Method, device, equipment and storage medium for data visualization

Abstract

The present invention provides a method, apparatus, device, and storage medium for data visualization, wherein the method comprises the steps of obtaining dimensional information of the data to be displayed; obtaining at least one dimensional combination based on the dimensional information; automatically searching and matching corresponding insight analysis solutions for each dimensional combination, and obtaining corresponding data visualization charts and data conclusions; wherein each dimensional combination corresponds to at least two data visualization charts, and the data visualization charts correspond to the data conclusions. The present invention aims to improve the problem that existing data visualization methods cannot quickly obtain data conclusions.

Description

Technical Field

[0001] The present invention relates to the field of data visualization, and in particular to a method, apparatus, device and storage medium for data visualization. Background Art

[0002] In today's scenarios, most users use statistical graphs, charts, infographics, and other tools to visualize data in order to convey information clearly and effectively. Generally speaking, data visualization involves encoding numerical data using points, lines, or bars to visually convey quantitative information. Existing commercially available technologies primarily fall into two categories: open source and commercially available analysis platforms.

[0003] Most open-source technologies come with a variety of simple or complex charts built in. Users can configure parameters, add existing data, and then select appropriate charts and styling rules for data visualization. Open-source technologies are more suitable for developers, but are less suitable for those working in data analysis and operations. They require a certain learning and development cost, and are still far from achieving the goal of quickly obtaining data conclusions.

[0004] Most commercially available analytics platforms have built-in analysis for various scenarios, such as funnel analysis, retention analysis, and path analysis. By configuring source data and attribute dimensions, scenario analysis can be visualized for further analysis. Commercial analytics platforms are more suitable for those working in data analysis and operations roles. They require configuration based on existing scenario analysis. However, they are unable to meet the needs of certain analytical tasks, requiring manual data collection and analysis, and lacking the ability to perform coordinated analysis. Furthermore, due to the multi-dimensionality and complexity of data, many analysts become lost during this process, requiring repeated attempts to find analytical insights. Summary of the Invention

[0005] In view of the above shortcomings of the prior art, the present invention provides a method, apparatus, device and storage medium for data visualization to improve the problem that the existing data visualization method cannot quickly obtain data conclusions.

[0006] To achieve the above and other related purposes, the present invention provides a method for data visualization, comprising the steps of:

[0007] Get the dimension information of the data to be displayed;

[0008] Acquire at least one dimension combination based on the dimension information;

[0009] Automatically retrieve and match the corresponding insight analysis solutions based on each dimension combination, and obtain corresponding data visualization charts and data conclusions;

[0010] Among them, each dimension combination corresponds to at least two data visualization charts, and the data visualization charts correspond to the data conclusions.

[0011] In one embodiment of the present invention, obtaining dimension information of data to be displayed includes the steps of: obtaining the data to be displayed and performing data attribute and data value separation processing; obtaining the data type based on the data attribute; dividing the data dimension according to the data type to obtain dimension information.

[0012] In one embodiment of the present invention, obtaining at least one dimension combination based on the dimension information includes the steps of: automatically matching the dimension combination according to the dimension information of the data.

[0013] In one embodiment of the present invention, obtaining at least one dimension combination based on the dimension information includes the steps of: receiving dimension combination configuration information and generating a dimension combination, wherein the dimension combination configuration information is manually configured according to one's own needs.

[0014] In one embodiment of the present invention, the dimension information includes data dimension.

[0015] In one embodiment of the present invention, the dimensional information further includes a caliber dimension, and obtaining at least one dimensional combination based on the dimensional information includes the steps of obtaining a plurality of caliber dimensions in a customized manner to form a caliber dimension combination.

[0016] In one embodiment of the present invention, when the number of dimensions corresponding to the data to be displayed is greater than a set threshold, obtaining at least one dimension combination based on the dimensional information includes the steps of: normalizing the data values ​​corresponding to the corresponding dimensions so that the data values ​​corresponding to each corresponding dimension are in the range of (0, 1); and performing dimensionality reduction processing on the normalized data to obtain a dimension combination.

[0017] In one embodiment of the present invention, the method further includes the step of triggering an event on the target chart to change the filter data.

[0018] The present invention also provides a device for data visualization, comprising:

[0019] A dimension information acquisition unit, configured to acquire dimension information of data to be displayed;

[0020] a dimension combination acquiring unit, configured to acquire at least one dimension combination based on the dimension information;

[0021] The processing unit automatically retrieves and matches the corresponding insight analysis plan based on each dimension combination, and obtains the corresponding data visualization charts and data conclusions; wherein, each dimension combination corresponds to at least two data visualization charts, and the data visualization charts correspond to the data conclusions.

[0022] The present invention further provides an electronic device, comprising:

[0023] one or more processors;

[0024] The storage device is used to store one or more programs, and when the one or more programs are executed by the one or more processors, the electronic device implements the method for data visualization.

[0025] The present invention also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor of a computer, the computer is caused to execute the method for data visualization.

[0026] In the data visualization method of the present invention, each dimension combination can automatically match the insight analysis plan and obtain at least two visualized charts and data conclusions. The visualized charts correspond to the data conclusions, so the data conclusions can be quickly obtained, which makes it easier for observers to grasp the content expressed by the data visualization charts. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0028] Figure 1 is a flow chart of a method for data visualization according to the present invention;

[0029] Figure 2 A flow chart of obtaining dimension information of data to be displayed according to the present invention;

[0030] Figure 3 A schematic diagram of a method for data visualization according to the present invention;

[0031] Figure 4 A schematic diagram of the invention's self-configured X-axis and Y-axis dimensions;

[0032] Figure 5 It is a diagrammatic representation of the caliber dimensions of the present invention;

[0033] Figure 6 It is a visualization diagram of the two-dimensional coordinate points of the present invention;

[0034] Figure 7 This is a schematic diagram of the process of event triggering of the present invention;

[0035] Figure 8A visual schematic diagram of event triggering of the present invention;

[0036] Figure 9 This is a structural block diagram of the device for data visualization of the present invention.

[0037] Component number description

[0038] 100. Dimension information acquisition unit; 200. Dimension combination acquisition unit; 300. Processing unit. DETAILED DESCRIPTION

[0039] The following describes the embodiments of the present invention through specific examples. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following examples and the features in the examples can be combined with each other unless they conflict. It should also be understood that the terms used in the examples of the present invention are for the purpose of describing specific embodiments, not for the purpose of limiting the scope of protection of the present invention. The test methods for which specific conditions are not specified in the following examples are generally carried out under conventional conditions or under the conditions recommended by the manufacturers.

[0040] It should be noted that the terms such as "upper", "lower", "left", "right", "middle" and "one" cited in this specification are only for the convenience of description and are not used to limit the scope of implementation of the present invention. Changes or adjustments to their relative relationships should be regarded as the scope of implementation of the present invention without substantially changing the technical content.

[0041] See also Figures 1 to 3 The present invention provides a method for data visualization, comprising the steps of:

[0042] See also Figure 2 , obtaining dimensional information of the data to be displayed; the dimensional information includes data dimensions. The specific implementation of this step includes the following steps: obtaining the data to be displayed by reading data from common data formats such as txt, excel, json, etc., or from an existing database. Separating the data attributes from the data values, and obtaining the data type based on the data attributes. Specifically, the process may include separating the data attributes from the data values ​​to be displayed, analyzing the data value attributes through code, and obtaining the basic data type. Specific data dimension division is performed based on different data types to obtain dimensional information.

[0043] See also Figure 3, obtaining at least one dimension combination based on the dimension information. Obtaining at least one dimension combination based on the dimension information includes at least one of the following implementations, wherein the first implementation includes the step of automatically matching the dimension combination according to the dimension information of the data.

[0044] Based on the divided data dimensions, we filter the dimension combinations and match the analysis intent, and use trend analysis, correlation, distribution, clustering, and outlier detection to automatically gain insights into the data. The specific process can be:

[0045] 1) Based on type dimension matching, analysis can be performed from a single dimension or from multiple dimensions. To avoid excessive dimension combinations, the system limits the number and quality of dimension combinations.

[0046] The combinations of dimensions generally include "time series x numerical type", "numerical type x numerical type", "category x numerical type", "geographic type x geographic type", "category x category x category type", etc.

[0047] 2) Automatically retrieve and match insight analysis solutions based on matched dimension combinations

[0048] Taking "time series x numerical type" as an example, visualization can generally be performed using area charts, line charts, bar charts, LOLLIPOP charts, etc. For time series, the system will use the MK test in machine learning to analyze the continuous increase or decrease trend of time series data, obtain trend variables, significance test p-values, variances, and slopes, etc., calculate and sort the weights of result sets with obvious characteristics, and match and select appropriate charts and display the core data conclusions.

[0049] See also Figure 4 The second implementation method of obtaining at least one dimension combination based on the dimension information includes the following steps: receiving dimension combination configuration information and generating a dimension combination, wherein the dimension combination configuration information is manually configured according to one's own needs.

[0050] Therefore, in addition to using the visualization charts recommended by the system, the method described in the present invention can also interactively operate other dimensions and indicators. By dragging any attribute to the x or y axis, the system will recommend an appropriate visualization chart based on the manually specified dimension. For indicator data, other aggregation methods can be adopted, namely sum, mean, maximum, minimum, median, variance, and standard deviation, to further expand the dimensions of data analysis.

[0051] See also Figure 5, the dimensional information also includes a caliber dimension, and the third implementation method of obtaining at least one dimensional combination based on the dimensional information includes the steps of: obtaining several caliber dimensions in a customized manner to form a caliber dimension combination. Therefore, in addition to the numerical dimension, the present invention also provides a caliber dimension, which can take several attributes to define a joint attribute, such as the interest caliber is a combination of sharing clicks and page browsing clicks. By defining different caliber dimensions, the user's behavior path can be defined: cognition-interest-purchase-loyalty, that is, the entire user behavior closed loop is defined. For user behavior, the system provides a conversion funnel analysis, which can analyze the layered conversion rates of different activities, such as cognition-interest, interest-loyalty, etc.

[0052] See also Figure 6 In one embodiment of the present invention, when the number of dimensions corresponding to the data to be displayed is greater than a set threshold, the fourth implementation method of obtaining at least one dimension combination based on the dimension information includes the steps of: normalizing the data values ​​corresponding to the corresponding dimensions so that the data values ​​corresponding to each corresponding dimension are in the range of (0, 1); performing dimensionality reduction processing on the normalized data to obtain a dimension combination.

[0053] Therefore, the present invention can also implement debuggable insight analysis solutions: For most scenarios, common insight analysis solutions are sufficient. However, for complex multi-dimensional data, ordinary insight analysis solutions are no longer sufficient. The system also provides debuggable insight analysis solutions. Taking "numeric xN" as an example, parallel coordinate charts, radar charts, etc. can generally be used for visualization. However, when there are too many dimensions, these charts are obviously not suitable. The system provides the following options:

[0054] 1) Normalize the selected dimensional data and scale the data of different dimensions to the range of (0, 1) for easy comparison and processing; use t-SNE to reduce the dimension of the normalized N-dimensional data to obtain the two-dimensional coordinate points of each data set after dimensionality reduction;

[0055] 2) Use the LOF outlier detection algorithm to analyze the data after dimensionality reduction and obtain some abnormal outliers.

[0056] 3) It also provides advanced parameter settings, such as learning rate, maximum number of iterations, distance metric and other properties, to facilitate data analysts to preview data visualization results after adjusting the algorithm parameters.

[0057] like Figure 6 As shown in the figure, after dimensionality reduction of data of several dimensions in a data set, the effect is displayed on a two-dimensional plane, from which some characteristics of data aggregation can be found.

[0058] Based on each dimension combination, the corresponding insight analysis plan is automatically retrieved and matched, and the corresponding data visualization charts and data conclusions are obtained; wherein, each dimension combination corresponds to at least two data visualization charts, the data visualization charts correspond to the data conclusions, and the correspondence between the data visualization charts and the data conclusions can be one-to-one.

[0059] See also Figure 7 and Figure 8 , triggering an event on the target chart to change the filtered data. After selecting several visualization analysis views for a dataset, you can leverage the Observer pattern in the design pattern to trigger an event on a particular view to change the filtered data. Upon receiving the notification, the parent component then distributes the event to notify other views to update. This approach supports linking multiple views for coordinated analysis, making it easier to explore data and obtain results, as shown in the figure.

[0060] In the data visualization method of the present invention, each dimension combination can automatically match the insight analysis plan and obtain at least two visualized charts and data conclusions. The visualized charts correspond to the data conclusions, so the data conclusions can be quickly obtained, which makes it easier for observers to grasp the content expressed by the data visualization charts.

[0061] See also Figure 9 The present invention also provides a device for data visualization, including: a dimension information acquisition unit 100, a dimension combination acquisition unit 200 and a processing unit 300; wherein, the dimension information acquisition unit 100 is used to obtain dimension information of data to be displayed; the dimension combination acquisition unit 200 obtains at least one dimension combination based on the dimension information; the processing unit 300 automatically retrieves and matches the corresponding insight analysis plan based on each dimension combination, and obtains corresponding data visualization charts and data conclusions; wherein, each dimension combination corresponds to at least two data visualization charts, and the data visualization charts correspond to the data conclusions.

[0062] The present invention further provides an electronic device, comprising:

[0063] one or more processors;

[0064] The storage device is used to store one or more programs, and when the one or more programs are executed by the one or more processors, the electronic device implements the method for data visualization.

[0065] The present invention also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor of a computer, the computer is caused to execute the method for data visualization.

[0066] Beneficial effects of the present invention:

[0067] 1. Before analyzing the data, patterns or anomalies in the data are detected in advance. This is achieved by combining traditional statistical analysis with algorithms such as LOF anomaly analysis in machine learning. Existing technologies generally only perform simple statistical analysis.

[0068] 2. During data exploration, the most credible visualization charts can be matched and recommended, reducing various obstacles that arise during data exploration. This is achieved primarily through the use of custom rule-based weighted matching and machine learning algorithms. Existing technologies have similar approaches, also employing machine learning algorithms to aid in screening, but use conventional built-in visualization charts for matching, which lacks flexibility.

[0069] 3. This invention provides a customizable and debuggable insight analysis solution. For complex, multi-dimensional data, this debuggable insight analysis solution meets the needs of data analysts for specific tasks. Existing technical solutions generally only provide style editing on existing charts and lack solutions or tools for in-depth analysis and mining.

[0070] In summary, the present invention effectively overcomes some practical problems in the prior art and thus has high utilization value and use significance.

[0071] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the present invention. Anyone skilled in the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by one of ordinary skill in the art without departing from the spirit and technical principles disclosed herein are intended to be covered by the claims of the present invention.

Claims

1. A method for data visualization, characterized in that Including steps: Get the dimension information of the data to be displayed; Acquire at least one dimension combination based on the dimension information; Automatically retrieve and match the corresponding insight analysis solutions based on each dimension combination, and obtain corresponding data visualization charts and data conclusions; Among them, each dimension combination corresponds to at least two data visualization charts, and the data visualization charts correspond to the data conclusions; Acquiring at least one dimension combination based on the dimension information includes the steps of: automatically matching dimension combinations according to the dimension information of the data; automatically matching dimension combinations according to the dimension information of the data includes: screening dimension combinations and matching analysis intentions according to the divided data dimensions, and using trend analysis, correlation, distribution, clustering and outlier detection to automatically gain insights into the data.

2. The method for data visualization according to claim 1, characterized in that Obtaining dimension information of data to be displayed includes the following steps: obtaining the data to be displayed and performing data attribute and data value separation processing; obtaining the data type based on the data attribute; and dividing the data dimension according to the data type to obtain dimension information.

3. The method for data visualization according to claim 2, characterized in that Acquiring at least one dimension combination based on the dimension information includes the steps of: receiving dimension combination configuration information and generating a dimension combination, wherein the dimension combination configuration information is manually configured according to one's own needs.

4. The method for data visualization according to claim 1, characterized in that The dimension information includes data dimension.

5. The method for data visualization according to claim 4, characterized in that The dimension information also includes a caliber dimension. Acquiring at least one dimension combination based on the dimensional information includes the steps of: acquiring a plurality of caliber dimensions in a customized manner to form a caliber dimension combination.

6. The method for data visualization according to claim 1, characterized in that The method further includes the step of triggering an event on a target chart to change the filtered data.

7. A device for data visualization, characterized in that include: A dimension information acquisition unit, configured to acquire dimension information of data to be displayed; a dimension combination acquiring unit, configured to acquire at least one dimension combination based on the dimension information; A processing unit automatically searches and matches a corresponding insight analysis solution based on each dimension combination, and obtains corresponding data visualization charts and data conclusions; wherein each dimension combination corresponds to at least two data visualization charts, and the data visualization charts correspond to the data conclusions; Among them, obtaining at least one dimension combination based on the dimension information includes the steps of: automatically matching the dimension combination according to the dimension information of the data; the automatic matching of the dimension combination according to the dimension information of the data includes: screening the dimension combination and matching the analysis intention according to the divided data dimensions, and using trend analysis, correlation, distribution, clustering and outlier detection to automatically gain insights into the data.

8. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, enables the electronic device to implement the method for data visualization according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that A computer program is stored thereon, and when the computer program is executed by a processor of a computer, the computer is caused to execute the method for data visualization according to any one of claims 1 to 6.

Images