A data comprehensive situation display method based on a recommendation model

Abstract

The application discloses a data comprehensive situation display method based on a recommendation model, constructs a data analysis recommendation model for data classification and fusion according to a special topic or a label dimension for global data in a background database, customizes a statistical analysis front visual chart plug-in, and binds and displays a front visual chart plug-in display framework and special topic data associated by the background data analysis recommendation model, customizes an embedded method for changing and displaying the display framework structure and layout of the front visual chart plug-in with time stages, integrates and analyzes the background data associated with the recommendation model to bind the front visual plug-in display framework for display and a front visual plug-in display framework dynamic change method, and realizes the process of dynamically pushing the background data to the front and visually customizing and displaying in the whole time flow and cycle.

Description

Technical Field

[0001] This invention belongs to the field of data front-end visualization technology, and in particular, it is a method for displaying the comprehensive data situation over time or with focus based on a tag-based thematic data recommendation model customized for specific fields and applications. Background Technology

[0002] Data visualization technology is a scientific and technological study of the visual representation of data. It is defined as information extracted in a certain summary form, including various attributes and variables of the corresponding information units.

[0003] Data visualization primarily aims to clearly and effectively convey and communicate information through graphical means. However, current data visualization technologies often fail to achieve a good balance between design and functionality, resulting in flashy but impractical data visualizations. On the one hand, they fail to achieve the main purpose of data visualization, failing to comprehensively present the data as a whole and effectively convey and communicate information; on the other hand, they fail to meet the customized needs of specific users for specific applications. This is the contradiction and shortcoming of existing visualization technologies. Summary of the Invention

[0004] The purpose of this invention is to provide a method for displaying the comprehensive data situation based on a tag-based thematic data recommendation model that is specifically tailored to changes in time or focus for specific applications in specific fields.

[0005] The technical solution for implementing this invention is: a method for comprehensive data situation presentation based on a recommendation model, comprising the following steps:

[0006] Step 1: Construct a data analysis and recommendation model for global data in the backend database, which classifies and integrates data by topic or tag dimension;

[0007] Step 2: Customize the statistical analysis front-end visualization chart plugin, and link and bind the front-end visualization chart plugin display framework with the topic data associated with the back-end data analysis recommendation model for display;

[0008] Step 3: Based on the progression of time or the change in visual focus, customize the display framework structure and layout of the front-end visualization chart plugin so that it can change with the time stage.

[0009] Step 4: Integrate and analyze the backend data associated with the recommendation model, bind it to the frontend visualization plugin display framework, and implement the dynamic change method of the frontend visualization plugin display framework to realize the process of dynamically pushing backend data to the frontend and customizing visualization display throughout the entire time process and cycle.

[0010] The specific steps for constructing the analysis and recommendation model that classifies and integrates data by topic or tag dimension in step 1 are as follows:

[0011] (1.1) For all data tables and data existing in the background database, traverse all data table names and use algorithms such as longest substring matching, string similarity calculation, and text classification to cluster tables with similar associations such as parent-child structure table names, similar or identical classification relationships;

[0012] (1.2) For classifications completed based on table names, select the classification result with the largest weight as the correct classification cluster conclusion; after all correct classifications are completed, repeat the operation of customizing clear labels or topics for the classification until each completely correct classification has clear and non-overlapping labels or topics.

[0013] (1.3) In the tag-theme dataset that has been correctly classified, reorganize the data tables in the classification dataset so that one or more data tables are arranged in a clear and ordered parallel or parent-child structure relationship;

[0014] (1.4) For any keyword input from the outside, perform character matching, similarity calculation or other equivalent classification algorithms on the keyword and the label thematic set of all classification sets. Take the data set of the classification with the highest calculated value that is higher than a certain threshold (0.7 is set here, that is, the value is considered to be equivalent match if it is higher than 0.7) and output the data.

[0015] Step (1.2) For the classification completed based on the table name, further checks and verifications are performed using rules based on recognized domain expertise. For tables that have been classified and clustered under the automated algorithm, if the classification meets the verification rules, then the classification is considered to have a correct classification result. If one or more tables under the classification belong to this category and only partially meet the verification rules, then they are reclassified and clustered under a directly related category (this category can be an existing category or a newly added category). If the classification does not meet the verification rules at all, then the classification is removed, and all tables under the original classification are reclassified under a directly related category (this category can be an existing category or a newly added category).

[0016] Based on the comprehensive verification of the classification results, the classification result with the highest weight is selected as the correct classification cluster conclusion.

[0017] The steps in step 2 to link and bind the front-end visualization chart plugin display framework with the back-end thematic data are as follows:

[0018] (2.1) For a dataset under a specific tag topic, the statistical presentation format of the data tables in the dataset is divided as follows: all entities in a data table are presented in an enumerated manner; for entities in a data table, the entity data is classified and summarized based on the differences in the key field values ​​(which have only a few fixed state values) in the data table; for entities in multiple data tables, the total values ​​of each data table are statistically compared; for entities in a data table, when new entities are added or existing entities change their key field values ​​(which have only a few fixed state values) over time, the values ​​of each key node that changes are statistically compared; other possible cases are similar.

[0019] (2.2) According to the display formats listed in step (2.1), customize the corresponding front-end visualization chart plugins. For single-table entity listing, display in table format; for single-table category summary statistics, display in pie chart format for comparing proportions and display in bar chart format for displaying numerical statistics; for multi-table total value statistics, display in pie chart format for comparing proportions and display in bar chart format for displaying numerical statistics; for single-table entity key field value change statistics, display in bar chart or line chart format; and so on for other possible cases, and formulate the corresponding chart visualization display formats.

[0020] (2.3) Adapt a dataset of a tag topic to a pre-defined front-end chart plugin visualization framework, while paying attention to whether each data table has the same common upper-level structure. In actual chart customization, adjust the display dimension values ​​and the multi-level changes of the chart plugin, and customize the data drill-up and drill-down capabilities.

[0021] (2.4) To meet special needs, the method of customizing the visualization chart plugin display framework for specific data is retained, which is used to replace the fixed data-adaptive visualization display content.

[0022] The steps in step 3 for the front-end visualization chart plugin to display changes in the framework's structure, layout, and content over time are as follows:

[0023] (3.1) For different applications, clarify the full set of content of the time stage or focus of the specific application. In the case of a specific application, there may be different front-end display requirements for different time stages or focus changes. Correspondingly, it is necessary to determine the statistical content that needs to be displayed on the front end under different time stages or under different focus conditions.

[0024] (3.2) For specific applications, the visualization chart plugin display framework displayed on the front end is adjusted and customized for each different time stage or focus to meet the unique needs of the application.

[0025] (3.3) The different front-end display visualization frameworks under different time stages or focus situations are dynamically changed by taking the time stage or focus as the dimension of change.

[0026] Step 4 involves integrating backend data, binding it to the frontend visualization plugin framework, and dynamically updating the focus of the frontend visualization plugin framework over time. The steps are as follows:

[0027] (4.1) For a specific application, once the time phase or focus set of the application is clear, develop a front-end visualization customization of the application-related visual plugin that can display the structural layout content of the framework changes with the time phase or focus.

[0028] (4.2) For a specific application, select and match the tag topic data set based on the relevant keywords entered according to the content requirements of the application;

[0029] (4.3) For specific applications, the selected tag topic data is associated with the front-end visualization chart plugin display framework for adaptive display. The front-end visualization chart plugin display framework is adjusted for each time stage or different focus situation of the application.

[0030] (4.4) Integrate the above steps to complete the implementation of a data-based recommendation model-based comprehensive data situation display method for specific applications.

[0031] The significant advantages of this invention compared to existing technologies are:

[0032] The present invention will now be described in further detail with reference to the accompanying drawings.

[0033] Beneficial effects: Compared with existing technologies, the rail transit data analysis method based on big data provided by this invention has the following advantages:

[0034] 1. This method can construct a data analysis and recommendation model based on tags and topics to meet the different needs of different users in specific fields and applications, and realize the dynamic visualization of the comprehensive data situation under complex application requirements.

[0035] 2. Construct a data tag-based thematic analysis and recommendation model to make the visualized data collection more clearly categorized and focused, and to convey and communicate information more clearly and effectively;

[0036] 3. Able to identify the complexity of a specific application in a specific field, analyze the changes in the application across time stages, focus, and other dimensions, and develop different framework structures for visualizing these changes.

[0037] 4. Integrate the analysis and recommendation of tag-based thematic data with complex processes of specific fields and applications to create a dynamic visualization of the overall data situation for complex and specific applications;

[0038] 5. The method of customizing visualization chart plugins for specific data display needs is retained, which can be used to replace fixed data-adaptive visualization content. Attached Figure Description

[0039] Figure 1 This is a flowchart of the data comprehensive situation display method based on the recommendation model of the present invention;

[0040] Figure 2 This is a flowchart illustrating the interaction between the front-end page and the back-end database. Detailed Implementation

[0041] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0042] The following is an example of an implementation of the present invention, and the purpose and features of the present invention can also be seen from the description of the example. It should be understood that the examples described herein are merely for illustration and explanation of the present invention and are not intended to limit the present invention.

[0043] A method for comprehensive data situation visualization based on a recommendation model includes the following steps:

[0044] Step 1: Construct a data analysis and recommendation model for global data in the backend database, which classifies and integrates data by topic or tag dimension;

[0045] (1.1) For all data tables and data existing in the background database, traverse all data table names and use algorithms such as longest substring matching, string similarity calculation, and text classification to cluster tables with similar associations such as parent-child structure table names, similar or identical classification relationships;

[0046] (1.2) For classifications completed based on table names, select the classification result with the largest weight as the correct classification cluster conclusion; after all correct classifications are completed, repeat the operation of customizing clear labels or topics for the classification until each completely correct classification has clear and non-overlapping labels or topics.

[0047] For classifications based on table names, further checks and verifications are performed using rules based on recognized domain expertise. For tables that are classified and clustered using automated algorithms, if the classification meets the verification rules, it is considered to have a correct classification result. If one or more tables under a classification only partially meet the verification rules, they are reclassified and clustered into a directly related classification (which can be an existing classification or a newly added one). If the classification does not meet the verification rules at all, the classification is removed, and all tables under the original classification are reclassified into a directly related classification (which can be an existing classification or a newly added one).

[0048] Based on the comprehensive verification of the classification results, the classification result with the highest weight is selected as the correct classification cluster conclusion.

[0049] (1.3) In the tag-theme dataset that has been correctly classified, reorganize the data tables in the classification dataset so that one or more data tables are arranged in a clear and ordered parallel or parent-child structure relationship;

[0050] (1.4) For any keyword input from the outside, perform character matching, similarity calculation or other equivalent classification algorithms on the keyword and the label thematic set of all classification sets. Take the data set of the classification with the highest calculated value that is higher than a certain threshold (0.7 is set here, that is, the value is considered to be equivalent match if it is higher than 0.7) and output the data.

[0051] Step 2: Customize the statistical analysis front-end visualization chart plugin, and link and bind the front-end visualization chart plugin display framework with the topic data associated with the back-end data analysis recommendation model for display;

[0052] 2.1) For a dataset under a specific tag topic, the statistical presentation format of the data tables within that dataset is categorized as follows: All entities in a single data table are presented as an enumeration; for entities within a single data table, the data is categorized and summarized based on the differences in key field values ​​(which have only a few fixed state values); for entities across multiple data tables, the total values ​​from each table are compared; for entities in a single data table, if new entities are added or existing entities experience changes in key field values ​​(which have only a few fixed state values) over time, the values ​​at each key change node are statistically compared; other possible scenarios are similar.

[0053] (2.2) According to the display formats listed in step (2.1), customize the corresponding front-end visualization chart plugins. For single-table entity listing, display in table format; for single-table category summary statistics, display in pie chart format for comparing proportions and display in bar chart format for displaying numerical statistics; for multi-table total value statistics, display in pie chart format for comparing proportions and display in bar chart format for displaying numerical statistics; for single-table entity key field value change statistics, display in bar chart or line chart format; and so on for other possible cases, and formulate the corresponding chart visualization display formats.

[0054] (2.3) Adapt a dataset of a tag topic to a pre-defined front-end chart plugin visualization framework, while paying attention to whether each data table has the same common upper-level structure. In actual chart customization, adjust the display dimension values ​​and the multi-level changes of the chart plugin, and customize the data drill-up and drill-down capabilities.

[0055] (2.4) To meet special needs, the method of customizing the visualization chart plugin display framework for specific data is retained, which is used to replace the fixed data-adaptive visualization display content.

[0056] Step 3: Based on the progression of time or the change in visual focus, customize the display framework structure and layout of the front-end visualization chart plugin so that it can change with the time stage.

[0057] (3.1) For different applications, clarify the full set of content of the time stage or focus of the specific application. In the case of a specific application, there may be different front-end display requirements for different time stages or focus changes. Correspondingly, it is necessary to determine the statistical content that needs to be displayed on the front end under different time stages or under different focus conditions.

[0058] (3.2) For specific applications, the visualization chart plugin display framework displayed on the front end is adjusted and customized for each different time stage or focus to meet the unique needs of the application.

[0059] (3.3) The different front-end display visualization frameworks under different time stages or focus situations are dynamically changed by taking the time stage or focus as the dimension of change.

[0060] Step 4: Integrate and analyze the backend data associated with the recommendation model, bind it to the frontend visualization plugin display framework, and dynamically change the frontend visualization plugin display framework to realize the process of dynamically pushing backend data to the frontend and customizing visualization display throughout the entire time process and cycle;

[0061] (4.1) For a specific application, once the time phase or focus set of the application is clear, develop a front-end visualization customization of the application-related visual plugin that can display the structural layout content of the framework changes with the time phase or focus.

[0062] (4.2) For a specific application, select and match the tag topic data set based on the relevant keywords entered according to the content requirements of the application;

[0063] (4.3) For specific applications, the selected tag topic data is associated with the front-end visualization chart plugin display framework for adaptive display. The front-end visualization chart plugin display framework is adjusted for each time stage or different focus situation of the application.

[0064] (4.4) Integrate the above steps to complete the implementation of a data-based recommendation model-based comprehensive data situation display method for specific applications.

[0065] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings and examples. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of the present invention.

Claims

1. A method for comprehensively displaying data trends based on a recommendation model, characterized in that: Includes the following steps Step 1: Construct a data analysis and recommendation model for global data in the backend database, which classifies and integrates data by topic or tag dimension; Step 2: Customize the statistical analysis front-end visualization chart plugin, and link and bind the front-end visualization chart plugin display framework with the topic data associated with the back-end data analysis recommendation model for display; Step 3: Based on the progression of time or the change in visual focus, customize the display framework structure and layout of the front-end visualization chart plugin so that it can change with the time stage. Step 4: Integrate and analyze the backend data associated with the recommendation model, bind it to the frontend visualization plugin display framework, and dynamically change the frontend visualization plugin display framework to realize the process of dynamically pushing backend data to the frontend and customizing visualization display throughout the entire time process and cycle; The specific steps for constructing the analysis and recommendation model that classifies and integrates data by topic or tag dimension in step 1 are as follows: (1.1) For all data tables and data existing in the background database, traverse all data table names and cluster tables with similar relationships such as parent-child structure table names, similar or identical classification relationships by longest substring matching, string similarity calculation, and text classification. (1.2) For classifications based on table names, select the classification result with the largest weight as the correct classification clustering conclusion; After all the correct categories are completed, repeat the process of creating specific tags or topics for each category until every completely correct category has specific tags or topics that do not overlap. (1.3) In the tag-theme dataset that has been correctly classified, reorganize the data tables in the classification dataset so that one or more data tables are arranged in a clear and ordered parallel or parent-child structure; (1.4) For any keyword input from the outside, perform character matching, similarity calculation or other equivalent classification algorithms on the keyword and the label thematic set of all classification sets, and output the data set of the classification with the highest calculated value that is higher than a certain threshold; The steps in step 2 to link and bind the front-end visualization chart plugin display framework with the back-end thematic data are as follows: (2.1) For a dataset under a specific tag topic, the data in the data table of the dataset is divided into statistical display formats: all entities in a data table are presented in an enumeration manner; For an entity in a data table, classify, summarize, and statistically analyze the entity data based on the different values ​​of the key fields in the data table; For entities in multiple data tables, the total values ​​of each data table are statistically compared; for entities in a single data table, when new entities are added or existing entities change their key field values ​​over time, the values ​​of each key change node are statistically compared. (2.2) According to the display formats listed in step (2.1), customize the corresponding front-end visualization chart plugins respectively. For single-table entity listing, display in table format; for single-table category summary statistics, display in pie chart format for comparing proportions and display in bar chart format for displaying numerical statistics; for multi-table total value statistics, display in pie chart format for comparing proportions and display in bar chart format for displaying numerical statistics; for single-table entity key field value change statistics, display in bar chart or line chart format; and formulate the corresponding chart visualization display formats. (2.3) Adapt a dataset of a tag topic to a pre-defined front-end chart plugin visualization framework, while paying attention to whether each data table has the same common upper-level structure. In actual chart customization, adjust the display dimension values ​​and the multi-level changes of the chart plugin, and customize the data drill-up and drill-down capabilities. (2.4) To meet special needs, the method of customizing the visualization chart plugin display framework for specific data is retained, which is used to replace the fixed data-adaptive visualization display content; The steps for the front-end visualization chart plugin in step 3 to display the structural layout and content changes of the framework over time are as follows: (3.1) For different applications, clarify the full set of content of the time stage or focus of the specific application. In the case of a specific application, there may be different front-end display requirements for different time stages or focus changes. Correspondingly, it is necessary to determine the statistical content that needs to be displayed on the front end under different time stages or under different focus conditions. (3.2) For specific applications, the visualization chart plugin display framework displayed on the front end is adjusted and customized for each different time stage or focus situation to meet the unique needs of the application. (3.3) The different front-end display visualization frameworks under different time stages or focus situations are dynamically changed by taking the time stage or focus as the dimension of change; Step 4 involves integrating backend data, binding it to the frontend visualization plugin framework, and dynamically updating the focus of the frontend visualization plugin framework over time. The steps are as follows: (4.1) For a specific application, after clarifying the time stage or focus set of the application, formulate a front-end visualization customization of the application-related visualization plugin to show the changes in the framework structure layout content as the time stage or focus changes; (4.2) For a specific application, select and match the tag-themed data set based on the relevant keywords entered according to the content requirements of the application; (4.3) For specific applications, the selected tag topic data is associated with the front-end visualization chart plugin display framework for adaptation and display. The front-end visualization chart plugin display framework is adjusted for each time stage or different focus situation of the application. (4.4) Integrate the above steps to complete the implementation of a data comprehensive situation display method based on data analysis recommendation model for specific applications.

2. The data comprehensive situation display method based on a recommendation model according to claim 1, characterized in that: Step (1.2) For the classification completed based on the table name, further checks and verifications are performed using rules based on recognized domain expertise. For tables that have completed classification and clustering under automated algorithms, if the classification meets the verification rules, then the classification is considered to have a correct classification result. If one or more tables in a category only partially meet the validation rules, they are reclassified and clustered into a directly related category; if a category does not meet the validation rules at all, the category is removed, and all tables in the original category are reclassified into a directly related category. Based on the comprehensive verification of the classification results, the classification result with the highest weight is selected as the correct classification cluster conclusion.

Images