Construction method of visual model based on stream-oriented computation
By building a dynamic weight distribution model and an adaptive layout algorithm, the shortcomings of streaming computing technology in visual model construction are solved, flexible display and real-time interaction of streaming data are achieved, and the adaptability of the system and user experience are improved.
Patent Information
- Application Number
- CN202510799116.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-16
- Publication Date
- 2025-09-26
AI Technical Summary
Existing streaming computing technologies are unable to meet users' needs for efficient, intuitive, and flexible dynamic data processing and real-time interactive display when building visualization models, especially in complex scenarios where their application capabilities are limited.
By designing data stream characteristic parameters, building a dynamic weight distribution model and adaptive layout algorithm, we can achieve real-time processing of streaming data and multi-layer nested ring chart display, dynamically adjust the weight and position of data points, and meet user interaction needs.
It enables highly flexible display and real-time interaction of streaming data, improves the performance of streaming computing systems, and provides an intuitive and efficient decision support tool suitable for complex business scenarios such as financial transaction monitoring, IoT device management, and meteorological data analysis.
Smart Images

Figure CN120705211A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of computer technology, and in particular to a method for constructing a visualization model based on streaming computing. Background Art
[0002] With the rapid development of streaming computing technology, its application in real-time data processing, analysis, and visualization is becoming increasingly widespread. With its efficient data processing capabilities and support for dynamic data streams, streaming computing has become an important technical means to meet real-time requirements. However, when combining streaming computing with visualization model construction, existing technologies still have significant limitations. In particular, the in-depth integration of dynamic data processing and real-time interactive display makes it difficult to meet users' needs for efficient, intuitive, and flexible visualization.
[0003] A search revealed that patent publication number CN113239081B provides a streaming data calculation method based on a detailed data source in Excel format. This method structures the data source's title names and indicator fields, calculates the data using a user-selected formula, and ultimately displays the results in a visual service terminal. However, this technical solution primarily focuses on optimizing the data calculation process and does not delve into how to build an efficient visualization model. Furthermore, this method relies on a specific data source format (such as Excel), limiting its applicability in diverse data scenarios. Its relatively simple visualization display method makes it difficult to meet the dynamic interaction requirements of complex scenarios. Another patent, publication number CN111177199B, proposes a streaming computing indicator generation system based on structured streaming. This system implements streaming computing indicator generation through a modular design, including a data source management module, an indicator logic definition module, and an output channel management module. While this system can generate streaming computing indicators based on user instructions and output them to designated channels, its focus is on optimizing indicator generation and data processing processes, lacking specific support for building visualization models. At the same time, the visualization function of the system is only part of the output channel, and does not fully consider the actual needs of users in real-time interaction and dynamic adjustment, resulting in limited application capabilities in complex scenarios.
[0004] The above issues demonstrate that existing streaming computing technologies are still insufficient when combined with visualization model building. In particular, a mature technical solution has yet to be developed for the coordinated mechanism between dynamic data processing and real-time interactive display. Therefore, a method that optimizes the coordinated mechanism between data processing and visualization display is urgently needed to enhance the system's flexibility and interactivity, thereby better meeting user needs for real-time data analysis and decision support. Summary of the Invention
[0005] In view of the deficiencies of the existing technology, the present invention provides a method for constructing a visualization model based on streaming computing, which solves the deficiencies of the existing technology.
[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A method for constructing a visualization model based on streaming computing, comprising the following steps:
[0007] Step S1: Design data stream characteristic parameters, including data update frequency, data dimension distribution, and user interaction priority;
[0008] Step S2: Based on the parameters designed in step S1, a dynamic weight distribution model is constructed and a visualization layout framework is initialized;
[0009] Step S3: Using the model constructed in step S2, the input streaming data is processed in real time to extract key feature values and generate preliminary visualization results;
[0010] Step S4: Analyze the eigenvalue distribution in the preliminary visualization results and divide the eigenvalues into core data points, auxiliary data points, and noise data points;
[0011] Step S5: Based on the analysis of the eigenvalue distribution in step S4, the visualization relationship of the streaming data is simplified into a multi-layer nested ring graph model;
[0012] Step S6: Based on the multi-layer nested ring graph model, an adaptive layout algorithm is introduced to dynamically adjust the position and weight of the eigenvalues;
[0013] Step S7: Establish a functional relationship between the weight parameters in the adaptive layout algorithm and the data stream characteristic parameters.
[0014] Optionally, in step S1, the data update frequency represents the number of changes in the data stream per unit time, the data dimension distribution describes the number of different dimensions in the data set and their distribution characteristics, and the user interaction priority is determined by user behavior analysis or manual configuration.
[0015] Optionally, in step S2, the dynamic weight allocation model calculates the dynamic weight of each data point using the following formula: ,in, Indicates the The dynamic weight of each data point, Indicates the data update frequency, Indicates the data dimension distribution, Indicates the user interaction priority, 、 and They are the corresponding weight coefficients, which are used to adjust the influence of each parameter on the final weight. 、 and Normalized mapping functions representing data update frequency, data dimension distribution, and user interaction priority, respectively.
[0016] Optionally, in step S3, streaming data is continuously input in the form of a time series, and the streaming data is parsed by a real-time processing algorithm to extract key feature values, which include device status information or information on whether a measurement value exceeds a threshold.
[0017] Optionally, in step S4, a clustering algorithm or a statistical analysis method is used to group the eigenvalues, and the eigenvalues are divided into core data points, auxiliary data points and noise data points.
[0018] Optionally, in step S5, the multi-layer nested annular graph model includes three layers of annular areas, the core data points are located in the innermost annular area, the auxiliary data points are located in the middle annular area, and the noise data points are located in the outermost annular area.
[0019] Optionally, in step S6, the adaptive layout algorithm achieves dynamic adjustment of the eigenvalue position and weight in the following ways: adjusting the refresh frequency of data points according to the data update frequency, adjusting the layout form of data points according to the data dimension distribution, and adjusting the display order and size of data points according to the user interaction priority.
[0020] Optionally, in step S7, the weight coefficient is determined by experiment or simulation. 、 and The value range of the weight coefficient is set according to the needs of the specific application scenario.
[0021] Optionally, the visualization layout framework initialized in step S2 includes a two-dimensional scatter plot or a multi-dimensional radar chart, and the initial layout form is selected according to the data dimension distribution.
[0022] Optionally, in step S6, the adaptive layout algorithm reflects the weight distribution of the data points by dynamically adjusting the color and width of the data points, and the color depth and width are proportional to the weight value.
[0023] Beneficial effects
[0024] The present invention achieves highly flexible display and real-time interaction of streaming data by introducing a dynamic weight distribution mechanism and an adaptive layout algorithm. In practical applications, this method can be widely applied to various complex business scenarios, such as financial transaction monitoring, IoT device management, meteorological data analysis, etc. For example, in a financial transaction monitoring scenario, the method of the present invention can process high-frequency transaction data in real time, extract core data points with abnormal price fluctuations, and display them in the form of a multi-layer nested ring graph to help users quickly identify market risks; in an IoT device management scenario, the method of the present invention can analyze the update frequency and dimensional distribution of sensor data to dynamically adjust the visual display form of the device status and improve the efficiency of equipment operation and maintenance; in a meteorological data analysis scenario, the method of the present invention can help users more intuitively understand weather trends by classifying and layering multi-dimensional meteorological data. Verified by the above application scenarios, the method of the present invention can not only significantly improve the performance of the streaming computing system, but also provide users with a more intuitive and efficient decision support tool. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] Figure 1 Schematic diagram of the process of the present invention;
[0026] Figure 2 It is a structural diagram of the multi-layer nested ring graph model. DETAILED DESCRIPTION
[0027] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on 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.
[0028] See also Figure 1 and Figure 2 The present invention provides a method for constructing a visualization model based on streaming computing, comprising the following steps:
[0029] Step S1: Design data stream characteristic parameters, including data update frequency, data dimension distribution, and user interaction priority;
[0030] Step S2: Based on the parameters designed in step S1, a dynamic weight distribution model is constructed and a visualization layout framework is initialized;
[0031] Step S3: Using the model constructed in step S2, the input streaming data is processed in real time to extract key feature values and generate preliminary visualization results;
[0032] Step S4: Analyze the eigenvalue distribution in the preliminary visualization results and divide the eigenvalues into core data points, auxiliary data points, and noise data points;
[0033] Step S5: Based on the analysis of the eigenvalue distribution in step S4, the visualization relationship of the streaming data is simplified into a multi-layer nested ring graph model;
[0034] Step S6: Based on the multi-layer nested ring graph model, an adaptive layout algorithm is introduced to dynamically adjust the position and weight of the eigenvalues;
[0035] Step S7: Establish a functional relationship between the weight parameters in the adaptive layout algorithm and the data stream characteristic parameters.
[0036] First, in step S1, data stream characteristic parameters need to be designed. These parameters include data update frequency, data dimension distribution, and user interaction priority. Data update frequency represents the number of changes in the data stream per unit time and is often used to measure the real-time requirements of the data. For example, in financial trading scenarios, high-frequency trading data may be updated thousands of times per second, while meteorological monitoring data may only be updated once per minute. Data dimension distribution describes the number of different dimensions in a dataset and their distribution characteristics, which is crucial for determining the complexity of the visualization model. For example, data from a sensor network may include multiple dimensions such as temperature, humidity, and air pressure, while product sales data from an e-commerce platform may only include two primary dimensions: sales volume and price. User interaction priority reflects the user's level of interest in specific data. This parameter can be determined through user behavior analysis or manual configuration. For example, in a monitoring system, users may pay more attention to abnormal data points than normal data points, so abnormal data points can be given a higher priority.
[0037] Next, in step S2, based on the data stream feature parameters designed in step S1, a dynamic weight allocation model is constructed and the visualization layout framework is initialized. The core of the dynamic weight allocation model is to generate a weight value for each data point by quantifying the data update frequency, data dimension distribution, and user interaction priority. Specifically, the weight value is calculated as follows: .in, Indicates the The dynamic weight of each data point, Indicates the data update frequency, Indicates the data dimension distribution, Indicates the user interaction priority, 、 and are the corresponding weight coefficients, which are used to adjust the influence of each parameter on the final weight. Function Normalize the data update frequency to the interval [0,1]. For example, for a data stream that updates 100 times per second, its normalized value may be 0.8. The mapping is performed based on the complexity of the data dimension distribution. For example, a data stream with 5 dimensions may be mapped to 0.6; the function Normalization is performed based on user interaction priority. For example, a high-priority data point might have a normalized value of 0.9. Using the above formula, the weight distribution of data points in the visualization model can be dynamically adjusted to achieve an adaptive display effect. Based on this, when initializing the visualization layout framework, an appropriate initial layout form can be selected based on the data dimensional distribution, such as a two-dimensional scatter plot or a multi-dimensional radar chart, and each data point can be assigned an initial position and size.
[0038] In step S3, the dynamic weight distribution model constructed in step S2 is used to process the input streaming data in real time, extract key feature values and generate preliminary visualization results. In practical applications, streaming data is usually continuously input in the form of time series, so it is necessary to use an efficient real-time processing algorithm to parse it. For example, in the IoT device monitoring scenario, the sensor data uploaded by the device may contain information such as timestamps, device IDs, and measurements. By parsing this information, key feature values can be extracted, such as whether the device status is abnormal, whether the measurement value exceeds the threshold, etc. The extracted key feature values are then input into the dynamic weight distribution model, and the dynamic weight of each data point is calculated according to the weight formula, and mapped to the visualization model to generate preliminary visualization results. The preliminary visualization results are usually presented in the form of simple charts, such as bar charts or line charts, to quickly show the overall trend of the data.
[0039] In step S4, the eigenvalue distribution in the preliminary visualization results needs to be analyzed, categorizing the eigenvalues into core data points, auxiliary data points, and noise data points. Core data points are those that have a significant impact on the current business scenario. For example, in stock trading monitoring, data points with price fluctuations exceeding a certain range can be considered core data points. Auxiliary data points are those that complement core data points, such as data points with small changes in trading volume but that can reflect market sentiment. Noise data points are those that are not significantly meaningful for business analysis, such as outliers caused by equipment failure. To achieve this classification, clustering algorithms or statistical analysis methods can be used to group eigenvalues. For example, the K-means clustering algorithm can be used to classify data points into three categories: core data points, auxiliary data points, and noise data points. By analyzing the eigenvalue distribution, key information in the data can be more clearly identified, laying the foundation for subsequent visualization optimization.
[0040] In step S5, based on the analysis results of the eigenvalue distribution in step S4, the visualization relationship of the streaming data is simplified into a multi-layer nested ring graph model. The multi-layer nested ring graph model is a visualization form suitable for displaying multi-level data relationships. Its core idea is to display different types of data points in layers through a ring structure. Figure 2 As shown in the figure, core data points are located in the innermost annular area, auxiliary data points are located in the middle annular area, and noise data points are located in the outermost annular area. The color and width of each annular area can be dynamically adjusted based on the weight of the data point. For example, data points with higher weights can be represented by darker colors and larger widths, while data points with lower weights can be represented by lighter colors and smaller widths. This layered display not only highlights the importance of core data points, but also effectively reduces the interference of noise data points on the overall visualization.
[0041] In step S6, an adaptive layout algorithm is introduced based on the multi-layer nested ring graph model to dynamically adjust the position and weight of the eigenvalues. The core goal of the adaptive layout algorithm is to adjust the position and weight of the data points in real time according to the changes in the characteristic parameters of the data stream, so as to ensure that the visualization model can accurately reflect the real-time changing characteristics of the data. Specifically, the adaptive layout algorithm is implemented through the following steps: first, the refresh frequency of the data points is adjusted according to the frequency of data update. For example, for data points with high frequency updates, the refresh interval can be shortened to ensure real-time performance; second, the layout form of the data points is adjusted according to the data dimension distribution. For example, a two-dimensional scatter plot can be used for low-dimensional data, while a three-dimensional stereogram or a multi-dimensional projection map can be used for high-dimensional data; finally, the display order and size of the data points are adjusted according to the user interaction priority. For example, high-priority data points can be placed in a more conspicuous position and displayed in an enlarged manner. Through the above adjustments, the visualization model can be more in line with the actual needs of the user.
[0042] In step S7, a functional relationship between the weight parameters in the adaptive layout algorithm and the data stream characteristic parameters is established to further optimize the display effect of the visualization model. Specifically, the calculation formula of the weight parameters has been given in step S2, that is, On this basis, the weight coefficient can be determined by experiment or simulation 、 and For example, in a financial transaction monitoring scenario, you can set 、 、 , to emphasize the impact of data update frequency on weight; in the meteorological monitoring scenario, you can set 3. 5. , to highlight the importance of data dimension distribution. In addition, the value of the weight coefficient can be continuously adjusted based on user feedback to further improve the adaptability and flexibility of the visualization model.
[0043] In summary, the present invention achieves highly flexible display and real-time interaction of streaming data by introducing a dynamic weight distribution mechanism and an adaptive layout algorithm. In practical applications, this method can be widely applied to various complex business scenarios, such as financial transaction monitoring, IoT device management, meteorological data analysis, etc. For example, in a financial transaction monitoring scenario, the method of the present invention can process high-frequency transaction data in real time, extract core data points with abnormal price fluctuations, and display them in the form of a multi-layer nested ring graph to help users quickly identify market risks; in an IoT device management scenario, the method of the present invention can analyze the update frequency and dimensional distribution of sensor data to dynamically adjust the visual display form of the device status and improve the efficiency of equipment operation and maintenance; in a meteorological data analysis scenario, the method of the present invention can help users understand weather change trends more intuitively by classifying and layering multi-dimensional meteorological data. Verified by the above application scenarios, the method of the present invention can not only significantly improve the performance of the streaming computing system, but also provide users with a more intuitive and efficient decision support tool.
[0044] Furthermore, the present invention uses a multidimensional feature mapping algorithm and Gaussian random distribution to model the dynamic characteristics of data streams, and combines this with real-time data streams to generate dynamic visualization components. In practical applications of traffic flow monitoring, this method can significantly improve the efficiency of data processing and visualization, and provide users with a more intuitive and flexible interactive experience.
[0045] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.
[0046] 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 method for constructing a visualization model based on streaming computing, characterized in that: The following steps are involved: Step S1: Design data stream characteristic parameters, including data update frequency, data dimension distribution, and user interaction priority; Step S2: Based on the parameters designed in step S1, a dynamic weight distribution model is constructed and a visualization layout framework is initialized; Step S3: Using the model constructed in step S2, the input streaming data is processed in real time to extract key feature values and generate preliminary visualization results; Step S4: Analyze the eigenvalue distribution in the preliminary visualization results and divide the eigenvalues into core data points, auxiliary data points, and noise data points; Step S5: Based on the analysis of the eigenvalue distribution in step S4, the visualization relationship of the streaming data is simplified into a multi-layer nested ring graph model; Step S6: Based on the multi-layer nested ring graph model, an adaptive layout algorithm is introduced to dynamically adjust the position and weight of the eigenvalues; Step S7: Establish a functional relationship between the weight parameters in the adaptive layout algorithm and the data stream characteristic parameters.
2. The method for constructing a visualization model based on streaming computing according to claim 1, characterized in that: In step S1, the data update frequency represents the number of changes in the data stream per unit time, the data dimension distribution describes the number of different dimensions in the data set and their distribution characteristics, and the user interaction priority is determined by user behavior analysis or manual configuration.
3. The method for constructing a visualization model based on streaming computing according to claim 1, characterized in that: In step S2, the dynamic weight allocation model calculates the dynamic weight of each data point using the following formula: ,in, Indicates the The dynamic weight of each data point, Indicates the data update frequency, Indicates the data dimension distribution, Indicates the user interaction priority, 、 and They are the corresponding weight coefficients, which are used to adjust the influence of each parameter on the final weight. 、 and Normalized mapping functions representing data update frequency, data dimension distribution, and user interaction priority, respectively.
4. The method for constructing a visualization model based on streaming computing according to claim 1, characterized in that: In step S3, streaming data is continuously input in the form of a time series, and the streaming data is parsed by a real-time processing algorithm to extract key feature values, which include device status information or information on whether the measurement value exceeds a threshold.
5. The method for constructing a visualization model based on streaming computing according to claim 1, characterized in that: In step S4, a clustering algorithm or a statistical analysis method is used to group the eigenvalues, and the eigenvalues are divided into core data points, auxiliary data points and noise data points.
6. The method for constructing a visualization model based on streaming computing according to claim 1, characterized in that: In step S5, the multi-layer nested annular graph model includes three layers of annular regions, the core data points are located in the innermost annular region, the auxiliary data points are located in the middle annular region, and the noise data points are located in the outermost annular region.
7. The method for constructing a visualization model based on streaming computing according to claim 1, characterized in that: In step S6, the adaptive layout algorithm achieves dynamic adjustment of the eigenvalue position and weight in the following ways: adjusting the refresh frequency of data points according to the data update frequency, adjusting the layout form of data points according to the data dimension distribution, and adjusting the display order and size of data points according to the user interaction priority.
8. The method for constructing a visualization model based on streaming computing according to claim 1, characterized in that: In step S7, the weight coefficient is determined by experiment or simulation. 、 and The value range of the weight coefficient is set according to the needs of the specific application scenario.
9. The method for constructing a visualization model based on streaming computing according to claim 1, characterized in that: The visualization layout framework initialized in step S2 includes a two-dimensional scatter plot or a multi-dimensional radar chart, and the initial layout form is selected according to the data dimension distribution.
10. The method for constructing a visualization model based on streaming computing according to claim 1, characterized in that: In step S6, the adaptive layout algorithm reflects the weight distribution of the data points by dynamically adjusting the color and width of the data points. The color depth and width are proportional to the weight value.
Citation Information
Patent Citations
Streaming Computing Index Generation System Based on Structured Flow
CN111177199B
A streaming data computing method
CN113239081B
