A Simple Method for Intelligent Template Orchestration of Integrating Table Model and Text2SQL Online Forms

Abstract

The present invention discloses a simple intelligent template layout method for integrating a table model and Text2SQL online forms, which includes the following methods: a method for generating personalized templates for thousands of people, a method for rendering employee online templates in combination with a table model, and a method for separately storing and saving data reported by employees online. The beneficial effects are as follows: The online form of the present invention is what you see is what you get, effectively solving the problem of repetitive module design and management work for designers, developers, and testers in the same presentation interface, without cognitive bias or understanding errors, reducing loss during work conversion, improving work efficiency, and enhancing the quality of the results; at the same time, based on the LLM model, industry and enterprise data are analyzed, arranged, and designed, and personalized employee templates for thousands of people are output through the Text2SQL function, with the manual efficiency increased by a hundred times and the effect improved geometrically; in addition, the template effect is visually evaluated online to eliminate the hallucinations of the large model.

Description

Technical Field

[0001] The present invention relates to the technical field of online form intelligent template arrangement methods, and particularly to a simple method for integrating a table model with Text2SQL for online form intelligent template arrangement. Background Art

[0002] All along, in the data form module of the IT software development industry, regardless of research and development and use in various industries, the documents and reports designed by R & D personnel through a large number of typesetting for various types all adopt traditional hard - coding technology when converting to IT technology research and development. Especially on related interfaces such as computers or mobile phones, it occupies one - third of the invalid interface space. Moreover, the formats, types written and the accuracy of the results presented by the content of each form are very low. It is necessary to repeatedly correct the writing and conduct inspection and testing, and the operation and use are complicated and inconvenient, seriously affecting work efficiency, resulting in high labor costs for IT - type R & D enterprises in various industries. Similarly, for the systems developed in this way, when enterprise employees in various industries use them, due to the need to fill in a large number of forms and reports correspondingly, it will lead to technical BUG problems in the developed technology or algorithm deficiencies during modification and upgrade, causing serious settlement impacts on various forms. These traditional manual filling methods have obvious problems such as low efficiency and easy errors.

[0003] Currently, in the data form module of the IT software development industry, there are online form tools. Although they can be used for multi - person online collaborative operations, they lack direct integration with the table model and cannot meet the convenient methods and tools for user work management, and are merely reduced to simple online excel tools.

[0004] With the development of information technology, digital form processing tools have gradually become popular, but some existing methods still have problems such as complex template design, inconvenient use, and insufficient customization.

[0005] When submitting and displaying a large amount of form data through the WEB page method, on the one hand, due to the limitations of the WEB page itself, the data display is incomplete and not intuitive, which is not suitable for business scenarios with financial requirements; on the other hand, for users who prefer data processing, managing data through a large number of offline excel electronic forms is more habitual and preferred. Summary of the Invention

[0006] The purpose of the present invention is to provide a simple method for integrating a table model with Text2SQL for online form intelligent template arrangement to solve the above problems.

[0007] The present invention achieves the above - mentioned purpose through the following technical solutions:

[0008] A simple method for intelligent template layout of integrating table models with Text2SQL online forms, which includes the following methods: the method of generating personalized templates for thousands of people, the method of rendering employee online templates in combination with table models, and the method of separating and saving employee online reported data; among the method of generating personalized templates for thousands of people, it includes basic templates differentiated by employee categories, an employee key indicator extraction module, an LLM business indicator template, an indicator business SQL module formed by Text2SQL, a visualization intervention module, and saving personalized templates for thousands of people; in the method of rendering employee online templates in combination with table models, it includes loading employee personalized templates, executing template business SQL to load the table model module, constructing structured data according to the template and styling data according to the template, and rendering the json data containing the table model, structured data, and styling data through a medium and finally outputting; in the method of separating and saving employee online reported data, it includes an employee data reporting and review module, a template data engine module for separating data from templates, saving the table model, and the LLM understanding the business and optimizing personalized templates.

[0009] Further, the method of generating personalized templates for thousands of people specifically includes the following steps:

[0010] Step 1: Configure basic templates by category, including data block structure and style;

[0011] Step 2: Train an LLM private model through enterprise and industry table models;

[0012] Step 3: Use the LLM private model to understand employee KPIs and form employee key prompts;

[0013] Step 4: Drive the LLM private model Text2SQL through the key prompts to generate the employee's proprietary business SQL;

[0014] Step 5: Integrate the employee's proprietary business SQL with the basic template to form complete template data that is individualized and personalized for the employee;

[0015] Step 6: Visually execute the employee's complete individualized template, judge the template execution effect, eliminate deficiencies due to large model hallucinations, perform visual layout, and form a template suitable for the employee individual;

[0016] Step 7: Save the employee's individual template for data reporting and display.

[0017] Further, the method of rendering employee online templates in combination with table models specifically includes the following steps:

[0018] Step 1: Load the employee's individualized template and extract the employee's exclusive business SQ;

[0019] Step 2: The data engine, combined with the table model, executes the exclusive business SQL to obtain the exclusive table model: own table model, subordinate table model, and approval table model;

[0020] Step 3: The data engine creates the header and the overall structure of rows and columns based on the individualized template. Driven by the dedicated table model, it creates multiple rows, multiple columns, and merges rows and columns to calculate and form structural data.

[0021] Step 4: For each title item in the data table header, configure a button. Click it to pop up the data aggregation of the column where the title item is located as a condition for filtering. After filtering, the list data content changes in a linked manner; data aggregation filtering includes: category selection of major categories and successively smaller categories, duplicate removal selection of data in this column, quantity / price formula selection, etc., and comprehensive condition selection;

[0022] Step 5: The data engine calculates the structural data and table model style according to the individualized template, including font: weight, color, centering, underline, bold and other style data;

[0023] Step 6: The data engine integrates the table model, structure data and style data to form json data;

[0024] Step 7: Output json data and present it to the user through front-end rendering.

[0025] Furthermore, the method for separating and saving the data submitted online by employees specifically includes the following steps:

[0026] Step 1: Employees submit comprehensive data through the front end, which includes: table model, structure data and style data;

[0027] Step 2: Data engine, extracting individualized employee templates;

[0028] Step 3: The data engine parses and decomposes the comprehensive data into a table model according to the individual templates, and discards the structure and style data;

[0029] Step 4: Data engine, follow up business rules to save table models (own business, approval data);

[0030] Step 5: The LLM model continues to understand the employee table model, outputs the business SQL through Text2SQL, and forms a new version of the template data;

[0031] Step 6: Visually manage new template data and directly publish it as an official template for employees.

[0032] The beneficial effects of the present invention are:

[0033] The WYSIWYG online form of the present invention effectively solves the problem of repetitive module design and management work faced by designers, developers, and testers in the same presentation interface, without cognitive bias or understanding errors, reduces the loss during work conversion, improves work efficiency, and enhances the quality of results. At the same time, based on the LLM model, it analyzes and arranges industry and enterprise data, and outputs personalized employee templates through the Text2SQL function, increasing the manual efficiency by a hundred times and achieving a geometric progression improvement in the effect. In addition, it visualizes and online executes the evaluation template effect to eliminate the hallucinations of the large model. Description of the Drawings

[0034] Figure 1 It is a flowchart of the production process of the personalized employee template in a simple fusion table model and Text2SQL online form intelligent template arrangement method of the present invention;

[0035] Figure 2 It is a flowchart of the rendering process of the employee online template combined with the table model in a simple fusion table model and Text2SQL online form intelligent template arrangement method of the present invention;

[0036] Figure 3 It is a flowchart of separating and saving the table model of the work online submission data through the template style in a simple fusion table model and Text2SQL online form intelligent template arrangement method of the present invention. Detailed Embodiments

[0037] A simple fusion table model and Text2SQL online form intelligent template arrangement method includes the following methods: a method for producing personalized templates, a method for rendering the employee online template combined with the table model, and a method for separating and saving the employee online submission data; among them, in the method for producing personalized templates, it includes a basic template due to employee category differences, an employee key indicator extraction module, an LLM business indicator template, an indicator business SQL module formed by Text2SQL, a visualization intervention module, and saving the personalized templates with thousands of faces; in the method for rendering the employee online template combined with the table model, it includes loading the personalized employee template, executing the template business SQL to load the table model module, constructing structured data according to the template and styling data according to the template style, and rendering and finally outputting the json data containing the table model, structured data, and styling data through the medium; in the method for separating and saving the employee online submission data, it includes an employee data submission and review module, a template data engine module for separating data from the template, saving the table model, and the LLM understanding the business and optimizing the personalized template.

[0038] In this embodiment, the method for producing personalized templates specifically includes the following steps:

[0039] Step 1: Configure the basic template by category, including data block structure and style;

[0040] Step 2: Train the LLM private model through the enterprise and industry table models;

[0041] Step 3: Understand employee KPIs through the LLM private model and form key employee prompts;

[0042] Step 4: Use the key words to drive the LLM private model Text2SQL to generate the employee's private business SQL;

[0043] Step 5: Integrate the employee-specific business SQL with the basic template to form complete template data that is individualized and personalized for the employee;

[0044] Step 6: Visually execute the complete individualized template of the employee, judge the execution effect of the template, eliminate the illusion of the large model, perform visual arrangement, and form a template suitable for individual employees;

[0045] Step 7: Save individual employee templates for data submission and presentation.

[0046] In this embodiment, the employee online template combined with the table model rendering method specifically includes the following steps:

[0047] Step 1: Load the employee's individual template and extract the employee's exclusive business SQ;

[0048] Step 2: The data engine, combined with the table model, executes the exclusive business SQL to obtain the exclusive table model: own table model, subordinate table model, and approval table model;

[0049] Step 3: The data engine creates the header and the overall structure of rows and columns based on the individualized template. Driven by the dedicated table model, it creates multiple rows, multiple columns, and merges rows and columns to calculate and form structural data.

[0050] Step 4: For each title item in the data table header, configure a button. Click it to pop up the data aggregation of the column where the title item is located as a condition for filtering. After filtering, the list data content changes in a linked manner; data aggregation filtering includes: category selection of major categories and successively smaller categories, duplicate removal selection of data in this column, quantity / price formula selection, etc., and comprehensive condition selection;

[0051] Step 5: The data engine calculates the structural data and table model style according to the individualized template, including font: weight, color, centering, underline, bold and other style data;

[0052] Step 6: The data engine integrates the table model, structure data and style data to form json data;

[0053] Step 7: Output JSON data, which is rendered by the front end and presented to the user.

[0054] In this embodiment, the method for separately storing the data reported online by employees specifically includes the following steps:

[0055] Step 1: Employees submit comprehensive data through the front end. The data includes: table model, structure data, and style data.

[0056] Step 2: The data engine extracts the individual templates of employees.

[0057] Step 3: The data engine parses the comprehensive data according to the individual templates to decompose the table model and discards the structure and style data.

[0058] Step 4: The data engine saves the table model (proprietary business, approved data) according to the business rules.

[0059] Step 5: The LLM model continuously understands the employee table model, outputs business SQL through Text2SQL, and forms new version template data.

[0060] Step 6: Visualize and manage the new template data, which can be directly published as the official template for employees.

[0061] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, 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 simple method for intelligent template layout of integrating table model and Text2SQL online form, characterized in that: It includes the following methods: a method for producing templates with different faces for different people, a method for rendering online templates for employees in combination with table models, and a method for separating and saving data submitted by employees online; the method for producing templates with different faces for different people includes basic templates due to differences in employee categories, employee key indicator extraction modules, LLM business indicator templates, indicator business SQL modules formed by Text2SQL, as well as visualization intervention modules, and templates for saving individualized templates with different faces for different people; the method for rendering online templates for employees in combination with table models includes loading individualized templates for employees, executing template business SQL to load table model modules, constructing structural data according to templates and style data according to template styles, and rendering json data containing table models, structural data, and style data through media and finally outputting them; the method for separating and saving data submitted by employees online includes employee data submission and review modules, template data engine modules for separating data from templates, and saving table models, and LLM understanding business and optimizing personalized templates. The method for producing templates with different faces for different people specifically includes the following steps: Step 1: Configure the basic template by category, including data block structure and style; Step 2: Train the LLM private model through the enterprise and industry table models; Step 3: Understand employee KPIs through the LLM private model and form key employee prompts; Step 4: Use the key words to drive the LLM private model Text2SQL to generate the employee's private business SQL; Step 5: Integrate the employee-specific business SQL with the basic template to form complete template data that is individualized and personalized for the employee; Step 6: Visually execute the complete individualized template of the employee, judge the execution effect of the template, eliminate the illusion of the large model, perform visual arrangement, and form a template suitable for individual employees; Step 7: Save individual employee templates for data submission and presentation.

2. The intelligent template layout method for a simple fusion table model and Text2SQL online form according to claim 1, characterized in that: The employee online template combined with the table model rendering method specifically includes the following steps: Step 1: Load the employee's individual template and extract the employee's exclusive business SQ; Step 2: The data engine, combined with the table model, executes the exclusive business SQL to obtain the exclusive table model: own table model, subordinate table model, and approval table model; Step 3: The data engine creates the header and the overall structure of rows and columns based on the individualized template. Driven by the dedicated table model, it creates multiple rows, multiple columns, and merges rows and columns to calculate and form structural data. Step 4: For each title item in the data table header, configure a button. Click it to pop up the data aggregation of the column where the title item is located as a condition for filtering. After filtering, the list data content changes in a linked manner; data aggregation filtering includes: category selection of major categories and successively smaller categories, duplicate removal selection of data in this column, quantity / price formula selection, etc., and comprehensive condition selection; Step 5: The data engine calculates the structural data and table model style according to the individualized template, including font: weight, color, centering, underline, bold and other style data; Step 6: The data engine integrates the table model, structure data and style data to form json data; Step 7: Output json data and present it to the user through front-end rendering.

3. A simple method for intelligent template layout of integrating table model and Text2SQL online form according to claim 1, characterized in that: The method for separately storing the data reported by employees online specifically includes the following steps: Step 1: Employees submit comprehensive data through the front end. The data includes: table models, structural data, and style data; Step 2: The data engine extracts the individual templates of employees; Step 3: The data engine parses the comprehensive data according to the individual templates to decompose the table models, and discards the structural and style data; Step 4: The data engine saves the table models according to the business rules, including the proprietary business and approved data; Step 5: The LLM model continuously understands the employee table models, outputs business SQL through Text2SQL, and forms new version template data; Step 6: Visually manage the new template data, which can be directly published as the official templates for employees.

Images