Training big data spatial-temporal distribution visual analysis system and method

Abstract

The invention relates to the technical field of artificial intelligence, in particular to a training big data spatio-temporal distribution visual analysis system and method, and the system comprises a training data preprocessing module, a training data range setting module, a spatio-temporal distribution visual display module and a training data traceability analysis module. Through key information extraction identification, data range setting according to needs, visual display combining time series and geographical distribution and double-layer traceability, data understanding is improved, a spatial-temporal distribution visual analysis means is provided for visually and efficiently analyzing and training big data, and the analysis efficiency of training big data can be effectively improved.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a system and method for visualizing and analyzing the spatiotemporal distribution of training big data. Background Technology

[0002] In high-intensity, standardized training scenarios such as emergency response drills (e.g., fire drills, disaster prevention drills), professional and technical training, and collaborative work training, training data generated by training institutions, participating units, and training system platforms, which reflect the entire training lifecycle, has the core characteristics of large volume, complex dimensions, and strong correlation.

[0003] With breakthroughs in IoT sensing and distributed storage technologies, massive amounts of training data can be collected and persistently stored at low cost, providing rich data sources for optimizing training methods. However, this training data is scattered. When using this data, trainers are like being in an ocean of information, requiring a great deal of time and effort to sort it out, making it difficult to quickly conduct analysis. Data visualization technology can transform complex big data into an intuitive and clear presentation, allowing users to quickly grasp key information and data trends. Currently, training data visualization is mostly presented in the form of numerical charts, providing some assistance for trainers in analyzing training data, but it is not intuitive and clear enough.

[0004] The aforementioned shortcomings cause trainers to become overwhelmed by an "information sea" and find it difficult to quickly discern patterns. There is an urgent need for an innovative solution that integrates spatiotemporal collaborative visualization and a multi-layered traceability architecture to improve the intuitiveness of training data analysis and the efficiency of decision support. Summary of the Invention

[0005] The main objective of this invention is to provide a system and method for visualizing and analyzing the spatiotemporal distribution of training big data, in order to solve the aforementioned technical problems.

[0006] To achieve the above objectives, the present invention provides a training big data spatiotemporal distribution visualization and analysis system, comprising: a training data preprocessing module, a training data range setting module, a spatiotemporal distribution visualization display module, and a training data source tracing and analysis module, wherein: The training data preprocessing module is used to receive the training data to be visualized and perform preprocessing operations, including data format conversion and key information extraction and identification. The key information extraction and identification includes extracting key information for visualization analysis from the training data and using large model technology to identify the extracted key information as related information according to a preset association template. The key information includes time, location, training institution, training activity, training method, and training system. The association template contains a statement structure of time, location, training institution, training activity, training method, and training system variables. The training data range setting module is used to determine filtering conditions based on the subject data type to obtain the filtered target dataset; it includes: selecting the subject data type based on user interaction instructions, wherein the subject data type includes training institution, training activity, training method, and training system; and dynamically activating the corresponding filtering conditions based on the selected subject data type; wherein the correspondence between the subject data type and the filtering conditions includes: Condition 1: When the main data type is the training institution, enable the following filtering conditions: country, establishment time range, and institution category. Condition 2: When the main data type is the training activity, enable the following filtering conditions: implementation time range, implementation geographical range, and training activity type. Condition 3.1: When the main data type is the training method, if the name of the method is not specified, enable the following filtering conditions: proposed time range, proposed geographical range; Condition 3.2: When the main data type is the training method, if the name of the method is specified, enable the following filtering conditions: application time range and application geographical range. Condition 4.1: When the main data type is the training system, if the system name is not specified, enable the following filtering conditions: time range of development and release, geographical range of development and release. Condition 4.2: When the main data type is the training system, if the system name is specified, enable the following filtering conditions: application time range and application geographical range. The spatiotemporal distribution visualization module is used to display the filtered target dataset by combining time series data with geographical distribution, including: Map display unit: Displays geographical locations related to the main data type on a two-dimensional digital map in the form of punctuation marks; Timeline display unit: Displays entity names and time information related to the main data type in a timeline format on the side of the two-dimensional digital map; Interactive response unit: In response to selecting a specific entity name on the timeline, its corresponding geographical location is highlighted on a two-dimensional digital map; Dynamic constraint unit: Makes the displayed content of the map and timeline subject to the filtering conditions of the training data range setting module; wherein, the main data types include training institutions, training activities, training methods and training systems, and the display logic dynamically adapts according to the selected main data types; The training data source tracing analysis module is used to perform hierarchical source tracing analysis on the main body of training data displayed in the spatiotemporal distribution visualization module, including: First-level tracing unit: In response to the tracing instruction for the training data subject, it displays the association information related to the data subject generated using the association template on the visualization interface; The second-layer tracing unit: in response to a tracing instruction for a specific entry in the associated information, displays the training data content to be visualized corresponding to that entry; Wherein: the main body of the training data is the entity presented in the spatiotemporal distribution visualization module, including specific training institutions, training activities, training methods or training systems; the association template is defined by the training data preprocessing module and is used to combine key information into semantic descriptions.

[0007] Preferably, the data format conversion transforms the imported training data into a format recognizable by the training data range setting module; the associated template includes at least one of the following semantic structures: "At [time][place], [training institution] conducted [training activities] using [training method] or [training system]"; "At [time][place], [training institution] was established"; "At [time][place], [training institution] proposed [training method]"; "At [time][place], [training institution] developed [training system]".

[0008] Preferably, when the main data type is a training institution, the map display unit marks the geographical location of the training institution, and the timeline display unit displays the name of the training institution and its establishment time; wherein, the dynamic constraint unit performs at least one of the following: when the country filter condition is enabled, only training institutions of that country are displayed; when the establishment time range filter condition is enabled, only training institutions established within that time range are displayed; when the institution category filter condition is enabled, only training institutions of that category are displayed.

[0009] Preferably, when the main data type is training activity, the map display unit marks the geographical location of the training activity; the timeline display unit displays the name of the training activity and its implementation time; the dynamic constraint unit performs at least one of the following: when the implementation time range filtering condition is enabled, only training activities within that time range are displayed; when the implementation geographical range filtering condition is enabled, only training activities within that geographical range are displayed; when the training activity type filtering condition is enabled, only training activities of that type are displayed.

[0010] Preferably, when the main data type is a training method, if no method name is specified: the map display unit marks the geographical location where the training method was proposed; the timeline display unit displays the name of the training method and its proposal time; the dynamic constraint unit performs at least one of the following: when the proposed time range filtering condition is enabled, only the training methods proposed within that time range are displayed; when the proposed geographical range filtering condition is enabled, only the training methods proposed within that geographical range are displayed. When the main data type is a training method, if a method name is specified: the map display unit's marker location includes the location where the specified method was proposed and its application location; the timeline display unit displays the time when the specified method was proposed and its application time; the dynamic constraint unit performs at least one of the following: when the application's time range filtering condition is enabled, only application events within that time range are displayed; when the application's geographical range filtering condition is enabled, only application events within that geographical range are displayed.

[0011] Preferably, when the main data type is a training system, if no system name is specified: the map display unit's marker position is the geographical location of the training system's development and release; the timeline display unit displays the training system's name and its development and release time; the dynamic constraint unit performs at least one of the following: when the development and release time range filtering condition is enabled, only training systems developed and released within that time range are displayed; when the development and release geographical range filtering condition is enabled, only training systems developed and released within that geographical range are displayed. When the main data type is a training system, if a system name is specified: the map display unit's marker location includes the development and release location and application location of the specified system; the timeline display unit displays the development and release time and application time of the specified system; the dynamic constraint unit performs at least one of the following: when the application's time range filtering condition is enabled, only application events within that time range are displayed; when the application's geographical range filtering condition is enabled, only application events within that geographical range are displayed.

[0012] Preferably, the triggering method of the first layer tracing unit includes: in response to the selection operation performed on the training data subject in the time axis display unit area of ​​the spatiotemporal distribution visualization display module, generating a first pop-up window covering the current interface, and displaying related information entries in the first pop-up window according to the related template.

[0013] Preferably, the triggering method of the second-layer tracing unit includes: in response to a selection operation performed on the associated information item displayed in the first-layer tracing unit, generating a second pop-up window that covers the first pop-up window and displaying the training data index list to be visualized associated with the information item; and in response to a selection operation on a specific data item in the index list, displaying the complete content of the training data to be visualized.

[0014] This invention also provides a method for visual analysis of the spatiotemporal distribution of training big data, implemented using the system described in any of the above claims, comprising the following steps: S1, Data preprocessing: Receive training data and perform preprocessing operations through the training data preprocessing module; S2, Set the training data range: Through the training data range setting module, select the main data type, dynamically enable filtering conditions, and obtain the target data range after filtering; S3, Spatiotemporal distribution visualization analysis: Visual analysis is performed through the spatiotemporal distribution visualization display module; S4, Hierarchical Source Analysis: Hierarchical source analysis is performed through the training data source analysis module, including: S41, perform the first-level source tracing analysis, perform source tracing selection operation on the main body of training data, and display the generated related information items in a pop-up window; S42, perform the second layer of source tracing analysis: perform source tracing selection operation on the related information items, and display the content of the training data to be visualized in a pop-up window.

[0015] The training big data spatiotemporal distribution visualization analysis system and method proposed in this application integrates information that traditionally requires comparison across multiple charts into a single spatiotemporal view through real-time linkage and highlighting of map points and timelines; by constructing a two-layer traceability architecture, it enhances the traceability capability of data and can effectively improve the efficiency of training data analysis. Attached Figure Description

[0016] Figure 1 This embodiment provides a schematic diagram of the module structure of a training big data spatiotemporal distribution visualization and analysis system. Figure 2 This embodiment provides a schematic diagram of the steps involved in a method for visualizing and analyzing the spatiotemporal distribution of training big data. Detailed Implementation

[0017] To better explain and facilitate understanding of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] like Figure 1As shown in the figure, this embodiment provides a training big data spatiotemporal distribution visualization analysis system, which includes: a training data preprocessing module, a training data range setting module, a spatiotemporal distribution visualization display module, and a training data source tracing analysis module.

[0019] Specifically, the training big data refers to a collection of structured and unstructured data generated by training institutions, participating units, and training system platforms in high-intensity, standardized training scenarios such as emergency response drills (e.g., fire drills, disaster prevention drills), professional and technical training, and collaborative operation training, reflecting the entire life cycle of training. The training data preprocessing module is used to receive the training data to be visualized and perform preprocessing operations, including data format conversion and key information extraction and identification. The key information extraction and identification includes extracting key information for visualization analysis from the training data and using large model technology to identify the extracted key information as related information according to a preset association template. The key information includes time, location, training institution, training activity, training method, and training system. The association template contains a statement structure of time, location, training institution, training activity, training method, and training system variables. The large model technology mentioned above is used for large model prompting engineering technology, and the associated template is the prompt template.

[0020] The training data range setting module is used to determine filtering conditions based on the subject data type to obtain the filtered target dataset; it includes: selecting the subject data type based on user interaction instructions, wherein the subject data type includes training institution, training activity, training method, and training system; and dynamically activating the corresponding filtering conditions based on the selected subject data type; wherein the correspondence between the subject data type and the filtering conditions includes: Condition 1: When the main data type is the training institution, enable the following filtering conditions: country, establishment time range, and institution category. Condition 2: When the main data type is the training activity, enable the following filtering conditions: implementation time range, implementation geographical range, and training activity type. Condition 3.1: When the main data type is the training method, if the name of the method is not specified, enable the following filtering conditions: proposed time range, proposed geographical range; Condition 3.2: When the main data type is the training method, if the name of the method is specified, enable the following filtering conditions: application time range and application geographical range. Condition 4.1: When the main data type is the training system, if the system name is not specified, enable the following filtering conditions: time range of development and release, geographical range of development and release. Condition 4.2: When the main data type is the training system, if the system name is specified, enable the following filtering conditions: application time range and application geographical range.

[0021] In addition, the correspondence between the main data type and the filtering conditions may also include: Condition 3.3: When the main data type is the training method, if the name of the method is not specified, enable the filtering condition: filter all training method data within the proposed time range and the proposed geographical range; Condition 4.3: When the main data type is the training system, if the name of the system is not specified, enable the filtering condition: filter all training system data within the time range and geographical range of the development and release.

[0022] The spatiotemporal distribution visualization module includes: displaying the filtered target data in a linked format of time axis and two-dimensional map, supporting interactive analysis. It is used to display the filtered target dataset in combination with time series and geographical distribution, including: Map display unit: Displays geographical locations related to the main data type on a two-dimensional digital map in the form of punctuation marks; Timeline display unit: Displays entity names and time information related to the main data type in a timeline format on the side of the two-dimensional digital map; Interactive response unit: In response to selecting a specific entity name on the timeline, its corresponding geographical location is highlighted on a two-dimensional digital map; Dynamic constraint unit: Makes the displayed content of the map and timeline subject to the filtering conditions of the training data range setting module; wherein, the main data types include training institutions, training activities, training methods and training systems, and the display logic dynamically adapts according to the selected main data types; The training data source tracing analysis module is used to perform hierarchical source tracing analysis on the main body of training data displayed in the spatiotemporal distribution visualization module, realizing two-layer data source tracing, and tracing the original data backward from the visualization results, including: First-level tracing unit: In response to the tracing instruction for the training data subject, it displays the association information related to the data subject generated using the association template on the visualization interface; Second-level tracing unit: In response to a tracing instruction for a specific entry in the associated information (e.g., one of the multiple associated information entries displayed by the first-level tracing), it displays the training data content to be visualized corresponding to that entry; Wherein: the main body of the training data is the entity presented in the spatiotemporal distribution visualization module, including specific training institutions, training activities, training methods or training systems; the association template is defined by the training data preprocessing module and is used to combine key information into semantic descriptions.

[0023] Preferably, the data format conversion includes converting the imported training data into a format recognizable by the training data range setting module; specifically, the recognizable format may be, for example, a standardized JSON format. The associated template contains at least one of the following semantic structures: "At [time][location], [training institution] conducted [training activities] using [training method] or [training system]"; "At [time][location], [training institution] was established"; "At [time][location], [training institution] proposed [training method]"; "At [time][location], [training institution] developed [training system]".

[0024] Furthermore, when the main data type is a training institution, the map display unit marks the geographical location of the training institution, and the timeline display unit displays the name of the training institution and its establishment time; wherein, the dynamic constraint unit performs at least one of the following: when the country filter condition is enabled, only training institutions of that country are displayed; when the establishment time range filter condition is enabled, only training institutions established within that time range are displayed; when the institution category filter condition is enabled, only training institutions of that category are displayed.

[0025] Specifically, the spatiotemporal distribution visualization module displays the geographical locations of each training institution as markers on a two-dimensional digital map. On the right side of the map, a timeline displays the names of different training institutions and their establishment dates. Clicking on a training institution's name on the timeline highlights its location on the map. By setting a country, the module displays only training institutions from that country, aiding users in analyzing the situation of training institutions from different countries. By setting an establishment time range, the module displays only training institutions established within that time range, aiding users in analyzing the temporal distribution of training institution establishments. By setting an institution category, the module displays only training institutions within that category, aiding users in analyzing the situation of training institutions within different categories.

[0026] Preferably, when the main data type is training activity, the map display unit marks the geographical location of the training activity; the timeline display unit displays the name of the training activity and its implementation time; the dynamic constraint unit performs at least one of the following: when the implementation time range filtering condition is enabled, only training activities within that time range are displayed; when the implementation geographical range filtering condition is enabled, only training activities within that geographical range are displayed; when the training activity type filtering condition is enabled, only training activities of that type are displayed.

[0027] Specifically, the spatiotemporal distribution visualization module displays the main geographical locations of different training activities as markers on a two-dimensional digital map. On the right side of the two-dimensional digital map, a timeline displays the names of different training activities and their implementation times. Clicking on a training activity name on the timeline highlights the location of that activity on the two-dimensional digital map. By setting an implementation time range, this module only displays training activities within that specified time range, assisting users in analyzing the temporal distribution of training activities. By setting an implementation geographical range, this module only displays training activities within a specified geographical area, assisting users in analyzing the geographical distribution of training activities. By setting a training activity type, this module only displays training activities of a specified type, assisting users in analyzing the implementation of different types of training activities.

[0028] Optionally, when the main data type is a training method, if no method name is specified: the map display unit marks the geographical location where the training method was proposed; the timeline display unit displays the name of the training method and its proposal time; the dynamic constraint unit performs at least one of the following: when the proposed time range filtering condition is enabled, only the training methods proposed within that time range are displayed; when the proposed geographical range filtering condition is enabled, only the training methods proposed within that geographical range are displayed. When the main data type is a training method, if a method name is specified: the map display unit's marker location includes the location where the specified method was proposed and its application location; the timeline display unit displays the time when the specified method was proposed and its application time; the dynamic constraint unit performs at least one of the following: when the application's time range filtering condition is enabled, only application events within that time range are displayed; when the application's geographical range filtering condition is enabled, only application events within that geographical range are displayed.

[0029] Specifically, the spatiotemporal distribution visualization module displays the geographical locations of different training methods as points on a two-dimensional digital map. On the right side of the map, a timeline displays the names of different training methods and their proposal dates. Clicking on a method name on the timeline highlights the location where that method was proposed on the map. Without specifying a method name, setting a time range for proposal will only display training methods proposed within that time range, aiding users in analyzing the temporal distribution of proposed methods. Similarly, setting a geographical range for proposal will only display methods proposed within that geographical range, aiding users in analyzing the geographical distribution of proposed methods. With a method name specified, the module displays only the proposal time and location, as well as the time and location of application. Setting an application time range will only display the application of a specified method within that time range, aiding users in analyzing the temporal distribution of application for a particular training method. Finally, setting a geographical range for application will only display the application of a specified method within that geographical range, aiding users in analyzing the geographical distribution of application for a particular training method.

[0030] Preferably, when the main data type is a training system, if no system name is specified: the map display unit's marker position is the geographical location of the training system's development and release; the timeline display unit displays the training system's name and its development and release time; the dynamic constraint unit performs at least one of the following: when the development and release time range filtering condition is enabled, only training systems developed and released within that time range are displayed; when the development and release geographical range filtering condition is enabled, only training systems developed and released within that geographical range are displayed. When the main data type is a training system, if a system name is specified: the map display unit's marker location includes the development and release location and application location of the specified system; the timeline display unit displays the development and release time and application time of the specified system; the dynamic constraint unit performs at least one of the following: when the application's time range filtering condition is enabled, only application events within that time range are displayed; when the application's geographical range filtering condition is enabled, only application events within that geographical range are displayed.

[0031] Specifically, the spatiotemporal distribution visualization module displays the geographical locations of different training systems developed and released using markers on a two-dimensional digital map. On the right side of the two-dimensional digital map, a timeline displays the names of different system methods and their development and release dates. Clicking on a system name on the timeline highlights the location of that training system's development and release on the two-dimensional digital map. Without setting a system name, by setting a development and release time range, this module only displays training systems developed and released within the specified time range, assisting users in analyzing the temporal distribution of training system development and releases. By setting a geographical range for development and release, this module only displays training systems proposed within a specified geographical area, assisting users in analyzing the geographical distribution of training system development and releases. With a system name set, this module only displays the development and release time and location of the specified system, as well as the time and location of its application. By setting an application time range, this module only displays the application of the specified system within the specified time range, assisting users in analyzing the application time distribution of a specific training system. By setting an application geographical range, this module only displays the application of the specified system within the specified geographical area, assisting users in analyzing the geographical distribution of the application of a specific training system.

[0032] Preferably, the triggering method of the first layer tracing unit includes: in response to the selection operation performed on the training data subject in the time axis display unit area of ​​the spatiotemporal distribution visualization display module, generating a first pop-up window covering the current interface, and displaying related information entries in the first pop-up window according to the related template.

[0033] Optionally, the triggering method of the second-layer tracing unit includes: in response to a selection operation performed on the associated information item displayed in the first-layer tracing unit, generating a second pop-up window that covers the first pop-up window and displaying the training data index list to be visualized associated with the information item; and in response to a selection operation on a specific data item in the index list, displaying the complete content of the training data to be visualized.

[0034] like Figure 2 As shown, this embodiment also provides a method for visualizing and analyzing the spatiotemporal distribution of training big data, implemented using the system described in any of the above embodiments, including the following steps: S1, Data preprocessing: Receive training data and perform preprocessing operations through the training data preprocessing module; S2, Set the training data range: Through the training data range setting module, select the main data type, dynamically enable filtering conditions, and obtain the target data range after filtering; S3, Spatiotemporal distribution visualization analysis: Visual analysis is performed through the spatiotemporal distribution visualization display module; S4, Hierarchical Source Analysis: Hierarchical source analysis is performed through the training data source analysis module, including: S41, perform the first layer of source tracing analysis, perform source tracing selection operation on the main body of training data, and display related items in a pop-up window; S42, perform the second layer of source tracing analysis: perform source tracing selection operation on the related information items, and display the content of the training data to be visualized in a pop-up window.

[0035] This embodiment provides a spatiotemporal distribution visualization analysis method for intuitive and efficient analysis of training big data by extracting and identifying key information, setting data ranges as needed, visualizing data by combining time series and geographical distribution, and enhancing data understanding through two-layer tracing. This method can effectively improve the analysis efficiency of training big data.

[0036] The above are merely specific application examples of the present invention and do not constitute any limitation on the scope of protection of the present invention. In addition to the above embodiments, the present invention may have other implementations. All technical solutions formed by equivalent substitution or equivalent transformation fall within the scope of protection claimed by the present invention.

Claims

1. A training big data spatio-temporal distribution visualization analysis system, characterized in that, The application relates to a training data visualization method and system. The application comprises a training data preprocessing module, a training data range setting module, a spatio-temporal distribution visualization display module and a training data traceability analysis module, wherein: The training data preprocessing module is used for receiving training data to be visualized and performing a preprocessing operation, wherein the preprocessing operation comprises data format conversion and key information extraction identification; the key information extraction identification comprises extracting key information for visualized analysis from the training data and identifying the extracted key information as associated information according to a preset association template by using a large model technology; wherein the key information comprises time, place, training institution, training activity, training method and training system; and the association template comprises a sentence structure of time, place, training institution, training activity, training method and training system variables; The training data range setting module is used for determining a screening condition according to a subject data type to obtain a screened target data set; the method comprises the following steps: selecting a subject data type based on a user interaction instruction, wherein the subject data type comprises a training institution, a training activity, a training method and a training system; and dynamically enabling a corresponding screening condition based on the selected subject data type; wherein the corresponding relationship between the subject data type and the screening condition comprises: Condition 1: when the subject data type is the training institution, a screening condition of country, establishment time range and institution category is enabled; Condition 2: when the subject data type is the training activity, a screening condition of implementation time range, implementation geographical range and training activity type is enabled; Condition 3.1: when the subject data type is the training method, if the name of the method is not specified, a screening condition of proposed time range and proposed geographical range is enabled; Condition 3.2: when the subject data type is the training method, if the name of the method is specified, a screening condition of applied time range and applied geographical range is enabled; Condition 4.1: when the subject data type is the training system, if the name of the system is not specified, a screening condition of development and publication time range and development and publication geographical range is enabled; Condition 4.2: when the subject data type is the training system, if the name of the system is specified, a screening condition of applied time range and applied geographical range is enabled; The spatio-temporal distribution visualization display module is used for displaying the screened target data set in combination with a time sequence and a geographical distribution, and comprises: A map display unit: a geographical position related to a subject data type is displayed in the form of a mark point on a two-dimensional digital map; A time axis display unit: entity names and time information related to the subject data type are displayed in the form of a time axis on the side of the two-dimensional digital map; An interactive response unit: in response to selecting a specific entity name on the time axis, the corresponding geographical position is highlighted on the two-dimensional digital map; A dynamic constraint unit: the display contents of the map and the time axis are constrained by the screening condition of the training data range setting module; wherein the subject data type comprises a training institution, a training activity, a training method and a training system, and the display logic is dynamically adapted according to the selected subject data type. The training data provenance analysis module is configured to perform hierarchical provenance analysis on a training data subject displayed in the spatiotemporal distribution visualization module, including: a first layer provenance unit configured to, in response to a provenance instruction on the training data subject, display, in a visualization interface, associated information related to the data subject and generated by using the association template; a second layer provenance unit configured to, in response to a provenance instruction on a specific item in the associated information, display the training data content to be visualized corresponding to the item; wherein the training data subject is an entity presented in the spatiotemporal distribution visualization module, including a specific training institution, training activity, training method or training system; the association template is defined by the training data preprocessing module and is used to combine key information into a semantic description. 2.The training big data spatio-temporal distribution visualization analysis system according to claim 1, wherein, The data format conversion refers to converting the imported training data into a format recognizable by the training data range setting module; the association template includes at least one of the following semantic structures: "In [time] [place], [training institution] carried out [training activity] by using [training method] or [training system]"; "In [time] [place], [training institution] was established"; "In [time] [place], [training institution] proposed [training method]"; "In [time] [place], [training institution] developed [training system]". 3.The training big data spatio-temporal distribution visualization analysis system of claim 1, wherein, When the subject data type is a training institution, the map display unit marks the location as the geographical location of the training institution, and the timeline display unit displays the name of the training institution and its establishment time; wherein the dynamic constraint unit performs at least one of the following: when the national screening condition is enabled, only the training institutions in the country are displayed; when the establishment time range screening condition is enabled, only the training institutions established within the time range are displayed; when the institution category screening condition is enabled, only the training institutions of the category are displayed. 4.The training big data spatio-temporal distribution visualization analysis system of claim 1, wherein, When the subject data type is a training activity, the map display unit marks the location as the geographical location of the training activity implementation; the timeline display unit displays the name of the training activity and its implementation time; the dynamic constraint unit performs at least one of the following: when the implementation time range screening condition is enabled, only the training activities within the time range are displayed; when the implementation geographical range screening condition is enabled, only the training activities within the geographical range are displayed; when the training activity type screening condition is enabled, only the training activities of the type are displayed. 5.The training big data spatio-temporal distribution visualization analysis system of claim 1, wherein, When the subject data type is a training method, if the method name is not specified: the map display unit marks the location as the geographical location where the training method is proposed; the timeline display unit displays the name of the training method and its proposal time; the dynamic constraint unit performs at least one of the following: when the proposed time range screening condition is enabled, only the training methods proposed within the time range are displayed; when the proposed geographical range screening condition is enabled, only the training methods proposed within the geographical range are displayed; When the subject data type is a training method, if a method name is specified: the map display unit marks the location of the specified method's proposal and its application; the timeline display unit displays the specified method's proposal time and its application time; the dynamic constraint unit performs at least one of the following: when the time range filter condition for application is enabled, only display application events within the time range; when the geographic range filter condition for application is enabled, only display application events within the geographic range. 6.The training big data spatio-temporal distribution visualization analysis system of claim 1, wherein, When the subject data type is a training system, if no system name is specified: the map display unit marks the geographic location of the training system's development and release; the timeline display unit displays the training system name and its development and release time; the dynamic constraint unit performs at least one of the following: when the time range filter condition for development and release is enabled, only display training systems developed and released within the time range; when the geographic range filter condition for development and release is enabled, only display training systems developed and released within the geographic range. When the subject data type is a training system, if a system name is specified: the map display unit marks the location of the specified system's development and release and its application; the timeline display unit displays the specified system's development and release time and its application time; the dynamic constraint unit performs at least one of the following: when the time range filter condition for application is enabled, only display application events within the time range; when the geographic range filter condition for application is enabled, only display application events within the geographic range. 7.The training big data spatio-temporal distribution visualization analysis system of claim 1, wherein, The triggering mode of the first layer traceability unit includes: in response to a selection operation performed on the training data subject in the timeline display unit area of the spatio-temporal distribution visualization display module, generating a first pop-up window covering the current interface, and displaying the associated information entries according to the association template in the first pop-up window. 8.The training big data spatio-temporal distribution visualization analysis system of claim 1, wherein, The triggering mode of the second layer traceability unit includes: in response to a selection operation performed on an associated information entry displayed by the first layer traceability unit, generating a second pop-up window covering the first pop-up window, and displaying the training data index list associated with the information entry; in response to a selection operation on a specific data item in the index list, displaying the complete content of the training data to be visualized.

9. A method for training spatio-temporal distribution visualization analysis of big data, characterized in that, The system implementation according to any one of claims 1-8, comprising the steps of: S1, data preprocessing: receiving training data and performing preprocessing operations through the training data preprocessing module; S2, setting the training data range: through the training data range setting module: selecting the subject data type, dynamically enabling the filter condition, and obtaining the target data range after filtering; S3, spatio-temporal distribution visualization analysis: performing visualization analysis through the spatio-temporal distribution visualization display module; S4, hierarchical traceability analysis: performing hierarchical traceability analysis through the training data traceability analysis module, which includes: S41, first layer traceability analysis, traceability selection operation on the training data subject, pop-up display of associated information entries; S42, performing second layer traceability analysis: performing traceability selection operation on the association information item, and displaying the content of the to-be-visualized training data in a pop-up window.

Images