study report and project application materials that pass the second threshold judgment and perform a third threshold judgment, and analyze the project automatic approval result according to the judgment result;

[0013] In step S100, the approved project materials in the historical projects are collected, and modular classification is performed according to the data functions and content sections of the feasibility study reports and project application materials in the project materials, and the data of each module is grouped and summarized, and the data format is set and stored in modules for structured data processing; the data in the project approval materials are classified and structured for storage, which is convenient for subsequent analysis of data relevance, and also unifies data standards, providing a unified structure data foundation for the training of subsequent project application materials data generation models;

[0014] Structure the feasibility study report data in the historically approved project materials to generate a data set Structuring project application materials to generate data sets

[0015] Among them, f1,f2,…,f m For each classification module of feasibility study report data, It is the data of each module in the feasibility study report; a1, a2, ..., a n Apply for material data for each classification module of the project, is the data of each module of the project application materials; m is the number of feasibility study report data classification modules, and n is the number of project application materials data classification modules; modularizing the data to facilitate the subsequent multi-dimensional judgment and analysis of the project application materials data;

[0016] In step S200, the feasibility study report of the current project and the project application materials are used to generate project structured data using the data classification and processing method in step S100;

[0017] Obtaining the requirement information of each sub-data of each content section of the project application materials in the current project requirements, setting a first threshold value for each module sub-data in the project application materials, and then judging whether each module sub-data meets the data standard in the project requirements information;

[0018] Perform a first threshold judgment on the sub-data of each module, and determine the project with sub-data that fails the first threshold judgment as failed in the automatic preliminary review, and feed back the sub-data that fails the first threshold judgment for manual review; for the project with all sub-data that passes the first threshold judgment, output an automatic preliminary review pass signal, and carry out the subsequent review process;

[0019] In step S300, the data association between the sub-data of each module in the historical project application materials and the sub-data of each module in the feasibility study report is analyzed, and the sub-data of each module in the feasibility study report that affects the data change is screened out for the sub-data of each module in the project application materials. The multivariate linear regression method is used to train the project application material data generation model using the data set F and the data set A described in step S100. For any sub-data y in the project application materials A , the data automatic generation model is as follows:

[0020] y A =α A *X A +C A ±∈ A ;

[0021] Among them, X A Sub-data for project application materials in the feasibility study report A Sub-data vector affected by data changes, α A For X A Each parameter in corresponds to a linear fitting parameter vector, C A For sub-data y A The linear fitting constant parameter, ∈ A For sub-data y A Linear fitting range parameters;

[0022] Subdata y A Linear fitting range parameter ∈ A When selecting values, the actual data of the project application materials in data set A is used to select values ​​so that the data with a proportion of p in data set A falls within ∈ A y when the preceding symbols are positive and negative A When using the model to predict project application material sub-data, use ∈ A When y is positive or negative, A The mean value is taken as the model prediction data of the sub-data;

[0023] By setting item p to determine the sub-data linear fitting range parameters, the influence of abnormal data on the matching results of subsequent data modules is filtered out in all sub-data;

[0024] In step S400, the structured data of the feasibility study report that passes the first threshold in step S200 is obtained, and the predicted values ​​of the sub-data models of each module in the project application materials are calculated and analyzed by the project application materials data generation model, and then compared and analyzed with the actual project application materials, and the matching degree between the predicted data of the project application materials and the actual data is calculated by module classification; for any module a in the project application materials, A , according to the formula:

[0025]

[0026] in, For data module a A The matching degree between the predicted data and the actual data, i is the data module a A Data number, n i For data module a A Medium data volume, For data module a A Neutron data i Actual value, For data module a A Neutron data i Prediction value, ∈ i For data module a A Neutron data i Linear fit range parameter, β i For data module a A Neutron data i The matching weight of

[0027] When calculating the matching degree between the predicted data and the actual data of the data module, the linear fitting range parameters of each sub-data in the data module are used as the metric parameters for calculating the matching degree of the data module to avoid errors in the evaluation results caused by missing data reference values ​​and abnormal data points;

[0028] Use the structured data of the current project feasibility study report to calculate the matching degree of all data modules in the project application materials and generate the matching degree matrix of the current project application materials Setting a second threshold for the matching degree of each module and performing a second threshold judgment;

[0029] The projects with data module matching degree less than the second threshold value are judged as failed automatic review, and the module data that failed automatic review is fed back for manual review; for projects with all module matching degrees greater than the second threshold value of the corresponding module, an automatic review pass signal is output, and the subsequent review process is carried out;

[0030] In the second threshold judgment, the modular classification data is judged for the second time, and the matching degree of each data module is calculated by comprehensively analyzing the deviation value between the predicted data and the actual data of each data module sub-data. Then, according to the matching degree judgment result, the projects with low matching degree of data modules are screened out, so as to realize the review and screening of the data of each content section of the project application materials;

[0031] In step S500, a matching evaluation coefficient is assigned to each data module in the project application materials according to the project type, and a matching evaluation coefficient matrix of the project application materials is generated. Then, according to the formula: Match = M × γ, calculate the comprehensive matching degree Match of the project feasibility study report;

[0032] Sort the matching degree data of all the project application materials of the current project and the corresponding feasibility study reports from large to small, set a third threshold for the passing ratio of all the application materials of the current project and make a third threshold judgment;

[0033] Among the sorted projects, all projects whose project application materials are ranked within the third threshold from largest to smallest are judged to have passed the automatic final review, and a subsequent manual selection process is carried out; all projects whose project application materials are ranked outside the third threshold are judged to have failed the automatic final review, and a signal to resubmit the project application materials is output; the matching degree of each data module of the project application materials is comprehensively considered, the matching degree of the project feasibility report is calculated and analyzed, the project is evaluated and reviewed as a whole, and then by setting the third threshold, the project application materials that have passed the review are selected as the final result of the automatic review.

[0034] An artificial intelligence-based feasibility study report matching analysis system using the artificial intelligence-based feasibility study report matching analysis method described in the above technical solution, the system includes a data processing module, a project data matching analysis unit, and a threshold judgment module;

[0035] The data processing module is used to collect the approved project feasibility study reports and project application materials in historical projects, and modularly classify the data according to the functions and content sections, and perform structured processing; the project data matching analysis unit is used to analyze the matching degree between the sub-data of the project application materials and the requirements of the current project, the matching degree between the predicted data of the project application materials data module and the actual data, and the comprehensive matching degree between the project application materials data automatically generated according to the project feasibility study report and the actual data of the project application materials; the threshold judgment module is used to perform a first threshold judgment on the sub-data of the project application materials, a second threshold judgment on the project application materials data module, and a third threshold judgment on the matching degree of the project feasibility study report;

[0036] The output end of the data processing module is connected to the project data matching degree analysis unit and the input end of the threshold judgment module; the output end of the project data matching degree analysis unit is connected to the input end of the threshold judgment module.

[0037] The data processing module includes a project approval data modularization unit and a project approval data structured unit;

[0038] The project approval data modularization unit is used to modularize and classify the project approval data according to its functions and content sections; the project approval data structured unit is used to structure the project approval data to generate a structured data set for the feasibility study report and a structured data set for the project application materials;

[0039] The output end of the project approval data modularization unit is connected to the input end of the project approval data structured unit; the project approval data structured unit is connected to the project data matching degree analysis unit and the input end of the threshold judgment module.

[0040] The threshold judgment module includes a first threshold judgment unit, a second threshold judgment unit, and a third threshold judgment unit;

[0041] The first threshold judgment unit is used to perform a first threshold judgment on the project application material sub-data according to the current project requirements; the second threshold judgment unit is used to perform a second threshold judgment on the data module according to the project application material data module matching calculation result; the third threshold judgment unit is used to perform a third threshold judgment on the project according to the comprehensive matching result of the project feasibility study report;

[0042] The input end of the first threshold judgment unit is connected to the output end of the data processing module; the input ends of the second threshold judgment unit and the third threshold judgment unit are connected to the output end of the project data matching degree analysis unit.

[0043] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0044] In the present invention, by analyzing the data correlation between each sub-data between the feasibility study report and the project application in the historical data, the project application automatic generation model is trained and the feasibility study report data submitted in the current project is used to automatically generate the project application data, the data connection between the project materials is fully considered, and a data basis is provided for the subsequent feasibility study report matching analysis; through multiple threshold judgments of different types of standards, threshold judgments are made on the sub-data in the project materials, the data module matching degree, and the comprehensive matching degree of the project feasibility study report, respectively, to provide a data basis for the automatic review of the project materials; in the project application material automatic generation model, the data fitting range parameter is introduced according to the historical approval project data, and then the matching degree of the project application material data generated by the feasibility study report data and the actual data is calculated and analyzed, and the historical data is fully utilized for project review and evaluation, so as to improve the interpretability and rationality of the review results. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0046] Figure 1 This is a flow chart of a feasibility study report matching analysis method based on artificial intelligence in the present invention;

[0047] Figure 2 It is a structural diagram of a feasibility study report matching analysis system based on artificial intelligence in the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] See also Figure 1-Figure 2 , the present invention provides a technical solution:

[0050] like Figure 1 As shown, the present invention provides a feasibility study report matching analysis method based on artificial intelligence, comprising the following analysis steps:

[0051] Step S100: Collect feasibility study reports and project application materials from historical project approval materials, perform modular classification and data structuring processing;

[0052] Step S200: Structuring the feasibility study report and project application materials of the current project, setting a first threshold for the structured data according to the requirements of the current project, and performing a first threshold judgment on the feasibility study report and project application materials of the current project;

[0053] Step S300: Analyze the relevance of sub-data of each module in the structured data of the feasibility study report and the project application materials, and then use multivariate linear fitting to train the project application materials data generation model, and use the structured data of the feasibility study report to generate corresponding project application materials data;

[0054] Step S400: Compare the automatically generated data of the project application materials with the actual data to calculate the matching degree of each data module between the feasibility study report and the project application materials, and perform a second threshold judgment on the current project application materials according to the matching results of each data module;

[0055] Step S500: Perform comprehensive matching calculation on the feasibility study report and project application materials that pass the second threshold judgment and perform a third threshold judgment, and analyze the project automatic approval result according to the judgment result;

[0056] In step S100, the approved project materials in the historical projects are collected, and modular classification is performed according to the data functions and content sections of the feasibility study reports and project application materials in the project materials, and the data of each module is grouped and summarized, and the data format is set and stored in modules for structured data processing;

[0057] Structure the feasibility study report data in the historically approved project materials to generate a data set Structuring project application materials to generate data sets

[0058] Among them, f1,f2,...,f m For each classification module of feasibility study report data, It is the data of each module in the feasibility study report; a1, a2, ..., a n Apply for material data for each classification module of the project, is the data of each module of the project application materials; m is the number of feasibility study report data classification modules, and n is the number of project application materials data classification modules;

[0059] In specific implementation, the information in the feasibility study report is divided into different modules according to the content section, and the information data in each module is stored in the same database table. The data in the module is then marked as the fields of the database. The different feasibility study reports and project application materials submitted in the current project are divided into modules and stored as records in each table in the database;

[0060] In step S200, the feasibility study report of the current project and the project application materials are used to generate project structured data using the data classification and processing method in step S100;

[0061] Obtaining the requirement information of each sub-data of each content section of the project application materials in the current project requirements, setting a first threshold value for each module sub-data in the project application materials, and then judging whether each module sub-data meets the data standard in the project requirements information;

[0062] Perform a first threshold judgment on the sub-data of each module, and determine the project with sub-data that fails the first threshold judgment as failed in the automatic preliminary review, and feed back the sub-data that fails the first threshold judgment for manual review; for the project with all sub-data that passes the first threshold judgment, output an automatic preliminary review pass signal, and carry out the subsequent review process;

[0063] In specific implementation, according to the project requirements, a project requirement threshold is set for each sub-data as the first threshold. For example, according to the project time requirements, project fund raising time and other data requirements, the first threshold is set for the project time and fund raising time fields in the structured sub-data to determine whether each sub-data in the project application materials meets the project requirement standards, and the application materials that do not meet the project requirements are screened out in the first round of automatic review;

[0064] In step S300, the data association between the sub-data of each module in the historical project application materials and the sub-data of each module in the feasibility study report is analyzed, and the sub-data of each module in the feasibility study report that affects the data change is screened out for the sub-data of each module in the project application materials. The multivariate linear regression method is used to train the project application material data generation model using the data set F and the data set A described in step S100. For any sub-data y in the project application materials A , the data automatic generation model is as follows:

[0065] y A =α A *X A +C A ±∈ A ;

[0066] Among them, X A Sub-data for project application materials in the feasibility study report A Sub-data vector affected by data changes, α A For X A Each parameter in corresponds to a linear fitting parameter vector, C A For sub-data y A The linear fitting constant parameter, ∈ A For sub-data y A Linear fitting range parameters;

[0067] Subdata y A Linear fitting range parameter ∈ A When selecting values, the actual data of the project application materials in data set A is used to select values ​​so that the data with a proportion of p in data set A falls within ∈ A y when the preceding symbols are positive and negative A When using the model to predict project application material sub-data, use ∈ A When y is positive or negative, A The mean value is taken as the model prediction data of the sub-data;

[0068] In a specific embodiment, when training the model, for any sub-data in the project application materials, the sub-data in the feasibility study report that have an impact on the data are selected for model training. During the training, the model is modified to: y A =α A *X A+C A Perform model training and parameter estimation. After the model is trained, set the data set ratio p and set the linear fitting range parameter ∈ according to the data in data set A. A , ensuring that when ∈ A When the positive and negative values ​​are taken before, the data with a proportion of p in the data set is in y A The value is included in the range and provides constraint criteria for subsequent data module matching calculation;

[0069] In step S400, the structured data of the feasibility study report that passes the first threshold in step S200 is obtained, and the predicted values ​​of the sub-data models of each module in the project application materials are calculated and analyzed by the project application materials data generation model, and then compared and analyzed with the actual project application materials, and the matching degree between the predicted data of the project application materials and the actual data is calculated by module classification; for any module a in the project application materials, A , according to the formula:

[0070]

[0071] in, For data module a A The matching degree between the predicted data and the actual data, i is the data module a A Data number, n i For data module a A Medium data volume, For data module a A Neutron data i Actual value, For data module a A Neutron data i Prediction value, ∈ i For data module a A Neutron data i Linear fit range parameter, β i For data module a A Neutron data i The matching weight of

[0072] Use the structured data of the current project feasibility study report to calculate the matching degree of all data modules in the project application materials and generate the matching degree matrix of the current project application materials Setting a second threshold for the matching degree of each module and performing a second threshold judgment;

[0073] When calculating the matching degree of each data module, if there is non-digital data in the module, determine whether the predicted result of the data is consistent with the actual result. If the data is consistent, determine The overall value is 0. When the data does not match, it is judged The overall value is 1, thereby achieving quantitative judgment of non-digital data;

[0074] The projects with data module matching degree less than the second threshold value are judged as failed automatic review, and the module data that failed automatic review is fed back for manual review; for projects with all module matching degrees greater than the second threshold value of the corresponding module, an automatic review pass signal is output, and the subsequent review process is carried out;

[0075] In a specific embodiment, the current project type and goal are analyzed, and the importance of the sub-data in each data module in the project progress is determined as the matching weight of each sub-data to calculate the matching degree of each data module, and then a comprehensive matching degree evaluation coefficient is assigned to each data module according to the importance of each content section data of the project application materials to the project implementation, and a project application material matching degree evaluation coefficient matrix γ is generated;

[0076] When setting the second threshold, for any data module in the project application materials, calculate The matching degree of the data module when the overall value is 0 is used as the matching degree of the data module when the sub-data in the data module is completely matched, and then the numerical range of the data module matching degree is determined, and a threshold point is selected in the numerical range as the second threshold, and the matching degree threshold judgment and screening are performed on the data modules in all project application materials;

[0077] In step S500, a matching evaluation coefficient is assigned to each data module in the project application materials according to the project type, and a matching evaluation coefficient matrix of the project application materials is generated. Then, according to the formula: Match = M × γ, calculate the comprehensive matching degree Match of the project feasibility study report;

[0078] Sort the matching degree data of all the project application materials of the current project and the corresponding feasibility study reports from large to small, set a third threshold for the passing ratio of all the application materials of the current project and make a third threshold judgment;

[0079] Among the ranked projects, all projects whose project application materials are ranked within the third threshold from largest to smallest are judged to have passed the automatic final review, and a subsequent manual selection process is carried out; all projects whose project application materials are ranked outside the third threshold are judged to have failed the automatic final review, and a signal for resubmitting the project application materials is output;

[0080] In a specific embodiment, a third threshold is set according to the number of application materials submitted for the current project that have passed the automatic review and re-examination stage. Among all the materials that have passed the automatic review and re-examination stage, a corresponding proportion of application materials are selected from large to small according to the comprehensive matching degree of the feasibility study report to pass the automatic review and final review stage, so as to realize the screening and review of project applications according to the matching degree between the feasibility study report and the project application materials. When the third threshold is actually set, it is generally ensured that the number of project materials that pass the automatic review and final review in the third threshold judgment screening is 3 to 5. When the number of application materials that pass the automatic review and re-examination stage is large, the proportion value of the third threshold can be appropriately reduced to ensure that the number of project materials that pass the final review does not exceed 10. When the number of application materials that pass the automatic review and re-examination stage is small, it is ensured that the number of project materials that pass the final review is not less than 3. If the number of application materials that pass the automatic review and re-examination stage is less than 3, it is determined that all of them have passed the automatic review and final review stage.

[0081] like Figure 2 As shown, the present invention provides a feasibility study report matching analysis system based on artificial intelligence, including a data processing module, a project data matching analysis unit, and a threshold judgment module;

[0082] The data processing module is used to collect the approved project feasibility study reports and project application materials in historical projects, and modularly classify the data according to the functions and content sections, and perform structured processing; the project data matching analysis unit is used to analyze the matching degree between the sub-data of the project application materials and the requirements of the current project, the matching degree between the predicted data of the project application materials data module and the actual data, and the comprehensive matching degree between the project application materials data automatically generated according to the project feasibility study report and the actual data of the project application materials; the threshold judgment module is used to perform a first threshold judgment on the sub-data of the project application materials, a second threshold judgment on the project application materials data module, and a third threshold judgment on the matching degree of the project feasibility study report;

[0083] The data processing module includes a project approval data modularization unit and a project approval data structured unit;

[0084] The project approval data modularization unit is used to modularize and classify the project approval data according to its functions and content sections; the project approval data structured unit is used to structure the project approval data to generate a structured data set for the feasibility study report and a structured data set for the project application materials;

[0085] The threshold judgment module includes a first threshold judgment unit, a second threshold judgment unit, and a third threshold judgment unit;

[0086] The first threshold judgment unit is used to perform a first threshold judgment on the project application material sub-data according to the current project requirements; the second threshold judgment unit is used to perform a second threshold judgment on the data module according to the matching calculation results of the project application material data module; the third threshold judgment unit is used to perform a third threshold judgment on the project according to the comprehensive matching results of the project feasibility study report.

[0087] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0088] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A feasibility study report matching analysis method based on artificial intelligence, characterized in that: The method comprises the following analysis steps: Step S100: Collect feasibility study reports and project application materials from historical project approval materials, perform modular classification and data structuring processing; Step S200: Structuring the feasibility study report and project application materials of the current project, setting a first threshold for the structured data according to the requirements of the current project, and performing a first threshold judgment on the feasibility study report and project application materials of the current project; Step S300: Analyze the relevance of sub-data of each module in the structured data of the feasibility study report and the project application materials, and then use multivariate linear fitting to train the project application materials data generation model, and use the structured data of the feasibility study report to generate corresponding project application materials data; Step S400: Compare the automatically generated data of the project application materials with the actual data to calculate the matching degree of each data module between the feasibility study report and the project application materials, and perform a second threshold judgment on the current project application materials according to the matching results of each data module; Step S500: Perform comprehensive matching calculation on the feasibility study report and project application materials that pass the second threshold judgment and perform a third threshold judgment, and analyze the project automatic approval result according to the judgment result; The step S100 includes the following contents: Collect approved project materials from historical projects, classify them into modules according to the data functions and content sections of the feasibility study reports and project application materials in the project materials, group and summarize the data of each module, set the data format and store them in modules for structured data processing; Structure the feasibility study report data in the historically approved project materials to generate a data set , structure the project application materials to generate a data set ; in, For each classification module of feasibility study report data, Provide data for each module of the feasibility study report; Apply for material data for each classification module of the project, Apply for data of each module of project materials; The number of data classification modules for the feasibility study report. Apply for the number of material data classification modules for the project; The step S200 includes the following contents: The feasibility study report of the current project and the project application materials are used to generate project structured data using the data classification and processing method in step S100; Obtaining the requirement information of each sub-data of each content section of the project application materials in the current project requirements, setting a first threshold value for each module sub-data in the project application materials, and then judging whether each module sub-data meets the data standard in the project requirements information; Perform a first threshold judgment on the sub-data of each module, and determine the project with sub-data that fails the first threshold judgment as failed in the automatic preliminary review, and feed back the sub-data that fails the first threshold judgment for manual review; for the project with all sub-data that passes the first threshold judgment, output an automatic preliminary review pass signal, and carry out the subsequent review process; The step S300 includes the following contents: Analyze the data association between each module sub-data in the historical project application materials and each module sub-data in the feasibility study report, filter out each module sub-data in the feasibility study report that affects the data change, and use the multivariate linear regression method to use the data set described in step S100 With the dataset Training project application materials data generation model, for any sub-data in the project application materials , the data automatic generation model is as follows: ； in, Sub-data for project application materials in the feasibility study report The sub-data vector affected by the data change, for Each parameter in corresponds to a linear fitting parameter vector, For sub-data The linear fitting constant parameters are: For sub-data Linear fitting range parameters; Subdata Linear fitting range parameters When taking values, according to the data set The actual data of the project application materials is taken to make the data set Middle proportion The data falls on When the leading symbols are positive and negative When using the model to predict the project application material sub-data, use When the positive and negative The mean value is taken as the model prediction data of the sub-data; The step S400 includes the following contents: Obtain the structured data of the feasibility study report that passes the first threshold in step S200, substitute it into the project application material data generation model to calculate and analyze the predicted values ​​of the sub-data models of each module in the project application material, and then compare and analyze it with the actual project application material, and calculate the matching degree between the predicted data of the project application material and the actual data by module classification; for any module in the project application material , according to the formula: ； in, For data module The matching degree between the predicted data and the actual data, For data module Data number, For data module Medium data volume, For data module Neutron data Actual value, For data module Neutron data Predicted value, For data module Neutron data Linear fitting range parameters, For data module Neutron data The matching weight of Use the structured data of the current project feasibility study report to calculate the matching degree of all data modules in the project application materials and generate the matching degree matrix of the current project application materials , setting a second threshold for the matching degree of each module and performing a second threshold judgment; The projects with data module matching degree less than the second threshold value are judged as failed automatic review, and the module data that failed automatic review is fed back for manual review; for projects with all module matching degrees greater than the second threshold value of the corresponding module, an automatic review pass signal is output, and the subsequent review process is carried out; The step S500 includes the following contents: Assign a matching evaluation coefficient to each data module in the project application materials according to the project type, and generate a matching evaluation coefficient matrix for the project application materials , and then according to the formula: , calculate the comprehensive matching degree of the project feasibility study report ; Sort the matching degree data of all the project application materials of the current project and the corresponding feasibility study reports from large to small, set a third threshold for the passing ratio of all the application materials of the current project and make a third threshold judgment; Among the sorted projects, all projects whose project application materials are ranked within the third threshold from large to small are judged to have passed the automatic final review, and the subsequent manual selection process is carried out; all projects whose project application materials are ranked outside the third threshold are judged to have failed the automatic final review, and a signal for resubmitting the project application materials is output.

2. A feasibility study report matching analysis system based on artificial intelligence using the feasibility study report matching analysis method based on artificial intelligence as described in claim 1, characterized in that: The system includes a data processing module, a project data matching degree analysis unit, and a threshold judgment module; The data processing module is used to collect the approved project feasibility study reports and project application materials in historical projects, and modularly classify the data according to the functions and content sections and perform structured processing; the project data matching degree analysis unit is used to analyze the matching degree between the sub-data of the project application materials and the current project requirements, the matching degree between the predicted data of the project application materials data module and the actual data, and the comprehensive matching degree between the project application materials data automatically generated according to the project feasibility study report and the actual data of the project application materials; the threshold judgment module is used to perform a first threshold judgment on the sub-data of the project application materials, a second threshold judgment on the project application materials data module, and a third threshold judgment on the matching degree of the project feasibility study report; The output end of the data processing module is connected to the project data matching degree analysis unit and the input end of the threshold judgment module; the output end of the project data matching degree analysis unit is connected to the input end of the threshold judgment module.

3. The feasibility study report matching analysis system based on artificial intelligence according to claim 2 is characterized by: The data processing module includes a project approval data modularization unit and a project approval data structuring unit; The project approval data modularization unit is used to modularize and classify the project approval data according to its functions and content sections; the project approval data structuring unit is used to structure the project approval data to generate a structured data set of a feasibility study report and a structured data set of project application materials; The output end of the project approval data modularization unit is connected to the input end of the project approval data structuring unit; the project approval data structuring unit is connected to the project data matching degree analysis unit and the input end of the threshold judgment module.

4. The artificial intelligence-based feasibility study report matching analysis system according to claim 2 is characterized by: The threshold judgment module includes a first threshold judgment unit, a second threshold judgment unit, and a third threshold judgment unit; The first threshold judgment unit is used to perform a first threshold judgment on the project application material sub-data according to the current project requirements; the second threshold judgment unit is used to perform a second threshold judgment on the data module according to the project application material data module matching calculation result; the third threshold judgment unit is used to perform a third threshold judgment on the project according to the comprehensive matching result of the project feasibility study report; The input end of the first threshold judgment unit is connected to the output end of the data processing module; The input ends of the second threshold judgment unit and the third threshold judgment unit are connected to the output end of the project data matching degree analysis unit.

Images