Public security alarm condition analysis system, method and equipment based on large model, and medium

Abstract

The invention discloses a public security alarm condition analysis system, method and device based on a large model and a medium. The alarm condition analysis task planning module analyzes an alarm condition analysis request through a pre-trained alarm condition analysis large model in combination with the alarm condition analysis request, target alarm condition data and a task chain planning text prompt prompt, obtains a specific alarm condition analysis treatment task chain, makes full use of the intelligent analysis capability of the model, and improves the alarm condition analysis efficiency. A reasonable task chain is generated according to different alarm analysis requests, and task nodes and an execution sequence are defined. And the alarm condition analysis task execution module schedules the corresponding alarm condition analysis task execution sub-modules according to the execution sequence of the alarm condition analysis treatment task chain. Each alarm analysis task execution sub-module performs special processing by means of a large model, information in alarm data is deeply mined from multiple angles, deep understanding of alarm content is realized, potential risk points and valuable intelligence clues can be found, and missing of important clues in a traditional mode is avoided.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence, and particularly to a public security police situation analysis system, method, device, and medium based on a large model. Background Art

[0002] In today's society, with the rapid economic development and continuous changes in social structure, the frequency and complexity of various police situations are increasing day by day, which poses a severe challenge to the police situation handling ability of the public security department.

[0003] In the actual operation of current police situation handling, clue research and judgment analysis, as a key link in police situation handling work, its accuracy and efficiency directly affect the effect of police situation disposal. At present, when conducting clue research and judgment work, researchers mainly adopt two common methods. One is to manually set a keyword library for specific special events. Based on their own experience and preliminary understanding of the event, researchers screen out keywords that may be related to the case, and then search for these keywords in a vast amount of police situation information. The other method is to conduct preliminary statistical analysis with the help of natural language processing (NLP) technology. The NLP system can automatically process police situation texts and extract some potential key information and clues.

[0004] However, both of these methods have obvious defects. For the method of manually setting keywords, researchers need to spend a lot of time and effort to determine the keywords, and the accuracy of the keywords largely depends on the experience of the researchers and their understanding of the case. Once the keywords are set inaccurately, it may lead to deviations in search results and missed important clues. Even if the keywords are set relatively accurately, researchers still need to view the search results one by one and screen out valuable clues from them, which is extremely cumbersome and time-consuming.

[0005] For the automatic analysis method relying on the NLP system, although it improves the analysis efficiency to a certain extent, due to the limited analysis ability of the NLP system, it may not be able to accurately identify some deep-level intelligence associations. In addition, researchers also need to manually review and screen the analysis results of the NLP system, which undoubtedly increases the complexity and workload of the work. Moreover, whether it is manually setting keywords or relying on the automatic analysis of the NLP system, it is difficult to avoid the situation of missing some deep-level intelligence associations, thus causing the public security department to miss some important clues and affecting the effect of police situation handling.

[0006] In summary, the existing police situation handling methods have obvious deficiencies in terms of efficiency and accuracy. Therefore, it is necessary to explore a more efficient and accurate police situation handling method to cope with the increasingly complex police situation and continuously improving policing needs. Summary of the Invention

[0007] The present application provides a public security police situation analysis system, method, device and medium based on a large model, which is used to solve the problems of low efficiency and poor accuracy in the existing police situation handling methods.

[0008] In a first aspect, the present application provides a public security police situation analysis system based on a large model, and the system includes:

[0009] A receiving module, configured to receive a police situation analysis request submitted by a user for target police situation data;

[0010] A police situation analysis task planning module, configured to parse the police situation analysis request through a pre-trained police situation analysis large model based on the police situation analysis request, the target police situation data, and a task chain planning text prompt prompt, so as to obtain a specific police situation analysis governance task chain for the police situation analysis request; wherein, the specific police situation analysis governance task chain includes each task node and the execution order of each task node;

[0011] A police situation analysis task execution module, configured to process the target police situation data by scheduling corresponding police situation analysis task execution sub-modules according to the execution order; wherein, each police situation analysis task execution sub-module includes: a description extraction sub-module for extracting police situation description elements from the target police situation data, a clue extraction sub-module for extracting police situation clue elements from the target police situation data, and a graph construction sub-module for constructing a graph relationship between the police situation description elements and the police situation clue elements; for each police situation analysis task execution sub-module, the police situation analysis task execution sub-module obtains a special processing result of the target police situation data by the police situation analysis large model based on the target police situation data and a corresponding task execution prompt;

[0012] A storage module, configured to store the special processing results of each police situation analysis task execution sub-module;

[0013] An output module, configured to output the special processing results of each police situation analysis task execution sub-module.

[0014] In a second aspect, the present application further provides a public security police situation analysis method based on a large model, and the method includes:

[0015] Receiving a police situation analysis request submitted by a user for target police situation data;

[0016] Through a pre-trained large model for police situation analysis, based on the police situation analysis request, the target police situation data, and the task chain planning text prompt, the police situation analysis request is parsed to obtain a specific police situation analysis and governance task chain; wherein, the specific police situation analysis and governance task chain includes each task node and the execution order of each task node.

[0017] According to the execution order, schedule the corresponding police situation analysis task execution sub-module to process the target police situation data; wherein, each police situation analysis task execution sub-module includes: a description extraction sub-module for extracting police situation description elements from the target police situation data, a clue extraction sub-module for extracting police situation clue elements from the target police situation data, and a graph construction sub-module for constructing the graph relationship between the police situation description elements and the police situation clue elements; for each police situation analysis task execution sub-module, the police situation analysis task execution sub-module obtains a special processing result of the target police situation data by the police situation analysis large model based on the target police situation data and the corresponding task execution prompt.

[0018] Store and output the special processing results of each police situation analysis task execution sub-module.

[0019] In a third aspect, the present application provides a computer device, which includes a processor, and the processor is used to implement the steps of the above-mentioned public security police situation analysis method based on a large model when executing a computer program stored in a memory.

[0020] In a fourth aspect, the present application provides a computer-readable storage medium, which stores a computer program, and the computer program is used to implement the steps of the above-mentioned public security police situation analysis method based on a large model when executed by a processor.

[0021] The beneficial effects of the present application are as follows:

[0022] 1. The police situation analysis task planning module parses the police situation analysis request through a pre-trained large model for police situation analysis, combines the police situation analysis request, the target police situation data, and the task chain planning text prompt, and obtains a specific police situation analysis and governance task chain. This task planning method based on a large model can make full use of the intelligent analysis ability of the model, generate a reasonable task chain according to different police situation analysis requests, and clarify each task node and execution order. Compared with the traditional method, it greatly improves the accuracy and systematicness of task planning, and avoids the blindness and incompleteness of manual planning.

[0023] 2. The police situation analysis task execution module schedules the corresponding police situation analysis task execution sub - modules to process the target police situation data according to the execution order of the police situation analysis and governance task chain. Each police situation analysis task execution sub - module has a clear division of labor, extracts police situation description elements and police situation clue elements from the target police situation data respectively, and constructs the graph relationship between them. Each police situation analysis task execution sub - module uses the police situation analysis large - model to perform special processing based on the target police situation data and the corresponding task execution prompt. This modular and intelligent processing method not only improves the efficiency of data processing, but also can deeply excavate the information in the police situation data from multiple angles, transcends the traditional keyword retrieval mode, realizes the deep - level understanding of the police situation content, helps to discover potential risk points and valuable intelligence clues, and avoids the omission of important clues in the traditional way.

[0024] 3. The storage module is used to store the special processing results of each police situation analysis task execution sub - module, and centrally and orderly save the processed important information. This enables the public security department to conveniently access and use these data in subsequent work, providing strong support for the further analysis of police situations, summarizing experience, and handling similar cases.

[0025] 4. Through the cooperation between the various modules in this system, the automatic review of police situation data can be realized, replacing the cumbersome manual review process, greatly shortening the review cycle, and reducing the waste of human resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0027] Figure 1 It is a schematic structural diagram of a public security police situation analysis system based on a large - model provided by an embodiment of the present application;

[0028] Figure 2 It is a schematic process diagram of a public security police situation analysis based on a large - model provided by an embodiment of the present application;

[0029] Figure 3 It is a schematic structural diagram of a computer device provided by an optional embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0030] To make the objectives, technical solutions, and advantages of this application clearer, the following will further describe this application in detail in conjunction with the accompanying drawings. Apparently, the described embodiments are only a part of the embodiments of this application, rather than all of them. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of this application.

[0031] To improve the efficiency and quality of public security police situation analysis, this application provides a public security police situation analysis system, method, device, and medium based on a large model.

[0032] Embodiment 1:

[0033] This application provides a public security police situation analysis system based on a large model. Figure 1 FIG. is a schematic structural diagram of a public security police situation analysis system based on a large model provided by an embodiment of this application. The system includes:

[0034] A receiving module 11, configured to receive a police situation analysis request submitted by a user for target police situation data.

[0035] A police situation analysis task planning module 12, configured to parse the police situation analysis request through a pre-trained police situation analysis large model based on the police situation analysis request, the target police situation data, and a task chain planning text prompt prompt to obtain a specific police situation analysis governance task chain for the police situation analysis request; wherein, the specific police situation analysis governance task chain includes each task node and the execution order of each task node.

[0036] A police situation analysis task execution module 13, configured to process the target police situation data by scheduling corresponding police situation analysis task execution sub-modules according to the execution order; wherein, each police situation analysis task execution sub-module includes: a description extraction sub-module for extracting police situation description elements from the target police situation data, a clue extraction sub-module for extracting police situation clue elements from the target police situation data, and a graph construction sub-module for constructing a graph relationship between the police situation description elements and the police situation clue elements; for each police situation analysis task execution sub-module, the police situation analysis task execution sub-module obtains a special processing result of the police situation analysis task execution sub-module for the target police situation data through the police situation analysis large model based on the target police situation data and a corresponding task execution prompt.

[0037] A storage module 14, configured to store the special processing results of each police situation analysis task execution sub-module.

[0038] An output module 15, configured to output the special processing results of each police situation analysis task execution sub-module.

[0039] This large model-based public security police situation analysis system aims to use advanced large model technology to efficiently and accurately analyze public security police situation data and provide valuable decision-making support for the public security department. Through the collaborative work of multiple functional modules, the system realizes the full-process automated processing from the receipt of the police situation analysis request to the output of the analysis results. The system mainly includes a receiving module 11, a police situation analysis task planning module 12, a police situation analysis task execution module 13, a storage module 14, and an output module 15. The functions of each module are introduced in detail below:

[0040] (1) Receiving module 11

[0041] The receiving module 11 is the entry point for the system to interact with users. Its main function is to receive the police situation analysis request submitted by users for the target police situation data. In practical applications, the receiving module 11 can be implemented in various ways. For example, the system can provide a dedicated user interface, which can be a web-based web interface or a client software interface installed on the internal office computers of the public security department, without specific limitations here. Users fill in or select information such as the scope of the target police situation data to be analyzed and the focus of the analysis on this interface, and then click the submit button to send the police situation analysis request to the receiving module 11. For example, when on duty outside, police officers can log in to the system through the police communication APP, fill in the police situation information collected on the spot, including a brief description of the event and the initially judged case type, and then submit the request.

[0042] In one example, for the police situation analysis request initiated by users, the types of the police situation analysis request for the target police situation data include any one of the following: conducting a police situation analysis on specific police situation data, conducting real-time monitoring of police situation data, and conducting associated police situation analysis on a batch of historical police situation data stored in the database. For different types of police situation analysis requests, the receiving module 11 can perform targeted processing according to the request type. Exemplarily, for a request to conduct a police situation analysis on specific police situation data, the system focuses on the detailed features and potential risks of this specific police situation; for a request to conduct real-time monitoring of police situation data, the system will update the data in real time and promptly feedback abnormal situations; for a request to conduct associated police situation analysis on a batch of historical police situation data stored in the database, the system will mine the association relationships and rules among the data.

[0043] For example, if the user submits a request for crime analysis of specific crime data, the user needs to accurately fill in or select the unique identifier of the specific crime (such as the crime number) on the system interface, or locate the target crime data by describing in detail the key information of the crime (such as the occurrence time, location, case type, etc.). The receiving module 11 passes the crime analysis request to the subsequent crime analysis task planning module 12. For the request initiated by the user to perform real-time monitoring of crime data, the receiving module 11 can determine the monitoring scope and frequency according to the user's settings. The user can select the area to be monitored (such as a specific urban area, street, etc.), the type of crime (such as theft, robbery, traffic accident, etc.) and the monitoring time interval (such as every minute, every five minutes, etc.) on the system interface. After receiving the request, the receiving module 11 records these setting information and interacts with the data collection module to ensure that crime data is obtained in real time according to the requirements set by the user. When the user needs to perform associated crime analysis on a batch of historical crime data stored in the database, the user needs to set the analysis time range, crime type screening conditions, etc. on the system interface. After receiving the crime analysis request, the receiving module 11 can query the database for the batch of historical crime data indicated by the crime analysis request and pass both the batch of historical crime data and the crime analysis request to the crime analysis task planning module 12.

[0044] In a possible implementation manner, after receiving the crime analysis request, the receiving module 11 can perform a preliminary format check and integrity check on the crime analysis request to ensure the validity of the crime analysis request. If there are format errors or missing key information in the crime analysis request, the receiving module 11 will promptly prompt the user to make corrections.

[0045] (2) Crime analysis task planning module 12

[0046] The crime analysis task planning module 12 is one of the core modules of the system. It is responsible for parsing the crime analysis request submitted by the user and planning out a specific crime analysis and governance task chain. This module completes the task planning work through a pre-trained large crime analysis model, in combination with the crime analysis request, the target crime data, and the task chain planning text prompt prompt.

[0047] In this application, before the system is deployed, it is necessary to pre-train the major model for police situation analysis. During the pre-training process, a large amount of historical police situation data, relevant laws and regulations texts, and public security business knowledge documents can be used as training data. For example, various types of police situation data are widely collected, covering not only the data of common cases such as theft, robbery, and fraud, but also the data of special police situations such as traffic accidents and mass incidents. At the same time, relevant laws and regulations provisions, such as the Criminal Law of the People's Republic of China and the Law on Public Security Administration Punishments, and public security business operation specification documents, such as the process specification for receiving and handling police cases and the case investigation guidance manual, are collected to improve the robustness of the trained major model for police situation analysis. By learning these training data, the major model for police situation analysis can master the basic laws and knowledge of police situation analysis, providing a solid foundation for subsequent task planning.

[0048] In one example, the major model for police situation analysis is obtained by performing instruction fine-tuning on a general text understanding major model based on public security police situation industry knowledge samples; wherein, the types of the public security police situation industry knowledge samples include one or more of the following: police situation classification specifications, law enforcement systems, and hot police situations.

[0049] A general text understanding major model can be selected as the base model of the major model for police situation analysis, and then, based on the collected public security police situation industry knowledge samples, instruction fine-tuning is performed on this base model to make the fine-tuned major model more suitable for the field of public security police situation data analysis. Specifically, pre-training can be carried out on a large-scale general text data using a deep learning framework, so that the trained general text understanding major model has basic language understanding and generation capabilities. Then, based on the collected public security police situation industry knowledge samples, instruction fine-tuning is performed on the pre-trained general text understanding major model. During the fine-tuning process, the public security police situation industry knowledge samples and the corresponding instructions are input into the model, allowing the general text understanding major model to learn how to extract key information from the public security police situation industry knowledge samples according to the instructions, perform logical reasoning, and generate outputs that meet the requirements of public security police situation analysis. For example, an instruction "Based on the following police situation description, judge its belonging police situation classification" is given, and the corresponding police situation description samples are provided. The general text understanding major model gradually masters the ability to accurately judge the police situation classification according to the police situation description through continuous learning and parameter adjustment. In this way, the general text understanding major model can better adapt to the police situation analysis task and can more accurately understand and process information related to public security police situations.

[0050] In a possible implementation manner, these collected training samples can be cleaned, removing duplicate and incorrect data, unifying the data format, and ensuring the quality of the training samples.

[0051] After obtaining the pre-trained large model for police situation analysis based on the above embodiments, the pre-trained large model for police situation analysis can be directly deployed in the server cluster where the system is located, or the pre-trained large model for police situation analysis can be deployed on an independent server or cloud platform by means of a system call engine and called through a network interface.

[0052] When the police situation analysis task planning module 12 receives a police situation analysis request and target police situation data, the two can be concatenated with the task chain planning text prompt configured in advance in a specific format. For example, the task chain planning text prompt may include "Please plan detailed analysis task steps based on the following police situation analysis request and target police situation data, including but not limited to the data collection scope, analysis methods, and key attention indicators", and then the police situation analysis request and target police situation data are placed after the task chain planning text prompt to form a complete input text. The input text is input into the pre-trained large model for police situation analysis, and the large model for police situation analysis can perform semantic understanding and logical reasoning on the input text to determine the specific analysis direction and task details of the police situation analysis request. For example, if the user's police situation analysis request is to analyze the modus operandi of recent theft cases in a certain area, the large model for police situation analysis will determine, based on the relevant information in the input text, that it is necessary to collect the police situation data of all recent theft cases in the area, analyze the relevance of information such as the time, location, and suspect characteristics in these data, and compare the situation of similar theft cases in history and other task details. Finally, the large model for police situation analysis organizes these task details into a specific police situation analysis governance task chain in a certain logical order, and the specific police situation analysis governance task chain contains each task node and the execution order of each task node.

[0053] (3) Police situation analysis task execution module 13

[0054] The police situation analysis task execution module 13 is responsible for scheduling the corresponding police situation analysis task execution sub-modules to process the target police situation data according to the execution order of the specific police situation analysis governance task chain generated by the police situation analysis task planning module 12. Among them, the police situation analysis task execution sub-module includes: a description extraction sub-module for extracting police situation description elements from the target police situation data, a clue extraction sub-module for extracting police situation clue elements from the target police situation data, and a graph construction sub-module for constructing a graph relationship between the police situation description elements and the police situation clue elements. The following is an explanation of each police situation analysis task execution sub-module:

[0055] Description extraction submodule: The main function of the description extraction submodule is to extract police description elements from the target police data, such as time, place, person, event, etc. When actually executing the task, the description extraction submodule inputs the target police data and the corresponding task execution prompt into the police analysis model. The police analysis model uses natural language processing technology to analyze the text description in the police data, identify key information such as time expressions, place names, and person names, and extracts this information as police description elements. For example, for the police description "At 3 pm on January 15, 2025, at a convenience store near the city center square, a man wearing a black hat and a blue jacket robbed the store owner with a knife", the description extraction submodule can accurately extract the police description elements such as time "3 pm on January 15, 2025", location "a convenience store near the city center square", characters "a man wearing a black hat and a blue jacket" and "store owner", and event "robbery with a knife" through the big model.

[0056] Clue extraction submodule: The clue extraction submodule is used to extract police clue elements from the target police data, such as suspect characteristics, crime tools, and methods of committing crimes. This submodule also inputs the target police data and the corresponding task execution prompt into the police analysis model. The police analysis model uses deep mining technology to analyze and extract clue information from the police data. For example, in the above-mentioned robbery case, the clue extraction submodule can extract the suspect's height, body shape, accent and other characteristics, the characteristics of the crime tool "knife", and whether the crime method was a surprise attack or a prior scouting, and other police clue elements through the police analysis model.

[0057] Graph construction submodule: The task of the graph construction submodule is to use the knowledge graph construction technology of the police situation analysis model to associate the police situation description elements and police situation clue elements extracted by the description extraction submodule and the clue extraction submodule, and construct a graph reflecting the relationship between them. When constructing the graph, the graph construction submodule inputs the extracted police situation description elements, police situation clue elements and corresponding task execution prompts into the police situation analysis model. The police situation analysis model constructs a visual knowledge graph based on the logical relationship between elements, such as the relationship between characters, the relationship between events and locations, etc. Through this graph, public security personnel can intuitively understand the overall picture of the police situation and the relationship between the various elements, providing strong support for further analysis and decision-making.

[0058] In one example, an alarm / police knowledge base is pre-constructed, and various historical alarm data and their corresponding special processing results are stored in the alarm / police knowledge base. For any sub-module for executing an alarm analysis task, during the process of executing the task through the alarm analysis large model, the large model can fully refer to the vast amount of knowledge in the alarm / police knowledge base. By drawing on the valuable experience and knowledge in historical cases, the alarm analysis large model can process the current alarm analysis task more accurately and deeply, thereby significantly improving the accuracy and reliability of the processing results. This reference and learning mechanism based on historical data helps to enhance the overall analysis ability of cases and provides strong support for subsequent case detection and handling.

[0059] In a possible implementation manner, the system will regularly update the knowledge in the alarm / police knowledge base according to the data stored in the storage module 14. The storage module 14 continuously collects and accumulates the latest alarm data and processing results, and these new data include new situations, new problems, and new processing methods that emerge over time and with the change of the social environment. By regularly updating the alarm / police knowledge base, it can ensure that the knowledge in the knowledge base always remains timely and accurate, enabling it to better adapt to the changing characteristics of alarms and processing requirements. This dynamic update mechanism helps the alarm analysis large model to always obtain the most cutting-edge and effective reference information when processing tasks, thereby continuously improving the quality and efficiency of alarm analysis.

[0060] (IV) Storage module 14

[0061] The storage module 14 is used to store the special processing results of each sub-module for executing an alarm analysis task. In practical applications, the storage module 14 can adopt various storage methods, such as relational databases, non-relational databases, or file systems, etc. The storage module 14 will select a suitable storage method and data structure according to the characteristics of the data and usage requirements. For example, for structured alarm description elements and alarm clue elements, they can be stored in a relational database for efficient query and statistical analysis; for the constructed knowledge graph data, a graph database can be used for storage to better support graph traversal and query operations. When storing data, the storage module 14 will classify and index the data according to certain storage rules for subsequent rapid query and analysis.

[0062] (V) Output module 15

[0063] The output module 15 is responsible for outputting the special processing results of each police situation analysis task execution sub-module to the user. The output module 15 can implement the output function in various ways. For example, the system can provide a visual interface to display the analysis results to the user in the form of charts, maps, reports, etc., facilitating the user to intuitively understand the results of the police situation analysis. At the same time, the output module 15 can also support providing the analysis results in a specific data format (such as JSON, XML, etc.) to other relevant systems for further processing, realizing data sharing and collaborative work between systems.

[0064] It should be noted that each prompt adopted in this system (including the task chain planning text prompt, task execution prompt, service generation prompt, police situation data verification prompt, etc.) is a text instruction for guiding the large model to perform specific tasks. In this system, different types of prompts are used to guide the police situation analysis large model to complete different tasks, such as task chain planning, task execution, service generation, police situation data verification, etc. For example, the task chain planning text prompt is a text instruction for guiding the police situation analysis large model to plan the analysis task steps. It contains information such as the data collection scope, analysis method, and key attention indicators. For example, 'Please plan detailed analysis task steps according to the following police situation analysis request and target police situation data. The data collection scope is all police situation data in this area in the past month. The analysis methods are association analysis and trend analysis. The key attention indicators are the distribution of the crime time and the characteristics of the suspects'. The task execution prompt is a text instruction for guiding the police situation analysis task execution sub-module to perform special processing on the target police situation data with the help of the police situation analysis large model. The task execution prompts corresponding to different police situation analysis task execution sub-modules have different focuses to meet the specific task requirements of each police situation analysis task execution sub-module. The service generation prompt is a text instruction for guiding the police situation analysis large model to generate police situation analysis data related to the target service according to the special processing results of each police situation analysis task execution sub-module when the system receives the generation instruction of the target police situation data on the target service. The police situation data verification prompt is a text instruction for guiding the police situation analysis large model to perform automatic verification on the real-time police situation data pulled from the external police situation data collection device of the system. Its main purpose is to ensure the quality and accuracy of the police situation data input into the system. By performing integrity verification and accuracy verification on the data, abnormal values, error information, or incomplete data in the data can be quickly identified.

[0065] The beneficial effects of this application are as follows:

[0066] 1. The police situation analysis task planning module 12 parses the police situation analysis request through a pre-trained large model for police situation analysis, in combination with the police situation analysis request, target police situation data, and the task chain planning text prompt "prompt", to obtain a specific police situation analysis and governance task chain. This task planning method based on a large model can make full use of the intelligent analysis ability of the model, generate a reasonable task chain according to different police situation analysis requests, and clarify each task node and execution order. Compared with the traditional method, it greatly improves the accuracy and systematicness of task planning, and avoids the blindness and incompleteness of manual planning.

[0067] 2. The police situation analysis task execution module 13 schedules the corresponding police situation analysis task execution sub-modules to process the target police situation data according to the execution order of the police situation analysis and governance task chain. Each police situation analysis task execution sub-module has a clear division of labor, extracts the police situation description elements and police situation clue elements from the target police situation data respectively, and constructs the graph relationship between them. Each police situation analysis task execution sub-module uses the large model for police situation analysis and conducts special processing based on the target police situation data and the corresponding task execution "prompt". This modular and intelligent processing method not only improves the efficiency of data processing, but also can deeply mine the information in the police situation data from multiple perspectives, transcends the traditional keyword retrieval mode, realizes the deep understanding of the police situation content, helps to discover potential risk points and valuable intelligence clues, and avoids the omission of important clues in the traditional method.

[0068] 3. The storage module 14 is used to store the special processing results of each police situation analysis task execution sub-module, and centrally and orderly save the processed important information. This enables the public security department to conveniently consult and use these data in subsequent work, providing strong support for the further analysis of police situations, summarizing experience, and handling similar cases.

[0069] 4. Through the cooperation between the various modules in this system, the automatic review of police situation data can be realized, replacing the cumbersome manual review process, greatly shortening the review cycle, and reducing the waste of human resources.

[0070] Embodiment 2:

[0071] In order to provide more comprehensive police situation analysis services, on the basis of the above embodiment, in this application, the system further includes: a business analysis and generation module;

[0072] The business analysis and generation module is configured to, if a generation instruction for the target alarm data on a target service is received by the receiving module 11, obtain a service generation prompt corresponding to the generation instruction; and based on the service generation prompt and the special processing results of each alarm analysis task execution sub-module through the alarm analysis large model, obtain alarm analysis data of the target alarm data on the target service; wherein the target service includes one or more of the following: generation of an alarm handling summary, generation of an alarm overview summary, generation of an alarm police station express, generation of an alarm sensitive point identification report, generation of an alarm verification express, generation of an alarm special express, generation of an alarm review summary;

[0073] The storage module 14 is further configured to store the alarm analysis data of the target alarm data on the target service;

[0074] The output module 15 is further configured to output the alarm analysis data.

[0075] In this application, the system may further include a business analysis and generation module, which is a key module in the system responsible for generating alarm analysis data related to specific services. When the receiving module 11 receives a generation instruction for the target alarm data on a target service, the business analysis and generation module starts to work. The business analysis and generation module will, according to the received generation instruction, obtain a service generation prompt corresponding to the target service from a pre-set prompt library. For example, if the generation instruction is to generate an alarm handling summary, then the business analysis and generation module will obtain a prompt specifically for generating an alarm handling summary, which may contain "Please generate a concise and clear alarm handling summary based on the following alarm-related information, covering key information such as the handling process, measures taken, and results, etc.". After obtaining the service generation prompt, the business analysis and generation module will integrate it with the special processing results of each alarm analysis task execution sub-module. For example, information such as the alarm description elements extracted by the description extraction sub-module, the clue elements extracted by the clue extraction sub-module, and the knowledge graph constructed by the graph construction sub-module will be input into the alarm analysis large model together with the service generation prompt. The alarm analysis large model, through understanding and reasoning of this information, generates alarm analysis data of the target alarm data on the target service according to the requirements of the service generation prompt. For example, when generating an alarm overview summary, the alarm analysis large model can comprehensively consider various aspects of the alarm, extract key content, and generate an overview summary covering aspects such as the cause, process, result, and impact of the event.

[0076] Among them, the operations supported by the system include one or more of the following: generation of police situation handling summaries, generation of police situation overview summaries, generation of police situation express reports for police stations, generation of reports on identification of sensitive points in police situations, generation of express reports for police situation verification, generation of express reports for special police situations, and generation of police situation review summaries. The following are examples of processing for different target operations:

[0077] Generation of police situation handling summaries: Based on the information provided by the sub-module for executing police situation analysis tasks, the large police situation analysis model can sort out the process of police situation handling, including the time of receiving the alarm, the time of dispatching the police, on-site handling measures, the situation of suspect capture, etc., and generate a concise police situation handling summary to facilitate relevant personnel to quickly understand the key information of police situation handling.

[0078] Generation of police situation overview summaries: The large police situation analysis model can comprehensively consider the macro background of the occurrence of police situations, such as the local public security situation, recent social dynamics, etc., as well as the detailed development process of police situations, including the cause of the incident, the escalation points of conflicts, the response measures of all parties, etc. At the same time, it analyzes the impact of police situations on aspects such as the surrounding environment, social order, and public sentiment. Through comprehensive and in-depth analysis, it generates a rich and well-structured police situation overview summary to help decision-makers grasp the police situation trend from a macro perspective.

[0079] Generation of police situation express reports for police stations: For police situation data at the police station level, the large police situation analysis model can integrate the core information of police situations at the fastest speed and generate express reports to provide key information and handling suggestions, so as to meet the needs of rapid information transmission and collaborative work within the police station and ensure that police officers at all positions can quickly understand the situation and carry out subsequent work.

[0080] Generation of reports on identification of sensitive points in police situations: The large police situation analysis model can conduct in-depth analysis of police situation data, identify possible sensitive points, such as information related to special groups, major public opinion risks, etc., and generate detailed reports on identification of sensitive points to provide a basis for the public security department to take preventive measures in advance.

[0081] Generation of express reports for police situation verification: For police situation information that needs further verification, the large police situation analysis model can systematically sort out and deeply analyze relevant clues, generate express reports for police situation verification, provide clear guidance and reference for subsequent verification work, and enable verification personnel to target and improve work efficiency.

[0082] Generation of police situation review summaries: Combining the whole process information of police situation handling, the large police situation analysis model can summarize the experience and lessons in the process of police situation handling, analyze the successful points and existing problems, and generate police situation review summaries to provide reference for subsequent police situation handling work.

[0083] Generation of Special Report on Police Incidents: According to specific requirements of special work, the large model for police incident analysis can accurately extract information related to special work from police incident data. For example, in the special operation against telecom fraud, it focuses on extracting the modus operandi of telecom fraud-related police incidents, such as whether the fraud is carried out through phone calls, text messages or online platforms; the characteristics of the deceived population, including age range, occupation distribution, reasons for being deceived, etc.; and the fund flow, tracking the transfer path and final destination of the deceived funds, and whether illegal activities such as money laundering are involved. By extracting and analyzing these key information, a highly targeted special report on police incidents is generated to facilitate the smooth progress of special work.

[0084] After obtaining the police incident analysis data of the target police incident data in the target business based on the above embodiments, the police incident analysis data can be stored in the storage module 14 for subsequent analysis and query. At the same time, the police incident analysis data can also be output through the output module 15 so that users can view the police incident analysis data in a timely manner.

[0085] Embodiment 3:

[0086] In order to further improve the efficiency and data accuracy of the system in police incident analysis and processing, based on the above embodiments, in this application, the system further includes: a police incident analysis governance API protocol call module;

[0087] The police incident analysis governance API protocol call module is used to establish a connection with the police incident data collection device outside the system and regularly pull real-time police incident data from the police incident data collection device; and based on the police incident data verification prompt through the police incident analysis large model, automatically verify the obtained police incident data.

[0088] In this application, the system further includes an API protocol call module for police situation analysis and governance. This API protocol call module for police situation analysis and governance undertakes the key task of establishing connections with police situation data collection devices outside the system. These police situation data collection devices may include, but are not limited to, surveillance cameras, alarm terminals, intelligent sensors, etc. distributed in various corners of the city, which collect various types of police situation-related data in real time. It can regularly send requests to the police situation data collection devices to pull real-time police situation data, ensuring that the system can obtain the latest police situation dynamics in a timely manner. After obtaining the real-time police situation data, this API protocol call module for police situation analysis and governance can, with the help of a large model for police situation analysis, based on the police situation data verification prompt, carry out automated verification of the obtained police situation data. Among them, the automated verification work includes integrity verification and accuracy verification. Through this automated verification process, outliers, error information, or incomplete data in the data can be quickly identified, ensuring the quality and accuracy of the police situation data input into the system. Only the police situation data that has passed the verification will be used for subsequent police situation analysis and processing work, thus laying a solid data foundation for the efficient execution of police situation analysis tasks.

[0089] Through this API protocol call module for police situation analysis and governance, the latest police situation data can be received and processed immediately, ensuring that all police situations can be handled in a timely and effective manner, and improving the emergency response speed.

[0090] Embodiment 4:

[0091] Based on the same inventive concept, this application also provides a method for public security police situation analysis based on a large model. Figure 2 FIG. is a schematic diagram of the process of public security police situation analysis based on a large model provided for an embodiment of this application. The process includes:

[0092] S201: Receive a police situation analysis request submitted by a user for target police situation data.

[0093] S202: Through a pre-trained large model for police situation analysis, based on the police situation analysis request, the target police situation data, and the task chain planning text prompt, parse the police situation analysis request to obtain a specific police situation analysis and governance task chain; wherein, the specific police situation analysis and governance task chain includes each task node and the execution order of each task node.

[0094] S203: Schedule the corresponding sub-module for executing the police situation analysis task to process the target police situation data according to the execution order; wherein, each sub-module for executing the police situation analysis task includes: a description extraction sub-module for extracting the description elements of the police situation from the target police situation data, a clue extraction sub-module for extracting the clue elements of the police situation from the target police situation data, and a graph construction sub-module for constructing the graph relationship between the description elements and the clue elements of the police situation; for each sub-module for executing the police situation analysis task, the sub-module for executing the police situation analysis task obtains the special processing result of the sub-module for executing the police situation analysis task on the target police situation data through the large model for police situation analysis, based on the target police situation data and the corresponding task execution prompt.

[0095] S204: Store and output the special processing results of each sub-module for executing the police situation analysis task.

[0096] In the method for public security police situation analysis based on a large model of the present application, it is applied to a computer device, and the computer device can be an intelligent terminal, such as a computer, a robot, etc., or a server, such as an application server, a business server, etc.

[0097] Since the principle of solving problems of the above-mentioned method for public security police situation analysis based on a large model is similar to that of the system for public security police situation analysis based on a large model, the implementation of the above-mentioned method for public security police situation analysis based on a large model can refer to the embodiments of the system, and the repeated parts will not be described again.

[0098] Embodiment 5:

[0099] Please refer to Figure 3 , Figure 3 which is a schematic structural diagram of a computer device provided by an optional embodiment of the present application. As Figure 3 shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting each component, including a high-speed interface and a low-speed interface. Each component communicates with each other using different buses and can be installed on a common motherboard or installed in other ways according to needs. The processor can process instructions executed within the computer device, including instructions stored in the memory or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some optional implementation manners, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a set of blade servers, or a multi-processor system). Figure 3 In

[0100] The processor 10 may be a central processing unit, a network processor, or a combination thereof. Among them, the processor 10 may further include a hardware chip. The above-mentioned hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The above-mentioned programmable logic device may be a complex programmable logic device, a field programmable gate array, a generic array logic, or any combination thereof.

[0101] Among them, the memory 20 stores instructions executable by at least one processor 10, so that at least one processor 10 executes the method shown in the above embodiments.

[0102] The memory 20 may include a program storage area and a data storage area. Among them, the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created according to the use of a computer device presented by a kind of landing page of a small program, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some alternative embodiments, the memory 20 may optionally include a memory remotely set relative to the processor 10, and these remote memories may be connected to the computer device through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0103] The memory 20 may include a volatile memory, for example, a random access memory; the memory may also include a non-volatile memory, for example, a flash memory, a hard disk, or a solid-state drive; the memory 20 may also include a combination of the above types of memories.

[0104] The computer device further includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30, and the output device 40 may be connected through a bus or other means. Figure 3 Taking connection through a bus as an example.

[0105] The input device 30 can receive input digital or character information, and generate key signal inputs related to the user settings and function control of the computer device, such as a touch screen, a keypad, a mouse, a trackpad, a touchpad, a pointing stick, one or more mouse buttons, a trackball, a joystick, etc. The output device 40 may include a display device, an auxiliary lighting device (for example, an LED), and a tactile feedback device (for example, a vibration motor), etc. The above-mentioned display device includes but is not limited to a liquid crystal display, a light-emitting diode, a display, and a plasma display. In some alternative embodiments, the display device may be a touch screen.

[0106] Embodiment 6:

[0107] Based on the above embodiments, an embodiment of the present application further provides a computer-readable storage medium, in which a computer program executable by a processor is stored. When the program runs on the processor, the processor is caused to execute the following steps:

[0108] Receive a request for police situation analysis submitted by a user for target police situation data;

[0109] Through a pre-trained large police situation analysis model, based on the police situation analysis request, the target police situation data, and the task chain planning text prompt (prompt), parse the police situation analysis request to obtain the specific police situation analysis and governance task chain of the police situation analysis request; wherein, the specific police situation analysis and governance task chain includes each task node and the execution order of each task node;

[0110] According to the execution order, schedule the corresponding police situation analysis task execution sub-module to process the target police situation data; wherein, each police situation analysis task execution sub-module includes: a description extraction sub-module for extracting police situation description elements from the target police situation data, a clue extraction sub-module for extracting police situation clue elements from the target police situation data, and a graph construction sub-module for constructing a graph relationship between the police situation description elements and the police situation clue elements; for each police situation analysis task execution sub-module, the police situation analysis task execution sub-module obtains the special processing result of the target police situation data by the large police situation analysis model based on the target police situation data and the corresponding task execution prompt;

[0111] Store and output the special processing results of each police situation analysis task execution sub-module.

[0112] Since the principle of the above computer-readable storage medium for solving problems is similar to the method for public security police situation analysis based on a large model, the implementation of the above computer-readable storage medium can refer to the embodiments of the method, and the repeated parts will not be described again.

[0113] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these changes and modifications.

Claims

1. A public security police situation analysis system based on a large model, characterized in that: The system comprises: A receiving module, used to receive a warning analysis request for target warning data submitted by a user; The alarm analysis task planning module is used to parse the alarm analysis request through the pre-trained alarm analysis model, based on the alarm analysis request, the target alarm data and the task chain planning text prompt, and obtain the specific alarm analysis governance task chain of the alarm analysis request; wherein the specific alarm analysis governance task chain includes each task node and the execution order of each task node; A police situation analysis task execution module is used to dispatch the corresponding police situation analysis task execution submodule to process the target police situation data according to the execution order; wherein each of the police situation analysis task execution submodules includes: a description extraction submodule for extracting police situation description elements from the target police situation data, a clue extraction submodule for extracting police situation clue elements from the target police situation data, and a graph construction submodule for constructing a graph relationship between police situation description elements and police situation clue elements; for each of the police situation analysis task execution submodules, the police situation analysis task execution submodule obtains the special processing result of the police situation analysis task execution submodule on the target police situation data through the police situation analysis big model, based on the target police situation data and the corresponding task execution prompt; A storage module, used to store the special processing results of each of the police situation analysis task execution submodules; The output module is used to output the special processing results of each of the police situation analysis task execution sub-modules.

2. The system according to claim 1, characterized in that The type of the alarm analysis request for the target alarm data includes any of the following: alarm analysis on specific alarm data, real-time monitoring of alarm data, and associated alarm analysis on batch historical alarm data stored in a database.

3. The system according to claim 1, characterized in that The system also includes: a business analysis generation module; The business analysis generation module is used to obtain the business generation prompt corresponding to the generation instruction if the generation instruction of the target police situation data on the target business is received through the receiving module; through the police situation analysis big model, based on the business generation prompt and the special processing results of each of the police situation analysis task execution submodules, obtain the police situation analysis data of the target police situation data on the target business; wherein, the target business includes one or more of the following: generation of police situation handling summary, generation of police situation summary, generation of police situation police station express report, generation of police situation sensitive point identification report, generation of police situation verification express report, generation of police situation special express report, generation of police situation review summary; The storage module is also used to store the target alarm data on the target business alarm analysis data; The output module is also used to output the alarm analysis data.

4. The system according to claim 1, characterized in that The system also includes: a police situation analysis and governance API protocol calling module; The police situation analysis and governance API protocol calling module is used to establish a connection with the police situation data collection device outside the system, and regularly pull real-time police situation data from the police situation data collection device; through the police situation analysis big model, based on the police situation data verification prompt, the acquired police situation data is automatically verified.

5. The system according to claim 1, wherein: For each of the police situation analysis task execution submodules, the police situation analysis task execution submodule obtains the special processing result of the police situation analysis task execution submodule on the target police situation data based on the target police situation data and the corresponding task execution prompt through the police situation analysis big model, including: For each of the police situation analysis task execution submodules, the police situation analysis task execution submodule obtains the special processing results of the police situation analysis task execution submodule on the target police situation data through the police situation analysis big model, based on the target police situation data and the corresponding task execution prompt, combined with the knowledge in the pre-built police situation / police knowledge base; wherein the police situation / police knowledge base stores various historical police situation data and the special processing results corresponding to each of the historical police situation data.

6. The system according to claim 5, characterized in that The knowledge in the police situation / police knowledge base is updated regularly based on the data stored in the storage module.

7. The system according to claim 1, characterized in that The police situation analysis big model is obtained by fine-tuning the general text understanding big model based on the public security police situation industry knowledge samples; wherein the types of the public security police situation industry knowledge samples include one or more of the following: police situation classification specifications, law enforcement systems, and hot police situations.

8. A method for analyzing public security situations based on a large model, characterized in that: The method comprises: Receive a warning analysis request for target warning data submitted by a user; Through the pre-trained police situation analysis big model, based on the police situation analysis request, the target police situation data and the task chain planning text prompt, the police situation analysis request is parsed to obtain the specific police situation analysis governance task chain of the police situation analysis request; wherein the specific police situation analysis governance task chain includes each task node and the execution order of each task node; According to the execution order, the corresponding police situation analysis task execution submodule is dispatched to process the target police situation data; wherein each of the police situation analysis task execution submodules includes: a description extraction submodule for extracting police situation description elements from the target police situation data, a clue extraction submodule for extracting police situation clue elements from the target police situation data, and a graph construction submodule for constructing a graph relationship between police situation description elements and police situation clue elements; for each of the police situation analysis task execution submodules, the police situation analysis task execution submodule obtains the special processing result of the police situation analysis task execution submodule on the target police situation data through the police situation analysis big model, based on the target police situation data and the corresponding task execution prompt; Store and output the special processing results of each of the police situation analysis task execution sub-modules.

9. A computer device, characterized in that: The computer device includes a processor, which is used to implement the steps of the public security alarm situation analysis method based on a large model as described in claim 8 above when executing a computer program stored in a memory.

10. A computer-readable storage medium, characterized in that: It stores a computer program that can be executed by a computer device. When the program is run on the computer device, the computer device executes the steps of the public security alarm situation analysis method based on a large model as described in claim 8 above.

Images