Market supervision dynamic monitoring and intelligent risk early warning system and method based on big data

Abstract

The invention discloses a market supervision dynamic monitoring and intelligent risk early warning system and method based on big data, and relates to the technical field of monitoring and early warning. According to the invention, a market supervision task is dynamically monitored in a whole process through the dynamic monitoring module, an abnormal behavior or trend of a market subject is intelligently analyzed through the risk analysis module, and monitoring data is subjected to feature extraction and risk classification through a chaos immune optimization algorithm and a migration immune algorithm. Generating a risk early warning report including the risk level, the risk type and the corresponding disposal suggestion; based on the analysis result of the risk analysis module, the stock, increment, industry trend and regional distribution condition of market subjects are displayed in a visual mode, an early warning notification function is integrated, a risk early warning result is sent to a related supervision department, priority processing of different levels of risks and a multi-department collaborative early warning mechanism are supported, and the risk early warning efficiency is improved. And full-life-cycle dynamic supervision and intelligent risk early warning of the market subject are realized.

Description

Technical Field

[0001] The present invention relates to the field of monitoring and early warning technology, and in particular to a market supervision dynamic monitoring and intelligent risk early warning system and method based on big data. Background Art

[0002] At present, with the rapid development of social economy and the continuous expansion of market scale, the complexity and difficulty of market supervision are increasing. The supervision of market entities involves multiple data sources, including registration information, credit records, consumer complaints, food and drug supervision, etc. These data are scattered among different departments and platforms, lacking a unified integration mechanism, resulting in serious data island problems. At the same time, most current market supervision systems lack the ability to dynamically monitor market entities, especially the real-time changes in risk factors of market entities are difficult to grasp in a timely manner, and the early warning window for potential problems may be missed.

[0003] Traditional risk analysis methods are mainly based on static indicators, which cannot adapt to the dynamic changes in the risk characteristics of market entities. They have weak capabilities for collaborative risk assessment of cross-regional and cross-industry market entities and are prone to regulatory blind spots.

[0004] Therefore, there is a need for a market supervision dynamic monitoring and intelligent risk early warning system and method based on big data to achieve efficient integration of data resources, intelligent risk identification and prediction, and dynamic market supervision. Summary of the invention

[0005] In response to the above problems, the present invention proposes a market supervision dynamic monitoring and intelligent risk early warning system and method based on big data.

[0006] The present invention achieves the above-mentioned purpose through the following technical solutions:

[0007] A market supervision dynamic monitoring and intelligent risk early warning system based on big data, the system comprising:

[0008] The data collection module is used to collect market supervision data from multiple heterogeneous data sources, including market entity registration information, business behavior data, credit records, law enforcement inspection data, consumer complaint data, food and drug supervision data, and external data related to market supervision;

[0009] A data processing module is used to clean, convert and integrate the market supervision data, eliminate data redundancy and inconsistency, and generate unified structured data, which includes basic attribute information of market entities, dynamic monitoring indicator data and input data required for risk analysis;

[0010] A dynamic monitoring module, used for dynamic monitoring of the entire process of market supervision tasks based on the structured data;

[0011] The risk analysis module is used to conduct intelligent analysis on abnormal behaviors or trends of market entities based on the real-time monitoring results of the dynamic monitoring module, extract features and classify risks of monitoring data through the chaotic immune optimization algorithm and the migration immune algorithm, and generate a risk warning report including risk level, risk type and corresponding disposal suggestions;

[0012] The visualization display module is used to display the stock, growth, industry trend and regional distribution of market entities in a visual way based on the analysis results of the risk analysis module. The visualization methods include line charts, bar charts, geographic information system maps and report forms, which dynamically display risk warning information; and integrate the warning notification function to send the risk warning results to relevant regulatory departments via SMS, email or system push, supporting priority processing of risks at different levels and multi-department collaborative warning mechanism.

[0013] As a preferred solution of the present invention, the data collection module dynamically adjusts the collection priority according to the activity frequency and importance of the market subject, the activity frequency is calculated according to the real-time business behavior data of the market subject, and the importance is calculated according to the industry risk weight, credit score fluctuation and historical abnormal behavior frequency of the market subject;

[0014] By defining the trust and contribution of data sources, the collection ratio of data sources is dynamically optimized. The formula is:

[0015]

[0016] Where P v represents the priority acquisition probability of data source v; T v is the trustworthiness of data source v; C v is the contribution of data source v; V is the total number of data source types.

[0017] As a preferred solution of the present invention, the data processing module adjusts the sensitivity of the data cleaning rule by generating a dynamic threshold by introducing a chaotic mapping function. The calculation formula of the dynamic threshold is:

[0018] θ(t)=μ·θ(t-1)·(1-θ(t-1))+γ·f error (t);

[0019] In the formula, θ(t) is the dynamic threshold of the current iteration; θ(t-1) is the dynamic threshold of the previous iteration; μ is the chaos parameter; γ is the disturbance coefficient; f error (t) is the error function of the current iteration;

[0020] By introducing the semantic weight w s and context relevance cDynamically adjust the integration priority of data fields. The formula is:

[0021]

[0022] In the formula, S merge Score the match between the two tables: b ,G b ) is the set of fields to be integrated; B is the total number of fields; Sim(F b ,G b ) indicates field F b and G b The semantic similarity of s is the semantic weight; r c is contextual relevance.

[0023] As a preferred solution of the present invention, the full-process dynamic monitoring includes:

[0024] Monitoring of task execution: obtaining the task name, start time, end time, execution status, duration, and execution log information;

[0025] Monitoring of data volume changes: Based on the data table change rate, identify data tables with significant data volume differences, calculate the data table change rate and compare it with the preset threshold to determine if the change rate | V new -V old | / V old ×100% exceeds the preset threshold, it is considered to be significantly different, and a data table warning information is generated, which includes the name of the data table, the change rate, the abnormal type and the warning generation time; wherein, V new Indicates the amount of data in the current data table, V old Indicates the amount of data in the data table during the previous detection;

[0026] Database resource monitoring: monitors the database disk usage, including used space, remaining space, total space and utilization, and generates resource utilization warning information;

[0027] Dynamic monitoring of core indicators: Real-time acquisition of changes in core indicators of market entities, including the number of existing households, incremental trends, rate of change of registered capital, industry distribution trends and regional distribution trends.

[0028] As a preferred solution of the present invention, the risk analysis module includes:

[0029] A data preprocessing unit is used to obtain market entity monitoring data from the dynamic monitoring module and perform data cleaning and normalization processing;

[0030] Chaos immune optimization unit, used to extract features from pre-processed data through chaos immune optimization algorithm, optimize antibody population, and obtain preliminary risk fitness value;

[0031] Migration Immunity Optimization Unit, used to optimize migration and dynamically adapt risk factors across regions;

[0032] A risk level assessment unit is used to generate the risk level of the market entity based on the optimized risk data;

[0033] The risk type analysis unit is used to further analyze the risk types of market entities according to risk levels, including credit risk, abnormal behavior risk and abnormal growth risk;

[0034] The risk warning report generation unit is used to generate a visual risk warning report based on the risk level and risk type.

[0035] As a preferred solution of the present invention, a dynamic weight function ω(t) is introduced into the chaotic immune optimization algorithm unit, and an improved chaotic mapping formula is used to generate an antibody population, and the formula is:

[0036] A(t+1)=μ·w(t)·A(t)·(1-A(t))+γ·f(A(t));

[0037] Where, t is the current iteration number; A(t) is the current risk characteristic value of the market entity; A(t+1) is the risk characteristic value of the market entity after chaos optimization and disturbance at the next moment; μ is the chaos parameter; γ is the disturbance coefficient; f(A(t)) is the fitness function;

[0038] Calculate the fitness of each antibody: Where R n is the risk factor value, is the benchmark value, n is the index of the current risk factor; N is the total number of risk factors;

[0039] In the process of population evolution, combined with the risk factor matrix R of market entities f Adjust the mutation probability, the formula is:

[0040]

[0041] Where P m is the mutation probability of the current market entity; P m0 is the initial mutation probability; ||R f || is the Euclidean norm of the market entity risk factor;

[0042] The optimized antibody population A(t+1) is generated as the input of the migration immunity optimization unit.

[0043] As a preferred solution of the present invention, in the migration immunity optimization unit, market entities are divided according to industry or region to form distributed antibody nodes, and the migration probability is adjusted based on similarity and time decay:

[0044]

[0045] Where P trans (i,j,t) is the probability of antibody migration of market players from region i to region j; Sim(A i ,A j ) is the similarity of antibodies of market players in region i and region j; ∑ k≠i Sim(A i ,A k ) is the sum of similarities between region i and all other regions; λ is the time decay coefficient;

[0046] Recalculate the environmental fitness F for the migrated antibodies env : F env =a·R growth +β·R deviation , where R growth represents the growth rate, R deviation represents the risk deviation; α and β are the weight coefficients of growth rate and risk deviation respectively;

[0047] The optimized antibody population and fitness data are transmitted to the risk level assessment unit.

[0048] As a preferred solution of the present invention, in the risk level assessment unit, the formula for generating the risk level of the market entity is:

[0049] f multi (x) = x1·f risk (x)+x2·f priority (x);

[0050] In the formula, f multi (x) is the final optimized value; f risk (x) is the risk deviation; f priority (x) is the priority of regulatory tasks; x1 and x2 are the weight coefficients of risk assessment and regulatory task priority, respectively; x represents the characteristic variable of the current market entity;

[0051] Risk levels are divided according to the final optimization value:

[0052] When the final optimized value f multi (x) When the set high-risk threshold is exceeded, the market entity is classified as a high-risk entity and requires priority supervision or emergency intervention measures;

[0053] When the final optimized value fmulti (x) When the market entity is between the high risk threshold and the low risk threshold, it is classified as a medium risk entity and requires regular monitoring and appropriate supervision;

[0054] When the final optimized value f multi (x) When the risk is less than the set low-risk threshold, the market entity is classified as a low-risk entity and only requires routine supervision.

[0055] A method for early warning of a market supervision dynamic monitoring and intelligent risk early warning system based on big data, the method comprising:

[0056] Collect market supervision data from multiple heterogeneous data sources, clean, convert and integrate them, eliminate data redundancy and inconsistency, and generate unified structured data;

[0057] Based on the structured data, the market supervision task is dynamically monitored throughout the entire process, and according to the real-time monitoring results of the dynamic monitoring module, the abnormal behaviors or trends of the market entities are intelligently analyzed. The monitoring data is feature extracted and risk classified through the chaotic immune optimization algorithm and the migration immune algorithm, and a risk warning report including risk level, risk type and corresponding disposal suggestions is generated;

[0058] Based on the analysis results of the risk analysis module, the stock, growth, industry trends and regional distribution of market entities are displayed in a visual way, and risk warning information is displayed dynamically; and the warning notification function is integrated to send the risk warning results to relevant regulatory departments via SMS, email or system push, supporting priority processing of risks at different levels and multi-department collaborative warning mechanism.

[0059] The beneficial effects of the present invention are: it realizes efficient integration and structured processing of multi-source data, eliminates data redundancy and inconsistency, and provides a high-quality data foundation for dynamic monitoring and risk analysis; it improves the real-time monitoring capability of market supervision tasks, and can dynamically capture the stock, increment, industry distribution trend and regional distribution of market entities, and realizes comprehensive coverage of full-process supervision; based on the chaotic immune optimization algorithm and the migration immune algorithm, it accurately extracts the risk characteristics of market entities, intelligently generates risk levels and early warning reports, and provides scientific risk assessment support for regulatory authorities; it intuitively displays the dynamic changes and risk distribution of market entities through multi-dimensional visualization, integrates priority processing and multi-department collaborative early warning mechanism, and significantly improves supervision efficiency and disposal effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0061] in:

[0062] Figure 1 It is a system modular structure diagram of the present invention;

[0063] Figure 2 4 is a flow chart of a method in an embodiment of the present invention. DETAILED DESCRIPTION

[0064] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0065] like Figure 1 FIG. 1 is an embodiment of the present invention, which provides a market supervision dynamic monitoring and intelligent risk early warning system based on big data, including:

[0066] (1) Data acquisition module

[0067] Used to collect market supervision data from multiple heterogeneous data sources, including market entity registration information, business behavior data, credit records, law enforcement inspection data, consumer complaint data, food and drug supervision data, and external data related to market supervision.

[0068] The data acquisition module further includes:

[0069] Dynamic data priority collection mechanism: dynamically adjust the collection priority according to the activity frequency and importance of market entities. The activity frequency is calculated based on the real-time business behavior data of market entities, and the importance is calculated based on the industry risk weight, credit score fluctuation and historical abnormal behavior frequency of market entities;

[0070] Quality-aware optimization collection mechanism based on multi-source data fusion: By defining the trust and contribution of data sources, the collection ratio of data sources is dynamically optimized. The formula is:

[0071]

[0072] Where P vrepresents the priority acquisition probability of data source v; T v is the trustworthiness of data source v, calculated based on the historical collection accuracy and delay rate of the data source; C v is the contribution of data source v, which is dynamically calculated based on the importance of the collected data to the market entity monitoring indicators; V is the total number of data source types, that is, the number of different data source types involved in the collection decision, such as market entity registration information, credit records, consumer complaint data, etc.

[0073] The data dynamic priority collection mechanism integrates the activity frequency and importance of market entities, breaking through the fixed limitations of traditional collection order; the quality-aware optimization collection mechanism dynamically allocates the collection ratio through the trust and contribution of the data source, avoiding the uniform processing method of conventional collection. Through the above mechanisms, the data collection module can give priority to the collection of high-quality and high-priority data when the data volume is large and the quality is uneven, improving the real-time performance and monitoring accuracy of the system.

[0074] (2) Data processing module

[0075] It is used to clean, format convert and integrate market supervision data, eliminate data redundancy and inconsistency, and generate unified structured data. The structured data includes basic attribute information of market entities, dynamic monitoring indicator data and input data required for risk analysis.

[0076] The data processing module further includes:

[0077] Data cleaning method based on chaos optimization: The sensitivity of data cleaning rules is adjusted by introducing a chaotic mapping function to generate a dynamic threshold. The calculation formula of the dynamic threshold is:

[0078] θ(t)=μ·θ(t-1)·(1-θ(t-1))+γ·f error (t);

[0079] In the formula, θ(t) is the dynamic threshold of the current iteration, which is used to adjust the sensitivity of the data cleaning rule; θ(t-1) is the dynamic threshold of the previous iteration, which is used to recursively calculate the current dynamic threshold; μ is the chaos parameter, which is used to control the dynamic range of the threshold change, and is usually taken in the range of (0,4); γ is the disturbance coefficient, which is used to dynamically adjust the rule update amplitude; f error (t) is the error function of the current iteration, which is calculated based on the redundancy rate, inconsistency rate or outlier ratio of the data and is used to dynamically correct the threshold;

[0080] Format conversion and integration algorithm based on semantic matching: By introducing the semantic weight w s and context relevance c Dynamically adjust the integration priority of data fields. The formula is:

[0081]

[0082] In the formula, S merge The match score for merging two data tables is used to measure the priority of field matching in the two tables:

[0083] (F b ,G b ) is the set of fields to be integrated, F b , G b is the kth field in the field set to be merged, which comes from two different data tables; B is the total number of fields, that is, the number of fields that need to be matched in the data tables to be merged; Sim(F b ,G b ) indicates field F b and G b The semantic similarity of is calculated using a specific semantic matching algorithm (such as cosine similarity based on word vectors); w s is the semantic weight, which reflects the importance of the field in the monitoring task and is dynamically adjusted according to the impact of the field on the core indicators; r c It is context dependency, which reflects the logical dependency between a field and its context fields.

[0084] Chaos optimization introduces dynamic threshold cleaning rules, combined with real-time error rate adjustment, to improve the flexibility of large-scale data processing; semantic matching and context relevance dynamically adjust the field integration priority, breaking through the traditional static integration method based on format matching. Through the above improvements, the data processing module can improve the flexibility of data cleaning and the accuracy of format conversion when processing large-scale heterogeneous data, effectively reduce redundancy and consistency problems, and provide high-quality input data for subsequent risk analysis.

[0085] (3) Dynamic monitoring module

[0086] It is used to dynamically monitor the entire process of market supervision tasks based on the structured data. The entire process of dynamic monitoring includes:

[0087] Monitoring of task execution: obtaining the task name, start time, end time, execution status, duration, and execution log information;

[0088] Monitoring of data volume changes: Based on the data table change rate, identify data tables with significant data volume differences, calculate the data table change rate and compare it with the preset threshold to determine if the change rate | V new -V old | / V old ×100% exceeds the preset threshold, it is considered to be significantly different, and a data table warning information is generated. The data table warning information includes the name of the data table, the change rate, the abnormal type, and the warning generation time; among them, V newIndicates the data volume of the current data table, that is, the number of records or storage capacity of the data table at the time of the most recent detection; V old Indicates the data volume of the data table at the last detection, that is, the number of records or storage capacity of the data table at the last monitoring;

[0089] By comparing V new and V old , can quantify the degree of change of the data table between two monitorings. If the change rate exceeds the preset threshold, the data table is considered to have changed significantly, thus triggering an early warning. For example: a data table has a record number of V in the last round of monitoring. old =10000, the number of records in this round of monitoring becomes V new =12000, the change rate is 12000-10000| / 10000×100%=20%. If the preset threshold of the change rate is 15%, the data table will be marked as "significant difference" and generate an early warning message;

[0090] Database resource monitoring: monitors the database disk usage, including used space, remaining space, total space and utilization, and generates resource utilization warning information;

[0091] Dynamic monitoring of core indicators: Real-time acquisition of changes in core indicators of market entities. The core indicators of market entities include the number of existing households, incremental trends, rate of change of registered capital, industry distribution trends and regional distribution trends.

[0092] (4) Risk Analysis Module

[0093] It is used to conduct intelligent analysis of abnormal behaviors or trends of market entities based on the real-time monitoring results of the dynamic monitoring module, extract features and classify risks of monitoring data through chaos immune optimization algorithm and migration immune algorithm, and generate risk warning reports including risk level, risk type and corresponding disposal suggestions.

[0094] In one specific embodiment, the risk analysis module includes:

[0095] A data preprocessing unit is used to obtain market entity monitoring data from the dynamic monitoring module and perform data cleaning and normalization processing;

[0096] Chaos immune optimization unit, used to extract features from pre-processed data through chaos immune optimization algorithm, optimize antibody population, and obtain preliminary risk fitness value;

[0097] In the chaotic immune optimization unit, the dynamic weight function ω(t) is introduced, and the improved chaotic mapping formula is used to generate the antibody population. The formula is:

[0098] A(t+1)=μ·w(t)·A(t)·(1-A(t))+γ·f(A(t));

[0099] In the formula, t is the current iteration number, which indicates the algorithm optimization process. The more iterations, the smaller the mutation probability, and the optimization gradually converges. A(t) is the current risk characteristic value of the market subject, which is the comprehensive risk performance of the current subject at the monitoring moment obtained by combining the risk assessment model. A(t+1) is the risk characteristic value of the market subject after chaos optimization and disturbance at the next moment, which is used to predict the risk level at future moments. μ is the chaos parameter, which controls the sensitivity of the risk characteristic change of the market subject and avoids local optimality by adjusting the optimization step size. γ is the disturbance coefficient, which is used to dynamically adjust the antibody generation accuracy to ensure that the risk details of the market subject can be more accurately reflected. f(A(t)) is the fitness function, which is calculated based on the risk deviation of the market subject and indicates the difference between the actual risk level of the current subject and the expected benchmark value.

[0100] Calculate the fitness of each antibody: Where R n is the risk factor value, is the benchmark value, n is the index of the current risk factor, indicating the nth specific risk indicator; N is the total number of risk factors, indicating the number of all risk indicators monitored. For example, if the risk factors monitored by the system include credit score, registered capital change rate, abnormal behavior deviation, etc., there are 5 factors in total, then N = 5;

[0101] In the process of population evolution, combined with the risk factor matrix R of market entities f Adjust the mutation probability, the formula is:

[0102]

[0103] Where P m is the mutation probability of the current market entity, which is used in the dynamic optimization algorithm to adjust the risk characteristics of the market entity and improve the search accuracy; m0 is the initial mutation probability, which indicates the initial disturbance intensity of the risk characteristics of market entities at the beginning of the algorithm; ||R f || is the Euclidean norm of the market entity risk factor, indicating the comprehensive risk level of the entity;

[0104] The optimized antibody population A(t+1) is generated as the input of the migration immunity optimization unit.

[0105] Migration Immunity Optimization Unit, used to optimize migration and dynamically adapt risk factors across regions;

[0106] In the migration immunity optimization unit, market entities are divided according to industry or region to form distributed antibody nodes, and the migration probability is adjusted based on similarity and time decay:

[0107]

[0108] Where P trans (i,j,t) is the probability of antibody migration of market players from region i to region j, indicating the possibility that the optimization information of market players in a certain region can be applied to other regions; Sim(A i ,A j ) is the similarity between the antibodies of market entities in region i and region j, which is used to measure the correlation between the risk characteristics of the two regions, for example, by calculating the cosine similarity through the risk indicators of the entities;

[0109] ∑ k≠i Sim(A i ,A k ) is the sum of similarities between region i and all other regions, ensuring the normalization of the migration probability; λ is the time decay coefficient, which controls the decrease of the migration probability over time. A higher value indicates a rapid decrease in the influence of migration;

[0110] Recalculate the environmental fitness F for the migrated antibodies env : F env =a·R growth +β·R deviation , where R growth represents the growth rate, R deviation represents the risk deviation; α and β are the weight coefficients of growth rate and risk deviation respectively;

[0111] The optimized antibody population and fitness data are transmitted to the risk level assessment unit.

[0112] The risk level assessment unit is used to generate the risk level of the market entity based on the optimized risk data. The formula is:

[0113] f multi (x) = x1·f risk (x)+x2·f priority (x);

[0114] In the formula, f multi (x) is the final optimized value; f risk (x) is the risk deviation; f priority (x) is the priority of regulatory tasks; x1 and x2 are the weight coefficients of risk assessment and regulatory task priority, respectively; x represents the characteristic variable of the current market entity;

[0115] Risk levels are divided according to the final optimization value:

[0116] When the final optimized value f multi (x) When a market entity exceeds the set high-risk threshold, it is classified as a high-risk entity, which may have serious credit problems, abnormal growth or abnormal behavior, and requires priority supervision or emergency intervention measures;

[0117] When the final optimized value f multi (x) When the market entity is between the high risk threshold and the low risk threshold, it is classified as a medium risk entity and may require regular monitoring and appropriate supervision;

[0118] When the final optimized value f multi (x) When the risk is less than the set low-risk threshold, the market entity is classified as a low-risk entity, which is usually characterized by stable operations or a good credit record and only requires routine supervision.

[0119] The risk type analysis unit is used to further analyze the risk types of market entities according to risk levels, including credit risk, abnormal behavior risk and abnormal growth risk;

[0120] The risk warning report generation unit is used to generate a visual risk warning report based on the risk level and risk type.

[0121] (5) Visualization display module

[0122] It is used to display the stock, growth, industry trend and regional distribution of market entities in a visual way based on the analysis results of the risk analysis module. The visualization methods include line charts, bar charts, geographic information system maps and report forms, which dynamically display risk warning information; and integrates the warning notification function to send the risk warning results to relevant regulatory departments via SMS, email or system push, supporting priority processing of risks at different levels and multi-department collaborative warning mechanism.

[0123] like Figure 2 As shown, another embodiment of the present invention provides an early warning method of a market supervision dynamic monitoring and intelligent risk early warning system based on big data, comprising the following steps:

[0124] S1: Collect market supervision data from multiple heterogeneous data sources, clean, convert and integrate them, eliminate data redundancy and inconsistency, and generate unified structured data;

[0125] S2: Based on structured data, the market supervision task is dynamically monitored throughout the entire process. According to the real-time monitoring results of the dynamic monitoring module, the abnormal behaviors or trends of market entities are intelligently analyzed. The monitoring data is feature extracted and risk classified through the chaotic immune optimization algorithm and the migration immune algorithm, and a risk warning report including risk level, risk type and corresponding disposal suggestions is generated.

[0126] S3: Based on the analysis results of the risk analysis module, the stock, growth, industry trends and regional distribution of market entities are displayed in a visual way, and risk warning information is displayed dynamically; and the warning notification function is integrated to send the risk warning results to relevant regulatory departments via SMS, email or system push, supporting priority processing of risks at different levels and multi-department collaborative warning mechanism.

[0127] In summary, the present invention collects multidimensional data related to market supervision from multi-source heterogeneous data sources, realizes the unified integration of scattered data, solves the problem of data islands in traditional supervision systems, and significantly improves the comprehensiveness and availability of data; the data processing module cleans, converts and integrates the collected data, eliminates data redundancy and inconsistency, and generates unified structured data, which lays a solid data foundation for subsequent dynamic monitoring and risk analysis, and improves data processing efficiency and accuracy; the dynamic monitoring module dynamically monitors the entire process of market supervision tasks based on structured data, covering task execution, data volume changes, database resource monitoring, and real-time changes in core indicators of market entities (such as stock, increment, industry distribution trends, etc.), effectively improving the ability to capture dynamic changes in market entities and enhancing real-time supervision. The risk analysis module combines the chaotic immune optimization algorithm and the migration immune algorithm to extract and classify the monitoring data of market entities, accurately identify the abnormal behavior or risk trend of market entities, and dynamically generate risk warning reports including risk level, risk type and corresponding disposal suggestions, providing scientific and intelligent risk assessment tools for regulatory authorities; the visualization display module dynamically displays the stock, increment, industry trend and regional distribution of market entities in the form of line charts, bar charts, geographic information system maps and reports, intuitively presents the risk distribution and change trend, and also integrates the early warning notification function, which can send the risk warning results to relevant regulatory authorities through SMS, email or system push, supports the priority processing of risks at different levels and the multi-department collaborative early warning mechanism, and effectively optimizes the allocation efficiency of regulatory resources. The present invention solves the problems of data dispersion, insufficient dynamic monitoring, delayed risk warning and poor visualization effect of the existing market supervision system through modular design and intelligent algorithms, realizes the dynamic supervision and intelligent risk warning of the whole life cycle of market entities, and provides technical support for the precise decision-making of regulatory authorities.

[0128] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A market supervision dynamic monitoring and intelligent risk early warning system based on big data, characterized in that: The system includes: a data collection module for collecting market supervision data from multi-source heterogeneous data sources, including registration information of market entities, business behavior data, credit records, law enforcement inspection data, consumer complaint data, food and drug supervision data, and external data related to market supervision; A data processing module is used to clean, convert and integrate the market supervision data, eliminate data redundancy and inconsistency, and generate unified structured data, which includes basic attribute information of market entities, dynamic monitoring indicator data and input data required for risk analysis; A dynamic monitoring module, used for dynamic monitoring of the entire process of market supervision tasks based on the structured data; The risk analysis module is used to conduct intelligent analysis on abnormal behaviors or trends of market entities based on the real-time monitoring results of the dynamic monitoring module, extract features and classify risks of monitoring data through the chaotic immune optimization algorithm and the migration immune algorithm, and generate a risk warning report including risk level, risk type and corresponding disposal suggestions; The visualization display module is used to display the stock, growth, industry trend and regional distribution of market entities in a visual way based on the analysis results of the risk analysis module. The visualization methods include line charts, bar charts, geographic information system maps and report forms, which dynamically display risk warning information; and integrate the warning notification function to send the risk warning results to relevant regulatory departments via SMS, email or system push, supporting priority processing of risks at different levels and multi-department collaborative warning mechanism.

2. According to the big data-based market supervision dynamic monitoring and intelligent risk early warning system of claim 1, it is characterized in that: The data collection module dynamically adjusts the collection priority according to the activity frequency and importance of the market subject, the activity frequency is calculated based on the real-time business behavior data of the market subject, and the importance is calculated based on the industry risk weight, credit score fluctuation and historical abnormal behavior frequency of the market subject; By defining the trust and contribution of data sources, the collection ratio of data sources is dynamically optimized. The formula is: Where P v represents the priority acquisition probability of data source v; T v is the trustworthiness of data source v; C v is the contribution of data source v; V is the total number of data source types.

3. According to the big data-based market supervision dynamic monitoring and intelligent risk early warning system of claim 1, it is characterized in that: The data processing module adjusts the sensitivity of the data cleaning rule by introducing a chaotic mapping function to generate a dynamic threshold. The calculation formula of the dynamic threshold is: θ(t)=μ·θ(t-1)·(1-θ(t-1))+γ·f error (t); In the formula, θ(t) is the dynamic threshold of the current iteration; θ(t-1) is the dynamic threshold of the previous iteration; μ is the chaos parameter; γ is the disturbance coefficient; f error (t) is the error function of the current iteration; By introducing the semantic weight w s and context relevance c Dynamically adjust the integration priority of data fields. The formula is: In the formula, S merge Score the match between the two tables: b ,G b ) is the set of fields to be integrated; B is the total number of fields; Sim(F b ,G b ) indicates field F b and G b The semantic similarity of s is the semantic weight; r c is contextual relevance.

4. According to the big data-based market supervision dynamic monitoring and intelligent risk early warning system of claim 1, it is characterized in that: The full-process dynamic monitoring includes: Monitoring of task execution: obtaining the task name, start time, end time, execution status, duration, and execution log information; Monitoring of data volume changes: Based on the data table change rate, identify data tables with significant data volume differences, calculate the data table change rate and compare it with the preset threshold to determine if the change rate | V new -V old | / V old ×100% exceeds the preset threshold, it is considered to be significantly different, and a data table warning information is generated, which includes the name of the data table, the change rate, the abnormal type and the warning generation time; wherein, V new Indicates the amount of data in the current data table, V old Indicates the amount of data in the data table during the previous detection; Database resource monitoring: monitors the database disk usage, including used space, remaining space, total space and utilization, and generates resource utilization warning information; Dynamic monitoring of core indicators: Real-time acquisition of changes in core indicators of market entities, including the number of existing households, incremental trends, rate of change of registered capital, industry distribution trends and regional distribution trends.

5. According to the big data-based market supervision dynamic monitoring and intelligent risk early warning system of claim 1, it is characterized in that: The risk analysis module includes: A data preprocessing unit is used to obtain market entity monitoring data from the dynamic monitoring module and perform data cleaning and normalization processing; Chaos immune optimization unit, used to extract features from pre-processed data through chaos immune optimization algorithm, optimize antibody population, and obtain preliminary risk fitness value; Migration Immunity Optimization Unit, used to optimize migration and dynamically adapt risk factors across regions; The risk level assessment unit is used to generate the risk level of the market entity based on the optimized risk data; the risk type analysis unit is used to further analyze the risk type of the market entity based on the risk level, including credit risk, abnormal behavior risk and abnormal growth risk; The risk warning report generation unit is used to generate a visual risk warning report based on the risk level and risk type.

6. According to claim 5, a market supervision dynamic monitoring and intelligent risk early warning system based on big data is characterized in that: In the chaotic immune optimization algorithm unit, a dynamic weight function ω(t) is introduced, and an improved chaotic mapping formula is used to generate an antibody population, and the formula is: A(t+1)=μ·w(t)·A(t)·(1-A(t))+γ·f(A(t)); In the formula, t is the current iteration number; A(t) is the current risk characteristic value of the market entity; A(t+1) is the risk characteristic value of the market entity after chaos optimization and disturbance at the next moment; μ is the chaos parameter; γ is the perturbation coefficient; f(A(t)) is the fitness function; calculate the fitness of each antibody: Where R n is the risk factor value, is the benchmark value, n is the index of the current risk factor; N is the total number of risk factors; In the process of population evolution, combined with the risk factor matrix R of market entities f Adjust the mutation probability, the formula is: Where P m is the mutation probability of the current market entity; P m0 is the initial mutation probability; ||R f || is the Euclidean norm of the market entity risk factor; The optimized antibody population A(t+1) is generated as the input of the migration immunity optimization unit.

7. According to claim 6, a market supervision dynamic monitoring and intelligent risk early warning system based on big data is characterized in that: In the migration immunity optimization unit, market entities are divided according to industry or region to form distributed antibody nodes, and the migration probability is adjusted based on similarity and time decay: Where P trans (i,j,t) is the probability of antibody migration of market players from region i to region j; Sim(A i ,A j ) is the similarity of antibodies of market players in region i and region j; ∑ k≠i Sim(A i ,A k ) is the sum of similarities between region i and all other regions; λ is the time decay coefficient; Recalculate the environmental fitness F for the migrated antibodies env :F env =α·R growth +β·R deviation , where R growth represents the growth rate, R deviation represents the risk deviation; α and β are the weight coefficients of growth rate and risk deviation respectively; The optimized antibody population and fitness data are transmitted to the risk level assessment unit.

8. According to claim 7, a market supervision dynamic monitoring and intelligent risk early warning system based on big data is characterized in that: In the risk level assessment unit, the formula for generating the risk level of the market entity is: f multi (x)=ξ1·f risk (x)+ξ2·f priority (x); In the formula, f multi (x) is the final optimized value; f risk (x) is the risk deviation; f priority (x) is the priority of regulatory tasks; ξ1 and ξ2 are the weight coefficients of risk assessment and regulatory task priority, respectively; x represents the characteristic variable of the current market entity; Risk levels are divided according to the final optimization value: When the final optimized value f multi (x) When the set high-risk threshold is exceeded, the market entity is classified as a high-risk entity and requires priority supervision or emergency intervention measures; When the final optimized value f multi (x) When the market entity is between the high risk threshold and the low risk threshold, it is classified as a medium risk entity and requires regular monitoring and appropriate supervision; When the final optimized value f multi (x) When the risk is less than the set low-risk threshold, the market entity is classified as a low-risk entity and only requires routine supervision.

9. The early warning method of a market supervision dynamic monitoring and intelligent risk early warning system based on big data according to any one of claims 1 to 8, characterized in that: The method comprises: Collect market supervision data from multiple heterogeneous data sources, clean, convert and integrate them, eliminate data redundancy and inconsistency, and generate unified structured data; Based on the structured data, the market supervision task is dynamically monitored throughout the entire process, and according to the real-time monitoring results of the dynamic monitoring module, the abnormal behaviors or trends of the market entities are intelligently analyzed. The monitoring data is feature extracted and risk classified through the chaotic immune optimization algorithm and the migration immune algorithm, and a risk warning report including risk level, risk type and corresponding disposal suggestions is generated; Based on the analysis results of the risk analysis module, the stock, growth, industry trends and regional distribution of market entities are displayed in a visual way, and risk warning information is displayed dynamically; and the warning notification function is integrated to send the risk warning results to relevant regulatory departments via SMS, email or system push, supporting priority processing of risks at different levels and multi-department collaborative warning mechanism.

Images