Basic-level governance law enforcement scene intelligent research and judgment method and storage medium

Abstract

The invention relates to the technical field of information, in particular to a knowledge graph-based grassroots governance law enforcement scene intelligent research and judgment method and a storage medium. A technical scheme capable of dynamically studying and judging, predicting risks and optimizing decision support is provided by combining a knowledge graph, a graph neural network GNN and a plurality of reasoning algorithms, so that the intelligence and informatization level in a grassroots governance law enforcement process is improved, the analysis, studying and judging and decision-making capabilities of law enforcement departments are enhanced, complex law enforcement challenges are dealt with, and the law enforcement risk prediction and risk prediction are realized. And effective implementation of social public safety and environmental protection is ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, specifically a grassroots governance and law enforcement scene intelligent research and judgment method and storage medium. In particular, it relates to a grassroots governance and law enforcement scene intelligent research and judgment method based on a knowledge graph and a storage medium. BACKGROUND

[0002] In the process of grassroots governance and law enforcement, the amount of information is large and complex, and traditional manual processing methods cannot efficiently and accurately analyze and research various law enforcement scenes and problems. In the prior art, although some systems can provide basic data statistics and analysis, they lack deep knowledge association and intelligent research and judgment capabilities, and cannot fully utilize big data and artificial intelligence technology for effective decision support.

[0003] For example, the prior art solution: an event processing method and device based on a knowledge graph (related patent document: CN110489569A). This solution uses traditional database technology to classify and statistically analyze urban governance data through rules, and the main steps include: 1. Data collection: structured data is collected in a fixed format. 2. Data analysis: use SQL statements to classify and query data. 3. Result display: display statistical results in table or simple chart form. The shortcomings are: insufficient information correlation: only simple classification and statistics can be performed, and dynamic correlation of multi-dimensional data or deep relationship mining cannot be performed. Insufficient dynamicity: relies on predefined rules, lacks dynamic expansion analysis capability, and cannot handle complex scenarios. Insufficient prediction capability: no intelligent analysis function, cannot implement event prediction and risk assessment. Insufficient technical innovation: completely based on traditional database technology, without using graph or intelligent technology. SUMMARY

[0004] The technical solution of the present application proposal: a grassroots governance and law enforcement scene intelligent research and judgment method based on a knowledge graph, combines a knowledge graph, a graph neural network (GNN), and multiple reasoning algorithms to provide a technical solution capable of dynamic research and judgment, risk prediction, and optimized decision support, and the main steps include: Data collection and preprocessing: multi-source data fusion, including monitoring data, complaint records, laws and regulations, historical law enforcement records, etc.

[0005] Knowledge graph construction: identify entities (such as pollution sources, enterprises) and relationships (such as the association between pollution sources and laws and regulations), and form a complete data network.

[0006] Intelligent research and judgment: use GNN and reasoning algorithms to perform deep analysis of the knowledge graph, implement event prediction and risk assessment for complex scenarios.

[0007] Decision support: generate optimized suggestions and visual reports to provide efficient and scientific decision-making basis for management departments.

[0008] The technical advantages of this application proposal are: First, multi-dimensional data association and analysis. The proposal associates multi-source data through knowledge graphs, which can extract complex relationships and provide a richer semantic basis for analysis. Data that is difficult to dynamically process and model with existing technologies (such as surveillance videos and complaint records) can be effectively utilized in this solution. Second, dynamic intelligent analysis and judgment capabilities. Based on GNN and inference algorithms, complex scenario analysis is achieved, and risk events are dynamically predicted. Compared with existing technologies, it can adapt to a variety of scenarios and provide accurate prediction results. Third, high scalability. The design of the knowledge graph supports multi-scenario applications and is suitable for environmental monitoring, noise nuisance, illegal buildings and other fields. The system continuously improves its analysis capabilities through data updates and model optimization. Fourth, comprehensive decision support, which not only analyzes and predicts events, but also provides risk assessments, optimization suggestions and visualization reports to help management departments make scientific decisions.

[0009] Compared with existing technologies, this application proposal has made up for the deficiencies in information relevance, dynamic prediction capabilities and scalability through technological innovation, and proposed a more efficient and intelligent solution for grassroots governance and law enforcement scenarios.

[0010] The main purpose of this invention is to improve the intelligence and informatization level of grassroots governance and law enforcement through technological innovation, enhance the analysis, judgment and decision-making capabilities of law enforcement departments, so as to cope with complex law enforcement challenges and ensure the effective implementation of social public security and environmental protection.

[0011] 1. Improve the intelligence level of grassroots governance and law enforcement: By constructing and utilizing an intelligent analysis and judgment system based on knowledge graphs, the present invention aims to enhance the intelligent analysis and judgment capabilities in grassroots governance and law enforcement scenarios, enabling law enforcement agencies to analyze and respond to complex law enforcement scenarios more accurately and efficiently.

[0012] 2. Realize cross-departmental and cross-domain data sharing and analysis: This invention breaks the data silos between different departments through knowledge graph technology, realizes cross-departmental and cross-domain data association and sharing, and thus provides more comprehensive and accurate information support for law enforcement decision-making.

[0013] 3. Improve law enforcement efficiency and decision-making accuracy: By introducing artificial intelligence technology, especially graph neural networks and inference algorithms, this invention can provide real-time event prediction, risk assessment and optimized decision-making support in complex and changing law enforcement environments, thereby significantly improving law enforcement efficiency and decision-making accuracy.

[0014] 4. Provide innovative and forward-looking law enforcement technical solutions: By optimizing relevant calculation methods and formulas, the present invention proposes an innovative intelligent analysis method, which can provide forward-looking and feasible technical solutions for future grassroots governance and law enforcement fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 For the method flowchart of the present application; Figure 2 For the knowledge graph construction flowchart; Figure 3 For the intelligent research and judgment flowchart. DETAILED DESCRIPTION

[0016] To solve the above technical problems, a knowledge graph-based grassroots governance law enforcement scene intelligent research and judgment method, system and storage medium are provided, which solve the above-mentioned deficiencies of the prior art.

[0017] To achieve the above purposes, the technical scheme adopted by the present application is: A knowledge graph-based grassroots governance law enforcement scene intelligent research and judgment method, comprising: Step 1, data acquisition and preprocessing Step 1.1, data source Data acquisition is the basis of the entire system, and its sources include but are not limited to law enforcement records, monitoring videos, sensor data, public reporting information, historical case data, relevant laws and regulations of public security, environmental protection, urban management and other departments. Data types are diverse, including structured data (such as database records), semi-structured data (such as log files), unstructured data (such as text, images, videos, etc.).

[0018] Step 1.2, data preprocessing Further, in order to ensure data quality, the collected data needs to be cleaned, de-duplicated, and format-converted for preprocessing. The specific steps include: Step 1.2.1, data cleaning: remove noise and error information in the data, such as correcting spelling errors, filling missing values, etc.

[0019] Step 1.2.2, data de-duplication: eliminate duplicate data to ensure data uniqueness.

[0020] Step 1.2.3, format conversion: convert data of different formats to a unified format for subsequent processing.

[0021] Step 1.2.4, data normalization: standardize the data, such as unit conversion, time format unification, etc.

[0022] Step 2, knowledge graph construction Step 2.1, semantic analysis Further, natural language processing (NLP) technology is used to analyze the semantics of the data, extract key entities and relationships. The main steps include: Step 2.1.1, Entity Recognition: Identify key entities such as people, places, events, etc. from the text. For example, identify pollution sources, law enforcement officers, enforcement times, etc. in environmental law enforcement data.

[0023] Step 2.1.2, Relation Extraction: Identify relationships between entities. For example, the emission relationship between a pollution source and a company, the handling relationship between a law enforcement officer and a case.

[0024] Step 2.1.3, Attribute Extraction: Identify attribute information of entities. For example, the specific location of a pollution source, the emission amount, etc.

[0025] Step 2.2, Knowledge Graph Association Further, based on the extracted entities and relationships, a grassroots governance knowledge graph is constructed. Graph databases (such as Neo4j) are used for storage and management. The construction process of the knowledge graph includes: Step 2.2.1, Graph Structure Design: Determine the node (entity) and edge (relationship) structure of the knowledge graph.

[0026] Step 2.2.2, Data Import: Import preprocessed data into the graph database to form the knowledge graph.

[0027] Step 2.2.3, Graph Query Optimization: Optimize the query performance of the graph database to ensure efficient querying under large-scale data.

[0028] Step 3, Intelligent Research and Judgment Method Step 3.1, Graph Neural Network (GNN) Further, graph neural networks (GNN) are used to conduct deep learning on the knowledge graph to capture complex semantic relationships. The main steps include: Step 3.1.1, Graph Embedding, map the nodes and edges in the knowledge graph to a low-dimensional vector space for deep learning.

[0029] Step 3.1.2, Model Training, use GNN model to train graph embedding to capture semantic information in graph structure.

[0030] Step 3.1.3, Feature Extraction, extract node and edge features from the trained model for subsequent research and reasoning.

[0031] Step 3.2, Reasoning Algorithm Step 3.2.1, Rule Reasoning, reasoning based on rules in the knowledge graph.

[0032] Step 3.2.2, Probabilistic Reasoning, reasoning based on probabilistic models.

[0033] Step 3.2.3. Deep reasoning, combining deep learning models and knowledge graphs, for complex semantic reasoning.

[0034] Event occurrence probability algorithm where, is the probability of event under knowledge graph and parameter set . is the weight of the th feature, is the th feature function, which depends on parameter .

[0035] : Event to be judged, e.g., a pollution event occurring in a certain area.

[0036] : Knowledge graph, including all related entities and relationships.

[0037] : Parameter set, including parameters for adjusting feature functions.

[0038] : Feature weight, indicating the importance of each feature in the reasoning process.

[0039] : Feature function, representing the influence of a specific feature on the occurrence of event , depending on parameter .

[0040] Risk assessment algorithm where, is the overall risk assessment value, is the probability of event under knowledge graph and parameter set, is the weight of event , is the impact coefficient of event .

[0041] : Overall risk assessment value, used to measure the risk level of the entire law enforcement scene.

[0042] : Each possible event in the knowledge graph.

[0043] : Event The weight indicates the importance of the event.

[0044] :event The impact coefficient reflects the possible consequences of an event.

[0045] Decision support algorithms in, For the best decision-making solution, For decision-making exist Utility in context, For the situation In the knowledge graph and parameters The probability under the set, For the situation The weight of .

[0046] : The best decision-making option, such as choosing a certain law enforcement measure.

[0047] : The set of all possible decisions.

[0048] :decision making In the situation The utility under the condition of , which indicates the effect of the decision in a specific situation.

[0049] : Possible scenarios, derived based on knowledge graph reasoning.

[0050] : Situation The probability in the knowledge graph reflects the possibility of a situation occurring.

[0051] : Situation The weight indicates the importance of the situation.

[0052] Parameter Explanation Feature weights and : Weights are set based on historical data and expert experience to represent the importance of different features and events. Weights can be optimized through model training to achieve optimal results.

[0053] characteristic function : Indicates the impact of specific features on the occurrence of an event. The function form can be designed according to the specific application scenario, such as linear function, nonlinear function, etc. Parameters Adjusting the characteristic function to better fit the actual data.

[0054] Influence coefficient : Represents the potential consequences of an event, such as economic losses, environmental impacts, etc. The influence coefficient can be set through expert evaluation and historical data.

[0055] Context weight : Represents the importance of different contexts, set by expert evaluation. Context weight can help the model better support decision-making.

[0056] Step 4, decision support and optimization Step 4.1, decision support Further, based on the research and judgment results, generate a decision support report and provide optimization suggestions. Main content includes: Step 4.1.1, event prediction: Predict future possible events and their impact. For example, predict the future possible pollution events in a certain area and their impact range.

[0057] Step 4.1.2, risk assessment: Assess the risk level in the current law enforcement scene. For example, assess the illegal emission risk of a certain enterprise.

[0058] Step 4.1.3, optimization suggestions: Based on the research and judgment results, provide optimized law enforcement solutions. For example, suggest strengthening the monitoring intensity of a certain area or taking specific law enforcement actions.

[0059] Step 4.2, visualization tools Further, through visualization tools, display the knowledge graph and research and judgment results to assist law enforcement personnel in decision-making. Main functions include: Step 4.2.1, graph browsing: Intuitively display entities and relationships in the knowledge graph, support node and edge query and analysis.

[0060] Step 4.2.1, data visualization: Show statistical analysis results of data, such as pollution source distribution map, risk assessment map, etc.

[0061] Step 4.2.1, result display: Show research and judgment results and decision support report, support dynamic interaction and custom display.

[0062] Step 5, comprehensive evaluation and feedback adjustment Step 5.1, effect evaluation After implementing intelligent research and decision support, the system tracks the effects of various urban management issues (such as illegal construction, noise disturbance, and road occupation), evaluates the actual implementation effects and improvements, and determines the accuracy of system predictions and the effectiveness of decision measures by comparing expected results with actual implementation results. For example, the system records show that in area A, through increased patrols, the increase of illegal construction has been successfully curbed.

[0063] Step 5.2, Model Feedback and Optimization: According to the results of effect evaluation, the system feeds back the actual law enforcement data to the knowledge graph and adjusts the parameter settings of the prediction model accordingly. By optimizing the weights and adjusting the characteristic function, the system continuously improves the accuracy of future predictions and the effectiveness of decision support. For example, for the case of night-time noise disturbance, the system optimizes the parameters of the noise prediction model based on actual noise monitoring data to improve the response speed and effectiveness of night-time law enforcement.

[0064] Further, in actual application, based on the knowledge graph and intelligent research algorithm, different types of law enforcement scenes can be analyzed and processed. For example, in environmental law enforcement, the future pollution risk of a certain area can be predicted and the illegal emission probability of enterprises can be evaluated based on monitoring data and historical case data; in public security law enforcement, the pattern of criminal activities can be analyzed and high-risk areas can be predicted based on case records and monitoring videos. Through the above calculation formula, the research results can be quantified to generate specific decision support reports to help law enforcement personnel develop scientific and effective law enforcement strategies.

[0065] Further, a computer-readable storage medium is provided, which stores a computer-readable program that, when invoked, performs a knowledge graph-based intelligent research and judgment method for grassroots governance law enforcement scenes as described above.

[0066] The following description is used to disclose the present application so that those skilled in the art can implement the present application. The preferred embodiments in the following description are only as examples, and other obvious modifications can be made by those skilled in the art.

[0067] Embodiment One: The present embodiment provides an intelligent research and judgment system for environmental pollution law enforcement in industrial parks based on a knowledge graph, aiming to improve the early warning capability for environmental pollution events and optimize the allocation of law enforcement resources. Referring to Figure 1 The overall process of the method is as follows: 1. Data collection and preprocessing: Data Sources: Collect pollutant emission monitoring data from 10 enterprises within the industrial park, monitoring data from 5 environmental monitoring stations within the park, public reporting information, and law enforcement record data over the past 3 years. Emission monitoring data mainly includes sulfur dioxide (SO2), nitrogen oxides (NOx), volatile organic compounds (VOC), etc., and monitoring station data includes air quality index (AQI), PM2.5 concentration, etc.

[0068] Data Preprocessing: Clean, deduplicate and format the collected data. For abnormal values in the monitoring data (such as SO2 concentration suddenly increasing to 1000mg / m³), they are removed. The data format is unified to YYYY-MM-DD HH:MM:SS to ensure data consistency.

[0069] 2. Knowledge Graph Construction Entity Recognition: Extract key entities from enterprise emission monitoring data, such as enterprises (A Company, B Company, C Company, etc.), pollutants (SO2, NOx, VOC), and emission equipment (chimney, drain, etc.).

[0070] Extract monitoring indicators from environmental monitoring station data, such as PM2.5 concentration, SO2 concentration, and monitoring station location.

[0071] Extract involved enterprises (such as A Company), penalties (such as fines, production restrictions), and rectification situations from historical law enforcement records.

[0072] Relationship Extraction: Establish the relationship between enterprises and their emissions in the knowledge graph (such as "A Company emits SO2").

[0073] Establish the relationship between emissions and environmental quality (such as "SO2 exceeds the standard leading to air quality decline").

[0074] Establish the correlation between public reports and actual pollution incidents.

[0075] Graph Optimization: Based on entities and relationships, construct a knowledge graph to associate all enterprises, pollutants, monitoring data, law enforcement records, etc.

[0076] Optimize the graph structure to ensure efficient query performance when dealing with complex multi-source data.

[0077] Calculation of Event Occurrence Probability In this example, the system calculates the probability of A Company's SO2 emission exceeding the standard in the next week based on historical data and current trends. The specific steps are as follows: Parameter Settings: Feature Weight of Historical Emission Records = 0.7, weight of recent emission trend = 0.3.

[0078] Feature function Reflects the historical probability of A company exceeding the standard twice a month in the past year (0.2), feature function Reflects the recent emission frequency and exceeding trend (0.3).

[0079] According to the calculation results, the probability of A company exceeding the SO2 emission standard in the next week is 23%.

[0080] Risk assessment In this example, the system assesses the pollution risk of the entire industrial park, and the specific steps are as follows: Single enterprise risk calculation: The probability of A company exceeding the SO2 emission standard is , the probability of B company exceeding the NOx emission standard is , and the probability of C company exceeding the VOC emission standard is .

[0081] Overall risk value: The overall pollution risk of the entire industrial park is 0.1535, which belongs to moderate risk.

[0082] Decision support and optimization Scenario set , , are respectively: : A company continues to exceed the emission standard, leading to deterioration of air quality.

[0083] : B company rectifies and meets the emission standard.

[0084] : C company's emission is stable and the risk is controllable.

[0085] Utility function definition represents the utility of decision under (e.g. the utility of the fine is the probability of reducing emissions, and the utility of the rectification notice is the speed of reducing risk).

[0086] In this example, based on the risk assessment results, the system provides the following decision support for the environmental protection department: Among them, is the best decision scheme, For decision In Context, the utility, For context In the knowledge graph And the probability of parameter Set, For context The weight.

[0087] Candidate decision set A: : Increase A company monitoring frequency (cost: 2 people / day, expected to reduce risk 30%).

[0088] : Notice of rectification to B company (cost: 1 person / day, expected to reduce risk 20%).

[0089] : Regular inspection of C company (cost: 0.5 person / day, expected to reduce risk 10%).

[0090] Context probability and weight: (High risk context weight high) Utility calculation under different scenarios: To (Increase monitoring): Similarly, the comprehensive utility value of And is calculated, and the decision corresponding to the maximum value is selected.

[0091] Output optimal decision and resource allocation: Combined with existing graph data and accumulated experience knowledge, prefer to choose to increase monitoring frequency and publish rectification notice as decision scheme ( ).

[0092] Best decision scheme (Increase A company monitoring frequency), the highest comprehensive utility value.

[0093] Resource allocation suggestion: Deploy additional monitoring equipment during peak emission periods at Company A (night and weekends) and arrange for law enforcement personnel to conduct daily patrols three times a day.

[0094] Link environmental monitoring stations to transmit SO2 concentration data to the system in real time, triggering an early warning when the standard is exceeded.

[0095] For Company B, issue a rectification notice to reduce NOx emissions within a specified time limit.

[0096] Emergency response measures: Advance notice to residential areas downwind of possible air quality deterioration.

[0097] Strengthen air quality monitoring during high-risk periods to ensure the accuracy of real-time monitoring data.

[0098] Comprehensive evaluation and feedback adjustment Effect evaluation: After one week of implementation, monitoring shows that SO2 over-standard events at Company A have been effectively controlled, and after one week, over-standard events at Company A have decreased by 50%, with the risk value reduced to R=0.12.

[0099] NOx emissions at Company B have significantly decreased, and overall pollution levels have been reduced.

[0100] Model feedback and optimization: Based on actual effects, fine-tune the feature weights and feature functions of Company A's emission events to enhance the system's monitoring accuracy for high-risk enterprises.

[0101] Parameter adjustment: Update feature weights Adjust the historical data weight from 0.7 to 0.75 to strengthen the influence of historical patterns. Optimize the feature function Introduce an exponential decay factor for recent emission trends: . Where is the decay coefficient, is the time interval.

[0102] In summary, through the application of this embodiment, the environmental protection department can effectively predict and respond to environmental pollution risks in industrial parks, improve law enforcement efficiency, optimize resource allocation, and provide strong technical support for maintaining urban environmental quality.

[0103] Example Two: In the city management of a certain city, the city management department often faces complex governance problems, such as frequent illegal construction, increasing night noise disturbances, and traffic congestion caused by road occupation in commercial areas. These problems seriously affect the normal operation of the city and the quality of life of residents. Therefore, an intelligent research and judgment system based on knowledge graph is introduced into the city management department to improve the early warning ability and resource allocation efficiency for the above problems.

[0104] Step One: Data Collection and Preprocessing 1. Data Source: Surveillance video data: The city has installed more than 500 high-definition cameras on major roads and key areas to monitor road conditions in real time, generating approximately 1TB of video data daily.

[0105] Resident complaint data: Resident complaints are collected through the citizen hotline and official WeChat platform, with an average of about 200 complaints received every day, covering noise nuisance, illegal construction, and occupying the road for business.

[0106] Historical law enforcement records: law enforcement records over the past five years, including the specific handling of various cases, punishment measures and rectification results.

[0107] Urban planning data: This includes the latest urban planning maps, building permits, road design drawings, etc., provided by the Municipal Planning Bureau, with a data volume of hundreds of GB.

[0108] 2. Data preprocessing: Data cleaning: Remove invalid footage from surveillance videos due to equipment failures and complaint records with duplicate timestamps. Merge duplicate records in historical law enforcement records to ensure data uniqueness.

[0109] Format conversion: All data is standardized to the time format YYYY-MM-DD HH:MM:SS, and video data is converted into analyzable image segments. Urban planning data is converted to GeoJSON format to facilitate spatial analysis.

[0110] Data deduplication: Multiple complaints related to the same event are merged to reduce data redundancy. For example, multiple noise complaints at the same location during the same time period are merged into a single record.

[0111] Step 2: Knowledge graph construction 1. Entity Recognition and Relation Extraction: Building Illegal Activities: Identify building illegal activities by analyzing resident complaints and historical law enforcement records, identifying their location, type, involved personnel, and associated building permits.

[0112] Noise nuisance incidents: Extract related entities of noise nuisance incidents based on the time, location, and type of noise source (such as construction noise and entertainment venue noise) of noise complaints.

[0113] Road occupation: Combine surveillance videos and residents’ complaints to identify the specific location, time and relevant responsible persons of road occupation.

[0114] 2. Relationship Building: The relationship between illegal buildings and land use rights: Illegal buildings are linked to land use rights information in urban planning data to determine their legality.

[0115] Noise disturbance and time period relationship: Analyze the high-frequency time periods of noise disturbance events and compare them with the noise control time in the city management regulations to establish the relationship between time periods and noise disturbance.

[0116] Road occupation and traffic impact relationship: Through monitoring data, analyze the actual impact of road occupation on traffic flow, and establish the correlation between road occupation behavior and traffic congestion.

[0117] 3. Atlas optimization: After completing entity recognition and relationship establishment, optimize the atlas to ensure the efficiency of the atlas structure in data query and complex scenario analysis. Specific optimization measures include reducing irrelevant nodes, merging duplicate relationships, and labeling key nodes.

[0118] Step three: Intelligent research and risk assessment 1. Illegal construction prediction and assessment: Predict whether there will be new illegal construction in a certain city within the next month.

[0119] Historical records: In the past year, there have been 20 illegal construction incidents in this city, of which 15 occurred in the densely built A area.

[0120] Current trends: In the past two months, A area has received 5 reports of unapproved construction activities.

[0121] Among them, =0.6 is the weight of historical data, =0.4 is the weight of current trends, =0.75 represents the historical data of illegal construction occurrence rate, =0.5 represents the current trend of illegal construction risk.

[0122] The probability of illegal construction in A area within the next month is: The result shows that the possibility of illegal construction in A area within the next month is 65%.

[0123] Overall risk assessment: event weight (illegal construction has a large impact on urban planning), impact coefficient (illegal construction may cause safety hazards).

[0124] 2. Noise disturbance event risk assessment: Assess the risk of noise disturbance events in a residential area within the next week.

[0125] Complaint records: In the past three months, this residential area has received 30 complaints of nighttime noise disturbances, mainly concentrated between 10 pm and 2 am.

[0126] Current noise monitoring data: In the past week, the average nighttime noise level in this area has exceeded 60 dB, approaching the upper limit of the city's noise control regulations (50 dB).

[0127] Knowledge graph (noise complaint records, monitoring data), parameters ( , ).

[0128] Overall risk assessment where the weight of complaint records , the weight of noise monitoring data ; the noise risk of historical complaints = 0.6, the current noise over-standard rate g.

[0129] The result shows that the risk of noise disturbance events in this residential area in the next week is 54.6%.

[0130] 3. Risk prediction of road occupation causing traffic congestion: Predict the likelihood of a commercial district experiencing traffic congestion due to road occupation during peak hours.

[0131] Monitoring records: In the past month, the commercial district has experienced 10 traffic congestion incidents caused by road occupation during the early morning peak (7:00-9:00) on weekends.

[0132] Current monitoring data: In the past week, there has been an increasing trend in road occupation, especially on Fridays and Saturdays.

[0133] where the weight of historical records , the weight of current trends ; the rate of traffic congestion caused by historical road occupation , the growth rate of current trends .

[0134] Calculation result: The probability of the commercial district experiencing traffic congestion due to road occupation during the early morning peak in the next week is: The result shows that the likelihood of the commercial district experiencing traffic congestion due to road occupation during the early morning peak in the next week is 50%.

[0135] Overall risk assessment: event weight , impact coefficient .

[0136] Step four: decision support and optimization 1. Decision support and measures for illegal construction: Generate optimal decision scheme Candidate decision set A: : Increase patrol frequency (cost: 2 people / day, expected to reduce the probability of illegal construction by 40%).

[0137] : Speed up legal approval (cost: 1 person / day, expected to reduce the probability of illegal construction by 20%).

[0138] Utility function and scenario probability: : Illegal construction is not found : Illegal construction is not found ， ； ， 。

[0139] Comprehensive utility calculation: Output decision (Increase patrol frequency).

[0140] Law enforcement recommendations: The system recommends strengthening patrols during high noise disturbance time periods (10 pm to 2 am), focusing on checking noise emissions at construction sites and entertainment venues.

[0141] Emergency response: Upon receiving a noise disturbance complaint, the system will automatically dispatch nearby law enforcement personnel for handling and impose fines and rectification on multiple violations.

[0142] 2. Decision support and measures for noise disturbance: Candidate decision set A : Strengthen night patrols (cost: 3 people / night, expected to reduce complaints by 30%).

[0143] : Mandatory installation of soundproofing facilities (Cost: 50,000 RMB per location, expected to reduce complaints by 50%).

[0144] Utility function and scenario probabilities: : Construction noise duration : Entertainment venue noise ， ； ， 。

[0145] Overall utility calculation: Equilibrium decision: Output decision Choose the one with lower cost (Enhance patrols).

[0146] Enforcement recommendation: Based on the prediction results, it is recommended that the city management department increase the frequency of patrols in area A within the next month, especially in densely built-up areas. In addition, during the peak period of construction activities, more law enforcement personnel should be arranged for on-site supervision to strictly control the occurrence of newly built illegal buildings.

[0147] Preventive measures: Strengthen building approval management in area A, speed up the approval progress of legal buildings, to reduce the phenomenon of illegal buildings caused by delayed approval.

[0148] 3. Decision support and measures for road occupation: Candidate decision set A : Increase patrols during peak hours (Cost: 2 people / day, expected to reduce 30% of road occupation).

[0149] : Set up temporary stall area (Cost: 30,000 RMB, expected to reduce 50% of road occupation).

[0150] Utility function and scenario probabilities: : Road occupation causes congestion ( ) : Road occupation does not affect traffic ( ) ， ， Output decision: Output decision (Set up temporary stall area) Enforcement Recommendations: To address traffic congestion caused by street vendors occupying the road, the system recommends increasing patrols during peak weekend hours to strictly enforce the law against vendors illegally occupying the road. Furthermore, during peak hours, it is recommended to set up temporary parking areas to reduce street vendors occupying the road.

[0151] Long-term measures: Plan more reasonable vendor operating areas in commercial areas to reduce their impact on traffic, and establish a long-term management mechanism to ensure smooth traffic.

[0152] Step 5: Comprehensive evaluation and feedback adjustment 1. Effect evaluation: Illegal Construction: Following implementation, new illegal construction in Area A decreased by 60%. Through the system's prediction and early warning capabilities, Area A successfully curbed the occurrence of new illegal construction within the next month. System records indicate that construction activity in the area has been effectively controlled.

[0153] Noise nuisance: After strengthening patrols, noise nuisance complaints decreased by 40%, and noise levels dropped to 55dB, especially during peak nighttime hours. The system shows that the situation of excessive noise has been significantly improved.

[0154] Road occupation: After enforcement actions during peak hours, road occupation was reduced by 30% and traffic congestion was alleviated (40%), demonstrating the effectiveness of the system’s enforcement.

[0155] 2. Model feedback and optimization: Based on actual law enforcement results, the system feeds new data back into the knowledge graph and readjusts the parameters of the prediction model based on the latest data. For example, within Area A, the frequency of inspections in densely populated areas could be appropriately increased to further improve prediction accuracy.

[0156] Illegal buildings Parameter adjustment: increase historical weight To 0.7, reduce the trend weight To 0.3. Update the impact coefficient (Increase in safety complaints due to illegal construction).

[0157] Noise nuisance Parameter adjustment: reduce historical weight To 0.6, increase the weight of the current trend To 0.4. Optimize the influence coefficient (Due to improved resident satisfaction).

[0158] Road occupation Parameter adjustment: reduce historical weight To 0.4, increase the current trend weight To 0.6. Optimize the impact coefficient Due to the improvement of economic activity.

[0159] Further, the present scheme also proposes a computer readable storage medium having stored thereon a computer readable program which, when invoked, performs a knowledge graph-based grassroots governance law enforcement scene intelligent research and judgment method as described above. It can be understood that the storage medium can be a magnetic medium, such as a floppy disk, a hard disk, a magnetic tape; an optical medium such as a DVD; or a semiconductor medium such as a solid state disk (SSD) and the like.

[0160] In summary, through the application of the present embodiment, the urban management department can accurately and efficiently deal with common problems in urban management, improving the intelligent level and law enforcement effect of urban management. At the same time, through continuous feedback and optimization, the prediction ability and decision support effect of the system are continuously improved.

[0161] The specific technical effects are as follows: 1. Data correlation and semantic analysis based on knowledge graph Cross-department data sharing: Use knowledge graph technology to realize cross-department and cross-field data sharing and correlation analysis, breaking down data silos. For example, integrate the data of public security, environmental protection, urban management and other departments in a knowledge graph to form a global view.

[0162] Semantic analysis: Through natural language processing and knowledge graph construction, semantic analysis of complex data is realized to identify key entities and relationships in the data. For example, automatically identify key information such as pollution sources and illegal behavior in law enforcement records.

[0163] 2. Combination of graph neural network and reasoning algorithm Graph Neural Network (GNN): Use graph neural network to learn knowledge graph in depth to capture complex semantic relationships. For example, through the GNN model, identify potential associations between different pollution sources and predict possible future pollution events.

[0164] Reasoning algorithm: Combine rule-based reasoning, probabilistic reasoning and deep reasoning to make intelligent research and judgment on law enforcement scenes. For example, predict the probability of illegal emissions by a certain enterprise; based on a deep learning model, identify patterns of criminal activity.

[0165] 3. Real-time response and optimization Real-time response: Through efficient data processing and analysis, provide real-time law enforcement scene research and decision support. For example, monitor pollution data in real time and respond quickly to sudden pollution incidents.

[0166] Method optimization: based on the judgment result, provide optimized law enforcement scheme, improve the accuracy and effectiveness of law enforcement decision. For example, suggest adjusting law enforcement strategy, strengthening supervision of high-risk enterprises.

[0167] The above shows and describes the basic principles, main features and advantages of the present application. Those skilled in the art should understand that the present application is not limited to the above examples, and the above examples and descriptions in the specification are only the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements fall within the scope of the claimed present application. The scope of protection claimed by the present application is defined by the appended claims and their equivalents.

[0168] Noun explanation NLP (Natural Language Processing): a technology that allows computers to understand and process human language, used in text analysis, speech recognition, etc.

[0169] GNN (Graph Neural Network): a neural network that processes graph-structured data, used for node classification, link prediction, etc.

[0170] AI (Artificial Intelligence): a technology that simulates human intelligence, covering machine learning, deep learning, etc., used to solve complex problems.

[0171] IoT (Internet of Things): a network of physical devices that allows devices to communicate with each other, used in smart homes, industrial automation, etc.

[0172] API (Application Programming Interface): an interface for system interaction and data exchange, used to integrate third-party applications and develop services.

[0173] SSD (Solid State Drive): a device used for data storage, based on flash memory technology rather than traditional magnetic disks.

Claims

1. Intelligent analysis and judgment method for grassroots governance and law enforcement scenarios, characterized by include: Data collection and preprocessing: Multi-source data integration, including monitoring data, complaint records, laws and regulations, and historical law enforcement records; Knowledge graph construction: identifying entities and relationships to form a complete data network; Intelligent analysis: Using GNN and reasoning algorithms, we conduct in-depth analysis of knowledge graphs to achieve event prediction and risk assessment in complex scenarios; Decision support: Generate optimization suggestions and visual reports to provide decision-making basis for management departments.

2. The intelligent analysis and judgment method for grassroots governance law enforcement scenarios according to claim 1 is characterized by include: Step 1: Data collection and preprocessing Step 1.

1. Data Source Data collection sources include, but are not limited to, law enforcement records from public security, environmental protection, urban management, and other departments, surveillance videos, sensor data, public reporting information, historical case data, and relevant laws and regulations; data types include database records of structured data, log files of semi-structured data, and text, images, or videos of unstructured data; Step 1.2: Data preprocessing Clean, remove duplicates, and convert the collected data into preprocessing format: Step 1.2.

1. Data cleaning: Remove noise and erroneous information from the data, including correcting spelling errors and filling in missing values; Step 1.2.2, data deduplication: eliminate duplicate data; Step 1.2.3, format conversion: convert data in different formats into a unified format; Step 1.2.4, Data Normalization: Standardize the data, including unit conversion and time format unification; Step 2: Knowledge Graph Construction Step 2.1: Semantic Analysis Use natural language processing (NLP) technology to perform semantic analysis on data and extract key entities and relationships: Step 2.1.1, Entity Recognition: Identify key entities of people, places, and events from the text; Step 2.1.2, Relationship Extraction: Identify the relationships between entities; Step 2.1.3, attribute extraction: identify the attribute information of the entity; Step 2.2: Knowledge Graph Association Build a grassroots governance knowledge graph based on the extracted entities and relationships; The graph database Neo4j is used for storage and management. The construction of the knowledge graph includes: Step 2.2.

1. Graph structure design: The nodes of the knowledge graph are entities, and the edges of the knowledge graph are entity relationships; Step 2.2.2, Data import: Import the preprocessed data into the graph database to form a knowledge graph; Step 2.2.3, graph query optimization: optimize database queries through knowledge graph; Step 3: Intelligent analysis method Step 3.1, Graph Neural Network: Use Graph Neural Network (GNN) to perform deep learning on the knowledge graph to capture complex semantic relationships: Step 3.1.1: Graph embedding, mapping nodes and edges in the knowledge graph into a low-dimensional vector space; Step 3.1.2: Model training: Use the GNN model to train graph embedding to capture the semantic information in the graph structure. Step 3.1.3: Feature extraction: extract node and edge features from the trained model. Step 3.2: Inference Algorithm Step 3.2.1: Rule reasoning, reasoning based on the rules in the knowledge graph; Step 3.2.2: Probabilistic reasoning, reasoning based on the probabilistic model; Step 3.2.3: Deep reasoning, combining deep learning models and knowledge graphs to perform complex semantic reasoning; Event probability algorithm in, For events In the knowledge graph and parameter collection The probability of For the The weight of the feature, For the characteristic function, which depends on the parameters ; : Events to be investigated and judged, such as pollution incidents in a certain area; : Knowledge graph, including all relevant entities and relationships; : parameter set, including parameters used to adjust the characteristic function; : Feature weight, which indicates the importance of each feature in the reasoning process; : Feature function, which represents the specific feature of the event The degree of impact depends on the parameters ; Risk Assessment Algorithm in, is the overall risk assessment value, For events In the knowledge graph and parameters The probability under the set, For events The weight of For events The influence coefficient of : Overall risk assessment value, used to measure the risk level of the entire law enforcement scenario; : Every possible event in the knowledge graph; :event The weight indicates the importance of the event; :event The impact coefficient reflects the possible consequences of the event; Decision support algorithms in, For the best decision-making solution, For decision-making exist Utility in context, For the situation In the knowledge graph and parameters The probability under the set, For the situation The weight of : The best decision-making option, such as choosing a certain law enforcement measure; : The set of all possible decisions; :decision making exist Contextual utility, which indicates the effect of the decision in a specific context; : Possible scenarios, derived based on knowledge graph reasoning; : Situation The probability in the knowledge graph reflects the possibility of a situation occurring; : Situation The weight of , which indicates the importance of the situation; Feature weights and : Set through historical data and expert evaluation to indicate the importance of different features and events; weights are optimized through training models; characteristic function : Indicates the impact of specific features on the occurrence of events. The function form is designed according to the specific application scenario, linear function, nonlinear function; parameter Used to adjust the characteristic function to better fit the actual data; Influence coefficient : Indicates the possible consequences of an event, such as economic losses, environmental impact, etc. The impact coefficient is set through expert evaluation and historical data; Contextual Weight : Indicates the importance of different scenarios and is set through expert evaluation; scenario weights help the model make decision support; Step 4: Decision support and optimization Step 4.1: Decision Support Generate a decision support report based on the research and judgment results and provide optimization suggestions: Step 4.1.

1. Event prediction: predict possible future events and their impacts, and predict possible pollution events in a certain area and their impact range; Step 4.1.2, Risk Assessment: Assess the risk level in the current law enforcement scenario; Step 4.1.3, Optimization Suggestions: Based on the research and judgment results, provide optimized law enforcement solutions; Step 4.2, Visualization Tools Visual tools are used to display knowledge graphs and analysis results to assist law enforcement officers in making decisions: Step 4.2.

1. Graph browsing: intuitively display entities and relationships in the knowledge graph, and support query and analysis of nodes and edges; Step 4.2.

1. Data visualization: Display the statistical analysis results of the data; Step 4.2.

1. Result display: Display the research and judgment results and decision support reports, supporting dynamic interaction and customized display; Step 5: Comprehensive evaluation and feedback adjustment Step 5.1: Effect evaluation After implementing intelligent analysis and decision support, compare the expected results with the actual execution results; Step 5.2: Model feedback and optimization: Based on the results of the effectiveness evaluation, the system feeds actual law enforcement data into the knowledge graph and adjusts the parameter settings of the prediction model accordingly. By optimizing weights and adjusting characteristic functions, the system continuously improves the accuracy of future predictions and the effectiveness of decision support. In practical applications, different types of law enforcement scenarios are analyzed and processed based on knowledge graphs and intelligent analysis algorithms.

3. A computer-readable storage medium according to the method for intelligent analysis of grassroots governance and law enforcement scenarios according to claim 1, on which a computer-readable program is stored, and when the computer-readable program is called, the method for intelligent analysis of grassroots governance and law enforcement scenarios as described above is executed.

Images