Self-adaptive chart display method and system based on metadata driving

Abstract

The invention relates to the technical field of data visualization, in particular to an adaptive chart display method and system based on metadata driving, a front-end and back-end separation architecture is adopted, and the overall process comprises the steps that a user inputs an SQL query statement on a front-end interface, a back-end executes query operation after receiving a request, metadata analysis is conducted, and the SQL query statement is obtained; performing chart type decision through a rule engine, formatting data and then transmitting the data to a front end, and completing rendering display of a chart by the front end; the method has the advantages that from simple classification data comparison to complex multi-dimensional data analysis, the system can adapt to most conventional data analysis scenes, and the data visualization requirements of enterprises in different business scenes are met.

Description

Technical Field

[0001] The present invention relates to the field of data visualization technology, and in particular to a metadata-driven adaptive chart display method and system. Background Art

[0002] As a key link in data analysis, data visualization's core technologies cover areas such as data feature analysis and chart type decision-making.

[0003] 1. For data feature analysis, column type identification is fundamental and can be achieved through SQL parsing and data value detection. SQL parsing primarily analyzes features such as grouping, while data value detection infers the type based on the first row of data values. This approach is suitable for scenarios without explicit metadata, relies heavily on data integrity, and is susceptible to outliers. Data distribution analysis, including statistical features of numerical columns, the number of unique values ​​in categorical columns, and continuity testing of time columns, is key to determining chart applicability.

[0004] 2. In terms of chart type decision-making technology, rule-based decision engines are commonly used. The core is the mapping of data features and chart types. These engines are based on simple column type combinations, but lack in-depth analysis of data distribution details (such as the number of unique values ​​in categorical columns and outliers in numerical columns). Furthermore, they lack priority strategies for conflict scenarios, resulting in rough decision-making. Summary of the Invention

[0005] The purpose of the present invention is to provide a metadata-driven adaptive chart display method and system to solve the problems raised in the above background technology.

[0006] To achieve the above-mentioned objectives, the present invention provides the following technical solutions: a metadata-driven adaptive chart display method, which adopts a front-end and back-end separation architecture. The overall process includes: the user enters an SQL query statement in the front-end interface, the back-end executes the query operation after receiving the request, performs metadata parsing, makes a chart type decision through the rule engine, formats the data and passes it to the front-end, and the front-end completes the rendering and display of the chart; wherein, the metadata parsing layer interacts with the data source with the help of JDBC to obtain key information of the database column type, and generates column metadata containing detailed column information in combination with the data value verification mechanism; the rule engine layer executes the chart type decision algorithm based on the column metadata and in-depth analysis of the data distribution characteristics, stores a series of preset chart mapping rules, and makes decisions by comprehensively considering the column type combination, data distribution characteristics, and keyword matching multi-dimensional factors; the visual rendering layer uses the ECharts visualization framework to dynamically generate high-quality visual charts according to the chart type and data passed by the back-end, and supports responsive layout and rich interactive functions.

[0007] Preferably, the metadata parsing layer adopts a two-layer detection mechanism to identify column types. The specific process is: using JDBC to interact with the data source to obtain the metadata of the database result set, including column name, column type, and column quantity information; combined with the data value verification mechanism, regular expressions are used to match common time formats for columns that may store time data, and for columns that may store numerical data, data values ​​are attempted to be converted into numerical types to confirm the true type of the column, and column metadata containing detailed column information is generated to provide an accurate data basis for subsequent rule decisions.

[0008] Preferably, the rule engine layer uses the column metadata generated by the metadata parsing layer to perform an efficient chart type decision algorithm by deeply analyzing the data distribution characteristics, and comprehensively considers the following multi-dimensional factors to make decisions:

[0009] Column type combination: When there are time columns and numeric columns, select the line chart LINE if the data is a dense time series; select the area chart AREA if the data is not a dense time series; when there are categorical columns and numeric columns, select the horizontal bar chart BAR_HORIZONTAL if the number of unique values ​​in the categorical column is greater than 20; select the vertical bar chart BAR_VERTICAL if the number of unique values ​​in the categorical column is less than or equal to 20;

[0010] Data distribution characteristics: The number of unique values ​​in the categorical column affects the chart type selection, and the outlier detection of the numerical column affects the chart type decision. If there are significant outliers in the numerical column, choose a box plot; otherwise, choose a histogram.

[0011] Keyword matching: When the column name or annotation contains specific keywords such as "ratio" or "percent", pie charts are preferred to show the proportional distribution of data.

[0012] Preferably, the system has powerful data source adaptation capabilities, seamlessly connecting to traditional relational databases such as MySQL and PostgreSQL, as well as diverse data sources such as Excel files and RESTful APIs. By designing a unified interface to obtain metadata and business data, when a new data source needs to be added, developers only need to implement the relevant interface to quickly integrate the new data source.

[0013] Preferably, the system supports flexible configuration of chart mapping rules through the UI interface. Users can specify classification columns or adjust the priority of existing rules according to business needs. The system also provides a plug-in interface to allow third-party developers to extend new chart types. During the data processing process, the system establishes a complete exception handling mechanism. If a column of data is all NULL, it will be marked as UNKNOWN type and will not participate in the chart type decision. When metadata parsing fails, the system automatically falls back to data value detection mode to ensure the continuous and stable operation of the system.

[0014] A system for adaptive chart display method driven by metadata adopts a front-end and back-end separation architecture, including a front-end interface and a back-end processing module; the front-end interface is used to receive SQL query statements input by users and render and display charts processed by the back-end; the back-end processing module sequentially performs query operations, metadata parsing, chart type decision and data formatting; wherein, the metadata parsing layer interacts with the data source with the help of JDBC to obtain key information of database column types, and combines the data value verification mechanism to generate column metadata containing detailed column information; the rule engine layer executes a chart type decision algorithm based on column metadata and in-depth analysis of data distribution characteristics, and the algorithm comprehensively considers column type combinations, data distribution characteristics, and keyword matching multi-dimensional factors; the visualization rendering layer uses the ECharts visualization framework to dynamically generate high-quality visualization charts according to the chart types and data transmitted by the back-end, and supports responsive layout and rich interactive functions.

[0015] Preferably, the metadata parsing layer adopts a two-layer detection mechanism to identify column types. The specific process is: obtain the metadata of the database result set through JDBC interaction with the data source, and the metadata includes column name, column type, and column quantity information; at the same time, combined with the data value verification mechanism, for columns that may store time data, use regular expressions to match common time formats, such as yyyy-MM-dd, yyyy / MM / ddHH:mm:ss; for columns that may store numerical data, try to convert the data value into a numerical type to confirm the true type of the column, and finally generate column metadata ColumnMeta containing detailed column information to provide an accurate data basis for subsequent rule decisions.

[0016] Preferably, the rule engine layer uses the column metadata generated by the metadata parsing layer to perform an efficient chart type decision algorithm by deeply analyzing the data distribution characteristics, and comprehensively considers the following multi-dimensional factors to make decisions:

[0017] Column type combination: When there are time columns and numeric columns, select the line chart LINE if the data is a dense time series; select the area chart AREA if the data is not a dense time series; when there are categorical columns and numeric columns, select the horizontal bar chart BAR_HORIZONTAL if the number of unique values ​​in the categorical column is greater than 20; select the vertical bar chart BAR_VERTICAL if the number of unique values ​​in the categorical column is less than or equal to 20;

[0018] Data distribution characteristics: The number of unique values ​​in the categorical column affects the chart type selection, and the outlier detection of the numerical column affects the chart type decision. If there are significant outliers in the numerical column, choose a box plot; otherwise, choose a histogram.

[0019] Keyword matching: When the column name or annotation contains specific keywords such as "ratio" or "percent", pie charts are preferred to show the proportional distribution of data.

[0020] Preferably, the system has a strong data source adaptation capability. It obtains metadata and business data by designing a unified interface, and seamlessly connects to traditional relational databases such as MySQL and PostgreSQL, as well as Excel files and RESTful API diversified data sources. When a new data source needs to be added, developers only need to implement the relevant interface to quickly integrate the new data source.

[0021] Preferably, the system supports flexible configuration of chart mapping rules through the UI interface. Users can specify classification columns or adjust the priority of existing rules according to business needs. The system also provides a plug-in interface to allow third-party developers to extend new chart types. During the data processing process, the system establishes a complete exception handling mechanism. If a column of data is all NULL, it will be marked as UNKNOWN type and will not participate in the chart type decision. When metadata parsing fails, the system automatically falls back to data value detection mode to ensure the continuous and stable operation of the system.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] The metadata-driven adaptive chart display method and system proposed in this invention, with its metadata-priority strategy and dual-layer detection mechanism, significantly reduces the time column misjudgment rate from 22% to 8%, and achieves an overall chart matching accuracy of 92%, enabling data to be presented in the most appropriate chart form, providing users with accurate and intuitive data visualization results.

[0024] The fully automated process completely eliminates the tedious steps of manual configuration, significantly shortens the average time required for data visualization, greatly improves work efficiency, and enables users to quickly obtain data insights.

[0025] From simple classification data comparison to complex multi-dimensional data analysis (such as heat maps and stacked chart displays), the system can adapt to most common data analysis scenarios and meet the data visualization needs of enterprises in different business scenarios.

[0026] Non-technical users do not need to master complex chart configuration knowledge. They can easily generate professional and beautiful visual charts by simply entering SQL query statements, which effectively lowers the technical threshold for data visualization and makes data visualization an efficient tool that can be used by personnel from all departments of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a system architecture diagram of the present invention;

[0028] Figure 2 This is a flow chart of column type identification according to the present invention. DETAILED DESCRIPTION

[0029] In order to clearly and completely describe the objectives and technical solutions of the present invention and make the advantages more clearly understood, the embodiments of the present invention are further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are part of the embodiments of the present invention, not all of them, and are only used to explain the embodiments of the present invention, not to limit the embodiments of the present invention. All other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0030] In embodiment 1, the present invention provides a technical solution: a metadata-driven adaptive chart display method, which adopts a front-end and back-end separation architecture. The overall process includes: the user enters an SQL query statement in the front-end interface, the back-end executes the query operation after receiving the request, performs metadata parsing, makes a chart type decision through the rule engine, formats the data and passes it to the front-end, and the front-end completes the rendering and display of the chart; wherein, the metadata parsing layer interacts with the data source with the help of JDBC to obtain key information of the database column type, and combines the data value verification mechanism to generate column metadata containing detailed column information; the rule engine layer executes the chart type decision algorithm based on the column metadata and in-depth analysis of the data distribution characteristics, stores a series of preset chart mapping rules, and makes decisions by comprehensively considering the column type combination, data distribution characteristics, and keyword matching multi-dimensional factors; the visualization rendering layer uses the ECharts visualization framework to dynamically generate high-quality visualization charts according to the chart type and data passed by the back-end, and supports responsive layout and rich interactive functions.

[0031] The metadata parsing layer uses a two-layer detection mechanism to identify column types. The specific process is: using JDBC to interact with the data source to obtain the metadata of the database result set, including column name, column type, and column quantity information; combined with the data value verification mechanism, regular expressions are used to match common time formats for columns that may store time data, and for columns that may store numerical data, data values ​​are converted to numerical types to confirm the true type of the column, and column metadata containing detailed column information is generated to provide an accurate data basis for subsequent rule decisions.

[0032] The rule engine layer, based on the column metadata generated by the metadata parsing layer, executes an efficient chart type decision algorithm by deeply analyzing the data distribution characteristics, and makes decisions by comprehensively considering the following multi-dimensional factors: Column type combination: When there are time columns and numerical columns, if the data is a dense time series, select the line chart LINE; if the data is not a dense time series, select the area chart AREA; when there are categorical columns and numerical columns, if the number of unique values ​​in the categorical column is greater than 20, select the horizontal bar chart BAR_HORIZONTAL; if the number of unique values ​​in the categorical column is less than or equal to 20, select the vertical bar chart BAR_VERTICAL; Data distribution characteristics: The number of unique values ​​in the categorical column affects the chart type selection, and outlier detection in the numerical column affects the chart type decision. If there are significant outliers in the numerical column, select a box plot; otherwise, select a histogram; Keyword matching: When the column name or annotation contains specific keywords such as "ratio" or "percent", pie charts are preferred to display the proportional distribution of data.

[0033] The system has powerful data source adaptation capabilities, seamlessly connecting to traditional relational databases such as MySQL and PostgreSQL, as well as diverse data sources such as Excel files and RESTful APIs. By designing a unified interface to obtain metadata and business data, when a new data source needs to be added, developers only need to implement the relevant interface to quickly integrate the new data source.

[0034] The system supports flexible configuration of chart mapping rules through the UI interface. Users can specify classification columns or adjust the priority of existing rules according to business needs. The system also provides a plug-in interface, allowing third-party developers to expand new chart types. During the data processing process, the system establishes a complete exception handling mechanism. If a column of data is all NULL, it will be marked as UNKNOWN type and will not participate in the chart type decision. When metadata parsing fails, the system automatically falls back to data value detection mode to ensure the continuous and stable operation of the system.

[0035] In the second embodiment, based on the first embodiment, a system for metadata-driven adaptive chart display method is proposed, including:

[0036] The user enters the SQL query statement on the front-end interface. After receiving the request, the back-end executes the query operation, then performs metadata parsing (using a combination of JDBC and value verification), and then uses the rule engine to decide the chart type. The data is formatted and passed to the front-end, and finally the front-end completes the rendering and display of the chart.

[0037] (1) Metadata parsing layer. The metadata parsing layer is the foundation of the entire system. It interacts with the data source via JDBC to obtain key information such as database column types, which serves as an important basis for subsequent column type identification and chart type decisions. The metadata parsing layer also incorporates a data value validation mechanism to perform format validation on data values ​​and further confirm the true type of the column. Column metadata containing detailed column information is generated to provide an accurate data foundation for subsequent decision-making.

[0038] (2) Rule Engine Layer. The rule engine layer is the core of decision-making. It executes an efficient chart type decision algorithm based on the column metadata generated by the metadata parsing layer and the in-depth analysis of data distribution characteristics. This layer stores a series of preset chart mapping rules. These rules comprehensively consider multiple factors such as column type combinations, data distribution characteristics, and keyword matching.

[0039] (3) Visualization rendering layer. The visualization rendering layer is responsible for visually displaying the chart types and processed data determined by the rule engine layer. Using the ECharts visualization framework, high-quality visualization charts are dynamically generated based on the chart types and data passed by the backend. Furthermore, to enhance the user experience, the visualization rendering layer also supports responsive layouts and rich interactive features, providing more complete details and rich information.

[0040] This fully automated operation method greatly improves the efficiency of data visualization and lowers the user's operation threshold.

[0041] Details are as follows:

[0042] (1) Metadata-driven column type identification (two-tier detection mechanism)

[0043] Process such as Figure 2 As shown in the figure, JDBC is used to interact with the data source and obtain key information such as database column types. JDBC provides a unified interface that allows the system to easily connect to different database types, such as MySQL and PostgreSQL. It can obtain metadata about database result sets, including column names, column types, and the number of columns. This metadata is crucial for subsequent column type identification and chart type decisions.

[0044] However, relying solely on metadata may not accurately determine the type of certain columns. For example, some string type columns may actually store time data or numerical data. Therefore, the metadata parsing layer also incorporates a data value verification mechanism. By performing format verification on the data value, the true type of the column is further confirmed. For example, for columns that may store time data, regular expressions are used to match common time formats, such as `yyyy-MM-dd`, `yyyy / MM / ddHH:mm:ss`, etc.; for columns that may store numerical data, attempts are made to convert the data value to a numerical type to determine whether it is a numerical column. In this way, the metadata parsing layer is able to generate column metadata (ColumnMeta) containing detailed column information, providing an accurate data basis for subsequent rule decisions.

[0045] (2) Multi-dimensional chart type decision engine

[0046] The rules engine is the system's decision-making core. Based on the column metadata generated by the metadata parsing layer, it deeply analyzes data distribution characteristics and executes an efficient chart type decision algorithm. This layer stores a series of preset chart mapping rules, which comprehensively consider multiple factors such as column type combinations, data distribution characteristics, and keyword matching.

[0047] The combination of column types is an important factor in decision-making. For example, when there are both time columns and numerical columns, a line chart or area chart is often appropriate to show data trends. When there are both categorical columns and numerical columns, a column chart or bar chart is more suitable for comparing the numerical differences between different categories.

[0048] Data distribution characteristics are also a key factor considered by the rule engine. For example, the number of unique values ​​in a categorical column can influence the choice of chart type. When a categorical column has too many unique values, using a bar chart may result in a crowded chart display. In this case, the system will automatically select a bar chart for vertical display to improve the readability of the chart. In addition, outlier detection in numeric columns can also influence the decision on chart type. If a numeric column has significant outliers, using a boxplot can better highlight abnormal data; otherwise, a histogram can more intuitively display the data distribution.

[0049] Keyword matching is another important dimension of the rule engine. When a column name or comment contains specific keywords such as "ratio" or "percent," the system will prioritize pie charts to display the proportional distribution of the data.

[0050] The rule engine layer comprehensively considers these multi-dimensional factors and selects the most suitable display method for the current data from a variety of chart types to ensure that the chart can accurately and intuitively present data characteristics.

[0051] The decision rules are shown in the following table:

[0052]

[0053]

[0054] Table 1: Description of the decision rules in the diagram

[0055] (3) Expansion and adaptability

[0056] The system boasts powerful data source adaptability, seamlessly integrating with traditional relational databases like MySQL and PostgreSQL, as well as diverse data sources like Excel files and RESTful APIs. By designing a unified interface for acquiring metadata and business data, developers can quickly integrate new data sources simply by implementing the relevant interfaces when new data sources are needed, significantly improving the system's scalability and versatility.

[0057] To meet the personalized needs of diverse users, the system supports flexible configuration of chart mapping rules through the UI. For example, users can specify classification columns or adjust the priority of existing rules based on their business needs. Furthermore, the system provides a plug-in interface, allowing third-party developers to add new chart types such as radar charts and Sankey diagrams, continuously enriching the system's functionality and application scenarios.

[0058] During data processing, the system has established a comprehensive exception handling mechanism. If a column of data is entirely NULL, it is marked as UNKNOWN and excluded from chart type decisions, preventing errors caused by abnormal data. If metadata parsing fails, the system automatically falls back to data value detection mode, ensuring continued stable operation and the smooth functioning of data visualization.

[0059] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A metadata-driven adaptive chart display method, characterized by: It adopts a front-end and back-end separation architecture, and the overall process includes: the user enters the SQL query statement in the front-end interface, the back-end executes the query operation after receiving the request, performs metadata parsing, makes chart type decisions through the rule engine, formats the data and passes it to the front-end, and the front-end completes the rendering and display of the chart; among them, the metadata parsing layer uses JDBC to interact with the data source to obtain key information of the database column type, and combines the data value verification mechanism to generate column metadata containing detailed column information; the rule engine layer executes the chart type decision algorithm based on the column metadata and in-depth analysis of the data distribution characteristics, stores a series of preset chart mapping rules, and makes decisions by comprehensively considering the column type combination, data distribution characteristics, and keyword matching multi-dimensional factors; the visual rendering layer uses the ECharts visualization framework to dynamically generate high-quality visual charts according to the chart type and data passed by the back-end, and supports responsive layout and rich interactive functions.

2. The metadata-driven adaptive chart display method according to claim 1, characterized in that: The metadata parsing layer uses a two-layer detection mechanism to identify column types. The specific process is: using JDBC to interact with the data source to obtain the metadata of the database result set, including column name, column type, and column quantity information; combined with the data value verification mechanism, regular expressions are used to match common time formats for columns that may store time data, and for columns that may store numerical data, data values ​​are converted to numerical types to confirm the true type of the column, and column metadata containing detailed column information is generated to provide an accurate data basis for subsequent rule decisions.

3. The metadata-driven adaptive chart display method according to claim 2, characterized in that: The rule engine layer uses the column metadata generated by the metadata parsing layer to perform an efficient chart type decision algorithm by deeply analyzing the data distribution characteristics. It makes decisions based on the following multi-dimensional factors: Column type combination: When there are time columns and numeric columns, select the line chart LINE if the data is a dense time series; select the area chart AREA if the data is not a dense time series; when there are categorical columns and numeric columns, select the horizontal bar chart BAR_HORIZONTAL if the number of unique values ​​in the categorical column is greater than 20; select the vertical bar chart BAR_VERTICAL if the number of unique values ​​in the categorical column is less than or equal to 20; Data distribution characteristics: The number of unique values ​​in the categorical column affects the chart type selection, and the outlier detection of the numerical column affects the chart type decision. If there are significant outliers in the numerical column, choose a box plot; otherwise, choose a histogram. Keyword matching: When column names or comments contain specific keywords such as "ratio" and "percent", pie charts are preferred to show the proportional distribution of data.

4. The metadata-driven adaptive chart display method according to claim 3, characterized in that: The system has powerful data source adaptation capabilities, seamlessly connecting to traditional relational databases such as MySQL and PostgreSQL, as well as diverse data sources such as Excel files and RESTful APIs. By designing a unified interface to obtain metadata and business data, when a new data source needs to be added, developers only need to implement the relevant interface to quickly integrate the new data source.

5. The metadata-driven adaptive chart display method according to claim 4, characterized in that: The system supports flexible configuration of chart mapping rules through the UI interface. Users can specify classification columns or adjust the priority of existing rules according to business needs. The system also provides a plug-in interface, allowing third-party developers to expand new chart types. During the data processing process, the system establishes a complete exception handling mechanism. If a column of data is all NULL, it will be marked as UNKNOWN type and will not participate in the chart type decision. When metadata parsing fails, the system automatically falls back to data value detection mode to ensure the continuous and stable operation of the system.

6. A system for the metadata-driven adaptive chart display method according to claim 5, characterized in that: It adopts a front-end and back-end separation architecture, including a front-end interface and a back-end processing module; the front-end interface is used to receive SQL query statements input by the user and render and display the charts processed by the back-end; the back-end processing module sequentially performs query operations, metadata parsing, chart type decisions, and data formatting; among them, the metadata parsing layer uses JDBC to interact with the data source to obtain key information of database column types, and combines the data value verification mechanism to generate column metadata containing detailed column information; the rule engine layer executes the chart type decision algorithm based on the column metadata and in-depth analysis of data distribution characteristics. The algorithm comprehensively considers column type combinations, data distribution characteristics, and keyword matching multi-dimensional factors; the visualization rendering layer uses the ECharts visualization framework to dynamically generate high-quality visualization charts based on the chart types and data passed by the back-end, and supports responsive layout and rich interactive functions.

7. A system according to claim 6, characterized in that: The metadata parsing layer uses a two-layer detection mechanism to identify column types. The specific process is: through JDBC interaction with the data source, the metadata of the database result set is obtained. The metadata includes column name, column type, and column quantity information; at the same time, combined with the data value verification mechanism, for columns that may store time data, regular expressions are used to match common time formats, such as yyyy-MM-dd, yyyy / MM / ddHH:mm:ss; for columns that may store numerical data, attempts are made to convert the data value into a numerical type to confirm the true type of the column, and finally column metadata ColumnMeta containing detailed column information is generated to provide an accurate data basis for subsequent rule decisions.

8. A system according to claim 7, characterized in that: The rule engine layer uses the column metadata generated by the metadata parsing layer to perform an efficient chart type decision algorithm by deeply analyzing the data distribution characteristics. It makes decisions based on the following multi-dimensional factors: Column type combination: When there are time columns and numeric columns, select the line chart LINE if the data is a dense time series; select the area chart AREA if the data is not a dense time series; when there are categorical columns and numeric columns, select the horizontal bar chart BAR_HORIZONTAL if the number of unique values ​​in the categorical column is greater than 20; select the vertical bar chart BAR_VERTICAL if the number of unique values ​​in the categorical column is less than or equal to 20; Data distribution characteristics: The number of unique values ​​in the categorical column affects the chart type selection, and the outlier detection of the numerical column affects the chart type decision. If there are significant outliers in the numerical column, choose a box plot; otherwise, choose a histogram. Keyword matching: When column names or comments contain specific keywords such as "ratio" and "percent", pie charts are preferred to show the proportional distribution of data.

9. A system according to claim 8, characterized in that: The system has powerful data source adaptation capabilities. It obtains metadata and business data by designing a unified interface, and seamlessly connects to traditional relational databases such as MySQL and PostgreSQL, as well as diverse data sources such as Excel files and RESTful APIs. When a new data source needs to be added, developers only need to implement the relevant interface to quickly integrate the new data source.

10. A system according to claim 9, characterized in that: The system supports flexible configuration of chart mapping rules through the UI interface. Users can specify classification columns or adjust the priority of existing rules according to business needs. The system also provides a plug-in interface, allowing third-party developers to expand new chart types. During the data processing process, the system establishes a complete exception handling mechanism. If a column of data is all NULL, it will be marked as UNKNOWN type and will not participate in the chart type decision. When metadata parsing fails, the system automatically falls back to data value detection mode to ensure the continuous and stable operation of the system.

Images