An event handling method and system based on large models and flowables

Abstract

This invention discloses an event handling method and system based on a large model and flowable, relating to the field of data processing technology; it includes: Step 1: generating an event knowledge graph using a large model; Step 2: automatically adjusting the event handling process; Step 3: automatically allocating events using a large model; Step 4: performing intelligent analysis of massive events: analyzing massive event data through a large model to provide multi-dimensional analytical data; using the analytical capabilities of the large model to intelligently analyze the handling process, statistically analyzing the handling time and completion rate of each participating department during the event handling process, and providing statistical data.

Description

Technical Field

[0001] This invention discloses an event handling method and system based on a large model and flowable, relating to the field of data processing technology. Background Technology

[0002] The main functions of an incident handling system are to aggregate incidents, coordinate their handling, and perform statistical analysis. However, existing incident handling systems still have some problems, such as:

[0003] The sources of events are complex and diverse, making standardized governance difficult. Some third-party system data may not have clear type fields, only descriptive information. This makes it difficult to transform raw data into standard data. Developers often need to manually analyze key information and update it continuously, resulting in low accuracy and efficiency.

[0004] The event handling process configured in the workflow engine is often static and cannot be dynamically adjusted based on the outcome of the event handling without modifying the process.

[0005] Merging events is difficult because the same event may come from different systems and different time periods. Merging such events often requires manual analysis and comparison, which is inefficient and inaccurate.

[0006] Current incident handling systems involve many repetitive tasks requiring human intervention, are not flexible enough, often require the involvement of professionals for adjustments, and have limited capabilities for analyzing massive amounts of data. Summary of the Invention

[0007] This invention addresses the problems of existing technologies by providing an event handling method and system based on large models and flowable technologies. It uses large model technology to analyze event elements, automatically identify, classify, and merge events, automatically optimize and adjust the handling process based on the analysis of the event handling process, and uses large model technology to analyze massive amounts of data to provide diverse statistical data.

[0008] The specific solution proposed in this invention is as follows:

[0009] This invention provides an event handling method based on a large model and flowable, comprising:

[0010] Step 1: Generate an event knowledge graph using a large model: Analyze existing events using a large model to generate structured event data. Extract key descriptive information, location, and source of the events from the structured event data, and associate them with standard event types to generate an event knowledge graph.

[0011] Step 2: Automatically adjust the event handling process: Configure the handling process for various types of events using Flowable, reserve common handling nodes during the handling process, and dynamically adjust the handling personnel or departments based on conditions.

[0012] Step 3: Automated Event Allocation Using a Large-Scale Model: For events requiring coordinated handling, the large-scale model analyzes historical handling of similar events. Using the NSGA-II multi-objective optimization algorithm, Pareto solutions are found, and corresponding labels and scores are assigned based on different hierarchical levels. Updates are performed periodically. Coordinated events are automatically categorized based on an event knowledge graph. All handling processes are located according to the categories, and the process with the highest overall score is automatically matched and allocated.

[0013] Step 4: Conduct intelligent analysis of massive events: Analyze massive event data through large models to provide multi-dimensional analytical data; utilize the analytical capabilities of large models to conduct intelligent analysis of the handling process, and statistically analyze the handling time and completion rate of each participating department during the event handling process, providing statistical data.

[0014] Furthermore, step 1 of the event handling method based on large models and flowable includes: for events with clear event classifications, creating a simple mapping relationship based on the original classification and source; for events without clear event classifications, extracting key metadata, performing natural language segmentation, creating event relationships, and adding them to the event knowledge graph; as the aggregated events accumulate, automatically improving the event knowledge graph, and increasing the matching success rate and accuracy.

[0015] Furthermore, step 2 of the aforementioned event handling method based on a large model and flowable includes:

[0016] When configuring the handling process for various types of events using Flowable, design a handling process for each type of event, and configure a general handling node based on task monitoring at key handling nodes in each process. This node monitors key parameter variables in the task configuration and dynamically modifies the person or department responsible for handling the event based on these key parameter variables.

[0017] When storing the configured handling process, additional scoring quantification data is configured to save the score value of each general handling node for subsequent calculation of the overall score of the process.

[0018] Furthermore, in step 3 of the aforementioned event handling method based on a large model and flowable, the historical handling of similar events is analyzed based on the large model. For the event handling of each node, the NSGA-II multi-objective optimization algorithm is used to comprehensively consider the handling time, completion rate, and return rate to obtain the Pareto solution, which is the optimal handling node. The score of the optimal handling node is set to 1, and the scores of the next lower level are adjusted down, and so on. All nodes are marked with scores and updated periodically. Based on the scores of the handling nodes, a comprehensive analysis of each process is performed.

[0019] This invention also provides an event handling system based on a large model and flowability, including a graph generation module, a process adjustment module, a distribution module, and an analysis module.

[0020] The knowledge graph generation module utilizes a large model to generate an event knowledge graph: it analyzes existing events using a large model to generate structured event data, extracts key descriptive information, location, and source of events from the structured event data, and associates these with standard event types to generate an event knowledge graph.

[0021] The process adjustment module automatically adjusts the event handling process: it uses flowable to configure the handling process for various types of events, reserves common handling nodes during the handling process, and dynamically adjusts the handling personnel or departments based on conditions.

[0022] The allocation module utilizes a large model for automatic event allocation: For events requiring coordinated handling, it analyzes historical handling of similar events based on the large model, uses the NSGA-II multi-objective optimization algorithm to find Pareto solutions, and assigns corresponding labels and scores based on different levels, updating periodically. Based on an event knowledge graph, it automatically classifies coordinated events, finds all handling processes according to the classification, and automatically matches the process with the highest comprehensive score for automatic allocation.

[0023] The analysis module performs intelligent analysis of massive events: it analyzes massive event data through large models to provide multi-dimensional analysis data; it uses the analytical capabilities of large models to intelligently analyze the handling process, and provides statistical data on the handling time and completion rate of each participating department during the event handling process.

[0024] Furthermore, the graph generation module of the event handling system based on a large model and flowable creates a simple mapping relationship based on the original classification and source for events with clear event classifications. For events without clear event classifications, it extracts key metadata, performs natural language segmentation, creates event relationships, and adds them to the event knowledge graph. As the accumulated events accumulate, the event knowledge graph is automatically improved, increasing the matching success rate and accuracy.

[0025] Furthermore, the process adjustment module of the event handling system based on a large model and flowable, when configuring the handling process for various types of events using flowable, designs a handling process for each type of event, and configures a general handling node based on task monitoring at key handling nodes in each process. This node is used to monitor key parameter variables of the task configuration and dynamically modify the person or department responsible for handling the event based on these key parameter variables.

[0026] When storing the configured handling process, additional scoring quantification data is configured to save the score value of each general handling node for subsequent calculation of the overall score of the process.

[0027] Furthermore, the analysis module of the event handling system based on a large model and flowable performs historical handling analysis of similar events based on the large model. For the event handling of each node, the NSGA-II multi-objective optimization algorithm is used to comprehensively consider the handling time, completion rate, and return rate to obtain the Pareto solution, which is the optimal handling node. The score of the optimal handling node is set to 1, and the scores of the next lower level are adjusted down, and so on. All nodes are marked with scores and updated periodically. Based on the scores of the handling nodes, a comprehensive analysis of each process is performed.

[0028] The advantages of this invention are:

[0029] (1) Intelligent event classification: Based on the generated event knowledge graph, events are automatically classified, which greatly reduces the burden on grassroots staff, requires less manual intervention, and improves efficiency and accuracy.

[0030] (2) Automatic adjustment of incident handling process: By monitoring and analyzing the incident handling process, the process parameters are continuously optimized and adjusted based on the actual incident handling process results, and the nodes in the handling are automatically adjusted, which improves the efficiency and handling rate of incident handling, and reduces backtracking and handling time.

[0031] (3) Automatic allocation of collaborative events: After a collaborative event arrives, the event is automatically classified through the event knowledge graph. The optimal handling process is automatically selected from the candidate handling processes, eliminating the intermediate manual selection step, greatly reducing the burden on grassroots staff, and improving the accuracy of selection, so that the event can be handled better.

[0032] (4) Intelligent analysis of massive events: Based on the big model technology, it performs intelligent analysis on massive event data and event handling process, providing multi-dimensional event analysis results. It can be used for statistical display, intelligent predictive analysis, departmental assessment, difficult event analysis and other scenarios, and can perform real-time personalized data analysis. Attached Figure Description

[0033] Figure 1 This is a schematic diagram illustrating the automatic adjustment of the node score calculation and processing flow involved in the method of this invention.

[0034] Figure 2 This is a schematic diagram of the process selection and automatic event dispatch involved in the method of the present invention. Detailed Implementation

[0035] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.

[0036] Example 1

[0037] This invention provides an event handling method based on a large model and flowable, comprising:

[0038] Step 1: Generate an event knowledge graph using a large model: Analyze existing events using a large model to generate structured event data. Extract key descriptive information, location, and source of the events from the structured event data and associate them with standard event types to generate an event knowledge graph.

[0039] Based on a large model, the system analyzes the currently aggregated event data, automatically generates structured event data, extracts key information from the structured event data, such as classification, description, source, and location, and uses the standard event types as the result set to create an event relationship graph.

[0040] For events with clearly defined event categories, a simple mapping relationship is created based on the original category and source. For events without clearly defined event categories, key metadata is extracted, natural language segmentation is performed, event relationships are created, and the events are added to the knowledge graph. As the aggregated events accumulate, the knowledge graph is automatically improved, increasing the matching success rate and accuracy.

[0041] Step 2: Automatically adjust the event handling process: Use Flowable to configure the handling process for various types of events, reserve common handling nodes in the handling process, and dynamically adjust the handling personnel or departments according to the conditions.

[0042] Based on a flowable workflow, a handling process is designed for each type of event. Multiple processes can be designed, and key handling nodes within each process are configured with general handling nodes based on task monitoring. Key parameter variables are configured for the monitoring tasks, and the person or department responsible for handling the event can be dynamically modified based on these parameter values. When storing the configured processes, additional scoring quantification data is configured to save the score value of each handling node, which is used to calculate the overall score of the process based on a certain algorithm.

[0043] Step 3: Automatic event allocation using a large model: For events requiring coordinated handling, the historical handling of similar events is analyzed based on the large model. The Pareto solution is found based on the NSGA-II multi-objective optimization algorithm, and corresponding labels and scores are assigned based on different levels. The results are updated periodically. The coordinated events are automatically classified based on the event knowledge graph. All handling processes are found according to the classification, and the process with the highest comprehensive score is automatically matched and allocated.

[0044] This process involves analyzing the historical event handling process using a large model. For each node's event handling, the NSGA-II multi-objective optimization algorithm is used to comprehensively consider handling time, completion rate, and return rate to obtain the Pareto solution, which represents the optimal handling node. These nodes are assigned a score of 1, and the scores of the next lower level are adjusted downwards, and so on, with all nodes marked with scores and updated periodically. Based on the node scores, each process is comprehensively analyzed to select the optimal nodes. An attempt is made to automatically adjust parameters to match nodes to the actual process nodes, achieving the goal of automatically adjusting the process.

[0045] In addition, after a collaborative event arrives at the event handling system, it first automatically classifies the event based on the event knowledge graph. For the classified events, it selects all relevant processes from the configured processes, and then calculates the total score based on the score of each node in each process and a certain weight. The process with the highest score is then automatically selected for automatic allocation.

[0046] Step 4: Intelligent Analysis of Massive Events: Analyze massive amounts of event data using a large-scale model to provide multi-dimensional analytical data. Leveraging the analytical capabilities of the large-scale model, intelligently analyze the handling process, statistically analyzing the handling time and completion rate of each participating department during the event handling process, providing statistical data. This involves aggregating and handling massive amounts of event data. Based on the advantages of the large-scale model in data analysis, multi-dimensional intelligent analysis is performed on this data. For example, it can analyze hot events based on dimensions such as time, region, department, and system; predict future event trends based on historical event trends; and assist users in proactively responding. Intelligent analysis based on event handling can provide performance data, such as the event completion rate, on-time completion rate, and completion time for each department. It can also provide intelligent reminders and displays for difficult-to-solve events. Based on the large-scale model, intelligent real-time personalized data analysis can also be performed.

[0047] Example 2

[0048] This invention also provides an event handling system based on a large model and flowability, including a graph generation module, a process adjustment module, a distribution module, and an analysis module.

[0049] The knowledge graph generation module utilizes a large model to generate an event knowledge graph: it analyzes existing events using a large model to generate structured event data, extracts key descriptive information, location, and source of events from the structured event data, and associates these with standard event types to generate an event knowledge graph.

[0050] The process adjustment module automatically adjusts the event handling process: it uses flowable to configure the handling process for various types of events, reserves common handling nodes during the handling process, and dynamically adjusts the handling personnel or departments based on conditions.

[0051] The allocation module utilizes a large model for automatic event allocation: For events requiring coordinated handling, it analyzes historical handling of similar events based on the large model, uses the NSGA-II multi-objective optimization algorithm to find Pareto solutions, and assigns corresponding labels and scores based on different levels, updating periodically. Based on an event knowledge graph, it automatically classifies coordinated events, finds all handling processes according to the classification, and automatically matches the process with the highest comprehensive score for automatic allocation.

[0052] The analysis module performs intelligent analysis of massive events: it analyzes massive event data through large models to provide multi-dimensional analysis data; it uses the analytical capabilities of large models to intelligently analyze the handling process, and provides statistical data on the handling time and completion rate of each participating department during the event handling process.

[0053] The information interaction and execution process between the modules in the above system are based on the same concept as the method embodiment of the present invention, and the specific details can be found in the description in the method embodiment of the present invention, and will not be repeated here.

[0054] Similarly, the advantages of the system of the present invention are:

[0055] (1) Intelligent event classification: Based on the generated event knowledge graph, events are automatically classified, which greatly reduces the burden on grassroots staff, requires less manual intervention, and improves efficiency and accuracy.

[0056] (2) Automatic adjustment of incident handling process: By monitoring and analyzing the incident handling process, the process parameters are continuously optimized and adjusted based on the actual incident handling process results, and the nodes in the handling are automatically adjusted, which improves the efficiency and handling rate of incident handling, and reduces backtracking and handling time.

[0057] (3) Automatic allocation of collaborative events: After a collaborative event arrives, the event is automatically classified through the event knowledge graph. The optimal handling process is automatically selected from the candidate handling processes, eliminating the intermediate manual selection step, greatly reducing the burden on grassroots staff, and improving the accuracy of selection, so that the event can be handled better.

[0058] (4) Intelligent analysis of massive events: Based on the big model technology, it performs intelligent analysis on massive event data and event handling process, providing multi-dimensional event analysis results. It can be used for statistical display, intelligent predictive analysis, departmental assessment, difficult event analysis and other scenarios, and can perform real-time personalized data analysis.

[0059] It should be noted that not all steps and modules in the above processes and system structures are mandatory; some steps or modules can be omitted as needed. The execution order of the steps is not fixed and can be adjusted as required. The system structures described in the above embodiments can be physical or logical structures. That is, 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 certain components in multiple independent devices.

[0060] The above-described embodiments are merely preferred embodiments provided to fully illustrate the present invention, and the scope of protection of the present invention is not limited thereto. Equivalent substitutions or modifications made by those skilled in the art based on the present invention are all within the scope of protection of the present invention. The scope of protection of the present invention is defined by the claims.

Claims

1. An event handling method based on a large model and flowable principles, characterized by: include: Step 1: Generate an event knowledge graph using a large model: Analyze existing events using a large model to generate structured event data. Extract key descriptive information, location, and source of the events from the structured event data, and associate them with standard event types to generate an event knowledge graph. Step 2: Automatically Adjust Event Handling Processes: Configure handling processes for various event types using Flowable. Reserve common handling nodes during the handling process and dynamically adjust the handling personnel or departments based on conditions. When configuring handling processes for various event types using Flowable, design a handling process for each type of event, and configure common handling nodes based on task monitoring for key handling nodes in each process. This is used to monitor key parameter variables in the task configuration and dynamically modify the handling personnel or departments based on these key parameter variables. When storing the configured handling processes, additional scoring quantification data is configured to save the score value of each common handling node for subsequent calculation of the overall process score. Step 3: Automatic event allocation using a large model: For events requiring coordinated handling, the historical handling of similar events is analyzed based on the large model. The Pareto solution is found based on the NSGA-II multi-objective optimization algorithm, and corresponding labels and scores are assigned based on different levels. The results are updated periodically. The coordinated events are automatically classified based on the event knowledge graph. All handling processes are found according to the classification, and the process with the highest comprehensive score is automatically matched and allocated. Step 4: Perform intelligent analysis of massive events: Analyze massive event data through large models to provide multi-dimensional analytical data; By leveraging the analytical capabilities of large-scale models, intelligent analysis of the handling process is conducted, and statistical data is provided on the handling time and completion rate of each participating department during the incident handling process.

2. The event handling method based on a large model and flowable as described in claim 1, characterized in that: Step 1 includes: for events with a clear event category, creating a simple mapping relationship based on the original category and source; for events without a clear event category, extracting key metadata, performing natural language segmentation, creating event relationships, and adding them to the event knowledge graph. As the aggregated events accumulate, the event knowledge graph is automatically improved, increasing the matching success rate and accuracy.

3. The event handling method based on a large model and flowable as described in claim 1, characterized in that: In step 3, based on the historical handling analysis of similar events using a large model, the NSGA-II multi-objective optimization algorithm is used to comprehensively consider the handling time, completion rate, and return rate for each node's event handling to obtain the Pareto solution, which is the optimal handling node. The score of the optimal handling node is set to 1, and the scores of the next lower level are adjusted downwards, and so on, to mark all nodes with scores and update them periodically. Based on the scores of the handling nodes, a comprehensive analysis of each process is performed.

4. An event handling system based on a large model and flowability, characterized by: It includes a map generation module, a process adjustment module, a distribution module, and an analysis module. The knowledge graph generation module utilizes a large model to generate an event knowledge graph: it analyzes existing events using a large model to generate structured event data, extracts key descriptive information, location, and source of events from the structured event data, and associates these with standard event types to generate an event knowledge graph. The process adjustment module automatically adjusts the event handling process: it uses flowable to configure the handling process for various types of events, reserves common handling nodes in the handling process, and dynamically adjusts the handling personnel or departments according to conditions. When the process adjustment module uses flowable to configure the handling process for various types of events, it designs a handling process for each type of event, and configures common handling nodes based on task monitoring in key handling nodes in each process. This is used to monitor key parameter variables of the task configuration and dynamically modify the handling personnel or departments based on the key parameter variables. When storing the configured handling process, the process adjustment module additionally configures scoring quantification data and saves the score value of each general handling node for subsequent calculation of the overall score of the process. The allocation module utilizes a large model for automatic event allocation: For events requiring coordinated handling, it analyzes historical handling of similar events based on the large model, uses the NSGA-II multi-objective optimization algorithm to find Pareto solutions, and assigns corresponding labels and scores based on different levels, updating periodically. Based on an event knowledge graph, it automatically classifies coordinated events, finds all handling processes according to the classification, and automatically matches the process with the highest comprehensive score for automatic allocation. The analysis module performs intelligent analysis of massive events: it analyzes massive event data through large models to provide multi-dimensional analysis data; it uses the analytical capabilities of large models to intelligently analyze the handling process, and provides statistical data on the handling time and completion rate of each participating department during the event handling process.

5. The event handling system based on a large model and flowability according to claim 4, characterized in that: For events with clear event categories, the event knowledge graph generation module creates simple mapping relationships based on the original category and source. For events without clear event categories, it extracts key metadata, performs natural language segmentation, creates event relationships, and adds them to the event knowledge graph. As the aggregated events accumulate, the event knowledge graph is automatically improved, increasing the matching success rate and accuracy.

6. The event handling system based on a large model and flowability according to claim 4, characterized in that: The analysis module analyzes the historical handling of similar events based on a large model. For the event handling at each node, the NSGA-II multi-objective optimization algorithm is used to comprehensively consider the handling time, completion rate, and return rate to obtain the Pareto solution, which is the optimal handling node. The score of the optimal handling node is set to 1, and the scores of the next lower level are adjusted down, and so on. All nodes are marked with scores and updated periodically. Based on the scores of the handling nodes, a comprehensive analysis of each process is performed.

Images