Large social governance risk early warning model

Abstract

The invention discloses a social governance risk early warning large model, which comprises a data acquisition unit, an event comparison unit, an early warning analysis table generation unit, an early warning research and judgment unit, an early warning disposal unit and an interaction unit, and is characterized in that the data acquisition unit is used for generating a detailed data set of key personnel; the event comparison unit is used for generating a comparison event data set; the early warning analysis table generation unit is used for generating a key personnel comparison early warning analysis table; the early warning research and judgment unit is used for constructing an early warning research and judgment agent; the early warning processing unit is used for building an early warning processing agent; and the interaction unit is used for displaying the key personnel risk research and judgment situation overview map and the risk disposal strategy in real time, and building an intelligent interaction question and answer window through the early warning research and judgment agent and the early warning disposal agent. According to the invention, by introducing big data analysis and artificial intelligence technologies, intelligent early warning, research and judgment and disposal management of key personnel are realized, and the efficiency and level of social governance are improved.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a large-scale model for early warning of social governance risks. Background Technology

[0002] With rapid societal development, public safety and social governance face increasing challenges. In particular, traditional early warning and response methods are no longer sufficient to meet the demands for efficiency and precision in managing certain special groups (such as individuals with severe mental disorders or those involved in litigation).

[0003] Therefore, utilizing big data analytics and artificial intelligence (AI) technologies for intelligent early warning and assessment of key personnel has become an important means of improving social governance efficiency. Currently, some social governance systems based on big data and AI technologies exist. These systems typically collect and analyze data from multiple departments to conduct risk assessments and early warnings for specific groups. However, these systems still have shortcomings in handling complex analytical tasks, generating detailed risk assessment reports, and implementing intelligent question-and-answer and response management. For example, some systems can only provide simple risk level classifications and cannot deeply analyze personnel behavioral characteristics and risk trends; some systems lack intelligent question-and-answer functions and cannot respond promptly to the query needs of management personnel; and some systems are inefficient in response management and cannot quickly and effectively handle early warning events. Summary of the Invention

[0004] The purpose of this invention is to provide a large-scale social governance risk early warning model in order to solve the above-mentioned problems.

[0005] The present invention achieves the above objectives through the following technical solutions:

[0006] A large-scale social governance risk early warning model includes a data acquisition unit. This unit collects and aggregates data from personnel and key personnel reported by departments, generating a detailed dataset of key personnel. It acts as the system's "information collector," integrating key personnel information from different channels to provide a comprehensive data foundation for subsequent analysis.

[0007] The event comparison unit is used to collect and summarize event comparison data from various departments and form a comparison event dataset. This unit focuses on the integration of various event information, which serves as an important reference for analyzing the risks of key personnel.

[0008] The early warning analysis table generation unit, based on the detailed dataset of key personnel generated by the data acquisition unit and the dataset of comparison events generated by the event comparison unit, associates the key personnel with the feature points of the comparison events to generate a key personnel comparison early warning analysis table. It acts as a "linkage bridge," providing crucial clues for risk analysis by establishing connections between key personnel and events.

[0009] The early warning and analysis unit, based on the generated key personnel comparison and early warning analysis table, conducts in-depth analysis of the trends of key personnel appearing in compared events, generating a key personnel early warning analysis report and key personnel early warning analysis details. Simultaneously, it constructs a key personnel early warning knowledge base based on the key personnel comparison and early warning analysis table, key personnel early warning analysis report, and key personnel early warning analysis details, and builds an early warning and analysis intelligent agent in conjunction with an AI big data model. This unit is one of the system's "intelligent brains," providing intelligent support for risk early warning through data analysis and knowledge construction.

[0010] The early warning and response unit analyzes the risk level of key personnel based on the early warning analysis report, determines the risk response strategy for key personnel, and generates a risk assessment and assignment form for key personnel. Based on the risk assessment and assignment form for key personnel and combined with the AI ​​big data model, it constructs an intelligent early warning and response agent. This unit is responsible for translating the early warning results into specific response actions and is a key link in risk prevention and control.

[0011] The interactive unit, based on the early warning assessment and early warning processing intelligent agents formed by the early warning assessment and processing units, displays a real-time overview of the risk assessment situation for key personnel and risk management strategies. It also establishes an intelligent interactive Q&A window through these two intelligent agents. This provides users (such as social governance personnel) with an intuitive and convenient interactive interface, facilitating their access to information and communication.

[0012] Preferably, the collected personnel data and the key personnel data reported by departments both include name, ID number, gender, date of birth, mobile phone number, special population information, and risk level information. These rich information dimensions help to comprehensively characterize the features of key personnel and provide multi-faceted data support for risk assessment. For example, special population information can help determine whether a key personnel belongs to a specific sensitive group, while risk level information directly reflects their potential risk level.

[0013] Preferably, the event data for comparison includes events involving large affected populations, events attracting public attention, events with high media exposure, and events with serious social consequences. These event types reflect different attributes and impacts of events, illustrating potential risk scenarios in social governance from various perspectives. For example, events involving large affected populations can help assess the scope of an event's impact, events attracting public attention and those with high media exposure reflect the event's influence on society, and events with serious social consequences directly indicate the severity of the event.

[0014] Preferably, the key personnel and comparison event feature points include those using the personnel's ID number or contact number as unique feature points. The ID number and contact number are unique and identifiable, enabling accurate association between key personnel and individuals in relevant events, ensuring the accuracy and reliability of data analysis. For example, the ID number can be used to accurately match the activity trajectories and related information of key personnel in different events.

[0015] Preferably, the key personnel early warning analysis report includes daily, weekly, and monthly reports. Reports at different time periods can meet different decision-making needs, facilitating relevant departments to promptly understand short-term changes and long-term trends in the risks associated with key personnel. For example, daily reports can be used to promptly grasp daily risk dynamics, while monthly reports help with monthly summaries and the formulation of long-term strategies.

[0016] Preferably, the key personnel early warning analysis details include personnel profiles, behavioral analysis, and risk warning levels. The personnel profiles depict the characteristics of key personnel from multiple dimensions, the behavioral analysis delves into their behavioral patterns, and the risk warning levels clearly define their risk severity, providing detailed and specific information for risk assessment and handling. For example, behavioral analysis can identify abnormal behavioral patterns of key personnel, and combined with the personnel profiles and risk warning levels, it provides a basis for developing targeted handling strategies.

[0017] Preferably, the key personnel risk assessment and assignment form includes recommendations for handling key personnel risk assessments, which are categorized into three levels: general risk (advisory and educational talks), medium risk (forced removal), and high risk (tracking and mandatory isolation). This differentiated approach based on risk level ensures more scientific, reasonable, and effective measures, guaranteeing appropriate responses under different risk conditions.

[0018] Preferably, the early warning processing unit further includes issuing task assignment orders for risk assessment of key personnel, and dynamically updating the early warning processing unit based on the progress of risk handling tasks reported by the issuing unit, generating an early warning processing progress report. This function enables real-time tracking and feedback of the risk handling process, which helps to adjust the handling strategy in a timely manner and ensure that the risk is effectively controlled. For example, when it is found that the progress of a risk handling task is slow, the relevant unit can be urged to speed up the progress or adjust the handling method in a timely manner.

[0019] A working method for a large-scale social governance risk early warning model includes the following steps:

[0020] S1. Upload detailed information of key personnel through two methods: data collection and reporting, and data collection and summarization of key personnel data from both the collected personnel and the departments, and generate a detailed dataset of key personnel.

[0021] S2. Simultaneously, based on the event data reported by each department, the comparative event data from each department is collected and summarized to form a comparative event dataset. These two steps are the foundation of the entire method, ensuring that comprehensive and accurate data is available for subsequent analysis.

[0022] S3. Based on the detailed dataset of key personnel and the dataset of comparison events, associate the key personnel with the personnel involved in the comparison events using their ID numbers or contact information as unique features, and generate a key personnel comparison early warning analysis table. This step establishes the connection between key personnel and events, providing crucial evidence for risk analysis;

[0023] S4. Based on the generated key personnel comparison and early warning analysis table, analyze the trends of key personnel in comparison events, generate a key personnel early warning analysis report and key personnel early warning analysis details, and construct a key personnel early warning knowledge base based on relevant information. Combine this with an AI big data model to build an early warning judgment intelligent agent. Through this step, data analysis and intelligent methods are used to deeply explore potential risks.

[0024] S5. Based on the key personnel early warning analysis report, analyze the risk level of key personnel, determine the risk handling strategy for key personnel, and generate a key personnel risk assessment and assignment form. Combine this with an AI big data model to build an intelligent early warning and handling agent. This step transforms the early warning results into specific handling actions to achieve risk prevention and control.

[0025] S6. Based on the early warning assessment and early warning processing units, the early warning assessment and early warning processing intelligent agents are formed, displaying a real-time overview of the risk assessment situation for key personnel and risk processing strategies. An intelligent interactive Q&A window is also established through these two intelligent agents. This step provides users with an intuitive and convenient interactive interface, facilitating information acquisition and communication.

[0026] Preferably, in steps S4 and S5, the AI ​​large-scale model adopts either the ERNIE 3.5-8K non-inference large-scale model or the DeepSeek-R1 full-fledged inference large-scale model. These advanced large-scale models have powerful data analysis and processing capabilities, providing strong support for the system's intelligent analysis and decision-making. For example, the ERNIE 3.5-8K non-inference large-scale model performs well in natural language processing and knowledge understanding, while the DeepSeek-R1 full-fledged inference large-scale model has advantages in inference and prediction. The appropriate model can be selected according to specific needs to improve system performance.

[0027] The beneficial effects of this invention are as follows:

[0028] 1. This invention integrates multi-source data through a data acquisition unit and an event comparison unit, enabling a comprehensive grasp of information on key personnel and related events; based on this, correlation analysis and in-depth judgment improve the accuracy and comprehensiveness of risk warnings, avoiding risk omissions caused by data dispersion and simplistic analysis;

[0029] 2. This invention utilizes large-scale AI models to construct intelligent agents for early warning analysis and early warning response, thereby enhancing the system's intelligence level. These intelligent agents can quickly and accurately analyze large amounts of data, identify potential risk trends, provide a scientific basis for risk assessment and response, and improve the efficiency and scientific rigor of social governance.

[0030] 3. This invention formulates differentiated risk management strategies based on the risk levels of key personnel, making the management measures more scientific and reasonable, effectively addressing risks of different degrees, improving the effectiveness of risk prevention and control, and reducing social governance costs. The real-time tracking and feedback of the risk management task progress by the early warning processing unit helps to adjust the management strategy in a timely manner, ensuring that risks are effectively controlled and enhancing the timeliness and effectiveness of risk management. Moreover, the real-time display and intelligent interactive question-and-answer window provided by the interactive unit facilitates information access and communication for social governance staff, improves work efficiency, and promotes collaboration among departments.

[0031] 4. The overall system of this invention adopts multi-model collaboration and uses big data analysis and artificial intelligence technology to deeply mine and analyze the detailed information of key personnel; by deploying non-reasoning and reasoning large models, it handles routine intent recognition tasks and complex analysis tasks respectively, ensuring the comprehensiveness and depth of risk assessment; compared with traditional systems that can only provide simple risk level classification, this invention can more accurately identify and assess the risks of key personnel, providing a more reliable basis for subsequent handling and management.

[0032] 5. This invention enables real-time intelligent question and answer, improving work efficiency. The system introduces intelligent question and answer functionality, providing question and answer capabilities based on the original large model and achieving streaming output. This means that managers can query the information they need through the system at any time without waiting for manual replies or reviewing a large number of paper documents. The real-time intelligent question and answer function greatly improves work efficiency, enabling managers to make decisions and respond more quickly, thereby improving the overall social governance effectiveness.

[0033] 6. This invention optimizes the handling and management process and improves response speed. The system realizes rapid response and effective handling of early warning events through an intelligent handling and management system, including functions such as risk handling task issuance and risk handling task progress tracking, to ensure that early warning events are handled in a timely manner. Compared with the cumbersome handling process and untimely response of traditional systems, this invention can initiate the handling procedure more quickly, reduce the occurrence and expansion of potential risks, thereby improving the response speed and effectiveness of social governance.

[0034] 7. This invention provides a visual display to enhance decision support. The system displays an overview of the risk assessment situation of key personnel, a warning list, and risk assessment profiles through a large-scale visual screen for city governance. This information is presented to managers in an intuitive and comprehensive way, making it easy for them to quickly understand the current risk situation and the progress of handling. The visual display function provides managers with more intuitive and comprehensive decision support, enabling them to grasp the dynamics of risks more accurately and make more scientific decisions.

[0035] 8. This invention improves data governance efficiency and ensures data quality. The system obtains detailed information on key personnel through two methods: data collection and reporting, and data reporting from various departments. It then performs data governance and comparative analysis. Through an efficient data governance process, the system can ensure the accuracy and integrity of the data, providing reliable data support for subsequent risk assessment and management. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 This is the system architecture diagram of the present invention.

[0038] Figure 2 This is a schematic diagram of the working principle of the system of this invention.

[0039] Figure 3 This is a diagram of the multi-model collaborative working system architecture in this invention.

[0040] Figure 4 This is a flowchart of the workflow of the present invention. Detailed Implementation

[0041] The following is in conjunction with the appendix Figure 1-4 The technical solution of the present invention will be further explained below:

[0042] Example 1

[0043] like Figure 1-4 As shown, the social governance risk early warning model employs advanced technologies such as the Qianfan all-in-one machine, Docker operating system, and Model Builder to construct its overall hardware architecture. One Qianfan all-in-one machine with 8 cards deploys a non-inference-type large model (such as ERNIE 3.5-8K) for handling routine intent recognition tasks; another Qianfan all-in-one machine with 8 cards deploys an inference-type large model (such as DeepSeek R1 full-fledged version) for handling complex analysis tasks. In addition, the system also includes a city-wide big data platform, a key personnel risk assessment intelligent agent, and a business data governance module.

[0044] Specifically, such as Figure 2As shown, the data acquisition unit and the event comparison unit are respectively used to collect and summarize data from personnel and key personnel reported by departments, generating a detailed dataset of key personnel and to collect and summarize comparison event data from various departments, forming a comparison event dataset. In other words, personnel data can be obtained through on-site visits and questionnaires by grassroots staff. For example, community workers regularly visit key personnel within their jurisdiction to record their latest living conditions and behavioral dynamics. Data on key personnel reported by departments is compiled and reported by relevant departments, such as civil affairs departments, according to a unified data format and standards. To ensure the accuracy and completeness of the data, a data review mechanism can be established to strictly review the reported data, such as checking whether the required fields are complete and whether the data format is correct.

[0045] It is important to note that the collected data should be stored in a dedicated database, using either a relational database such as MySQL or a non-relational database such as MongoDB, depending on the characteristics and requirements of the data. The database should possess strong data security and scalability, with strict access permissions set so that only authorized personnel can access and modify the data. Regular backups should be performed to prevent data loss. The data should be categorized and managed according to the different attributes and sources of key personnel to facilitate subsequent queries and retrieval.

[0046] Furthermore, each department collects and compares event data through its own business systems and monitoring methods. For example, one department collects data on serious social incidents through a case management system, the publicity department collects data on events of public concern and media exposure through a monitoring system, and the civil affairs department collects data on the scale of affected populations through a relief management system. To ensure the timeliness and consistency of the data, a data sharing platform is established, whereby each department uploads the collected event data to the sharing platform in real time, achieving centralized management and sharing of the data.

[0047] The collected personnel data and key personnel data reported by departments include name, ID number, gender, date of birth, mobile phone number, special groups, and risk level information. The compared event data includes events affecting large populations, events of public concern, events with media exposure, and serious social incidents. The collected event data may have issues such as inconsistent formats and data duplication, requiring preprocessing. First, the data is cleaned to remove duplicates and invalid data. Then, data of different formats is converted to a format that the system can recognize and process. For example, the event time format recorded by different departments is standardized to a standard date and time format. Simultaneously, the event data undergoes standardization, normalizing indicators of different magnitudes to facilitate subsequent analysis and comparison, handling minor input errors, and ensuring the accuracy of correlations.

[0048] Specifically, such as Figure 2 As shown, the early warning analysis table generation unit is used to associate key personnel with the feature points of the comparison events based on the detailed dataset of key personnel generated by the data acquisition unit and the comparison event dataset generated by the event comparison unit, thus generating a key personnel comparison early warning analysis table. In other words, after successful association, the relevant information is organized to generate the key personnel comparison early warning analysis table. The table's columns can include basic information about the key personnel (name, gender, ID number, etc.), relevant information about the comparison events (event type, occurrence time, impact level, etc.), and the unique associated feature point (ID number or contact number). To facilitate viewing and analysis, the table is sorted and grouped, for example, sorted by event occurrence time or grouped by the risk level of the key personnel. Simultaneously, the data in the table is visualized, such as generating bar charts and line charts, to intuitively display the correlation and trends between key personnel and comparison events.

[0049] like Figure 2 As shown, the early warning and assessment unit is used to analyze the trends of key personnel in compared events based on the generated key personnel comparison and early warning analysis table. In other words, it uses data analysis algorithms, such as time series analysis and cluster analysis, to analyze the trends of key personnel in compared events. For example, time series analysis observes the frequency changes of key personnel's involvement in various events over different time periods to determine whether their risk trend is rising or falling. Cluster analysis is used to classify key personnel according to behavioral patterns and risk characteristics for more targeted risk assessment. Combined with machine learning algorithms, such as decision trees and random forests, the unit learns from the historical data and current event data of key personnel to predict their probability of appearing in compared events and the degree of risk in the future.

[0050] This system generates key personnel early warning analysis reports, which are produced daily, weekly, and monthly based on trend analysis results. The reports include an overall risk profile of key personnel, the distribution of personnel at different risk levels, and changes in risk trends. For example, the daily report mainly presents the various events involving key personnel that day and the changes in risk; the weekly report summarizes and analyzes the risks of key personnel within a week; and the monthly report assesses the risk situation of key personnel from a more macro perspective. The reports are presented in a combination of charts and text, facilitating quick understanding of key information by relevant personnel.

[0051] The system includes detailed early warning analysis of key personnel. This is achieved by integrating multi-source data on key personnel and utilizing data mining techniques to construct personnel profiles. These profiles cover the basic attributes, behavioral habits, and social relationships of key personnel, such as analyzing their daily activity trajectories and frequency of contact with other individuals. In-depth analysis of the behavioral data of key personnel during comparative events uncovers their behavioral patterns and regularities, generating behavioral analysis reports. Combining trend analysis results with personnel profiles and behavioral analysis, the system determines the risk warning level for key personnel, detailing the sources of risk and their potential impact.

[0052] Specifically, such as Figure 2 and Figure 4 As shown, a key personnel early warning knowledge base is constructed based on the key personnel comparison and early warning analysis table, the key personnel early warning analysis report, and the key personnel early warning analysis details. Then, an early warning judgment intelligent agent is built based on this knowledge base and a large AI model. In other words, key information is extracted and organized from the key personnel comparison and early warning analysis table, the key personnel early warning analysis report, and the key personnel early warning analysis details to construct the key personnel early warning knowledge base. The knowledge base is stored and managed in the form of a knowledge graph, facilitating knowledge retrieval and reasoning. Combined with a large AI model, such as the selected ERNIE3.5-8K non-reasoning large model or the DeepSeek-R1 full-fledged reasoning large model, the knowledge in the knowledge base is learned and understood to build the early warning judgment intelligent agent. The intelligent agent can quickly perform risk assessment and trend prediction based on the input key personnel and event information, and provide corresponding suggestions and decision support.

[0053] like Figure 2-4 As shown, the early warning and response unit analyzes the risk level of key personnel based on the early warning analysis report. That is, based on the information in the early warning analysis report, the unit comprehensively considers the key personnel's level of participation in the compared event, the severity of the event, and risk trend factors, and uses the analytic hierarchy process (AHP) and fuzzy comprehensive evaluation method to determine the risk level of the key personnel. Specifically, the AHP determines the weights of different factors, and then the fuzzy comprehensive evaluation method is used to quantitatively assess the risk of key personnel, resulting in general, medium, and high risk levels.

[0054] Then, risk management strategies for key personnel are determined, and corresponding strategies are formulated based on the risk level. For general risks, professional personnel are assigned to conduct persuasive and educational talks with key personnel, including disseminating legal and regulatory information and guiding behavioral norms. The talks and the feedback from key personnel are recorded for follow-up. For medium risks, when taking compulsory removal measures, the legality and safety of the action are ensured, and a detailed removal plan is developed in advance, including the action time, personnel arrangements, and emergency measures. For high risks, compulsory isolation and tracking are implemented. A dedicated tracking team is established, and location technology is used to monitor the location information of key personnel in real time. Isolation measures are strictly implemented in accordance with relevant laws, regulations, and procedures to ensure that the risk is effectively controlled.

[0055] like Figure 3 and Figure 4 As shown, a key personnel risk assessment assignment form is generated simultaneously. Based on this form, an AI-powered early warning and response intelligent agent is constructed using a large-scale model. In other words, a key personnel risk assessment assignment form is generated according to the risk response strategy, clearly defining the content of the assigned task, the implementing unit, and the completion time. The system distributes the assignment tasks to the corresponding implementing units, who then promptly report the task receipt status. A task tracking mechanism is established to monitor task progress in real time, requiring implementing units to regularly report on the progress of risk handling tasks, such as measures taken and changes in the status of key personnel. The early warning and response unit is dynamically updated based on the progress feedback from implementing units. The progress of the tasks is analyzed and evaluated, such as determining whether the tasks are proceeding as planned and whether the expected results have been achieved. An early warning and response progress report is generated, recording the latest progress, problems encountered, and solutions, providing a reference for subsequent risk response decisions. If task progress is found to be unsatisfactory or the risk situation changes, the risk response strategy and assigned tasks are adjusted promptly to ensure effective risk management.

[0056] like Figure 2 As shown, the interactive unit is used to display a real-time overview of the risk assessment situation for key personnel and risk management strategies based on the early warning assessment and early warning processing intelligent agents formed by the early warning assessment and processing units. It also establishes an intelligent interactive Q&A window through these intelligent agents. In other words, based on the results of the early warning assessment and processing units, data visualization technology is used to create an overview of the risk assessment situation for key personnel. This overview map visually displays the risk distribution of key personnel in an intuitive graphical way; for example, different colors and sizes of icons represent the distribution of key personnel with different risk levels in different areas. It also displays the implementation status of risk management strategies, such as the measures already taken and the progress of ongoing tasks. Dynamic charts show changes in risk trends, such as increases or decreases in risk levels and the number of events involved, enabling relevant personnel to quickly understand the overall risk situation.

[0057] Furthermore, an intelligent interactive question-and-answer window is built based on the early warning and judgment intelligent agent and the early warning and response intelligent agent. Natural language processing technology is employed to enable the intelligent agent to understand the user's input questions and provide accurate answers based on knowledge in the knowledge base and analysis results. For example, a user can inquire about the risk situation of a key personnel and corresponding handling strategies, and the intelligent agent can quickly provide a detailed answer. Common questions are preset and optimized to improve the accuracy and efficiency of the answers. Simultaneously, user questions and the intelligent agent's answers are recorded to continuously improve the knowledge base and the intelligent agent's answer logic, enhancing the interactive experience.

[0058] It's also important to select the appropriate AI model based on the specific needs of the system and the characteristics of the data. If the system prioritizes understanding textual information and knowledge reasoning, such as analyzing textual materials related to key personnel, the ERNIE3.5-8K non-inference model is a suitable choice, as it possesses strong capabilities in natural language processing. If the system requires rapid prediction of risk trends and real-time decision support, the DeepSeek-R1 full-fledged inference model may be more suitable due to its excellent performance in inference and prediction tasks. Before selecting a model, performance tests should be conducted on both models using real-world data for simulation analysis. Compare their accuracy, efficiency, and other metrics in risk assessment and decision-making, and make a more reasonable choice based on the test results.

[0059] Regardless of the model chosen, targeted training using key personnel data and comparative event data from the system is essential. Data preprocessing, including data cleaning and labeling, is crucial to ensure the data meets the model's input requirements. During training, model parameters, such as learning rate and number of iterations, are adjusted to improve performance. Cross-validation and other methods are employed to evaluate the model's accuracy and generalization ability, preventing overfitting or underfitting. Regular model updates and optimizations are necessary; as new data accumulates, the model is retrained to adapt to the dynamic changes in social governance risks and maintain good performance.

[0060] Example 2

[0061] A method for early warning of social governance risks includes the following steps:

[0062] S1. Upload detailed information of key personnel through two methods: data collection and reporting, and data collection and summarization of key personnel data from both the collected personnel and the departments, and generate a detailed dataset of key personnel.

[0063] S2. Simultaneously, based on the event data reported by each department, the comparative event data from each department is collected and summarized to form a comparative event dataset. These two steps are the foundation of the entire method, ensuring that comprehensive and accurate data is available for subsequent analysis.

[0064] S3. Based on the detailed dataset of key personnel and the dataset of comparison events, associate the key personnel with the personnel involved in the comparison events using their ID numbers or contact information as unique features, and generate a key personnel comparison early warning analysis table. This step establishes the connection between key personnel and events, providing crucial evidence for risk analysis;

[0065] S4. Based on the generated key personnel comparison and early warning analysis table, analyze the trends of key personnel in comparison events, generate a key personnel early warning analysis report and key personnel early warning analysis details, and construct a key personnel early warning knowledge base based on relevant information. Combine this with an AI big data model to build an early warning judgment intelligent agent. Through this step, data analysis and intelligent methods are used to deeply explore potential risks.

[0066] S5. Based on the key personnel early warning analysis report, analyze the risk level of key personnel, determine the risk handling strategy for key personnel, and generate a key personnel risk assessment and assignment form. Combine this with an AI big data model to build an intelligent early warning and handling agent. This step transforms the early warning results into specific handling actions to achieve risk prevention and control.

[0067] S6. Based on the early warning assessment and early warning processing units, the early warning assessment and early warning processing intelligent agents are formed, displaying a real-time overview of the risk assessment situation for key personnel and risk processing strategies. An intelligent interactive Q&A window is also established through these two intelligent agents. This step provides users with an intuitive and convenient interactive interface, facilitating information acquisition and communication.

[0068] Preferably, in steps S4 and S5, the AI ​​large-scale model adopts either the ERNIE 3.5-8K non-inference large-scale model or the DeepSeek-R1 full-fledged inference large-scale model. These advanced large-scale models have powerful data analysis and processing capabilities, providing strong support for the system's intelligent analysis and decision-making. For example, the ERNIE 3.5-8K non-inference large-scale model performs well in natural language processing and knowledge understanding, while the DeepSeek-R1 full-fledged inference large-scale model has advantages in inference and prediction. The appropriate model can be selected according to specific needs to improve system performance.

[0069] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention.

Claims

1. A large-scale social governance risk early warning model, characterized in that: include: The data acquisition unit is used to collect and summarize data from personnel and key personnel reported by departments, and to generate detailed datasets of key personnel. The event comparison unit is used to collect and summarize the event comparison data from various departments and form a comparison event dataset; The early warning analysis table generation unit, based on the detailed dataset of key personnel formed by the data acquisition unit and the dataset of comparison events formed by the event comparison unit, associates the key personnel with the feature points of the comparison events to generate a key personnel comparison early warning analysis table. The early warning and judgment unit analyzes the trends of key personnel in the comparison events based on the generated key personnel comparison and early warning analysis table, generates a key personnel early warning analysis report and key personnel early warning analysis details, and constructs a key personnel early warning knowledge base based on the key personnel comparison and early warning analysis table, key personnel early warning analysis report and key personnel early warning analysis details. Based on the key personnel early warning knowledge base and combined with the AI ​​big model, an early warning and judgment intelligent agent is constructed. The early warning and response unit analyzes the risk level of key personnel based on the early warning analysis report, determines the risk response strategy for key personnel, and generates a risk assessment and assignment form for key personnel. Based on the risk assessment and task assignment of key personnel, an intelligent early warning and response system is constructed using a large-scale AI model. The system also includes an interactive unit that, based on the early warning assessment and early warning processing units, displays a real-time overview of the risk assessment situation for key personnel and risk management strategies. Furthermore, it establishes an intelligent interactive Q&A window through the early warning assessment and early warning processing units.

2. The large-scale social governance risk early warning model as described in claim 1, characterized in that, The collected personnel data and the key personnel data reported by departments both include name, ID number, gender, date of birth, mobile phone number, special groups, and risk level information.

3. The large-scale social governance risk early warning model as described in claim 1, characterized in that, The comparison event data includes events affecting large numbers of people, events that have garnered public attention, events that have received media exposure, and events that have caused serious social harm.

4. The large-scale social governance risk early warning model as described in claim 1, characterized in that, The key personnel and comparison event feature points include those with personnel ID numbers or contact numbers as unique feature points.

5. The large-scale social governance risk early warning model as described in claim 1, characterized in that, The key personnel early warning analysis report includes daily, weekly, and monthly reports.

6. The large-scale social governance risk early warning model as described in claim 1, characterized in that, The details of the key personnel early warning analysis include personnel profiles, behavioral analysis, and risk warning levels.

7. The large-scale social governance risk early warning model as described in claim 1, characterized in that, The risk assessment and assignment form for key personnel includes recommendations for risk assessment and assignment for key personnel, which are divided into three categories: general risk, persuasion and education; medium risk, forced removal; and high risk, tracking and forced isolation.

8. The large-scale social governance risk early warning model as described in claim 1, characterized in that, The early warning processing unit also includes the issuance of task assignment orders for risk assessment of key personnel, and the dynamic updating of the early warning processing unit based on the progress of risk handling tasks reported by the task assignment unit for risk assessment of key personnel, generating an early warning processing progress report.

Images