Front-end lightweight correlation analysis and calculation method and device

Abstract

The invention discloses a front-end lightweight correlation analysis and calculation method and device, and relates to the technical field of urban management information. Comprising the steps of 1, constructing a front-end lightweight correlation analysis and calculation device, 2, constructing a front-end equipment correlation relation pair according to longitude and latitude in metadata of accessed multi-source front-end sensing equipment through a lightweight correlation calculation engine module, and preprocessing collected sensing data to obtain standardized data; extracting feature vectors from the standardized data by using a lightweight dynamic graph neural network, and performing entity trajectory clustering analysis on the sensing data based on a DBSCAN clustering algorithm to generate various entity trajectories; on the basis of the entity co-occurrence frequency, the time proximity and the space proximity, the association strength between the entities is calculated, and an association analysis result is obtained; 3, converting an association analysis result into a visual decision support interface through a leader cockpit visual adaptation engine module; 4, building a configurable association analysis process by using a fuzzy scene matching mode through a dynamic configurable association rule chain module; and 5, performing real-time data synchronization under a high-load condition through a real-time data synchronization and performance optimization module, and performing calculation performance optimization and resource loading optimization.

Description

TECHNICAL FIELD

[0001] The application discloses a front-end lightweight correlation analysis calculation method and device, and relates to the technical field of city management information. BACKGROUND

[0002] With the deepening of the construction of smart cities, city managers need to process multi-source heterogeneous real-time data streams and quickly extract valuable information. The traditional back-end correlation analysis system has problems such as high response delay, large data transmission load and lagging analysis results, and it is difficult to meet the high requirements of leadership cockpit on real-time and interactivity.

[0003] Representative schemes in the prior art include: 1) a leadership cockpit visualization system based on digital twin technology, which monitors network device traffic and bandwidth utilization through a digital twin model and performs time series collaborative analysis, but this method has high computational complexity and is heavily dependent on back-end computing power; 2) a management cockpit software system based on a component-based designer, which supports designing UI interfaces through drag-and-drop operations, but lacks front-end lightweight correlation analysis capabilities; 3) existing correlation analysis methods such as text correlation analysis based on LSTM neural networks and knowledge graphs, which can perform semantic correlation analysis, but the model is too large to be deployed on the front end. SUMMARY

[0004] The present application provides a front-end lightweight correlation analysis calculation method and device to address the problems of the prior art, which is suitable for city-level big data real-time processing scenarios, including but not limited to municipal facility monitoring, public safety early warning, traffic management optimization, and resource allocation decision-making, and other comprehensive applications.

[0005] The specific scheme provided by the present application is:

[0006] The present application provides a front-end lightweight correlation analysis calculation method, comprising:

[0007] Step 1: Construct a front-end lightweight correlation analysis calculation device, which includes a lightweight correlation calculation engine module, a leadership cockpit visualization adaptation engine module, a dynamically configurable correlation rule chain module, and a real-time data synchronization and performance optimization module,

[0008] Step 2: Construct a front-end device correlation relationship pair according to the latitude and longitude in the metadata of the multi-source front-end sensing device accessed by the lightweight correlation calculation engine module, and preprocess the collected sensing data to obtain standardized data; use a lightweight dynamic graph neural network to extract a feature vector from the standardized data, perform entity trajectory clustering analysis on the sensing data based on a DBSCAN clustering algorithm, and generate various entity trajectories; calculate the correlation strength between entities based on entity co-occurrence frequency, temporal proximity, and spatial proximity, and obtain correlation analysis results;

[0009] Step 3: Transform the correlation analysis results into a visual decision support interface through the lead cockpit visualization adaptation engine module,

[0010] Step 4: Use fuzzy scenario matching to build a configurable correlation analysis process through the dynamically configurable correlation rule chain module,

[0011] Step 5: Perform real-time data synchronization under high load conditions and perform calculation performance optimization and resource loading optimization through the real-time data synchronization and performance optimization module.

[0012] Further, in step 2 of the front-end lightweight correlation analysis calculation method, the lightweight correlation calculation engine module collects the metadata of the multi-source front-end sensing devices, including device ID, latitude and longitude coordinates, and device type. During preprocessing, the collected sensing data is denoised, missing values are filled, and the data format is unified for preprocessing operations to form standardized data streams.

[0013] When calculating the correlation strength between entities, a time decay factor and a dynamic weight coefficient are introduced to update the correlation relationship strength in real time, ensuring the freshness and accuracy of the correlation relationship.

[0014] Further, in step 3 of the front-end lightweight correlation analysis calculation method, it specifically includes:

[0015] Step 31: Component-based designer construction:

[0016] Component visualization component library: provides various visualization components such as charts, tables, and maps, supporting departments to choose as needed,

[0017] Design drag-and-drop interface: freely combine visualization components through drag-and-drop operations to design personalized monitoring interfaces without writing code;

[0018] Step 32: Multi-dimensional data presentation:

[0019] Correlation relationship graph visualization: visually present entity correlation relationships in the form of force-directed graphs, heat maps, and Sankey diagrams, highlighting key correlation paths and abnormal correlation patterns,

[0020] Temporal and spatial data linkage analysis: supports time axis sliding and historical data backtracking to realize linkage analysis and display of temporal and spatial data;

[0021] Step 33: Interactive exploration analysis:

[0022] Drilling and filtering: multi-level data drilling and multi-dimensional filtering of data by time, region, and type,

[0023] Real-time early warning and prompt: set key indicator threshold, automatic early warning and prompt, for identifying problems and making decisions.

[0024] Further, the step 4 of the front-end lightweight correlation analysis calculation method specifically comprises:

[0025] Step 41: rule chain configuration management:

[0026] Assemble rule template library: contains a variety of pre-defined correlation rule templates, covering municipal management, public safety, traffic management,

[0027] Perform graphical rule editing: connect different rule nodes through a graphical interface to build complex correlation rule chains, support conditional branching, loop complex logic,

[0028] Step 42: fuzzy scenario matching:

[0029] Perform multi-factor weight allocation: allocate dynamic weights for event quantity, level multi-factor and time decay factor,

[0030] Perform suspected degree calculation: generate 0-1 source IP attack threat suspected degree value based on weighted cumulative results, support same source and same destination dimension display suspected degree ranking,

[0031] Step 43: incremental calculation and optimization:

[0032] Periodic incremental update: incremental calculation according to pre-designed calculation period, integrate new data and historical data, set event effective time range,

[0033] Dynamic adjustment of parameters: provide configurable item adjustment of related coefficients, continuously optimize with changes in urban management needs.

[0034] Further, the step 5 of the front-end lightweight correlation analysis calculation method specifically comprises:

[0035] Step 51: real-time data synchronization:

[0036] Perform WebSocket long connection: establish WebSocket long connection channel, realize real-time push of server data changes to the front-end,

[0037] Incremental data update: use incremental update mechanism, only synchronize changed data part, greatly reduce network transmission load,

[0038] Step 52: perform calculation performance optimization:

[0039] Use Web Worker multi-threaded calculation: allocate time-consuming calculation tasks to Web Worker threads to avoid blocking main threads and interface rendering,

[0040] Model quantization and lightweight: quantize and distill the machine learning model deployed in the front end, significantly reduce the model size and computing resource consumption,

[0041] Step 53: Resource loading optimization:

[0042] Tree Shaking application: use the static analysis features of ES modules to remove uncalled code during construction, effectively reducing the final package size,

[0043] Dynamic import and lazy loading: combine the dynamic import function of Webpack tool, split the non-core function module into independent chunk, load on demand, improve the first screen loading speed.

[0044] The application also provides a front-end lightweight correlation analysis computing device, comprising a lightweight correlation computing engine module, a leader cockpit visualization adaptation engine module, a dynamically configurable correlation rule chain module, and a real-time data synchronization and performance optimization module,

[0045] The lightweight correlation computing engine module constructs the front-end device correlation relationship pair according to the latitude and longitude in the metadata of the accessed multi-source front-end sensing device, pre-processes the collected sensing data to obtain standardized data, extracts feature vectors from the standardized data using a lightweight dynamic graph neural network, performs entity trajectory clustering analysis on the sensing data based on a DBSCAN clustering algorithm, and generates various entity trajectories; based on entity co-occurrence frequency, time proximity, and spatial proximity, the correlation strength between entities is calculated to obtain the correlation analysis result;

[0046] The leader cockpit visualization adaptation engine module converts the correlation analysis result into a visual decision support interface,

[0047] The dynamically configurable correlation rule chain module uses a fuzzy scenario matching method to configure a configurable correlation analysis process,

[0048] The real-time data synchronization and performance optimization module performs real-time data synchronization under high load conditions, and performs computing performance optimization and resource loading optimization.

[0049] Further, the lightweight correlation computing engine module of the front-end lightweight correlation analysis computing device collects metadata of multi-source front-end sensing devices, the metadata including device ID, latitude and longitude coordinates, and device type, and performs preprocessing, including denoising, filling missing values, and unifying data format preprocessing operations on the collected sensing data to form a standardized data stream;

[0050] When calculating the correlation strength between entities, a time decay factor and a dynamic weight coefficient are introduced to update the correlation relationship strength in real time, ensuring the freshness and accuracy of the correlation relationship.

[0051] Further, the leading cockpit visualization adaptation engine module of the front-end lightweight correlation analysis computing device executes the process, specifically including:

[0052] Step 31: Component-based designer construction:

[0053] Component visualization component library: Provide various visualization components of charts, tables, and maps, support various departments to choose as needed,

[0054] Design drag-and-drop interface: freely combine visualization components through drag-and-drop operations to design personalized monitoring interfaces without writing code;

[0055] Step 32: Multi-dimensional data presentation:

[0056] Correlation graph visualization: visually present entity correlation in the form of force-directed graph, heat map, and Sankey diagram, highlighting key correlation paths and abnormal correlation patterns,

[0057] Temporal and spatial data linkage analysis: support time axis sliding and historical data backtracking to realize linkage analysis and display of temporal and spatial data;

[0058] Step 33: Interactive exploration analysis:

[0059] Drilling and filtering: multi-level data drilling and multi-dimensional filtering of data by time, region, and type,

[0060] Real-time warning and prompt: set key indicator thresholds for automatic warning and prompt to identify problems and make decisions.

[0061] Further, the dynamic configurable correlation rule chain module of the front-end lightweight correlation analysis computing device executes the process, specifically including:

[0062] Step 41: Rule chain configuration management:

[0063] Build rule template library: contains various predefined correlation rule templates covering municipal management, public safety, and traffic management,

[0064] Graphical rule editing: connect different rule nodes through a graphical interface to build complex correlation rule chains, support conditional branching, and complex logic,

[0065] Step 42: Fuzzy scenario matching:

[0066] Multi-factor weight allocation: dynamically allocate weights to event quantity, level multi-factor, and time decay factor,

[0067] Suspected degree calculation: generate a 0-1 source IP attack threat suspected degree value based on the weighted cumulative result, and support the same source and same destination dimension display suspected degree ranking,

[0068] Step 43: incremental calculation and optimization:

[0069] Periodic incremental update: incremental calculation according to the pre-designed calculation period, integration of new data and historical data, setting of event effective time range,

[0070] Dynamic adjustment of parameters: provide configurable item adjustment related coefficient, and continuously optimize with the change of city management demand.

[0071] Further, the real-time data synchronization and performance optimization module execution process of the front-end lightweight correlation analysis calculation device specifically comprises:

[0072] Step 51: real-time data synchronization:

[0073] WebSocket long connection: establish a WebSocket long connection channel to realize real-time pushing of server data changes to the front end,

[0074] Incremental data update: incremental update mechanism is adopted, only the changed data part is synchronized, and the network transmission load is greatly reduced,

[0075] Step 52: calculation performance optimization:

[0076] Web Worker multi-threaded calculation: time-consuming calculation tasks are allocated to Web Worker threads to avoid blocking the main thread and interface rendering,

[0077] Model quantization and lightweight: the machine learning model deployed on the front end is quantized and distilled, which significantly reduces the model size and computing resource consumption,

[0078] Step 53: resource loading optimization:

[0079] Tree Shaking application: use the static analysis characteristics of ES modules to remove uncited code during construction, effectively reducing the final packaging size,

[0080] Dynamic import and lazy loading: combine the dynamic import function of Webpack tool, split non-core function modules into independent chunks, and load on demand to improve the first screen loading speed.

[0081] The beneficial aspects of the present application are:

[0082] Significant improvement in data processing efficiency: Through lightweight front-end computing, the data transmission round-trip time is reduced, the correlation analysis response speed is significantly improved, the front-end computing delay is effectively reduced, the real-time analysis and presentation of urban management data are realized, and the use experience and decision-making efficiency of the leader cockpit are greatly improved.

[0083] Significant reduction in computational complexity and resource consumption: Using improved lightweight dynamic graph neural network technology, the feature transformation and nonlinear activation operations in traditional convolution operations are removed, reducing the computational complexity by more than 40%, allowing the system to run stably in ordinary hardware environments, reducing the deployment and operation and maintenance costs of urban management information systems.

[0084] Enhance the flexibility and adaptability of the system: Through componentized designer and configurable association rule chain, support users to customize analysis interface and association rule according to specific needs, make the system adapt to the changing needs of different urban management scenarios, prolong the technical life cycle and use value of the system.

[0085] Improve the accuracy and practicality of correlation analysis: Through multi-factor dynamic weight distribution, time decay factor and fuzzy scene matching technology, improve the accuracy and practicality of correlation relationship mining, better identify potential risks and abnormal patterns in urban operation, provide more reliable decision support for urban managers.

[0086] Realize comprehensive visualization and interactive experience: Provide a variety of visualization components and interactive exploration functions, enable urban managers to intuitively and comprehensively understand the urban operation status, support multi-level data analysis from macro to micro, improve the informatization level and decision-making quality of urban management.

[0087] Guarantee high performance and stability of the system: Through Web Worker multi-threaded computing, Tree Shaking to eliminate dead code, dynamic import and lazy loading performance optimization strategies, the system can still maintain smooth and stable in high concurrency and large data volume, resource loading efficiency is significantly improved. BRIEF DESCRIPTION OF DRAWINGS

[0088] Figure 1 is a lightweight correlation computing engine module execution process schematic diagram.

[0089] Figure 2 is a leader cockpit visualization adaptation engine execution process schematic diagram.

[0090] Figure 3 is a dynamically configurable association rule chain module execution process schematic diagram.

[0091] Figure 4 is a real-time data synchronization and performance optimization module execution process schematic diagram. DETAILED DESCRIPTION

[0092] The application will be further described below in conjunction with the drawings and specific embodiments so that those skilled in the art can better understand the application and implement it. The embodiments are not intended to limit the application.

[0093] Embodiment 1

[0094] The application provides a front-end lightweight correlation analysis calculation method, comprising:

[0095] Step 1: Construct a front-end lightweight correlation analysis calculation device, which comprises a lightweight correlation calculation engine module, a leader cockpit visualization adaptation engine module, a dynamically configurable correlation rule chain module and a real-time data synchronization and performance optimization module.

[0096] Step 2: Construct a front-end device correlation relationship pair according to the latitude and longitude in the metadata of the accessed multi-source front-end sensing device through the lightweight correlation calculation engine module, and pre-process the collected sensing data to obtain standardized data; extract feature vectors from the standardized data using a lightweight dynamic graph neural network, perform entity trajectory clustering analysis on the sensing data based on a DBSCAN clustering algorithm, and generate various entity trajectories; calculate the correlation strength between entities based on entity co-occurrence frequency, time proximity and spatial proximity, and obtain correlation analysis results.

[0097] The metadata of the multi-source front-end sensing device is collected by the lightweight correlation calculation engine module, and the metadata includes device ID, latitude and longitude coordinates, and device type. When pre-processing, the pre-processing operations include denoising, filling missing values and unifying data formats for the collected sensing data to form a standardized data stream.

[0098] The dynamic graph data feature vectors are extracted from the standardized data using a lightweight dynamic graph neural network, which can include: converting dynamic data into dynamic graph structure data, where nodes represent city entities and edges represent spatio-temporal co-occurrence relationships; constructing a loss function; performing graph convolution operation on the dynamic graph structure data to extract features; removing the feature transformation linear layer and the nonlinear activation function in the convolution operation to reduce the computational complexity; determining whether the loss value converges or the number of iterations reaches the upper limit; and outputting the dynamic graph data features.

[0099] The lightweight dynamic graph neural network can introduce a symmetric normalized adjacency tensor, and the calculation method is as follows:

[0100] Where A is the original adjacency matrix, I is the identity matrix, and D is the degree matrix, which is used to enhance the model's ability to describe the cross-spatio-temporal correlation relationship in dynamic graph data, reduce the computational complexity, and improve the training efficiency.

[0101] When calculating the correlation strength between entities, a time decay factor and a dynamic weight coefficient are introduced to update the correlation strength in real time, ensuring the freshness and accuracy of the correlation.

[0102] Step 3: Transform the correlation analysis results into a visual decision support interface through the leader cockpit visualization adaptation engine module.

[0103] Further specific embodiments can include:

[0104] Step 31: Build a componentized designer:

[0105] Component visualization component library: Provide a variety of visualization components such as charts, tables, and maps, support departments to choose as needed,

[0106] Design drag-and-drop interface: freely combine visualization components through drag-and-drop operations to design personalized monitoring interfaces without writing code;

[0107] Step 32: Multi-dimensional data presentation:

[0108] Correlation graph visualization: visually present entity correlation in the form of force-directed graphs, heat maps, and Sankey diagrams, highlighting key correlation paths and abnormal correlation patterns,

[0109] Temporal and spatial data linkage analysis: support time axis sliding and historical data backtracking to realize temporal and spatial data linkage analysis and display;

[0110] Step 33: Interactive exploration analysis:

[0111] Drilling and filtering: multi-level data drilling and multi-dimensional filtering by time, region, and type,

[0112] Real-time warning and prompt: set key indicator thresholds for automatic warning and prompt to identify problems and make decisions.

[0113] Step 4: Use fuzzy scenario matching to build a configurable correlation analysis process through a dynamically configurable correlation rule chain module. Specific embodiments can include:

[0114] Step 41: Rule chain configuration management:

[0115] Build a rule template library: contains a variety of predefined correlation rule templates covering municipal management, public safety, and traffic management,

[0116] Graphical rule editing: connect different rule nodes through a graphical interface to build complex correlation rule chains, support conditional branching, and complex logic,

[0117] Step 42: Fuzzy scenario matching:

[0118] Multi-factor weight distribution is performed: dynamic weights are assigned to the number of events, multi-factor of level, and time decay factor,

[0119] Suspected degree calculation is performed: a 0-1 source IP attack threat suspected degree value is generated based on the weighted cumulative result, and the suspected degree ranking in the same source and same destination dimension is supported,

[0120] Step 43: Incremental calculation and optimization:

[0121] Periodic incremental update is performed: incremental calculation is performed according to the pre-designed calculation period, new data and historical data are integrated, and the effective time range of the event is set,

[0122] Dynamic adjustment of parameters: configurable items are provided to adjust the related coefficients, and the system is continuously optimized as the city management needs change.

[0123] Step 5: Real-time data synchronization and performance optimization module is performed under high load conditions, and calculation performance optimization and resource loading optimization are performed. Specifically, it can include:

[0124] Step 51: Real-time data synchronization is performed:

[0125] WebSocket long connection is performed: WebSocket long connection channel is established to realize real-time push of server data changes to the front end,

[0126] Incremental data update is performed: incremental update mechanism is adopted, only the changed data part is synchronized, and the network transmission load is greatly reduced,

[0127] Step 52: Calculation performance optimization is performed:

[0128] Web Worker multi-threaded calculation is adopted: time-consuming calculation tasks such as GNN model inference and large-scale clustering analysis are allocated to Web Worker threads to avoid blocking the main thread and interface rendering,

[0129] Model quantization and lightening are performed: machine learning models deployed on the front end are quantized and distilled to significantly reduce model size and computing resource consumption,

[0130] Step 53: Resource loading optimization is performed:

[0131] Tree Shaking application is performed: the static analysis feature of ES modules is used to remove unused code during construction, effectively reducing the final package size,

[0132] Dynamic import and lazy loading: combined with the dynamic import function of Webpack tool, non-core function modules such as complex chart library and historical analysis module are split into independent chunks and loaded on demand, improving the first screen loading speed.

[0133] Embodiment 2

[0134] The application also provides a front-end lightweight correlation analysis computing device, comprising a lightweight correlation computing engine module, a leader cockpit visualization adaptation engine module, a dynamically configurable correlation rule chain module and a real-time data synchronization and performance optimization module,

[0135] The lightweight correlation computing engine module constructs a front-end device correlation relationship pair according to the latitude and longitude in the metadata of the accessed multi-source front-end sensing device, pre-processes the collected sensing data to obtain standardized data, extracts a feature vector from the standardized data by using a lightweight dynamic graph neural network, performs entity trajectory clustering analysis on the sensing data based on a DBSCAN clustering algorithm, and generates various entity trajectories; and calculates the correlation strength between entities based on entity co-occurrence frequency, time proximity and spatial proximity to obtain a correlation analysis result.

[0136] The leader cockpit visualization adaptation engine module converts the correlation analysis result into a visual decision support interface,

[0137] The dynamically configurable correlation rule chain module uses a fuzzy scene matching method to establish a configurable correlation analysis process,

[0138] The real-time data synchronization and performance optimization module performs real-time data synchronization under high load conditions, and performs computing performance optimization and resource loading optimization.

[0139] The information interaction and execution process between the modules in the above device are based on the same concept as the method embodiments of the application, and the specific content can be referred to the description in the method embodiments of the application, which will not be repeated here.

[0140] Similarly, the advantages of the device of the application are:

[0141] Significantly improve data processing efficiency: through front-end lightweight computing, the data transmission round-trip time is reduced, the correlation analysis response speed is significantly improved, the front-end computing delay is effectively reduced, the real-time analysis and presentation of urban management data are realized, and the use experience and decision efficiency of the leader cockpit are greatly improved.

[0142] Significantly reduce computing complexity and resource consumption: the improved lightweight dynamic graph neural network technology is adopted, the feature transformation and nonlinear activation operation in the traditional convolution operation are removed, the computing complexity is reduced by more than 40%, the system can stably run in a normal hardware environment, and the deployment and operation and maintenance cost of the urban management information system is reduced.

[0143] Enhancing the flexibility and adaptability of the system: Through the component-based designer and configurable association rule chain, users can customize the analysis interface and association rules according to specific needs, making the system adaptable to the changing needs of different urban management scenarios, extending the technical life cycle and use value of the system.

[0144] Improving the accuracy and practicality of association analysis: Through multi-factor dynamic weight distribution, time decay factor and fuzzy scenario matching technology, the accuracy and practicality of association relationship mining are improved, which can better identify potential risks and abnormal patterns in urban operation and provide more reliable decision support for urban managers.

[0145] Realizing comprehensive visualization and interactive experience: Providing various visualization components and interactive exploration functions, urban managers can intuitively and comprehensively grasp the urban operation status, supporting multi-level data analysis from macro to micro, improving the informatization level and decision quality of urban management.

[0146] Ensuring high performance and stability of the system: Through Web Worker multi-threaded computing, Tree Shaking dead code elimination, dynamic import and lazy loading performance optimization strategies, the system can maintain smooth and stable performance under high concurrency and large data volume, and the resource loading efficiency is significantly improved.

[0147] It should be noted that not all steps and modules in the above processes and device structures are necessary, and some steps or modules can be omitted according to actual needs. The execution order of each step is not fixed and can be adjusted as needed. The system structure described in the above embodiments can be a physical structure or a logical structure, i.e., some modules may be implemented by the same physical entity, or some modules may be implemented by multiple physical entities, or they may be implemented by some components in multiple independent devices.

[0148] The above-described embodiments are only preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Any equivalent replacement or transformation made by those skilled in the art based on the present application is within the protection scope of the present application. The protection scope of the present application is subject to the claims.

Claims

1. A front-end lightweight association analysis computing method, characterized by Comprise: Step 1: build a front-end lightweight correlation analysis computing device, which comprises a lightweight correlation calculation engine module, a leader cockpit visualization adaptation engine module, a dynamically configurable correlation rule chain module and a real-time data synchronization and performance optimization module, Step 2: through the lightweight correlation calculation engine module, the latitude and longitude in the metadata of the accessed multi-source front-end sensing device is used to build the front-end device correlation relationship pair, and the collected sensing data is preprocessed to obtain standardized data; the standardized data is extracted by using a lightweight dynamic graph neural network to obtain a feature vector, the sensing data is analyzed by using a DBSCAN clustering algorithm to generate various entity trajectories; based on entity co-occurrence frequency, time proximity and spatial proximity, the correlation strength between entities is calculated to obtain the correlation analysis result; Step 3: through the leader cockpit visualization adaptation engine module, the correlation analysis result is converted into a visual decision support interface, Step 4: through the dynamically configurable correlation rule chain module, a configurable correlation analysis process is established by using a fuzzy scene matching method, Step 5: through the real-time data synchronization and performance optimization module, real-time data synchronization is performed under high load conditions, and calculation performance optimization and resource loading optimization are performed.

2. The method of claim 1, wherein the method is characterized by In step 2, the metadata of the multi-source front-end sensing device is collected by the lightweight correlation calculation engine module, and the metadata includes device ID, latitude and longitude coordinates, and device type. During preprocessing, denoising, missing value filling, and unified data format preprocessing operations are performed on the collected sensing data to form a standardized data stream; When calculating the correlation strength between entities, a time decay factor and a dynamic weight coefficient are introduced to update the correlation relationship strength in real time, ensuring the freshness and accuracy of the correlation relationship.

3. The method of claim 1, wherein the method is characterized by In step 3, it specifically includes: Step 31: componentized designer construction: Component visualization component library: provides various visualization components such as charts, tables and maps, and supports departments to select as needed, Design drag-and-drop interface: freely combine visualization components through drag-and-drop operations to design personalized monitoring interfaces without writing code, Step 32: multi-dimensional data presentation: Correlation relationship graph visualization: visually present entity correlation relationships in the form of force-directed graphs, heat maps and Sankey diagrams, highlighting key correlation paths and abnormal correlation patterns, Temporal and spatial data linkage analysis: supports time axis sliding and historical data backtracking to realize linkage analysis and display of temporal and spatial data; Step 33: interactive exploration analysis: Drilling and filtering: multi-level data drilling and multi-dimensional filtering of data by time, region and type, Real-time warning and prompt: set key indicator thresholds for automatic warning and prompt to identify problems and make decisions.

4. The method of claim 1, wherein the method is characterized by In step 4, it specifically includes: Step 41: rule chain configuration management: Establish a rule template library: contains various pre-defined correlation rule templates covering municipal management, public safety and traffic management, Graphical rule editing: connect different rule nodes through a graphical interface to build complex correlation rule chains, supporting conditional branching and loop complex logic, Step 42: fuzzy scene matching: Multi-factor weight distribution is performed: dynamic weights are assigned to the number of events, multi-factor and time decay factor, Suspected degree calculation is performed: a 0-1 source IP attack threat suspected degree value is generated based on the weighted cumulative result, and suspected degree ranking in the same source and same destination dimension is supported, Step 43: incremental calculation and optimization: Periodic incremental update is performed: incremental calculation is performed according to the pre-designed calculation period, new data and historical data are integrated, and the effective time range of the event is set, Dynamic adjustment of parameters: provide configurable item adjustment related coefficient, continuously optimize with the change of city management demand.

5. The method of claim 1, wherein the method is characterized by In step 5, specifically comprising: Step 51: real-time data synchronization is performed: WebSocket long connection is performed: WebSocket long connection channel is established to realize real-time pushing of server data changes to the front end, Incremental data update is performed: incremental update mechanism is adopted to synchronize only the changed data part, which greatly reduces the network transmission load, Step 52: calculation performance optimization is performed: Web Worker multi-threaded calculation is adopted: time-consuming calculation tasks are assigned to Web Worker threads to avoid blocking the main thread and interface rendering, Model quantization and lightening: the machine learning model deployed on the front end is quantized and distilled to significantly reduce the model size and computing resource consumption, Step 53: resource loading optimization is performed: Tree Shaking application: use the static analysis feature of ES module to remove uncited code during construction, effectively reducing the final package size, Dynamic import and lazy loading: combined with the dynamic import function of Webpack tool, non-core function modules are split into independent chunks and loaded on demand to improve the first screen loading speed.

6. A front-end lightweight correlation analysis computing device, characterized by It includes a lightweight correlation calculation engine module, a leader cockpit visualization adaptation engine module, a dynamically configurable correlation rule chain module, and a real-time data synchronization and performance optimization module, The lightweight correlation calculation engine module constructs the front-end device correlation relationship pair according to the latitude and longitude in the metadata of the accessed multi-source front-end sensing device, and pre-processes the collected sensing data to obtain standardized data; a feature vector is extracted from the standardized data using a lightweight dynamic graph neural network, entity trajectory clustering analysis is performed on the sensing data based on the DBSCAN clustering algorithm, and various entity trajectories are generated; based on entity co-occurrence frequency, time proximity and spatial proximity, the correlation strength between entities is calculated to obtain the correlation analysis result; The leader cockpit visualization adaptation engine module converts the correlation analysis result into a visual decision support interface, The dynamically configurable correlation rule chain module uses fuzzy scene matching to build a configurable correlation analysis process, The real-time data synchronization and performance optimization module performs real-time data synchronization under high load conditions, and performs calculation performance optimization and resource loading optimization.

7. The front-end lightweight correlation analysis computing device of claim 6, wherein The lightweight correlation calculation engine module collects metadata of multi-source front-end sensing devices, including device ID, latitude and longitude coordinates, and device type, and performs preprocessing, including denoising, missing value filling, and unified data format preprocessing operations on the collected sensing data to form a standardized data stream; When calculating the correlation strength between entities, a time decay factor and a dynamic weight coefficient are introduced to update the correlation strength in real time, ensuring the freshness and accuracy of the correlation.

8. The front-end lightweight correlation analysis computing device of claim 6, wherein The leader cockpit visualization adaptation engine module performs a process, specifically including: Step 31: Component-based designer construction: Component visualization component library: provides various visualization components such as charts, tables, and maps, allowing departments to select components as needed, Design drag-and-drop interface: freely combine visualization components through drag-and-drop operations to design personalized monitoring interfaces without writing code; Step 32: Multi-dimensional data presentation: Correlation graph visualization: visually present entity correlation in the form of force-directed graphs, heat maps, and Sankey diagrams, highlighting key correlation paths and abnormal correlation patterns, Temporal and spatial data linkage analysis: supports time axis sliding and historical data backtracking to achieve linkage analysis and display of temporal and spatial data; Step 33: Interactive exploration and analysis: Drilling and filtering: multi-level data drilling and multi-dimensional filtering based on time, region, and type, Real-time warning and prompt: set key indicator thresholds for automatic warning and prompt to identify problems and make decisions.

9. The front-end lightweight correlation analysis computing device of claim 6, wherein The dynamically configurable correlation rule chain module performs a process, specifically including: Step 41: Rule chain configuration management: Rule template library: contains various predefined correlation rule templates covering municipal management, public safety, and traffic management, Graphical rule editing: connect different rule nodes through a graphical interface to build complex correlation rule chains, supporting conditional branching and loop complex logic, Step 42: Fuzzy scenario matching: Multi-factor weight allocation: dynamically allocate weights to event quantity, level multi-factors, and time decay factors, Suspected degree calculation: generate a 0-1 source IP attack threat suspected degree value based on weighted cumulative results, supporting suspected degree ranking in the same source and same destination dimensions, Step 43: Incremental calculation and optimization: Periodic incremental update: incrementally calculate based on the pre-designed calculation period, integrate new data and historical data, and set event validity time range, Dynamic parameter adjustment: provide configurable item adjustment related coefficients for continuous optimization as city management needs change.

10. The front-end lightweight correlation analysis computing device of claim 1, wherein The real-time data synchronization and performance optimization module performs a process, specifically including: Step 51: Real-time data synchronization: WebSocket long connection: establishes a WebSocket long connection channel to achieve real-time data push from the server to the front end, Incremental data update: uses an incremental update mechanism to synchronize only the changed data, significantly reducing network transmission load, Step 52: Performance optimization: Web Worker multi-threaded calculation: assigns time-consuming calculation tasks to Web Worker threads to avoid blocking the main thread and interface rendering, Model quantization and lightweight: quantizes and distills machine learning models deployed on the front end, significantly reducing model size and computational resource consumption, Step 53: Resource loading optimization: Tree Shaking: Take advantage of the static analysis features of ES modules to remove unused code at build time, effectively reducing the final bundle size, Dynamic Import and Lazy Loading: Combine the dynamic import feature of Webpack to split non-core function modules into independent chunks, and load them on demand to improve the first screen loading speed.

Images