Prewarning method and device for mining hidden danger of contradictory dispute risk

Abstract

The invention provides a contradictory dispute risk hidden danger mining early warning method and device. The method comprises the following steps: collecting data; performing data preprocessing on the acquired information; modeling by using data, including event factor analysis and personnel factor analysis, for determining an assessment risk level and a risk portrait; and performing risk analysis by using the risk level and the risk portrait. Aiming at high-risk relationships or family contradictions, by analyzing causes of the high-risk relationships or family contradictions, basic-level related dispute handling laws and regulations and resolving suggestions and similar cases are provided, then accurate guidance is provided, potential contradictions and dispute risk hidden dangers are recognized, and the problem solving speed and accuracy are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent early warning systems, and in particular to a method and device for early warning of conflict and dispute risk hazards. Background Art

[0002] With the development of society, driven by various social pressures such as work and life, different types of conflicts and disputes continue to emerge, and furthermore, civil cases and homicide cases caused by conflicts and disputes are on the rise. Conflicts and disputes have caused serious social risks and hidden dangers, hindering the normal operation of social order. The current work of troubleshooting conflicts and disputes is facing the following pain points, which need to be solved urgently.

[0003] How to effectively discover the characteristics of new types of conflicts and disputes, and intelligently understand the intentions of both parties in conflict and dispute cases, will help conflict and dispute mediators increase the success rate of mediation and resolve risks in the bud. Most current research and inventions focus on hot event mining and basic information text analysis, such as the patent: a deep learning method for identifying causal relationships in conflict and dispute events. This patent has poor application effects in conflict and dispute cases and cannot accurately describe the semantic information of the text; such as the patent: a method, device and application of a classification model for conflict and dispute events. This patent only classifies the cases, but does not further study and judge the classified conflict and dispute cases, which leads to its poor application in dispute mediation. Summary of the invention

[0004] The technical problem to be solved by the present invention is how to accurately and quickly process high-frequency conflict and dispute mining models and high-risk conflict and dispute models to provide effective decision-making support for decision makers; in view of this, the present invention provides a conflict and dispute risk hidden danger mining and early warning method and device.

[0005] The technical solution adopted by the present invention is a method for early warning of conflict and dispute risk hazards, comprising:

[0006] Step S1, data collection;

[0007] Step S2, performing data preprocessing on the acquired information;

[0008] Step S3, using the data to perform modeling, including event factor analysis and personnel factor analysis, to determine the assessment risk level and risk profile;

[0009] Step S4, perform risk analysis using risk levels and risk profiles.

[0010] In one embodiment, step S2 specifically includes:

[0011] Clean the acquired data and unify the format of multi-source data;

[0012] Use the acquired data to construct a knowledge graph.

[0013] In one embodiment, the event factor analysis specifically includes:

[0014] Select a preset number of indicators as analysis factors, divide and weight each type of factor into multiple levels, and construct a factor matrix for multi-dimensional risk factor analysis. The indicators include: language, action, number of people, economy, emotion, time, petition, satisfaction with disposal results, online public opinion, involved fields / types, cross-jurisdiction, casualties, spatial field, probability of occurrence, and cross-industry / cross-department;

[0015] The CNN model is used to extract text features from the content and related situations in the conflicts and disputes and align them with the preset factor matrix.

[0016] In one embodiment, the personnel factor analysis specifically includes:

[0017] A preset number of indicators were selected as analysis factors, including: pre-trial mediation by the court, whether it involved sensitive police incidents, the number of times a person was involved in multiple police incidents, the number of times a person was involved in multiple cases, divorced or about to be divorced, and whether there was sensitive network behavior;

[0018] Based on the risk weights of each factor configured, the structural equation model is used to conduct risk assessment;

[0019] The structure of the risk assessment is output in a hierarchical manner.

[0020] In one embodiment, the risk portrait includes: personnel portrait, relationship portrait, and family portrait.

[0021] In one embodiment, step S4 specifically includes:

[0022] Use event factor analysis and personnel factor analysis risk labels to generate type labels and person labels in events as risk labels;

[0023] Utilizing the acquired data and the risk tags, the preset large model algorithm is used to output the risk status and disposal suggestions corresponding to the current time, and push them.

[0024] Another aspect of the present invention provides a conflict and dispute risk hidden danger mining and early warning device, comprising:

[0025] A collection unit configured for data collection;

[0026] A preprocessing unit, configured to perform data preprocessing on the acquired information;

[0027] A modeling unit is configured to use the data to perform modeling, including event factor analysis and personnel factor analysis, to determine the assessment risk level and risk profile;

[0028] The analysis unit is configured to perform risk analysis using risk levels and risk profiles.

[0029] Another aspect of the present invention provides an electronic device, comprising: 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, implements the steps of the method for mining and warning of conflict and dispute risks as described in any one of the above items.

[0030] Another aspect of the present invention further provides a computer storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for mining and warning of conflict and dispute risks as described in any of the above items are implemented.

[0031] By adopting the above technical solution, the present invention has at least the following advantages:

[0032] The present invention provides an efficient method for mining and early warning of conflict and dispute risks. For conflicts in high-risk relationships or families, by analyzing their causes, it provides the basis for relevant laws and regulations on dispute resolution at the grassroots level, suggestions for resolving them, and similar cases, thereby providing precise guidance, identifying potential conflicts and dispute risks, and improving the speed and accuracy of problem solving. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 1. A schematic diagram of a process flow of a method for mining and early warning of conflict and dispute risk hazards according to an embodiment of the present invention;

[0034] Figure 2 Schematic diagram of a conflict and dispute risk potential mining and early warning model according to an embodiment of the present invention;

[0035] Figure 3 A schematic diagram of data collection according to an embodiment of the present invention;

[0036] Figure 4 A schematic diagram of a data aggregation structure according to an embodiment of the present invention;

[0037] Figure 5 A schematic diagram of a knowledge graph according to an embodiment of the present invention;

[0038] Figure 6 A schematic diagram of large model intelligent analysis according to an embodiment of the present invention;

[0039] Figure 7 Schematic diagram of the composition structure of the conflict and dispute risk hidden danger mining and early warning device according to an embodiment of the present invention;

[0040] Figure 8 FIG. 1 is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0041] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined purpose, the present invention is described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0042] In the drawings, the thickness, size and shape of objects have been slightly exaggerated for ease of explanation. The drawings are only examples and are not drawn strictly to scale.

[0043] It should also be understood that the terms "comprises", "including", "having", "includes" and / or "comprising", when used in this specification, indicate the presence of the stated features, wholes, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, parts and / or combinations thereof. In addition, when expressions such as "at least one of..." appear after a list of listed features, they modify the entire listed features rather than modifying the individual elements in the list. In addition, when describing embodiments of the present application, "may" is used to mean "one or more embodiments of the present application". And, the term "exemplary" is intended to refer to an example or illustration.

[0044] As used herein, the terms "substantially," "approximately," and similar terms are used as terms of approximation, not degree, and are intended to account for the inherent variations in measurements or calculations that would be recognized by those of ordinary skill in the art.

[0045] Unless otherwise defined, all terms (including technical terms and scientific terms) used in this article have the same meaning as commonly understood by ordinary technicians in the field to which this application belongs. It should also be understood that terms (such as terms defined in commonly used dictionaries) should be interpreted as having the same meaning as their meaning in the context of the relevant technology, and will not be interpreted in an idealized or overly formal sense unless explicitly defined in this article.

[0046] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0047] The first embodiment of the present invention is a method for early warning of conflict and dispute risk hazards, such as Figure 1 As shown, the following steps are included:

[0048] Step S1, data collection;

[0049] Step S2, performing data preprocessing on the acquired information;

[0050] Step S3, using the data to perform modeling, including event factor analysis and personnel factor analysis, to determine the assessment risk level and risk profile;

[0051] Step S4, perform risk analysis using risk levels and risk profiles.

[0052] In this embodiment, step S2 specifically includes: performing data cleaning on the acquired data and unifying the format of multi-source data; and constructing a knowledge graph using the acquired data.

[0053] In this embodiment, the event factor analysis specifically includes:

[0054] Select a preset number of indicators as analysis factors, divide and weight each type of factor into multiple levels, and construct a factor matrix for multi-dimensional risk factor analysis. The indicators include: language, action, number of people, economy, emotion, time, petition, satisfaction with disposal results, online public opinion, involved fields / types, cross-jurisdiction, casualties, spatial field, probability of occurrence, and cross-industry / cross-department;

[0055] The CNN model is used to extract text features from the content and related situations in the conflicts and disputes and align them with the preset factor matrix.

[0056] In this embodiment, the personnel factor analysis specifically includes:

[0057] A preset number of indicators were selected as analysis factors, including: pre-trial mediation by the court, whether it involved sensitive police incidents, the number of times a person was involved in multiple police incidents, the number of times a person was involved in multiple cases, divorced or about to be divorced, and whether there was sensitive network behavior;

[0058] Based on the risk weights of each factor configured, the structural equation model is used to conduct risk assessment;

[0059] The structure of the risk assessment is output in a hierarchical manner.

[0060] In this embodiment, the risk portrait includes: personnel portrait, relationship portrait, and family portrait.

[0061] In this embodiment, step S4 specifically includes:

[0062] Use event factor analysis and personnel factor analysis risk labels to generate type labels and person labels in events as risk labels;

[0063] Using the acquired data and the risk tags, the preset large model algorithm is used to output the risk status and disposal suggestions corresponding to the current time, and push them.

[0064] Combination Figure 2 , the method provided in this embodiment will be described in detail step by step below.

[0065] Step S1, data collection.

[0066] like Figure 3 In order to solve the problem of multi-source heterogeneity of conflict and dispute data, this embodiment studies multimodal data collection and aggregation technology, relies on the public security big data center, efficiently aggregates structured and unstructured data related to conflicts and disputes from public security, government, courts, the Internet, etc., and builds a conflict and dispute business data resource library.

[0067] Step S2, performing data preprocessing on the acquired information.

[0068] Taking into account the wide range of data sources and high repetition of the data gathered by the model, in order to ensure data integrity and query efficiency, complete missing information, eliminate redundant data, abstract and summarize the characteristics of conflict and dispute data, design data cleaning methods, and ensure data accuracy and consistency, data cleaning, governance and mining are carried out through the public security big data modeling platform and knowledge graph technology to provide data support for model construction. Figure 4 This is a diagram of the data aggregation structure.

[0069] 1. Relying on the public security big data modeling platform, classify and model the multi-source conflict and dispute data, and carry out data governance to form a thematic database, such as:

[0070] (1) Management of personnel involved in police incidents: By extracting the identity information of personnel involved in the alarm content and police response content in the police incident data, and combining it with the number confidence analysis of the owner of the alarm number, a comprehensive sorting is formed to form a special database of personnel involved in police incidents; (2) Management of personnel police incident related relationship data: For personnel involved in police incidents, data management is carried out on multiple people being associated with the same police incident, and the number and type of police incident data associated between personnel are sorted out to form a special database of personnel-related police incidents; (3) Management of personnel marriage relationship data: Governance is carried out on data collected from the civil affairs department and the court, and classification statistics are conducted on personnel relationships such as marriage registration, divorce registration and divorce disputes to form a special database of personnel-related police incidents. (4) Data governance of mentally ill persons: for the mentally ill patients data collected from different data sources (provincial departments and municipal health and health commissions), unified governance is carried out to check for omissions and fill in the gaps in the data from different sources to form a special database of mentally ill persons; (5) Data governance of abnormal accommodation of local people: for the data of local registered residents, hotels, and online booking houses, association governance is carried out to discover the abnormal accommodation of local people in hotels or online booking houses, and form a special database of local abnormal accommodation people; (6) Data governance of people with non-fixed income: for the data of provident fund, human resources and social security personal insurance, etc., association governance is carried out to discover the people who do not have a fixed income in the local area, and form a special database of people with non-fixed income.

[0071] 2. Knowledge graph data mining

[0072] Based on the extracted multi-source heterogeneous data, knowledge units are constructed, and a complex network structure around events, people, and relationships is constructed through knowledge unit fusion. The internal correlation of conflicts and disputes is mined to form a conflict and dispute knowledge graph with a clear data structure. Specific graph data mining methods include: Figure 5 .

[0073] (1) Knowledge unit construction. Based on the requirements of the conflict and dispute hidden danger investigation model, a conflict and dispute graph database framework is constructed. The Bert algorithm is used to realize entity feature extraction and relationship extraction, and a complete knowledge graph dimension of entity concepts is established; (2) Knowledge fusion. Through information extraction, the goal of obtaining entity, relationship and entity attribute information is obtained. Knowledge fusion can eliminate concept ambiguity, eliminate redundant and erroneous concepts, ensure the uniqueness of knowledge, and thus ensure the quality of knowledge; (3) Knowledge reasoning. Through rule mining, the knowledge graph is completed and quality verified, link prediction, association relationship reasoning and conflict detection are carried out, that is, unknown facts or relationships are inferred based on existing facts or relationships in the graph.

[0074] 3. Platform data complementarity

[0075] During data preprocessing, data complementarity is achieved by using the big data modeling platform and the knowledge graph platform. First, the big data modeling platform is used to clean and govern multi-source conflict and dispute data. Then, the knowledge graph platform is used to perform algorithmic capabilities such as knowledge fusion and relationship reasoning on the governed data to extract and merge related disputes. Finally, the mined dispute data is fed back to various thematic libraries of the big data platform to complete the data preparation for conflict and dispute risk mining and modeling.

[0076] Step S3, using the data to build models, including event factor analysis and personnel factor analysis, to determine the assessment risk level and risk profile.

[0077] Data modeling is divided into event factor analysis and personnel factor analysis to assess risk levels and risk profiles. The specific assessment methods are as follows:

[0078] 1. Event Factor Analysis

[0079] Before conducting personnel risk modeling, it is necessary to conduct factor analysis on the events. Based on multi-channel conflict and dispute data, through the review of relevant domestic and foreign literature and field investigations and interviews, a combination of theory and practice is used to select indicators and construct an evaluation indicator system.

[0080] (1) Factor extraction and weighting. More than 15 indicators were selected as analysis factors, and each type of factor was divided into four levels and weighted, and a multi-dimensional risk factor analysis was carried out, including "language, action, number of people (scale), economy, emotion, time, petition, satisfaction with handling results, online public opinion, involved fields / types, cross-jurisdiction, casualties, spatial field, probability of occurrence, and cross-industry / cross-department";

[0081] (2) Text feature extraction: The CNN model is used to extract text features from the alarm content and police dispatch situation and align them with the designed factor matrix.

[0082] 2. Personnel Factor Analysis

[0083] In view of the multi-source data involving personal identity attributes, by analyzing the common characteristics of homicides caused by conflicts and disputes in recent years, combined with the experience of business experts and the structural equation model, factor extraction, factor weighting, risk assessment and level classification are completed.

[0084] (1) Factor extraction. Based on business experience, more than 20 indicators were selected as analysis factors to conduct multi-dimensional risk factor analysis, including "court pre-trial mediation, whether sensitive police incidents were involved, multiple police incidents and times involved by personnel, multiple cases and times involved by personnel, divorced or intending to divorce personnel, and whether there were sensitive network behaviors";

[0085] (2) Factor weighting. Refer to the conversion rate of disputes in each dimension into criminal cases and homicide cases in historical data, and determine the risk weight of each factor in combination with the dispute handling manpower and subjective experience;

[0086] (3) Risk assessment. Based on the risk weights of the factors, the structural equation model is selected for risk assessment. The following formula is the risk structural equation model:

[0087] Y=αX1+βX2+γX3+ε

[0088] X1=a1U1+a2U2+...+a n U n +ε1

[0089] X2=β1V1+β2V2+…+β n V n +ε2

[0090] X3=γ1W1+γ2W2+...+γ n W n +ε3

[0091] Y is the personnel risk assessment score, X1 is the personnel risk attribute, X2 is the relationship risk attribute, X3 is the behavior risk attribute, α, β, γ are the weights of the influence of each risk attribute on the personnel risk assessment score, ε is the error factor (the error term is not considered for the time being), and U, V, W are the observed variables of each risk attribute;

[0092] 4) Grading. According to the risk index classification: high risk (80-100 points), medium risk (60-79 points), low risk (less than 60 points), graded warnings are carried out. Risk personnel who reach the high-risk warning disposal level will be pushed according to the jurisdiction of the business unit, forming a list of personnel to be handled by the warning data, so that users can view the warning information conveniently.

[0093] 3. Risk profiling

[0094] (1) Personnel portrait. Based on factor analysis and other personnel attributes, personnel portraits are created. Based on the experience of business experts and sample learning, the risk factor coefficients of personnel conflicts and disputes are pushed;

[0095] (2) Relationship profiling. Based on the aggregated multi-source conflict and dispute data, knowledge graph technology is used to analyze the relationships between personnel, such as police relationships, marital relationships, romantic relationships, and family relationships;

[0096] (3) Family portrait. For family, marriage and emotional disputes, the police analysis of individuals is expanded to the family dimension. By using knowledge graph technology, the relationship between people is further expanded, and "family portrait" is conducted based on couples, divorced families, current families, etc., providing new ideas and methods for the analysis of conflict and dispute related fields and practical police operations.

[0097] 4. Intelligent analysis of large models

[0098] like Figure 6 The big model intelligent analysis mainly relies on the user inputting the question they want to query, and the big model analyzes the semantics of the question, gives the corresponding data query statement, queries the person-event relationship in the knowledge graph database, and provides the knowledge basis for the subsequent analysis of the big model. The big model needs to build training data for different databases for fine-tuning. The specific steps are as follows:

[0099] (1) Intent analysis: The big model performs semantic analysis on the questions raised by users, determines the type of question, and extracts entity information and time information from the question.

[0100] (2) Cypher generation: Based on the information extracted in the first step, the big model will select the corresponding database and graph, and then select the required nodes and attributes based on the entity information. This information will help the big model generate more accurate query statements.

[0101] (3) Cypher verification: Perform knowledge correction, static verification, and dynamic verification on the generated Cypher query statements. If a query statement reports an error, the error information is returned to the big model, allowing the big model to modify the statement and query again, continuously verifying the syntax errors.

[0102] (4) Intelligent analysis report: Search for relationships between people and events in the knowledge graph, and then obtain similar cases as examples based on the case type. This information is put into the big model for semantic understanding. Through the report writing big model, an intelligent analysis report is given to determine the development trend, severity and recommended disposal opinions of the event.

[0103] Step S4, perform risk analysis using risk levels and risk profiles.

[0104] Specifically include:

[0105] (1) Risk label generation. Based on various factor extraction and analysis, type labels and character labels are given in the event, such as whether the police incident is a sensitive incident or a high-risk incident. The generation of labels facilitates subsequent risk scoring and allows the large model to obtain features for intelligent analysis.

[0106] (2) Intelligent analysis report. In the previous step, relevant data was queried in the knowledge graph and an intelligent analysis report was given to determine the development trend, severity and recommended disposal opinions of the incident.

[0107] (3) Push for disposal. The risk score obtained through the label and the intelligent judgment report generated by the large model are used as a support for decision-making and sent to the relevant departments as reference information. Finally, the staff of the relevant departments make decisions based on this information. The risk score and judgment report greatly reduce the complexity of event handling, and humans only need to participate in the last step of the decision-making stage.

[0108] In summary, compared with the prior art, this embodiment has at least the following advantages:

[0109] 1) The present invention provides an efficient early warning model for mining conflict and dispute risks. By analyzing the causes of conflicts in high-risk relationships or families, the present invention provides the basis for the relevant laws and regulations on dispute resolution at the grassroots level, as well as suggestions for resolving them and similar cases, thereby providing precise guidance, identifying potential conflicts and dispute risks, and improving the speed and accuracy of problem solving.

[0110] 2) This embodiment can meet the actual application needs in the public security field, assist in the reasonable allocation of resources and reduce the occurrence of high-frequency and high-risk cases. In view of the huge number of historical police cases, complex classification, and difficulty in handling, it can effectively trace back police cases, discover risk points, and provide timely disposal feedback.

[0111] The second embodiment of the present invention corresponds to the first embodiment. This embodiment introduces a conflict and dispute risk hidden danger mining and early warning device, such as Figure 7 As shown, it includes the following components:

[0112] A collection unit configured for data collection;

[0113] A preprocessing unit, configured to perform data preprocessing on the acquired information;

[0114] A modeling unit is configured to use the data to perform modeling, including event factor analysis and personnel factor analysis, to determine the assessment risk level and risk profile;

[0115] The analysis unit is configured to perform risk analysis using risk levels and risk profiles.

[0116] In this embodiment, the preprocessing unit is further configured as follows:

[0117] Clean the acquired data and unify the format of multi-source data;

[0118] Use the acquired data to construct a knowledge graph.

[0119] In this embodiment, the event factor analysis specifically includes:

[0120] Select a preset number of indicators as analysis factors, divide and weight each type of factor into multiple levels, and construct a factor matrix for multi-dimensional risk factor analysis. The indicators include: language, action, number of people, economy, emotion, time, petition, satisfaction with disposal results, online public opinion, involved fields / types, cross-jurisdiction, casualties, spatial field, probability of occurrence, and cross-industry / cross-department;

[0121] The CNN model is used to extract text features from the content and related situations in the conflicts and disputes and align them with the preset factor matrix.

[0122] In this embodiment, the personnel factor analysis specifically includes:

[0123] A preset number of indicators were selected as analysis factors, including: pre-trial mediation by the court, whether it involved sensitive police incidents, the number of times a person was involved in multiple police incidents, the number of times a person was involved in multiple cases, divorced or about to be divorced, and whether there was sensitive network behavior;

[0124] Based on the risk weights of each factor configured, the structural equation model is used to conduct risk assessment;

[0125] The structure of the risk assessment is output in a hierarchical manner.

[0126] In this embodiment, the risk portrait includes: personnel portrait, relationship portrait, and family portrait.

[0127] In this embodiment, the analysis unit is further configured as follows:

[0128] Use event factor analysis and personnel factor analysis risk labels to generate type labels and person labels in events as risk labels;

[0129] Utilizing the acquired data and the risk tags, the preset large model algorithm is used to output the risk status and disposal suggestions corresponding to the current time, and push them.

[0130] A third embodiment of the present invention is an electronic device, such as Figure 8 As shown, it can be understood as a physical device, including a processor and a memory storing instructions executable by the processor. When the instructions are executed by the processor, the following operations are performed:

[0131] Step S1, data collection;

[0132] Step S2, performing data preprocessing on the acquired information;

[0133] Step S3, using the data to perform modeling, including event factor analysis and personnel factor analysis, to determine the assessment risk level and risk profile;

[0134] Step S4, perform risk analysis using risk levels and risk profiles.

[0135] The fourth embodiment of the present invention, the process of the conflict and dispute risk hidden danger mining early warning method of this embodiment is the same as the first, second or third embodiment, the difference is that, in terms of engineering implementation, this embodiment can be implemented by means of software plus a necessary general hardware platform, of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the method of the present invention can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM / RAM, disk, CD), including a number of instructions for a device to execute the method described in the embodiment of the present invention.

[0136] Through the description of the specific implementation methods, a deeper and more specific understanding of the technical means and effects adopted by the present invention to achieve the predetermined purpose should be obtained. However, the accompanying drawings are only for reference and illustration purposes and are not intended to limit the present invention.

Claims

1. A method for early warning of conflict and dispute risk hazards, characterized in that: include: Step S1, data collection; Step S2, performing data preprocessing on the acquired information; Step S3, using the data to perform modeling, including event factor analysis and personnel factor analysis, to determine the assessment risk level and risk profile; Step S4: Perform risk analysis using the risk level and the risk profile.

2. The method for mining and early warning of conflict and dispute risks according to claim 1 is characterized in that: The step S2 specifically includes: Clean the acquired data and unify the format of multi-source data; Use the acquired data to construct a knowledge graph.

3. The method for mining and early warning of conflict and dispute risks according to claim 1 is characterized in that: The event factor analysis specifically includes: Select a preset number of indicators as analysis factors, divide and weight each type of factor into multiple levels, and construct a factor matrix for multi-dimensional risk factor analysis. The indicators include: language, action, number of people, economy, emotion, time, petition, satisfaction with disposal results, involved fields / types, cross-jurisdiction, casualties, spatial field, probability of occurrence, and cross-industry / cross-department; The CNN model is used to extract text features from the content and related situations in the conflicts and disputes and align them with the preset factor matrix.

4. The method for mining and early warning of conflict and dispute risks according to claim 3 is characterized in that: The personnel factor analysis specifically includes: A preset number of indicators are selected as analysis factors, including: pre-trial mediation by the court, whether it involves sensitive police incidents, whether the personnel are involved in multiple police incidents and the number of times, multiple cases and the number of times, divorced or divorced personnel, whether it involves emotional disputes and domestic violence police incidents, whether it is a mental patient, whether it is a non-visitor, whether it is a military personnel, whether it is a key person with seven criminal records, whether it is a key resettlement and assistance personnel, whether it is a person with non-fixed income, whether it is a special care group, whether the locals have abnormal accommodation, and whether there is sensitive network behavior; Based on the risk weights of each factor configured, the structural equation model is used to conduct risk assessment; The structure of the risk assessment is output in a hierarchical manner.

5. The method for mining and early warning of conflict and dispute risks according to claim 4 is characterized in that: The risk portrait includes: personnel portrait, relationship portrait, and family portrait.

6. The method for mining and early warning of conflict and dispute risks according to claim 5 is characterized in that: The step S4 specifically includes: Use event factor analysis and personnel factor analysis risk labels to generate type labels and person labels in events as risk labels; Utilizing the acquired data and the risk tags, the preset large model algorithm is used to output the risk status and disposal suggestions corresponding to the current time, and push them.

7. A device for early warning of potential conflicts and disputes, characterized in that: include: A collection unit configured for data collection; A preprocessing unit, configured to perform data preprocessing on the acquired information; A modeling unit is configured to use the data to perform modeling, including event factor analysis and personnel factor analysis, to determine the assessment risk level and risk profile; The analysis unit is configured to perform risk analysis using risk levels and risk profiles.

8. An electronic device, characterized in that: The electronic device comprises: 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, the steps of the method for mining and early warning of conflict and dispute risks as described in any one of claims 1 to 6 are implemented.

9. A computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method for mining and early warning of conflict and dispute risks as described in any one of claims 1 to 6.

Images