Data visualization generation method and device based on natural language big model function-call

Abstract

The invention relates to a data visualization generation method and device based on a natural language big model function-call. The method comprises the following steps: acquiring metadata information from a data source, and creating a chart template; creating a data table visualization engine, determining a configuration template of the data table visualization engine, obtaining a visualization chart according to the chart configuration data, the request data structure and the configuration template, obtaining a visualization generation request input by a user, performing intention recognition through a natural language large model, and obtaining a visualization result. The method comprises the following steps: obtaining a data source, a data view or metadata information, storing the data source, the data view or the metadata information into a buffer area, identifying a chart type corresponding to a visualization generation request, obtaining a configuration template corresponding to the chart type, storing the configuration template into the buffer area, and carrying out parameter assembly and triggering in the buffer area by utilizing a natural language big model function-call capability. And obtaining visual data. By adopting the method, flexible visual generation can be realized.

Description

Technical Field

[0001] The present application relates to the field of visualization technology, and in particular to a method and device for generating data visualization based on a natural language large model function-call. Background Art

[0002] Data visualization generation is an important means for companies or institutions to analyze data and assist in decision-making. However, data visualization generation is cumbersome to configure, the process is complicated, and requires a certain level of professional knowledge. These shortcomings mean that data visualization not only requires professionals but also requires a lot of effort to design and configure data visualization reports, which limits the rapid generation and delivery of paradigm data charts, and indirectly affects the speed of data analysis and business decision-making.

[0003] Generally speaking, there are two ways to speed up the generation of data reports. The first method is to speed up the generation of reports by increasing manpower input and improving proficiency, but this method does not generally reduce the speed of data visualization generation and is more dependent on individuals.

[0004] The second method is to pre-configure various visualization templates and generate visualization reports based on the template configuration, thereby shortening the time for data visualization generation. However, at present, data requirements are various and the focus is different. Templates cannot predict requirements in advance, which has certain limitations in improving the speed of data visualization. Therefore, it is urgent to propose a data visualization generation method that can quickly generate visualization reports and cope with complex data requirements. Summary of the invention

[0005] Based on this, it is necessary to provide a method and device for generating data visualization based on a large natural language model function-call in response to the above technical problems.

[0006] A data visualization generation method based on a natural language large model function-call, the method comprising:

[0007] Get metadata information from data sources;

[0008] Creating a chart template; the chart template is created in a visual operation mode, the visual interface parses metadata information of the data view, operates the data view to a specified area after the parsing is completed, and stores the operation as chart configuration data, and assembles the selected specified area into a request data structure; the request data structure is used for statement query;

[0009] Creating a data table visualization engine, and determining a configuration template of the data table visualization engine, and obtaining a visualization chart according to the chart configuration data, the request data structure and the configuration template;

[0010] Obtain a visualization generation request input by the user, perform intent recognition through a natural language big model, obtain data source, data view or metadata information, and store the data source, data view or metadata information in a buffer, identify the chart type corresponding to the visualization generation request, obtain a configuration template corresponding to the chart type and store it in a buffer, use the function-call capability of the natural language big model to assemble and trigger parameters in the buffer, and obtain visualization data.

[0011] In one embodiment, the data source includes a relational database, a non-relational database, a structured file, and a request interface; further comprising: when the data source is a relational database, by setting a query logic view in a data connection pool, performing single table query, joint table query, and field calculation conversion on the relational database to obtain metadata information;

[0012] When the data source is a non-relational database, a structured file or a request interface, metadata information is extracted from the data of the non-relational database, the structured file or the request interface by setting fields.

[0013] In one embodiment, setting a field includes: field name, field description, type, default value, and separator.

[0014] In one embodiment, the visualization operation includes: dragging, pulling and dropping; the designated area includes: selected rows, columns and filtered fields. It also includes: dragging, pulling and dropping the data view to the selected rows, columns and filtered fields, storing the data generated by the operation as chart configuration data, and assembling the selected rows, columns and filtered fields into a request data structure.

[0015] In one of the embodiments, the configuration information in the configuration template includes: table type, table header information, table style, and data analysis;

[0016] The table types include: ordinary table, perspective chart, bar chart and line chart;

[0017] The table header information includes: column header information, row header information and grouping information;

[0018] The table style includes: table header color, table header font style, table body color and table body font style;

[0019] The data analysis includes: paging row number, row total configuration, column total configuration, and download settings.

[0020] In one of the embodiments, it further includes: when the data source is a relational database, using the request data structure to construct SQL through the SQL Builder function of the visualization framework, and then parsing, verifying, optimizing and executing, querying data from the relational database and assembling it into a two-dimensional matrix including column headers;

[0021] When the data source is a non-relational database, a structured file or a request interface, metadata information is extracted from the data of the non-relational database, the structured file or the request interface, and the metadata information is assembled into a two-dimensional data matrix.

[0022] In one of the embodiments, it also includes: according to the intention recognition of the visualization generation request, the buffer data is fine-tuned by using the natural language large model function-call method.

[0023] A data visualization generation device based on a natural language large model function-call, the device comprising:

[0024] A data acquisition module, used to obtain metadata information from a data source;

[0025] A chart template creation module is used to create a chart template; the chart template is created in a visual operation mode, the visual interface parses the metadata information of the data view, operates the data view to a specified area after the parsing is completed, and stores the operation as chart configuration data, and assembles the selected specified area into a request data structure; the request data structure is used for statement query;

[0026] A visualization engine creation module, used to create a data table visualization engine and determine a configuration template of the data table visualization engine, and obtain a visualization chart according to the chart configuration data, the request data structure and the configuration template;

[0027] The visualization generation module is used to obtain a visualization generation request input by a user, perform intent recognition through a natural language big model, obtain data source, data view or metadata information, and store the data source, data view or metadata information in a buffer, and identify the chart type corresponding to the visualization generation request, obtain the configuration template corresponding to the chart type and store it in a buffer, and use the natural language big model function-call capability to assemble and trigger parameters in the buffer to obtain visualization data.

[0028] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0029] Get metadata information from data sources;

[0030] Creating a chart template; the chart template is created in a visual operation mode, the visual interface parses metadata information of the data view, operates the data view to a specified area after the parsing is completed, and stores the operation as chart configuration data, and assembles the selected specified area into a request data structure; the request data structure is used for statement query;

[0031] Creating a data table visualization engine, and determining a configuration template of the data table visualization engine, and obtaining a visualization chart according to the chart configuration data, the request data structure and the configuration template;

[0032] Obtain a visualization generation request input by the user, perform intent recognition through a natural language big model, obtain data source, data view or metadata information, and store the data source, data view or metadata information in a buffer, identify the chart type corresponding to the visualization generation request, obtain a configuration template corresponding to the chart type and store it in a buffer, use the function-call capability of the natural language big model to assemble and trigger parameters in the buffer, and obtain visualization data.

[0033] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the following steps:

[0034] Get metadata information from data sources;

[0035] Creating a chart template; the chart template is created in a visual operation mode, the visual interface parses metadata information of the data view, operates the data view to a specified area after the parsing is completed, and stores the operation as chart configuration data, and assembles the selected specified area into a request data structure; the request data structure is used for statement query;

[0036] Creating a data table visualization engine, and determining a configuration template of the data table visualization engine, and obtaining a visualization chart according to the chart configuration data, the request data structure and the configuration template;

[0037] Obtain a visualization generation request input by the user, perform intent recognition through a natural language big model, obtain data source, data view or metadata information, and store the data source, data view or metadata information in a buffer, identify the chart type corresponding to the visualization generation request, obtain a configuration template corresponding to the chart type and store it in a buffer, use the function-call capability of the natural language big model to assemble and trigger parameters in the buffer, and obtain visualization data.

[0038] The above-mentioned data visualization generation method and device based on the natural language large model function-call provide rich channels for data acquisition by presetting multiple types of data sources and corresponding data views; construct a query data structure to achieve effective docking between data and chart templates. Build a visualization framework, and use the Ali Antv S2 engine and detailed configuration information to determine the chart style and function. The data visualization generation step adapts to different data sources to obtain data matrices and convert them into charts. The key is to use the large model function-call capability, iterate the link according to the user's multimodal request, accurately identify the data source and chart type, fine-tune the template configuration, optimize the assembly parameters to trigger data query and chart rendering, and generate visualization data that meets the intention. The overall efficiency of data visualization report generation is improved, complex data requirements are met, and personalized data display is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a flowchart of a method for generating data visualization based on a natural language large model function-call in one embodiment;

[0040] Figure 2 It is a structural block diagram of a data visualization generation device based on a natural language large model function-call in one embodiment;

[0041] Figure 3 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0043] In one embodiment, Figure 1 As shown, a data visualization generation method based on a natural language large model function-call is provided, including the following steps:

[0044] Step 102: Obtain metadata information from a data source.

[0045] For relational databases, metadata information includes field name, field type, field description, field default value, and field calculation logic (for calculation conversion fields). This information clearly describes the basic properties and characteristics of each field in the database. For example, the field name clarifies the identification of the data, the field type determines the format of the data (such as integer, string, date, etc.), and the field description explains the meaning of the field to help users understand the data content. For example, in a sales database, in the metadata information of the "Order Amount" field, the field name is "Order Amount", the field type may be "Numeric", and the field description is "Record the transaction amount of each order."

[0046] For non-relational databases, structured files, or specific request interfaces, metadata setting information specifies field names, field descriptions, field types, field default values, field separators, etc. Taking structured files (such as CSV files) as an example, the field separators (such as commas) in the metadata information are used to distinguish different fields in the file to ensure that the data can be correctly parsed when reading and processing.

[0047] Step 104, create a chart template.

[0048] The chart template is created using a visual operation method. The visual interface parses the metadata information of the data view. After the parsing is completed, the data view is operated to the specified area and the operation is stored as chart configuration data. The selected specified area is assembled into a request data structure; the request data structure is used for statement query.

[0049] Step 106, creating a data table visualization engine, and determining a configuration template of the data table visualization engine, and obtaining a visualization chart according to the chart configuration data, the request data structure and the configuration template.

[0050] Step 108, obtain the visualization generation request input by the user, perform intent recognition through the natural language big model, obtain the data source, data view or metadata information, and store the data source, data view or metadata information in the buffer, and identify the chart type corresponding to the visualization generation request, obtain the configuration template corresponding to the chart type and store it in the buffer, use the natural language big model function-call capability to assemble and trigger the parameters in the buffer, and obtain visualization data.

[0051] In the above-mentioned data visualization generation method based on the natural language large model function-call, by presetting various types of data sources and corresponding data views, rich channels are provided for data acquisition; the query data structure is constructed to achieve effective docking between data and chart templates. Build a visualization framework, use the Alibaba Antv S2 engine and detailed configuration information to determine the chart style and function. The data visualization generation step adapts to different data sources to obtain data matrices and convert them into charts. The key is to use the large model function-call capability, iterate the link according to the user's multimodal request, accurately identify the data source and chart type, fine-tune the template configuration, optimize the assembly parameters to trigger data query and chart rendering, and generate visualization data that meets the intention. The overall efficiency of data visualization report generation is improved, complex data requirements are met, and personalized data display is achieved.

[0052] In one of the embodiments, the data source includes a relational database, a non-relational database, a structured file, and a request interface; when the data source is a relational database, a query logic view is set in a data connection pool, and single-table query, joint-table query, and field calculation conversion are performed on the relational database to obtain metadata information; when the data source is a non-relational database, a structured file, or a request interface, the metadata information is extracted from the data of the non-relational database, structured file, or request interface by setting fields.

[0053] In one embodiment, setting a field includes: field name, field description, type, default value, and separator.

[0054] In another embodiment, the visualization operation method includes: dragging, pulling and dropping; the designated area includes: selected rows, columns and filtered fields; creating a data table visualization engine includes: dragging, pulling and dropping the data view to the selected rows, columns and filtered fields, storing the data generated by the operation as chart configuration data, and assembling the selected rows, columns and filtered fields into a request data structure.

[0055] Specifically, chart templates need to be created manually in the early stage. The creation of chart templates is a visual drag-and-drop process. The page will first parse the metadata information of the pre-selected data view. After the metadata fields are parsed, they can be dragged to areas such as rows, columns, and filters. These data will be stored in the chart configuration data in the form of objects converted to json. The selected rows, columns, and filter fields will be assembled into the request data structure as parameters for building query statements.

[0056] In one of the embodiments, the configuration information in the configuration template includes: table type, header information, table style and data analysis; table types include: ordinary table, perspective chart, bar chart and line chart; header information includes: column header information, row header information and grouping information; table style includes: header color, header font style, body color and body font style; data analysis includes: number of paging rows, row total configuration, column total configuration, download settings.

[0057] Specifically, the underlying layer of the data visualization framework uses Alibaba's Antv S2 data table visualization engine, and the chart construction is completed through the json configuration template stored in the database. The above configuration determines the overall style and appearance details of this type of data chart, and fine-tunes the parameters to meet the requirements of data users.

[0058] In one of the embodiments, when the data source is a relational database, the SQL is constructed using the SQL Builder function of the visualization framework using the request data structure, and then parsed, verified, optimized and executed to query data from the relational database and assemble it into a two-dimensional matrix containing column headers; when the data source is a non-relational database, structured file or request interface, metadata information is extracted from the data of the non-relational database, structured file or request interface, and the metadata information is assembled into a two-dimensional data matrix.

[0059] In another embodiment, based on the intent recognition of the visualization generation request, a natural language large model function-call method is used to fine-tune the buffer data.

[0060] Specifically, according to the user's incoming requests in multiple modes, the big model will identify their intent and implement link iteration. First, the big model will identify the data source and data view or metadata information based on the context passed in by the user. If it cannot be identified, the user will be prompted to continue inputting until it can be identified. After identification, the parameter will be stored in the cache area; then the big model will identify the chart type required by the user based on the context, obtain the configuration template corresponding to the type, and generate a backup to the buffer. The various configurations of the template can be fine-tuned according to the user's intent, and the template fine-tuning adopts the big model function-call method; finally, the big model optimizes the parameters of the buffer in a specific format, and uses the function-call capability to assemble and trigger the parameters, which ultimately triggers a series of methods such as data query and chart rendering, and finally generates a complex visualization data that meets the user's intent, and returns the preview link to the user, who can preview or download it.

[0061] It should be understood that although Figure 1The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0062] In one embodiment, Figure 2 As shown, a data visualization generation device based on a natural language large model function-call is provided, comprising: a data acquisition module 202, a chart template creation module 204, a visualization engine creation module 206 and a visualization generation module 208, wherein:

[0063] The data acquisition module 202 is used to acquire metadata information from a data source;

[0064] The chart template creation module 204 is used to create a chart template; the chart template is created by visual operation, the visual interface parses the metadata information of the data view, operates the data view to the specified area after the parsing is completed, and stores the operation as chart configuration data, and assembles the selected specified area into a request data structure; the request data structure is used for statement query;

[0065] A visualization engine creation module 206, used to create a data table visualization engine, and determine a configuration template of the data table visualization engine, and obtain a visualization chart according to the chart configuration data, the request data structure and the configuration template;

[0066] The visualization generation module 208 is used to obtain a visualization generation request input by a user, perform intent recognition through a natural language big model, obtain data source, data view or metadata information, and store the data source, data view or metadata information in a buffer, and identify the chart type corresponding to the visualization generation request, obtain the configuration template corresponding to the chart type and store it in a buffer, and use the natural language big model function-call capability to assemble and trigger parameters in the buffer to obtain visualization data.

[0067] In one embodiment, the data source includes a relational database, a non-relational database, a structured file, and a request interface; the data acquisition module 202 is also used to, when the data source is a relational database, set a query logic view in the data connection pool, perform single-table query, joint-table query, and field calculation conversion on the relational database to obtain metadata information; when the data source is a non-relational database, a structured file, or a request interface, extract metadata information from the data of the non-relational database, structured file, or request interface by setting fields.

[0068] In one embodiment, setting a field includes: field name, field description, type, default value, and separator.

[0069] In one embodiment, the visualization operation method includes: dragging, pulling and dropping; the designated area includes: selected rows, columns and filtered fields; the visualization engine creation module 206 is also used to drag, pull and drop the data view to the selected rows, columns and filtered fields, store the data generated by the operation as chart configuration data, and assemble the selected rows, columns and filtered fields into a request data structure.

[0070] In one of the embodiments, the configuration information in the configuration template includes: table type, table header information, table style, and data analysis;

[0071] The table types include: ordinary table, perspective chart, bar chart and line chart;

[0072] The table header information includes: column header information, row header information and grouping information;

[0073] The table style includes: table header color, table header font style, table body color and table body font style;

[0074] The data analysis includes: paging row number, row total configuration, column total configuration, and download settings.

[0075] In one of the embodiments, the visualization engine creation module 206 is further used to use the request data structure to construct SQL through the SQL Builder function of the visualization framework when the data source is a relational database, and then parse, verify, optimize and execute, query data from the relational database and assemble it into a two-dimensional matrix including column headers;

[0076] When the data source is a non-relational database, a structured file or a request interface, metadata information is extracted from the data of the non-relational database, the structured file or the request interface, and the metadata information is assembled into a two-dimensional data matrix.

[0077] In one of the embodiments, the visualization generation module 208 is further configured to fine-tune the buffer data by using a natural language large model function-call method according to the intent recognition of the visualization generation request.

[0078] For the specific limitations of the data visualization generation device based on the natural language large model function-call, please refer to the limitations of the data visualization generation method based on the natural language large model function-call above, which will not be repeated here. Each module in the above-mentioned data visualization generation device based on the natural language large model function-call can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0079] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a data visualization generation method based on a natural language large model function-call is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0080] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0081] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method in the above embodiment when executing the computer program.

[0082] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method in the above embodiment are implemented.

[0083] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0084] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0085] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A data visualization generation method based on a natural language large model function-call, characterized in that: The method comprises: Get metadata information from data sources; Creating a chart template; the chart template is created in a visual operation mode, the visual interface parses metadata information of the data view, operates the data view to a specified area after the parsing is completed, and stores the operation as chart configuration data, and assembles the selected specified area into a request data structure; the request data structure is used for statement query; Creating a data table visualization engine, and determining a configuration template of the data table visualization engine, and obtaining a visualization chart according to the chart configuration data, the request data structure and the configuration template; Obtain a visualization generation request input by the user, perform intent recognition through a natural language big model, obtain data source, data view or metadata information, and store the data source, data view or metadata information in a buffer, identify the chart type corresponding to the visualization generation request, obtain a configuration template corresponding to the chart type and store it in a buffer, use the function-call capability of the natural language big model to assemble and trigger parameters in the buffer, and obtain visualization data.

2. The method according to claim 1, characterized in that: The data sources include relational databases, non-relational databases, structured files, and request interfaces; Get metadata information from the data source, including: When the data source is a relational database, a query logic view is set in the data connection pool to perform single-table query, joint-table query, and field calculation conversion on the relational database to obtain metadata information; When the data source is a non-relational database, a structured file or a request interface, metadata information is extracted from the data of the non-relational database, the structured file or the request interface by setting fields.

3. The method according to claim 2, characterized in that Field settings include: field name, field description, type, default value, and separator.

4. The method according to claim 1, characterized in that: The visual operation methods include: dragging, dropping and dropping; the designated area includes: selected rows, columns and filtered fields; The step of creating a data table visualization engine includes: Drag, drop and drop the data view to the selected rows, columns and filtered fields, store the data generated by the operation as chart configuration data, and assemble the selected rows, columns and filtered fields into the request data structure.

5. The method according to claim 1, characterized in that: The configuration information in the configuration template includes: table type, table header information, table style and data analysis; The table types include: ordinary table, perspective chart, bar chart and line chart; The table header information includes: column header information, row header information and grouping information; The table style includes: table header color, table header font style, table body color and table body font style; The data analysis includes: paging row number, row total configuration, column total configuration, and download settings.

6. The method according to any one of claims 1 to 5, characterized in that: Obtaining a visual chart according to the chart configuration data, the request data structure and the configuration template, including: When the data source is a relational database, the request data structure is used to construct SQL through the SQL Builder function of the visualization framework, and then parsed, verified, optimized and executed to query data from the relational database and assemble it into a two-dimensional matrix containing column headers; When the data source is a non-relational database, a structured file or a request interface, metadata information is extracted from the data of the non-relational database, the structured file or the request interface, and the metadata information is assembled into a two-dimensional data matrix.

7. The method according to any one of claims 1 to 5, characterized in that: Before utilizing the function-call capability of the natural language large model to assemble and trigger parameters in the buffer to obtain visualization data, the method further includes: Based on the intent recognition of the visualization generation request, the buffer data is fine-tuned using the natural language large model function-call method.

8. A data visualization generation device based on a natural language large model function-call, characterized in that: The device comprises: A data acquisition module, used to obtain metadata information from a data source; A chart template creation module is used to create a chart template; the chart template is created in a visual operation mode, the visual interface parses the metadata information of the data view, operates the data view to a specified area after the parsing is completed, and stores the operation as chart configuration data, and assembles the selected specified area into a request data structure; the request data structure is used for statement query; A visualization engine creation module, used to create a data table visualization engine and determine a configuration template of the data table visualization engine, and obtain a visualization chart according to the chart configuration data, the request data structure and the configuration template; The visualization generation module is used to obtain a visualization generation request input by a user, perform intent recognition through a natural language big model, obtain data source, data view or metadata information, and store the data source, data view or metadata information in a buffer, and identify the chart type corresponding to the visualization generation request, obtain the configuration template corresponding to the chart type and store it in a buffer, and use the natural language big model function-call capability to assemble and trigger parameters in the buffer to obtain visualization data.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images