Company data and report analysis system for artificial intelligence large model

Abstract

The invention, which relates to the technical field of data analysis, discloses a company data and report analysis system for an artificial intelligence large model, comprising a data acquisition module, an industry term library module, a term library association module, a report generation module and a rationality analysis module. The system has the advantages that dynamic semantic analysis of cross-domain terms can be realized through the knowledge network constructed by the industry term library module and the term library association module, and accurate matching of report contents and user demands is ensured by using a three-layer generation architecture of demand unit-term mapping-rationality verification; meanwhile, complex industry term expressions can be understood and processed more meticulously, misunderstanding caused by fuzzy or ambiguous term meanings is avoided, the accuracy of understanding professional languages is improved, the accuracy and rationality of term use in the report can be ensured through the rationality analysis module, and the quality and credibility of the report are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and particularly to a company data and report analysis system for artificial intelligence large models. Background Art

[0002] Commercial enterprise data is collected, processed, sorted and published by commercial companies. It is developed as a valuable commodity, owned by the commercial companies, and published and sold by the commercial companies. Its purpose is to provide other commercial companies with demand with channels to obtain potential customers, help other enterprises develop effective customers, and provide data services for small and medium-sized enterprises. The forms of publication of enterprise data are diverse. Usually, it is stored in the form of EXCEL tables, provided in the form of databases, or there are also yellow page books as carriers. Powerful data providing companies will also independently develop customer management software to intelligently manage enterprise data. As a bridge between commercial individuals, enterprise data is useful for both the enterprises where the data is published and the data acquirers. When summarizing company data, reports are often used for display;

[0003] When common company data and report analysis systems are in use, there are difficulties in semantic understanding and parsing of industry terms, resulting in the inability to accurately grasp the meaning of terms when involving multi-field knowledge interaction. When encountering situations where the meaning of terms is ambiguous or polysemous, misunderstandings are likely to occur, affecting the accurate understanding and application of professional content. For this reason, we propose a company data and report analysis system for artificial intelligence large models. Summary of the Invention

[0004] The purpose of the present invention is to provide a company data and report analysis system for artificial intelligence large models.

[0005] To solve the problems raised in the above background art, the present invention provides the following technical solutions: A company data and report analysis system for artificial intelligence large models, including a data acquisition module, an industry term library module, a term library association module, a report generation module, and a rationality analysis module;

[0006] The data acquisition module collects company data, obtains report information with different evaluations, analyzes the reasons for different evaluations, establishes a report generation requirement unit, and uses the report generation requirement unit to collect the requirements of staff for generating reports to obtain report requirement information;

[0007] The industry term library module collects the meanings represented by different industry terms, records them to form a term database, then analyzes the term keywords in the term database, and at the same time, when obtaining report requirement information, retrieves the term keywords in the report requirement information;

[0008] The thesaurus association module extracts term keywords, analyzes the different meanings represented when different term keywords are used together, constructs a meaning association graph, obtains the node positions in the meaning association graph, arranges different term keywords at different node positions, and when different keyword combinations have different meanings, records the corresponding meaning association graph, and then inserts the represented meaning into the meaning association graph in the form of a label;

[0009] The report generation module extracts the positions where industry terms are used in the requirement information, analyzes the industry terms used in the requirement information, analyzes the term keywords in the industry terms, obtains the meanings represented by different industry terms, and generates a report form according to the report requirement information and the meanings represented by different industry terms;

[0010] The rationality analysis module extracts the industry terms used in the report form and the meanings represented by the industry terms, synchronously analyzes the application field of the report requirement information, obtains the report field, judges the usage situation of the industry terms and the meanings represented by the industry terms in the report field, calculates the rationality degree of the industry terms and the meanings represented by the industry terms in the report form, and obtains a rationality index.

[0011] As a further solution of the present invention: when the report requirement information in the data acquisition module is obtained, the staff drags the requirement information to the report generation requirement unit, then opens the requirement information in the report generation requirement unit, the report generation requirement unit will synchronously establish an effective area and a material area, and then enters the requirement information into the material area. The staff has the permission to edit the effective area. When the staff selects the opened requirement information, the selected information will be automatically transferred to the effective area.

[0012] As a further solution of the present invention: when the effective area in the data acquisition module is edited, the background of the requirement information in the material area will be set to a 25% gray background, then the background of the selected information will be automatically adjusted to a yellow background, and the different selected information will be underlined, and then numbered in the order of selection below the underline. At the same time, the same information in the effective area will be underlined, and the corresponding number will be recorded at the lower end of the corresponding underline.

[0013] As a further solution of the present invention: when the term keywords in the industry thesaurus module are obtained, let the term keyword be S 1 、S 2 、S 3 、……、S X ,let the report requirement information be B 1 、B 2 、B 3 、……、B C ,let the term keyword in the report requirement information be B关键词 :

[0014] B 关键词 = (S 1 , S 2 , S 3 , ……, S X ) ∩ (B 1 , B 2 , B 3 , ……, B C

[0015] Calculate the term keywords in the report requirement information according to the above formula.

[0016] As a further solution of the present invention: After obtaining the term keywords in the report requirement information, the sentence information where the term keywords are located will be synchronously extracted, a sentence table will be established, and then the sentence information will be recorded in the sentence table, and at the same time, the term keywords will be recorded in the sentence table corresponding to the sentence information.

[0017] As a further solution of the present invention: When obtaining the meaning association diagram in the term library association module, the same meaning units will be synchronously recommended, and the same meaning folders will be established in the same meaning units. Then, the meaning association diagrams with the same meaning will be recorded in the same same meaning folder, and the same meaning folder will be named with the same meaning.

[0018] As a further solution of the present invention: After the report form in the report generation module is generated, the report form will be displayed to the staff, and the staff will confirm whether the report form meets the requirements. When the report form meets the requirements, the report form will be output. When the report form does not meet the requirements, the report form will be regenerated, and at the same time, the staff has the permission to select the positions that need to be modified in the report form.

[0019] As a further solution of the present invention: After obtaining the report field, the rationality analysis module will collect all the report forms in the company, obtain the experimental report, extract the evaluation of the experimental report, collect the evaluation level, analyze the quantity of the evaluation level, and then convert the evaluation of the experimental report into the score of the experimental report in the form of 100 points. Let the quantity of the evaluation level be P 数量 , let the level of the evaluation of the experimental report in the evaluation level be P 等级 , let the score of the experimental report be S 评分 :

[0020]

[0021] Calculate the score of the experimental report according to the above formula.

[0022] As a further solution of the present invention: when the rationality analysis module determines the usage of industry terms and the meanings they represent in the report field, it will first retrieve the report field in the experimental report to obtain similar reports, then analyze the number of similar reports using the industry term in the similar reports, and extract the scores of the similar reports. Then, an effective report unit is established. When the score of the similar report is higher than 60, it will be recorded in the effective report unit. Then, set the rationality multiple of the number of similar reports in the effective report unit as Set the rationality index as H 指数 , set the number of similar reports recorded in the effective report unit as X 数量 :

[0023]

[0024] Calculate the rationality index according to the above formula.

[0025] Adopting the above technical solution, compared with the prior art, the beneficial effects of the present invention are as follows:

[0026] 1. The knowledge network constructed by the industry term library module and the term library association module of the present invention can realize the dynamic semantic parsing of cross-domain terms. By using a three-layer generation architecture of demand unit - term mapping - rationality verification, it ensures the precise matching of the report content with the user's needs. At the same time, it helps to more carefully understand and process complex industry term expressions, avoids misunderstandings caused by ambiguous or polysemous term meanings, improves the accuracy of understanding professional language, and uses the rationality analysis module to ensure the accuracy and rationality of term usage in the report, improving the quality and credibility of the report;

[0027] 2. The data acquisition module of the present invention can quickly integrate report requirements, improve data acquisition efficiency, enable staff to clearly distinguish valid information from other information, facilitate the screening and sorting of key content, reduce the error rate of information processing. The industry term library module can accurately identify professional terms in the requirement information, provide accurate basic data for subsequent term analysis and report generation, help to deeply understand the term meaning from the context perspective, avoid deviations caused by isolated understanding of terms, and enhance the comprehensiveness and accuracy of the industry term library module's understanding of terms;

[0028] 3. The term library association module of the present invention facilitates the quick search and call of term combinations with the same meaning, improves term management efficiency, optimizes the system's processing ability for complex term relationships, uses the report generation module to improve the accuracy and usability of report generation, and the rationality analysis module facilitates the intuitive comparison of the quality levels of different experimental reports, can objectively judge the rationality of the usage of industry terms in the report field, and provides a strong basis for report quality control. Description of the Drawings

[0029] Figure 1 This is a schematic diagram of the system process in an embodiment of the present invention. Specific embodiments

[0030] The following further describes the specific embodiments of the present invention in conjunction with the accompanying drawings. It should be noted here that the description of these embodiments is used to help understand the present invention, but does not limit the present invention.

[0031] In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0032] Embodiment 1:

[0033] Therefore, in order to effectively solve the above problems, the present application proposes a company data and report analysis system for artificial intelligence large models, as shown in the accompanying drawings of the specification Figure 1 shown, including a data collection module, an industry term library module, a term library association module, a report generation module, and a rationality analysis module;

[0034] The data collection module collects company data, obtains report information with different evaluations, analyzes the reasons for different evaluations, and establishes a report generation requirement unit. Using the report generation requirement unit, it collects the requirements of staff for generating reports and obtains report requirement information;

[0035] The industry term library module collects the meanings represented by different industry terms, records them to form a term database, then analyzes the term keywords in the term database, and at the same time, when obtaining report requirement information, retrieves the term keywords in the report requirement information;

[0036] The term library association module extracts term keywords, analyzes the different meanings represented when different term keywords are used together, constructs a meaning association graph, obtains the node positions in the meaning association graph, arranges different term keywords at different node positions, and when different keyword combinations result in different meanings, records the corresponding meaning association graph, and then inserts the represented meaning in the form of a label into the meaning association graph;

[0037] Adopt a hierarchical attention mechanism to assign weights to term nodes to achieve focused strengthening of key terms. When new associations between digital currency and carbon trading appear in the financial field, the system automatically generates an association path and verifies the confidence level within 24 hours;

[0038] The report generation module extracts the positions where industry terms are used in the requirement information, analyzes the industry terms used in the requirement information, analyzes the term keywords in the industry terms, obtains the meanings represented by different industry terms, and generates a report form according to the report requirement information and the meanings represented by different industry terms;

[0039] The rationality analysis module extracts the industry terms used in the report form and the meanings represented by the industry terms, synchronously analyzes the application fields of the report requirement information to obtain the report field, judges the usage of the industry terms and the meanings represented by the industry terms in the report field, calculates the rationality of the industry terms and the meanings represented by the industry terms in the report form, and obtains the rationality index;

[0040] The data acquisition module, the industry term library module, the term library association module, and the rationality analysis module can constitute a three-layer generation architecture of demand unit-term mapping-rationality verification;

[0041] At the same time, the domain adaptability index (DAI) and the term consistency coefficient (TCC) are introduced into the rationality analysis module to support real-time semantic compliance detection;

[0042] Specific work process: Collect company data, use the report generation demand unit to collect the requirements of staff to generate reports, collect the meanings represented by different industry terms, record them to form a term database, retrieve the term keywords in the report requirement information, analyze the different meanings represented when different term keywords are used together, and at the same time construct a meaning association graph, obtain the node positions in the meaning association graph, arrange different term keywords at different node positions, analyze the industry terms used in the requirement information, analyze the term keywords in the industry terms, and generate a report form according to the report requirement information and the meanings represented by different industry terms, analyze the application field of the report requirement information, and judge the usage of the industry terms and the meanings represented by the industry terms in the report field;

[0043] Furthermore, the knowledge network constructed by the industry term library module and the term library association module can realize the dynamic semantic parsing of cross-domain terms. Using the three-layer generation architecture of demand unit-term mapping-rationality verification, it ensures the precise matching of the report content with the user requirements. At the same time, it helps to more carefully understand and process complex industry term expressions, avoids misunderstandings caused by ambiguous or polysemous term meanings, improves the accuracy of understanding professional language, and uses the rationality analysis module to ensure the accuracy and rationality of term usage in the report, improving the quality and credibility of the report.

[0044] Embodiment 2:

[0045] On the basis of Embodiment 1, as shown in the accompanying drawings of the specification Figure 1As shown, when the report requirement information in the data acquisition module is obtained, the staff drags the requirement information into the report generation requirement unit, and then opens the requirement information in the report generation requirement unit. The report generation requirement unit will synchronously establish the effective area and the material area, and then enter the requirement information into the material area. The staff has the permission to edit the effective area. When the staff selects the opened requirement information, the selected information will be automatically transferred to the effective area;

[0046] For financial report requirements, the input formats and analysis frameworks for the relevant data of the balance sheet and income statement will be preset in the effective area; for market analysis report requirements, the layouts of methods such as market share and competitor analysis will be set. In this way, when the staff edits the effective area, they can quickly sort out and enter key information based on a more reasonable framework, greatly improving the efficiency and standardization of report generation;

[0047] When the effective area in the data acquisition module is edited, the background of the requirement information in the material area will be set to a 25% gray background, and then the background of the selected information will be automatically adjusted to a yellow background, and the different selected information will be underlined, and then numbered in the order of selection below the underline. At the same time, the same information in the effective area will be underlined, and the corresponding numbers will be recorded at the lower end of the corresponding underline;

[0048] When the term keywords in the industry terminology library module are obtained, let the term keywords be S 1 、S 2 、S 3 、……、S X , let the report requirement information be B 1 、B 2 、B 3 、……、B C , let the term keywords in the report requirement information be B 关键词 :

[0049] B 关键词 =(S 1 、S 2 、S 3 、……、S X )∩(B 1 、B 2 、B 3 、……、B C

[0050] Calculate the term keywords in the report requirement information according to the above formula;

[0051] After obtaining the term keywords in the report requirement information, the sentence information where the term keywords are located will be extracted synchronously, and a sentence table will be established. Then, the sentence information will be recorded in the sentence table, and at the same time, the term keywords will be recorded in the sentence table corresponding to the sentence information;

[0052] Specific work process: The staff drags the requirement information into the report generation requirement unit, opens the requirement information in the report generation requirement unit, and then enters the requirement information into the material area. When the staff selects the opened requirement information, the selected information will be automatically transferred to the valid area. The background of the requirement information in the material area is set to a 25% gray background, and then the background of the selected information is automatically adjusted to a yellow background. Calculate the term keywords in the report requirement information, extract the sentence information where the term keywords are located synchronously, and establish a sentence table. Then, the sentence information will be recorded in the sentence table, and at the same time, the term keywords will be recorded in the sentence table corresponding to the sentence information;

[0053] Furthermore, through the data collection module, the report requirements can be quickly integrated, the data collection efficiency can be improved, enabling the staff to clearly distinguish valid information from other information, facilitating the screening and sorting of key content, reducing the error rate of information processing. Through the industry term library module, professional terms in the requirement information can be accurately identified, providing accurate basic data for subsequent term analysis and report generation, helping to deeply understand the meaning of terms from the context perspective, avoiding deviations caused by isolated understanding of terms, and enhancing the comprehensiveness and accuracy of the industry term library module in understanding terms.

[0054] Embodiment Three:

[0055] Based on Embodiment Two, as shown in the accompanying drawings of the specification Figure 1 When the meaning association diagram in the term library association module is obtained, the same meaning units will be suggested synchronously, and a same meaning folder will be established in the same meaning unit. Then, the meaning association diagrams with the same meaning will be recorded in the same same meaning folder, and the same meaning folder will be named with the same meaning;

[0056] For some term combinations with similar meanings in different contexts, the system can automatically identify and cluster their association diagrams into sub-folders with similar meanings. At the same time, detailed semantic tags and descriptions will be generated for each same meaning folder and sub-folder, facilitating the staff to quickly understand and search for relevant term combinations, and further improving the refinement degree and retrieval efficiency of term management;

[0057] After the report form in the report generation module is generated, it will be displayed to the staff, and the staff will confirm whether the report form meets the requirements. When the report form meets the requirements, it will be output. When the report form does not meet the requirements, a new report form will be generated. At the same time, the staff has the right to select the positions that need to be modified in the report form;

[0058] After obtaining the report field, the rationality analysis module will collect all the report forms in the company, obtain the experimental reports, extract the evaluations of the experimental reports, collect the evaluation levels, analyze the number of evaluation levels, and then convert the evaluation of the experimental report into the score of the experimental report in the form of 100 points. Let the number of evaluation levels be P 数量 Let the level of the evaluation of the experimental report in the evaluation level be P 等级 Let the score of the experimental report be S 评分 :

[0059]

[0060] Calculate the score of the experimental report according to the above formula;

[0061] For example, if the evaluation of an experimental report is good, and the evaluation levels are poor, relatively poor, normal, good, excellent and extremely good, then the number of evaluation levels is 6, and the level of the evaluation of the experimental report in the evaluation level is 4. At this time The score of the experimental report is obtained as 66.7;

[0062] When the rationality analysis module judges the usage situation of the industry term and the meaning represented by the industry term in the report field, it will first search for the report field in the experimental report to obtain similar reports, then analyze the number of similar reports using the industry term in the similar reports, and extract the scores of the similar reports. Then, an effective report unit is established. When the score of the similar report is higher than 60, it will be recorded in the effective report unit. Then, let the rationality multiple of the number of similar reports in the effective report unit be Let the rationality index be H 指数 Let the number of similar reports recorded in the effective report unit be X 数量 :

[0063]

[0064] Calculate the rationality index according to the above formula;

[0065] Specific workflow: Suggest synonymous units, create synonymous folders within the synonymous units, then record the meaning association diagrams with the same meaning in the same synonymous folder, name the synonymous folder with the same meaning, display the report form to the staff, and have the staff confirm whether the report form meets the requirements. When the report form meets the requirements, output the report form, collect all the report forms within the company to obtain the experimental report, extract the evaluation of the experimental report, collect the evaluation grades, analyze the quantity of the evaluation grades, calculate the score of the experimental report. First, search for the report field in the experimental report to obtain similar reports, then analyze the number of similar reports using the industry terms in the similar reports, and extract the scores of the similar reports. Then establish an effective report unit. When the score of a similar report is higher than 60, record it in the effective report unit and calculate the rationality index;

[0066] Furthermore, the term library association module facilitates the quick search and call of term combinations with the same meaning, improves the efficiency of term management, optimizes the system's processing ability for complex term relationships, uses the report generation module to improve the accuracy and usability of report generation, and through the rationality analysis module, it is convenient to visually compare the quality levels of different experimental reports, objectively judge the rationality of the use of industry terms in the report field, and provide a strong basis for the control of report quality.

[0067] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art of this industry should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principle of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A company data and report analysis system for artificial intelligence big models, characterized by: It includes data collection module, industry term base module, term base association module, report generation module and rationality analysis module; The data collection module collects company data and obtains report information of different evaluations, analyzes the reasons for different evaluations in the report information, and establishes a report generation demand unit. The report generation demand unit is used to collect the requirements of staff for generating reports and obtain report demand information; The industry terminology library module collects the meanings of different industry terms and records them to form a terminology database, then analyzes the terminology keywords in the terminology database and retrieves the terminology keywords in the report requirement information when obtaining the report requirement information; The terminology library association module extracts terminology keywords, analyzes the different meanings represented by different terminology keywords when used together, and constructs a meaning association graph to obtain the node positions in the meaning association graph. Different terminology keywords are arranged at different node positions. When different meanings appear after different keyword combinations, the corresponding meaning association graph is recorded, and then the represented meanings are inserted into the meaning association graph in the form of tags. The report generation module extracts the locations where industry terms are used in the demand information, analyzes the industry terms used in the demand information, analyzes the term keywords in the industry terms, obtains the meanings represented by different industry terms, and generates a report table based on the report demand information and the meanings represented by different industry terms; The rationality analysis module extracts the industry terms used in the report form and the meanings represented by the industry terms, and simultaneously analyzes the application fields of the report demand information to obtain the report field, determines the usage of the industry terms and the meanings represented by the industry terms in the report field, calculates the rationality of the industry terms and the meanings represented by the industry terms in the report form, and obtains the rationality index.

2. The company data and report analysis system for artificial intelligence big model according to claim 1, characterized in that: When the report requirement information in the data collection module is obtained, the staff drags the requirement information to the report generation requirement unit, and then opens the requirement information in the report generation requirement unit. The report generation requirement unit will simultaneously establish the effective area and the material area, and then enter the requirement information into the material area. The staff has the authority to edit the effective area. When the staff selects the opened requirement information, the selected information will be automatically transferred to the effective area.

3. The company data and report analysis system for artificial intelligence big model according to claim 2, characterized in that: When editing the valid area in the data acquisition module, the background of the required information in the material area will be set to a 25% gray background, and then the background of the selected information will be automatically adjusted to a yellow background, and different selected information will be underlined, and then numbered in the order of selection on the lower side of the underline, and at the same time, the same information in the valid area will be underlined, and then the corresponding number will be recorded at the lower end of the corresponding underline.

4. The company data and report analysis system for artificial intelligence big model according to claim 1, characterized in that: When the term keywords in the industry term library module are obtained, the term keywords are set to S1, S2, S3, ..., S X , assuming that the report requirement information is B1, B2, B3, ..., B C , assuming that the term keyword in the report requirement information is B 关键词 : <h2 style=";text-align:left;direction:ltr">B<h2 style=";text-align:left;direction:ltr"> 关键词 <h2 style=";text-align:left;direction:ltr"> (S1, S2, S3,……, S)<h2 style=";text-align:left;direction:ltr"> X <h2 style=";text-align:left;direction:ltr"> )∩(B1、B2、B3、……、B<h2 style=";text-align:left;direction:ltr"> C <h2 style=";text-align:left;direction:ltr"> ) The key words in the report requirement information are calculated according to the above formula.

5. The company data and report analysis system for artificial intelligence big model according to claim 4, characterized in that: After obtaining the terminology keywords in the report requirement information, the sentence information where the terminology keywords are located will be extracted synchronously, and a sentence table will be established. The sentence information will then be recorded in the sentence table, and the terminology keywords will be recorded in the sentence table corresponding to the sentence information.

6. The company data and report analysis system for artificial intelligence big model according to claim 1, characterized in that: When the meaning association diagram in the term base association module is obtained, the same meaning unit will be synchronously suggested, and a same meaning folder will be established in the same meaning unit, and then the meaning association diagram with the same meaning will be recorded in the same same meaning folder, and the same meaning folder will be named with the same meaning.

7. The company data and report analysis system for artificial intelligence big model according to claim 1, characterized in that: After the report form in the report generation module is generated, the report form will be displayed to the staff, and the staff will confirm whether the report form meets the requirements. When the report form meets the requirements, the report form will be output. When the report form does not meet the requirements, the report form will be regenerated. At the same time, the staff has the authority to select the location in the report form that needs to be modified.

8. The company data and report analysis system for artificial intelligence big model according to claim 1, characterized in that: After obtaining the report field, the rationality analysis module will collect all report forms in the company, obtain the experimental report, extract the evaluation of the experimental report, collect the evaluation level, analyze the number of evaluation levels, and then convert the evaluation of the experimental report into the score of the experimental report in the form of 100 points. Let the number of evaluation levels be P 数量 , let the evaluation level of the experimental report be P 等级 , let the score of the experimental report be S 评分 : Calculate the score of the lab report according to the above formula.

9. The company data and report analysis system for artificial intelligence big model according to claim 8, characterized in that: When judging the usage of industry terms and the meanings represented by industry terms in the report field, the rationality analysis module will first search the report field in the experimental report to obtain similar reports, then analyze the number of similar reports that use the industry terms in the similar reports, extract the scores of the similar reports, and then establish a valid report unit. When the score of the similar report is higher than 60, it will be recorded in the valid report unit, and then the rationality multiple of the number of similar reports in the valid report unit is set to Let the reasonable index be H 指数 , let the number of similar reports recorded in the valid reporting unit be X 数量 : The rationality index is calculated according to the above formula.

Images