Warning condition data intelligent analysis system and method based on open source model

Abstract

The invention provides an alarm data intelligent analysis system and method based on an open source model, and the system comprises a data obtaining module which is used for obtaining multi-source alarm data; the data analysis module is used for analyzing the multi-source alarm data by using an open source model to obtain multi-dimensional features of the alarm data; the disposal method matching module is used for matching a corresponding disposal method according to the multi-dimensional features, pushing related alarm data to a corresponding disposal process according to the disposal method, and receiving an alarm disposal result in real time; the disposal result checking module is used for checking the disposal method and the alarm disposal result according to a preset disposal rule and selecting whether to execute a correction operation or not according to a checking result; and the clue recommendation module is used for generating a recommendation clue based on the multi-dimensional features, the disposal method, the alarm disposal result and the check result, and feeding back the recommendation clue to the disposal method matching module so as to dynamically optimize the disposal method matching in real time.

Description

Technical Field

[0001] This invention relates to the field of computer artificial intelligence technology, and in particular to an intelligent analysis system and method for police incident data based on an open-source model. Background Technology

[0002] With rapid societal development and continuous advancements in information technology, the volume of police incident data has experienced explosive growth. This data contains a wealth of information, such as relevant entities (people, vehicles, phone numbers, etc.), incident types, and response outcomes. This information is crucial for optimizing police response efforts and improving policing efficiency. In recent years, open-source models have demonstrated powerful capabilities in natural language processing and data mining, providing new technological means for the intelligent analysis of police incident data.

[0003] Among the existing police incident data processing technologies, some solutions have initially achieved intelligent police incident processing. These existing technological solutions mainly focus on the following aspects: Police report information extraction and classification: This involves using large open-source models to parse police report texts and extract key information such as the type of incident, information about the officers involved, and the details of the incident. For example, some studies have used LoRA technology to fine-tune large open-source language models (such as Qwen2-7B-Instruct and GLM-4-9B-Chat) to adapt to the language characteristics of the police report domain, thereby improving the accuracy and controllability of information extraction.

[0004] Police Incident Data Management and Analysis: Existing police incident big data management systems include multiple alarm terminals and a system server. Alarm terminals can issue alarms via telephone, SMS, or an app, and the system server processes the received alarm information. For example, a certain police incident big data management method includes the following steps: In response to an alarm signal from an alarm terminal, initial query information is sent to the alarm terminal. In response to initial feedback data from the alarm terminal, a police incident question-and-answer model is invoked, and guiding query information is sent to the alarm terminal.

[0005] The intelligent police incident handling system analyzes alarm information to obtain various attributes of the incident and encapsulates this attribute data. It selects appropriate officers based on the incident type and assigns dispatch tasks. The system also includes the function of receiving data uploaded by officers and generating historical incident data. For example, it analyzes received alarm information to obtain various attributes of the incident and encapsulates this attribute data. It selects potential officers based on the incident type and then chooses one officer from the pool of potential officers to handle the incident based on historical incident data. It receives data uploaded by officers during the dispatch, incident handling, record-keeping, incident feedback, and incident closure process and generates historical incident data.

[0006] Incident Element Extraction: Incident element extraction is performed using a large-scale model built within the public security domain. LoRA fine-tuning and cue word engineering are employed to improve model performance, and data augmentation techniques are used to expand the data and enhance model performance. For example, one study tested various lightweight open-source large-scale models and combined data augmentation methods to improve the F1 score of incident element extraction.

[0007] Intelligent security and crime analysis: Utilizing big data mining, artificial intelligence, and other technologies for crime situation prediction and analysis, early warning analysis of key personnel, and early warning analysis of key vehicles. For example, a certain crime data analysis model uses K-means algorithm, Naive Bayes algorithm, classification algorithm, clustering algorithm, etc., for multi-dimensional analysis and judgment to achieve early warning monitoring of key personnel and vehicles.

[0008] While these existing technological solutions provide a foundation and reference for the intelligent processing of police incident data, they still need further improvement and enhancement in terms of comprehensiveness of functions, depth and accuracy of analysis, and supervision and management of incident response results. Specifically, these shortcomings include the following: 1. Low efficiency in processing police incident data: Traditional police incident analysis relies heavily on human experience. Manual processing of police incident data is slow and cannot cope with the explosive growth in the amount of police incident data. As a result, the police incident analysis cycle is long and police decisions cannot be adjusted in a timely manner based on the latest police situation.

[0009] 2. Insufficient accuracy in police incident analysis: Human experience is insufficient to fully extract complex information from police incident data, such as police-related entities and their relationships, which leads to inaccurate identification of incident types, inappropriate matching of handling methods, and inaccurate verification of compliance of handling results, affecting the quality and effectiveness of police work.

[0010] 3. Lack of an effective supervision and management mechanism: The existing technology lacks a systematic supervision process, and the tracking and handling of non-compliant police response results and other important clues are inadequate, resulting in important police incidents not being handled in a timely and effective manner, which affects the execution and credibility of police work.

[0011] 4. Insufficient value mining of police incident data: Traditional methods cannot deeply analyze the potential information in police incident data, such as clues about frequently involved police entities and locations with high incidence of police incidents. As a result, police work lacks accurate clues and forward-looking support, making it impossible to effectively combat crime and carry out public security prevention and control. Summary of the Invention

[0012] The purpose of this invention is to provide an intelligent analysis system and method for police incident data based on an open-source model, aiming to solve the aforementioned problems in the prior art.

[0013] This invention provides an intelligent analysis system for police incident data based on an open-source model, comprising: A data acquisition module, connected to a data parsing module, is used to acquire multi-source police incident data and transmit the multi-source police incident data to the data parsing module; wherein, the multi-source police incident data includes historical police incident data and current police incident data; The data parsing module, connected to the data acquisition module, the handling method matching module, and the clue recommendation module, is used to parse the multi-source police incident data using an open-source model, obtain multi-dimensional features of the police incident data, and send the multi-dimensional features to the handling method matching module and the clue recommendation module. The handling method matching module is connected to the data parsing module, the handling result verification module, and the clue recommendation module. It is used to match the corresponding handling method according to the multi-dimensional features, push relevant police data to the corresponding handling process according to the handling method, receive the police handling results in real time, and transmit the handling method and the police handling results to the handling result verification module and the clue recommendation module. The handling result verification module is connected to the handling method matching module and the clue recommendation module. It is used to verify the handling method and the police incident handling result according to the preset handling rules, send the verification result to the clue recommendation module, and select whether to perform a correction operation based on the verification result. The clue recommendation module is connected to the data parsing module, the handling method matching module, and the handling result verification module. It is used to generate recommended clues based on the multi-dimensional features, handling methods, police incident handling results, and verification results, and to feed the recommended clues back to the handling method matching module for real-time dynamic optimization of handling method matching.

[0014] This invention provides an intelligent analysis method for police incident data based on an open-source model, comprising: The data acquisition module acquires multi-source police incident data and transmits the multi-source police incident data to the data parsing module; wherein, the multi-source police incident data includes historical police incident data and current police incident data; The data parsing module uses an open-source model to parse the multi-source police incident data, obtains multi-dimensional features of the police incident data, and sends the multi-dimensional features to the handling method matching module and the clue recommendation module. The handling method matching module matches the corresponding handling method based on the multi-dimensional features, pushes the relevant police data to the corresponding handling process according to the handling method, receives the police handling results in real time, and transmits the handling method and the police handling results to the handling result verification module and the clue recommendation module. The handling result verification module verifies the handling method and the police incident handling result according to the preset handling rules, sends the verification result to the clue recommendation module, and selects whether to perform a correction operation based on the verification result; The clue recommendation module generates recommended clues based on the multidimensional features, handling methods, police incident handling results, and verification results, and feeds the recommended clues back to the handling method matching module for real-time dynamic optimization of handling method matching.

[0015] This invention also provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of the above-described intelligent analysis method for police data based on an open-source model.

[0016] This invention also provides a computer-readable storage medium storing an information transmission implementation program, which, when executed by a processor, implements the steps of the above-described intelligent analysis method for police data based on an open-source model.

[0017] The embodiments of this invention can include the following beneficial effects: Based on existing technologies, these embodiments further innovate and expand the methods for analyzing and processing police incident data. Compared with existing technologies, these embodiments can not only efficiently acquire and parse police incident data, but also extract information on police-related entity elements through in-depth analysis and analyze their correlations, recommending clues such as people, vehicles, and mobile phone numbers frequently involved in police incidents, as well as locations with high incidence of incidents. Furthermore, these embodiments provide functions such as providing suggestions for handling police incidents and verifying the quality and compliance of handling results. For police incidents with non-compliant handling results and other important clues, they provide functions for supervision, issuance, receipt, feedback, completion processes, and statistical analysis. These innovations give these embodiments significant advantages in improving the efficiency of police incident analysis, enhancing the accuracy of analysis, optimizing police handling workflows, and assisting in the scientific decision-making of police affairs. They can effectively improve the efficiency and quality of police work and better maintain social security and stability. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in one or more embodiments of this specification or in 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 recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of an intelligent police data analysis system based on an open-source model, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the overall architecture of the intelligent analysis system for police incident data based on an open-source model, according to an embodiment of the present invention. Figure 3 This is a hardware device structure diagram according to an embodiment of the present invention; Figure 4 This is a flowchart of the alarm data parsing process according to an embodiment of the present invention; Figure 5 This is a flowchart of the matching and verification process for the processing method in this embodiment of the invention; Figure 6 This is a flowchart illustrating the supervision process according to an embodiment of the present invention; Figure 7 This is a flowchart of the intelligent analysis method for police incident data based on an open-source model, according to an embodiment of the present invention. Detailed Implementation

[0020] To enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of this document.

[0021] System Implementation Examples According to embodiments of the present invention, an intelligent analysis system for police incident data based on an open-source model is provided. Figure 1 This is a schematic diagram of an intelligent police data analysis system based on an open-source model, as described in an embodiment of the present invention. Figure 1 As shown, the intelligent analysis system for police incident data based on an open-source model according to an embodiment of the present invention specifically includes: The data acquisition module 10, connected to the data parsing module, is used to acquire multi-source alarm data and transmit the multi-source alarm data to the data parsing module; wherein, the multi-source alarm data includes historical alarm data and current alarm data; Data parsing module 12, connected to the data acquisition module, the handling method matching module, and the clue recommendation module, is used to parse the multi-source police incident data using an open-source model to obtain multi-dimensional features of the police incident data, and send the multi-dimensional features to the handling method matching module and the clue recommendation module. Specifically, it is used for: A finely tuned open-source natural language processing model is used to perform named entity recognition, keyword extraction, and classification on police incident texts to extract police-related entity elements and incident types; and relations are extracted from the police-related entity elements to construct a knowledge graph of the police-related entity elements and their relationships. An open-source speech recognition model is used to convert alarm recordings into text, and the converted text is then input into a natural language processing model for processing to obtain key alarm information. Using open-source image recognition and target detection models, at least one of the following elements—face, vehicle, license plate number, and scene—is identified from police photos or videos as the police-related entity element.

[0022] The handling method matching module 14 is connected to the data parsing module, the handling result verification module, and the clue recommendation module. It is used to match the corresponding handling method according to the multi-dimensional features, push the relevant police data to the corresponding handling process according to the handling method, receive the police handling results in real time, and transmit the handling method and the police handling results to the handling result verification module and the clue recommendation module. The handling result verification module 16, connected to the handling method matching module and the clue recommendation module, is used to verify the handling method and the police incident handling result according to preset handling regulations, send the verification result to the clue recommendation module, and select whether to perform a correction operation based on the verification result. Specifically, it is used for: Establish a police incident verification rule base; wherein, the police incident verification rule base is pre-set based on police incident handling specifications; The handling method and the handling result are automatically checked according to the pre-set handling rules in the police incident verification rule base. If the verification result shows compliance, no correction operation is required. If the verification result shows non-compliance, a verification report containing specific non-compliance items and modification suggestions is automatically generated according to the pre-set modification rules. The verification report is then fed back to the handling method matching module for real-time dynamic correction of the handling method matching.

[0023] The police incident handling guidelines include timeliness rules, element integrity rules, and logical consistency rules; The timeliness rules are used to check whether the police response time, on-site handling time, and case completion cycle exceed the prescribed thresholds. The element completeness rule is used to verify whether the essential handling elements are complete; the essential handling elements include at least one of the following: information records of the persons involved, uploaded evidence materials, and legal citations. The logical consistency rule is used to check whether there is a logical contradiction between the nature of the police incident, the handling measures, and the final result.

[0024] The clue recommendation module 18, connected to the data parsing module, the handling method matching module, and the handling result verification module, is used to generate recommended clues based on the multi-dimensional features, handling methods, police incident handling results, and verification results, and to feed the recommended clues back to the handling method matching module for real-time dynamic optimization of handling method matching. Specifically, it is used for: Based on the aforementioned multidimensional features, handling methods, police incident handling results, and verification results, at least one of the following mining analyses is performed: frequent pattern mining, spatiotemporal clustering analysis, and abnormal behavior identification; The frequent pattern mining involves using association rule algorithms based on historical police incident data to mine frequent itemsets that meet the minimum support and confidence requirements, generating high-frequency co-occurring combinations of police-related entities as recommendation clues; the combinations of police-related entities include at least one of personnel-personnel, personnel-vehicle, and personnel-location. The spatiotemporal clustering analysis is based on the time and geographical location information of the incidents. It uses a density clustering algorithm to identify the clusters of incidents that are highly concentrated in the spatiotemporal dimension, and identifies the corresponding regions and time periods as high-intensity incident points as recommended clues. The abnormal behavior identification involves identifying police officers or units whose handling results are consistently non-compliant based on the verification results, serving as clues for law enforcement quality supervision; and identifying abnormally active police-related entities as key clues for attention based on the frequency of police incident associations of individual entities within a preset time window.

[0025] The system further includes: The report generation module is connected to the data parsing module, the handling method matching module, the handling result verification module, and the clue recommendation module. It is used to generate a police incident analysis report based on the multidimensional features, handling methods, police incident handling results, verification results, and recommended clues, and to visualize the police incident analysis report. The supervision and management module is connected to the handling result verification module, clue recommendation module and statistical analysis module. It is used to manage the process of issuing, signing for, providing feedback on and completing supervision tasks for non-compliant police reports or important clues. The statistical analysis module, connected to the supervision and management module, is used to perform statistical analysis on the supervision status, generate statistical reports, and visualize the statistical reports.

[0026] The following describes the specific details of the intelligent analysis system for police incident data based on an open-source model, as described in this embodiment of the invention. Figure 2 The diagram (including the relationship between the software system and hardware device in the embodiments of the present invention, as well as the data flow and interaction relationship between the data acquisition module, data parsing module, disposal method matching module, disposal result verification module, clue recommendation module, report generation module, supervision and management module, and statistical analysis module) provides a detailed description of the above-mentioned technical solutions of the embodiments of the present invention.

[0027] I. System Architecture The intelligent police incident data analysis system used in this embodiment of the invention comprises two parts: a software system and a hardware device. The software system is deployed on a server and connects to multiple police incident data sources via a network to perform functions such as data acquisition, parsing, analysis, report generation, supervision and management, and statistical analysis. The hardware device is a dedicated police incident data processing terminal equipped with a high-performance processor, large-capacity memory, and a high-speed network interface, used to run the software system and provide a stable and reliable data processing and analysis environment.

[0028] II. Software System 1. Data Acquisition Module: Used to acquire police data to be analyzed from police data sources (such as police reporting systems, manually imported police data), and supports access to multiple data formats (such as text, tables).

[0029] 2. Data Analysis Module: Based on the characteristics of the police data to be analyzed, open-source models (such as natural language processing models, image recognition models, etc.) are used to analyze the type of such data (such as disputes, domestic violence, theft, fights, traffic accidents, etc.), the entity elements involved (such as the names, ID numbers, license plate numbers, mobile phone numbers, locations, etc. of the police personnel involved), and the relationships between entities (such as the relationship between the suspect and the victim, the relationship between the vehicle and the case, etc.).

[0030] 3. Handling Method Matching Module: After identifying the type of police report, the module matches the corresponding handling method (such as on-site mediation, case investigation, traffic control, etc.) according to the preset mapping rules between police report type and handling method, and pushes the police report data to the corresponding handling process.

[0031] 4. Response Result Verification Module: After identifying the response result, the module verifies its compliance according to pre-defined response regulations (such as response time requirements, evidence collection requirements, and applicable laws). For non-compliant responses, the module outputs a modification plan (such as supplementing evidence or reclassifying the incident) and feeds it back to the response process for correction.

[0032] 5. Clue Recommendation Module: This module comprehensively analyzes the frequency, type, and location of entity elements in police incident data and uses data mining algorithms (such as association rule mining and cluster analysis) to generate recommended clues, including clues about people, vehicles, and mobile phone numbers that are frequently involved in police incidents, as well as locations with high incidence of police incidents.

[0033] 6. Report Generation Module: This module dynamically analyzes discovered clues and generates a police incident analysis report by combining temporal, spatial, and typological information from the incident data. The report covers an overall overview of the incident, analysis of frequently involved entities, high-incidence areas and time periods, quality assessment of response results, and targeted work recommendations. Presented in visual charts and text descriptions, it provides intuitive evidence for police decision-making.

[0034] 7. Supervision and Management Module: This module provides management functions for issuing, signing for, providing feedback on, and completing supervision procedures for police incidents with non-compliant handling results and other important leads. It can automatically assign supervision tasks to relevant units or personnel, record signing times, feedback information, and completion status, ensuring that important police incidents are handled promptly and effectively.

[0035] 8. Statistical Analysis Module: This module performs statistical analysis on the supervision status of each unit, including indicators such as the number of supervised tasks, on-time completion rate, and overdue uncompleted tasks. It generates statistical reports and visualization charts to provide quantitative evaluation basis for police management work and help optimize police resource allocation and workflow.

[0036] III. Hardware Devices The hardware device is a dedicated police data processing terminal, such as... Figure 3 The diagram shows the external structure and internal component layout of the police data processing terminal, including the location and connection relationships of the processor, memory, network interface, security module, human-computer interaction interface, etc., and it has the following characteristics: High-performance processor: It adopts a multi-core high-performance processor, which can quickly process massive amounts of police data and meet the system's high efficiency requirements for data processing.

[0037] Large-capacity storage: Equipped with a large-capacity solid-state drive and memory to store alarm data, knowledge graphs, analysis models, statistical reports and other information, ensuring data read and write speed and stability during system operation.

[0038] High-speed network interface: Equipped with a high-speed Ethernet interface and wireless network module, it can conduct high-speed and stable data communication with alarm data sources, servers, user terminals and other devices to ensure the real-time performance and response speed of the system.

[0039] Security protection features: Built-in firewall, data encryption module and other security protection measures ensure the security of alarm data during transmission and storage, and prevent data leakage and tampering.

[0040] Human-computer interaction interface: Equipped with a high-resolution touch screen and physical buttons, it facilitates police officers to perform interactive operations such as system operation, data query, and report browsing, thereby improving the user experience.

[0041] The specific implementation process of this invention is as follows: 1. Implementation of the data acquisition module The data acquisition module employs various methods, including data interface integration, file import, and web crawling, to obtain police incident data from data sources such as the police incident reporting system, on-site law enforcement recording system, and video surveillance system. For example, it retrieves police incident text data from the police incident reporting system in real time via API interfaces and obtains video clip data related to the police incident from the video surveillance system.

[0042] 2. Implementation of the data parsing module The data parsing module utilizes open-source natural language processing models (such as BERT and GPT) to perform semantic parsing on police report text data, extracting key information such as the type of incident, information about the officers involved, and the course of events. It employs image recognition models (such as YOLO and SSD) to analyze police-related images and videos, identifying entity elements such as vehicle license plates and facial features. Simultaneously, a knowledge graph is constructed to store and associate the parsed entity elements and their relationships, providing structured data support for subsequent analysis. Specifically, as shown below... Figure 4 As shown, the process of parsing police incident data is described in detail, from data input to the steps of parsing the incident type, entity elements and relationships using an open-source model, as well as the knowledge graph construction process.

[0043] 3. Implementation of the treatment method matching module The handling method matching module establishes a knowledge base mapping police incident types to handling methods. Based on characteristics such as incident type keywords and event nature, it quickly determines the corresponding handling method through a matching algorithm. For example, for "theft" incidents, it matches handling methods such as case filing and investigation, and scene investigation; for "traffic accident" incidents, it matches handling methods such as traffic control and accident liability determination. The police incident data is then pushed to the corresponding handling process, such as automatically assigning it to relevant police departments for handling via a workflow engine.

[0044] 4. Implementation of the handling result verification module The incident handling result verification module establishes detailed incident handling regulations, covering aspects such as response time, evidence collection, and applicable laws. It utilizes an open-source model to perform compliance checks on the incident handling result text, such as using a rule engine to determine whether the response time complies with regulations and whether evidence collection is complete. For non-compliant incident handling results, it generates modification plans based on preset modification rules, such as prompting for supplementary evidence types or reclassifying the incident, and feeds this feedback back to the handling process for correction via message pushes and work order feedback.

[0045] in, Figure 5 It details the matching process for handling police incident data and the verification process for handling results.

[0046] 5. Implementation of the clue recommendation module The clue recommendation module collects information such as the frequency, type, and location of entity elements in police incident data and analyzes it using data mining algorithms. For example, it uses association rule mining algorithms to identify clues about police-related personnel and vehicles that frequently appear together, and clustering analysis algorithms to determine high-incidence areas for police incidents. Based on the analysis results, a list of recommended clues is generated to provide key directions for police investigations.

[0047] 6. Implementation of the report generation module The report generation module combines the time, location, and type of information from police incident data, and uses data visualization technology to generate incident analysis reports. The reports present overall incident trends, distribution of frequently involved entities, and high-incidence areas and time periods in chart form, while providing detailed descriptions of incident characteristics, response quality, and work recommendations in text form. The reports can be provided to police officers through system interface display and file export.

[0048] 7. Implementation of the Supervision and Management Module The supervision and management module automatically generates supervision tasks for police incidents with non-compliant handling results and other important clues. These tasks clearly specify the supervision content, responsible unit or personnel, and completion deadline, and are automatically distributed to relevant units or personnel via system messages. Relevant units or personnel acknowledge receipt of the supervision tasks in the system and provide timely progress feedback during processing, including measures taken, problems encountered, and estimated completion time. After a supervision task is completed, the system reviews the completion status to determine if the expected goals have been achieved. Tasks that do not meet the requirements are returned for reprocessing; completed tasks are recorded with the completion time and processing results.

[0049] like Figure 6 As shown, the complete process from task generation, issuance, signing, feedback, completion review, and statistical analysis is demonstrated, clearly defining the operational steps and data flow of each stage.

[0050] 8. Implementation of the statistical analysis module The statistical analysis module performs statistical analysis on the supervision status of each unit, with statistical indicators including the number of supervised tasks, on-time completion rate, and overdue uncompleted tasks. It generates statistical reports and visualization charts, such as bar charts displaying the number of supervised tasks for each unit and line charts showing the on-time completion rate trend over time, providing quantitative evaluation basis for police management work and helping to optimize police resource allocation and workflow.

[0051] In summary, the intelligent police data analysis system based on an open-source model proposed in this invention provides comprehensive and accurate decision support and effective execution supervision for police work through intelligent data processing and analysis methods. It has significant innovation and practicality, and can effectively improve the efficiency and quality of police work, which is of great significance for maintaining social security and stability.

[0052] For example, after introducing the intelligent police incident data analysis system of this invention, a certain unit achieved efficient processing and in-depth analysis of massive amounts of police incident data. The system obtains police incident data from the police reporting system, parses the incident type, involved entities and relationships using an open-source model, and matches corresponding handling methods. In the handling result verification stage, the system checks the compliance of the handling results according to preset regulations, and promptly provides feedback on modification plans for non-compliant handling results to ensure the standardization of police incident handling. The clue recommendation module comprehensively analyzes police incident data to generate clues such as people, vehicles, and mobile phone numbers frequently involved in police incidents, as well as locations with high incidence of police incidents, providing precise direction for police work. The police incident analysis reports generated by the report generation module provide strong support for police decision-making. Meanwhile, for police incidents with non-compliant handling results and other important clues, the supervision and management module realizes full-process management from the generation, issuance, signing, and feedback of supervision tasks to completion review. Through statistical analysis of the supervision status of various units, it optimizes the allocation of police resources and workflows, improving the overall efficiency of police work.

[0053] The intelligent police incident data analysis system based on an open-source model proposed in this invention aims to provide efficient and accurate decision support for police work through in-depth mining and analysis of police incident data, and to achieve supervision process management and statistical analysis of non-compliant police response results and other important clues. Specifically, it includes: 1. Improve the efficiency of police incident analysis: By introducing the powerful data processing capabilities of open source models, police incident data can be quickly parsed and analyzed, which can significantly shorten the police incident analysis cycle and enable police decisions to be adjusted more promptly based on the latest police situation.

[0054] 2. Enhance the accuracy of police incident analysis: By deeply mining the entity elements and their relationships in police incident data, accurately identify the types of police incidents and the methods of handling them, accurately verify the compliance of the handling results, provide more accurate clues and decision support for police work, and reduce misjudgments caused by human experience bias.

[0055] 3. Strengthen the supervision and management of police incidents: Establish a systematic supervision and management mechanism to ensure that police incidents with non-compliant results and other important clues are handled in a timely and effective manner, thereby improving the execution and credibility of police work, providing quantitative evaluation basis for police management work, and helping to optimize the allocation of police resources and work processes.

[0056] 4. Fully explore the value of police incident data: deeply analyze the potential information in police incident data, such as frequently involved police entities and high-incidence locations of police incidents, to provide powerful clues for crime fighting and public security prevention and control, enhance the initiative and foresight of police work, and better maintain social security and stability.

[0057] Based on the above analysis, it can be seen that the embodiments of the present invention have made systematic improvements to address the shortcomings of the prior art, and can effectively solve the problems of low efficiency, insufficient accuracy, imperfect supervision and management, and insufficient data value mining in existing police situation analysis methods.

[0058] Method Implementation Examples According to embodiments of the present invention, an intelligent analysis method for police incident data based on an open-source model is provided. Figure 7 This is a flowchart of the intelligent analysis method for police incident data based on an open-source model, as described in this embodiment of the invention. Figure 7 As shown, the intelligent analysis method for police incident data based on an open-source model according to an embodiment of the present invention specifically includes: Step S701: Obtain multi-source alarm data through the data acquisition module and transmit the multi-source alarm data to the data parsing module; wherein, the multi-source alarm data includes historical alarm data and current alarm data; Step S702: The multi-source police incident data is analyzed by the data parsing module using an open-source model to obtain multi-dimensional features of the police incident data, and the multi-dimensional features are sent to the handling method matching module and the clue recommendation module. Step S703: The handling method matching module matches the corresponding handling method according to the multi-dimensional features, pushes the relevant police data to the corresponding handling process according to the handling method, receives the police handling results in real time, and transmits the handling method and the police handling results to the handling result verification module and the clue recommendation module. Step S704: The handling result verification module verifies the handling method and the police incident handling result according to the preset handling regulations, sends the verification result to the clue recommendation module, and selects whether to perform the correction operation based on the verification result; Step S705: The clue recommendation module generates recommended clues based on the multi-dimensional features, handling methods, police incident handling results, and verification results, and feeds the recommended clues back to the handling method matching module to perform real-time dynamic optimization of handling method matching.

[0059] The method further includes: The report generation module generates a police incident analysis report based on the multidimensional features, handling methods, incident handling results, verification results, and recommended clues, and then visualizes the police incident analysis report. The supervision and management module manages the process of issuing, receiving, providing feedback on, and completing supervision tasks for non-compliant police reports or important leads. The statistical analysis module performs statistical analysis on the supervision situation, generates statistical reports, and then visualizes the statistical reports.

[0060] The following describes in detail the above-mentioned technical solutions of the present invention with reference to the specific circumstances of the intelligent analysis method for police incident data based on the open-source model in the embodiments of the present invention.

[0061] This invention proposes an intelligent analysis method for police incident data based on an open-source model, comprising the following steps: 1. Data Acquisition: Acquire the alarm data to be analyzed from one or more alarm data sources to ensure the integrity and accuracy of the data.

[0062] 2. Data Analysis: Based on the characteristics of the police incident data to be analyzed, an open-source model is used to analyze the type of this type of data, the entity elements involved, and the relationships between entities, providing a basic data structure for subsequent analysis.

[0063] 3. Matching of handling methods: After identifying the type of alarm data, the corresponding handling method is matched, and the alarm data is pushed to the corresponding handling process to achieve accurate diversion and handling of alarms.

[0064] 4. Verification of handling results: After identifying the handling results of the police incident, verify the compliance of the handling results according to the preset handling regulations. For non-compliant handling results, output modification plans and provide feedback for correction to ensure the standardization and legality of police incident handling.

[0065] 5. Clue Recommendation: By comprehensively analyzing the frequency, type, location, and other relationships of entity elements in the police incident data, data mining algorithms are used to generate recommended clues, providing key investigation directions for police work.

[0066] 6. Report Generation: Dynamically analyze the results of discovered clues and generate police situation analysis reports by combining multi-dimensional information from police incident data, providing comprehensive and accurate references for police decision-making.

[0067] 7. Supervision and Issuance: For police incidents with non-compliant handling results and other important clues, supervision tasks are generated, specifying the supervision content, responsible unit or personnel, completion deadline, and other information, and are automatically issued to relevant units or personnel.

[0068] 8. Receipt and Feedback: Relevant units or personnel shall sign for the supervised tasks within the specified time and provide timely feedback on the progress during the process. Feedback information shall include the measures taken, the problems encountered, and the estimated completion time.

[0069] 9. Completion Review: After the completion of the supervised tasks, review the completion status to determine whether the expected goals have been achieved. For tasks that have not met the requirements, return them for reprocessing. For completed tasks, record the completion time, processing results, and other information.

[0070] 10. Statistical Analysis: Conduct statistical analysis on the supervision of each unit, generate statistical reports and visualization charts, provide quantitative evaluation basis for police management work, and help optimize the allocation of police resources and work processes.

[0071] This invention proposes a police incident data analysis method based on an open-source model, which includes multi-source data acquisition, entity element extraction, compliance verification, clue recommendation, and report generation.

[0072] The physical elements include personnel, vehicles, mobile phone numbers, and the location of the incident; The compliance check automatically determines whether the police response results are compliant based on the police rule base. The clue recommendation uses a frequent pattern mining algorithm to identify high-frequency police-related entities; The report includes crime statistics, trend predictions, and handling recommendations.

[0073] This invention, through a combination of open-source models, automated processes, and intelligent recommendations, addresses key pain points in police incident data analysis. It significantly outperforms traditional methods in terms of efficiency, accuracy, compliance, and practicality, and has broad application prospects. Specifically, it includes: 1. Improve the efficiency and effectiveness of police officers' handling of emergency calls and lay a solid foundation for emergency call data. Based on the emergency response procedures, the emergency incident classification and handling procedures, the emergency incident feedback mechanism, and relevant laws and regulations, the system introduces AI big data model capabilities to recommend accurate emergency incident classification tags and emergency response suggestions for each type of emergency incident, thereby achieving efficient emergency incident handling and improving the efficiency of the handling process. At the same time, it provides quality review of emergency response results, automatically marks non-standard emergency response results, provides supervision and follow-up suggestions, ensures the quality of emergency incident data, and lays a solid foundation for emergency incident data.

[0074] 2. Intelligent recommendations for various clues related to illegal activities, crimes, and disputes, enabling early detection of potential extreme and malicious incidents and enhancing public well-being. By introducing AI big data model capabilities, the system can automatically extract and classify various police-related entity elements, uncover hidden information such as interpersonal relationships, timelines, and behavioral patterns within police data, focus on recurring police incidents, and intelligently recommend high-value clues for frequent police-related incidents involving multiple police officers, multiple police officers for a single incident, or multiple police officers in a single location. This information can be applied to various business scenarios such as investigating criminal leads and preventing and controlling key individuals involved in conflicts and disputes, helping to discover key clues that are easily overlooked by traditional manual analysis, preventing the occurrence of extreme individual malicious incidents, and improving the early warning and prevention capabilities of grassroots public security.

[0075] 3. Promote intelligent policing, enhance the public security's proactive defense capabilities, and maintain social stability. The large-scale model empowers the construction of a smart policing ecosystem, recommends tiered and categorized emergency response guidelines, and intelligently verifies emergency response results, enabling the handling of police incidents to shift from "experience-driven" to "data-driven." It proactively detects high-incidence incidents, high-risk locations, sensitive events and accidents, and requests for help from vulnerable groups, promoting the transformation of police work from "passive strikes" to "proactive defense," providing scientific and precise support for public security decision-making, and improving the overall intelligence level of police work.

[0076] The embodiments of the present invention are method embodiments corresponding to the system embodiments described above. The specific operations of each step can be understood by referring to the description of the system embodiments, and will not be repeated here.

[0077] In summary, compared with the prior art, the embodiments of the present invention have the following beneficial effects: 1. Improve the efficiency of police incident analysis: By leveraging the powerful data processing capabilities of open-source models, police incident data can be quickly parsed and analyzed, significantly shortening the police incident analysis cycle and enabling police decisions to be adjusted more promptly based on the latest police situation.

[0078] 2. Enhance the accuracy of police incident analysis: By deeply mining the entity elements and their relationships in police incident data, accurately identify the types of police incidents and the methods of handling them, accurately verify the compliance of the handling results, provide more accurate clues and decision support for police work, and reduce misjudgments caused by human experience bias.

[0079] 3. Optimize police response workflow: Based on the results of police incident analysis, provide targeted suggestions for police response work, help optimize police force allocation, adjust police response strategies, improve police response efficiency and quality, and enhance police work effectiveness.

[0080] 4. Facilitating Scientific Police Decision-Making: The generated police situation analysis reports present the situation and analysis results with intuitive visual charts and detailed text descriptions, providing comprehensive and scientific basis for police leadership decisions and promoting the transformation of police decision-making from experience-based to data-driven.

[0081] 5. Uncover the value of police incident data: Fully explore the potential information in police incident data, such as clues about frequently involved entities and locations with high incidence of police incidents, to provide powerful clues for crime fighting and public security prevention and control, enhance the initiative and foresight of police work, and better maintain social security and stability.

[0082] 6. Strengthen the supervision and management of police incidents: Through systematic supervision process management, ensure that police incidents with non-compliant results and other important clues are handled in a timely and effective manner, thereby improving the execution and credibility of police work; at the same time, conduct statistical analysis on the supervision situation to provide quantitative evaluation basis for police management work, help optimize the allocation of police resources and work processes, and improve the level of refinement of police management.

[0083] Device Example 1 This invention provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, performs the steps described in the method embodiment.

[0084] Device Example 2 This invention provides a computer-readable storage medium storing an information transmission implementation program, which, when executed by a processor, performs the steps described in the method embodiment.

[0085] The computer-readable storage media described in this embodiment include, but are not limited to, ROM, RAM, disk, or optical disk.

[0086] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An open source model-based intelligent analysis system for police case data, characterized in that, The system comprises: a data acquisition module connected with a data analysis module, configured to acquire multi-source police incident data and transmit the multi-source police incident data to the data analysis module; wherein the multi-source police incident data comprises historical police incident data and current police incident data; a data analysis module connected with the data acquisition module, a disposition method matching module and a clue recommendation module, configured to analyze the multi-source police incident data by using an open source model to obtain multi-dimensional features of the police incident data, and send the multi-dimensional features to the disposition method matching module and the clue recommendation module; a disposition method matching module connected with the data analysis module, a disposition result checking module and the clue recommendation module, configured to match a corresponding disposition method according to the multi-dimensional features, push relevant police incident data to a corresponding disposition process according to the disposition method, and receive police incident disposition results in real time, and transmit the disposition method and the police incident disposition results to the disposition result checking module and the clue recommendation module; a disposition result checking module connected with the disposition method matching module and the clue recommendation module, configured to check the disposition method and the police incident disposition results according to a preset disposition regulation, send the checking results to the clue recommendation module, and select whether to perform a correction operation according to the checking results; a clue recommendation module connected with the data analysis module, the disposition method matching module and the disposition result checking module, configured to generate a recommended clue based on the multi-dimensional features, the disposition method, the police incident disposition results and the checking results, and feed back the recommended clue to the disposition method matching module to dynamically optimize the disposition method matching in real time.

2. The system of claim 1, wherein, The system further comprises: a report generation module connected with the data analysis module, the disposition method matching module, the disposition result checking module and the clue recommendation module, configured to generate a police incident analysis report according to the multi-dimensional features, the disposition method, the police incident disposition results, the checking results and the recommended clue, and visually display the police incident analysis report; a supervision management module connected with the disposition result checking module, the clue recommendation module and a statistical analysis module, configured to manage the issuance, signing, feedback and settlement of supervision tasks for non-compliant police incidents or important clues; a statistical analysis module connected with the supervision management module, configured to statistically analyze the supervision situation, generate a statistical report, and visually display the statistical report.

3. The system of claim 1, wherein, The data analysis module is specifically configured to: perform named entity recognition, keyword extraction and classification on police incident text by using a fine-tuned open source natural language processing model, extract police-related entity elements and police incident types, and perform relationship extraction on the police-related entity elements to construct a knowledge graph of the police-related entity elements and their associated relationships; convert a police call recording into text by using an open source speech recognition model, and input the converted text into a natural language processing model for processing to obtain key police incident information; identify at least one of a face, a vehicle, a license plate number and a scene element from a police incident photo or video by using an open source image recognition and target detection model as the police-related entity elements.

4. The system of claim 1, wherein, The disposition result checking module is specifically configured to: establish a police case checking rule library, wherein the police case checking rule library is pre-set based on police case handling specifications; automatically check the handling method and the police case handling result according to the handling regulations pre-set in the police case checking rule library; if the checking result shows compliance, no correction operation is needed; if the checking result shows non-compliance, a checking report containing specific non-compliant items and modification suggestions is automatically generated according to pre-set modification rules, and the checking report is fed back to the handling method matching module to dynamically correct the handling method matching in real time.

5. The system of claim 4, wherein, The police case handling specifications include timeliness rules, element integrity rules and logical consistency rules; The timeliness rules are used to check whether the police response time, on-site handling time and case settlement period exceed the specified threshold; The element integrity rules are used to check whether the essential handling elements are complete; the essential handling elements include at least one of the following: information record of involved personnel, uploading of evidence materials and citation of legal provisions; The logical consistency rules are used to check whether there is a logical contradiction between police case classification, handling measures and final results.

6. The system of claim 1, wherein, The clue recommendation module is specifically configured to: based on the multi-dimensional features, handling methods, police case handling results and checking results, at least one of the following mining analyses is performed: frequent pattern mining, spatio-temporal clustering analysis and abnormal behavior identification; The frequent pattern mining is to mine frequent item sets meeting minimum support and confidence based on historical police case data using association rule algorithm, and generate high-frequency co-occurring police-related entity combinations as recommended clues; the police-related entity combinations include at least one of the following: person-person, person-vehicle and person-place; The spatio-temporal clustering analysis is to identify highly clustered police case clusters in the time and space dimensions based on the time and geographic location information of police cases, and determine the corresponding area and time period as a police case high-risk hotspot as a recommended clue; The abnormal behavior identification is to identify police officers or units with continuously non-compliant handling results based on the checking results as law enforcement quality supervision clues; and to identify abnormally active police-related entities based on the police case association frequency of individual entities within a pre-set time window as key focus clues.

7. An open source model-based intelligent analysis method for police case data, characterized in that, It includes: acquire multi-source police case data through a data acquisition module, and transmit the multi-source police case data to a data analysis module; wherein the multi-source police case data includes historical police case data and current police case data; analyze the multi-source police case data using an open source model through the data analysis module to obtain multi-dimensional features of the police case data, and send the multi-dimensional features to the handling method matching module and the clue recommendation module; match the corresponding handling method according to the multi-dimensional features through the handling method matching module, push related police case data to the corresponding handling process according to the handling method, and receive police case handling results in real time, and transmit the handling method and the police case handling results to the handling result checking module and the clue recommendation module; The disposition result checking module checks the disposition method and the police case disposition result according to preset disposition regulations, sends a checking result to the clue recommendation module, and selects whether to perform a correction operation according to the checking result; The clue recommendation module generates a recommended clue based on the multi-dimensional features, the disposition method, the police case disposition result, and the checking result, and feeds back the recommended clue to the disposition method matching module to perform real-time dynamic optimization on disposition method matching.

8. The method of claim 7, wherein, The method further includes: The report generation module generates a police case analysis report according to the multi-dimensional features, the disposition method, the police case disposition result, the checking result, and the recommended clue, and visually displays the police case analysis report; The supervision management module manages the issuance, signing, feedback, and settlement of supervision tasks for non-compliant police cases or important clues; The statistical analysis module statistically analyzes the supervision situation, generates a statistical report, and visually displays the statistical report.

9. An electronic device, comprising: It includes: A memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the open source model-based police case data intelligent analysis method according to any one of claims 7-8.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores an information transmission implementation program, and the program, when executed by the processor, implements the steps of the open source model-based police case data intelligent analysis method according to any one of claims 7-8.

Images