A financial risk level labeling method and system based on data mining

Abstract

The present invention relates to the technical field of financial risk assessment, and in particular to a method for labeling financial risk levels based on data mining. The method comprises the following steps: obtaining target financial product data; determining the business field of the target financial product based on the target financial product data, thereby obtaining the industry data to which the product belongs; obtaining data collection target city data; collecting economic indicator data of the target city based on the data collection target city data, thereby obtaining an economic indicator data set; mining the implicit economic operation law of the economic indicator data set, thereby obtaining economic operation law data; labeling the target financial product with preliminary risk factors based on the economic operation law data, thereby obtaining preliminary risk factor data. This solution can achieve risk level labeling of financial products and investors through data mining and multi-dimensional evaluation methods.

Description

Technical Field

[0001] The present invention relates to the technical field of financial risk assessment, and in particular to a financial risk level labeling method based on data mining. Background Art

[0002] Financial risk management is considered one of the most important issues that financial institutions and investors must face in the current financial field. Effective financial risk management requires comprehensive risk assessment and labeling of financial products and investment portfolios. Traditional financial risk assessment methods mainly rely on statistical models and expert experience, but these methods have some obvious limitations, such as high subjectivity and high model complexity. Traditional methods usually rely on static models and are difficult to adapt to the needs of dynamic changes in the financial market. Most existing financial risk assessment methods use traditional statistical methods, which often only perform qualitative grading from a single dimension and are difficult to perform multi-dimensional quantitative assessments. This leads to a lack of comprehensive understanding of financial risks. Summary of the invention

[0003] Based on this, it is necessary for the present invention to provide a financial risk level labeling method based on data mining to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a financial risk level labeling method based on data mining includes the following steps:

[0005] Step S1: Obtain target financial product data; determine the business field of the target financial product based on the target financial product data, thereby obtaining the industry data to which the product belongs;

[0006] Step S2: Acquire data of the target city for data collection; collect economic indicator data of the target city based on the data of the target city for data collection, thereby obtaining an economic indicator data set; mine the economic indicator data set for implicit economic operation laws, thereby obtaining economic operation law data; perform preliminary risk factor labeling on the target financial product based on the economic operation law data, thereby obtaining preliminary risk factor data;

[0007] Step S3: Perform asset association network analysis on the enterprise to which the target financial product belongs, thereby obtaining asset association network data; perform technology innovation capability assessment on the enterprise to which the target financial product belongs, thereby obtaining technology innovation capability assessment data; perform risk resistance assessment on the enterprise to which the target financial product belongs based on the asset association network data and technology innovation capability assessment data, thereby obtaining comprehensive risk resistance data of the enterprise;

[0008] Step S4: Collect enterprise operation data of the target financial product to obtain an enterprise operation data set; perform feature extraction and dimension reduction on the enterprise operation data set to obtain product operating risk feature data; perform product potential operating risk rating on the product operating risk feature data to obtain product potential risk data;

[0009] Step S5: construct a comprehensive risk factor network for the preliminary risk factor data and the product potential risk data, thereby obtaining product risk factor network architecture data; label the target financial product with a final risk level according to the product risk factor network architecture data, thereby obtaining final risk level data; perform multi-dimensional visualization processing on the final risk level data, thereby obtaining a product risk level chart;

[0010] Step S6: Collect relevant records of the target financial product account of the investor to obtain the investor account record data; analyze the personal asset distribution of the investor to obtain the personal asset distribution data; conduct credit risk assessment on the investor based on the investor account record data and personal asset distribution data to obtain the investor credit risk assessment data; send the investor credit risk assessment data to the enterprise to which the target financial product belongs.

[0011] The present invention can obtain the data of the industry to which the product belongs by acquiring the data of the target financial product and determining its business field. The value of this step is that it provides basic information for subsequent risk assessment, because the risk factors faced by financial products in different industries may be different. By acquiring the data of the industry to which the product belongs, the financial risk level can be more accurately evaluated and marked. By acquiring the economic indicator data of the target city for data collection and mining the implicit economic operation law, the economic operation law data can be obtained. These law data can be used to mark the preliminary risk factors of the target financial product, thereby providing more comprehensive information for risk assessment. By analyzing the economic operation law, the risk level of the financial product can be more accurately assessed. By performing asset association network analysis and technology innovation capability assessment on the enterprise to which the target financial product belongs, asset association network data and technology innovation capability assessment data can be obtained. These data can be used to assess the risk resistance of the enterprise and obtain the comprehensive risk resistance data of the enterprise. Considering that the risk of financial products is often related to the risk of the enterprise to which they belong, the value of this step is that it provides important indicators for assessing the risk of financial products. By collecting, extracting features and reducing the dimension of the operation data of the enterprise to which the target financial product belongs, the product operating risk feature data can be obtained. These data can be used to rate the potential operating risks of products, thereby obtaining the potential risk data of products. By constructing a comprehensive risk factor network for the preliminary risk factor data and the potential risk data of products, the correlation between different factors can be taken into account in the risk assessment. This can more comprehensively assess the risk level of the target financial product and improve the accuracy and reliability of the assessment. By marking the final risk level of the target financial product according to the product risk factor network architecture data, the comprehensive risk factors can be incorporated into the assessment results, making the assessment results more operational. By marking the risk factors of financial products, assessing the risk resistance of enterprises and analyzing the potential risks of products, the risk level of financial products can be comprehensively assessed from multiple perspectives. This helps financial institutions to understand the potential risks of products more accurately and adopt corresponding risk management strategies and measures. By collecting relevant records of investor accounts and analyzing the distribution of personal assets, the personal situation and asset status of investors can be understood. By conducting credit risk assessment on investors based on these data, the credit risk level of investors in financial products can be assessed. Sending the investor credit risk assessment data to the enterprise to which the target financial product belongs can allow the enterprise to have a more comprehensive understanding of the credit status of investors and better manage risks. The present invention realizes the level labeling of financial risks, and realizes two-way information transmission and interaction between investors and enterprises through credit risk analysis and intelligent labeling of investors. In this two-way information transmission and interaction process, financial institutions can not only better understand users, but also provide risk management and financial management advice to users, helping users make more intelligent financial decisions.At the same time, users can also provide their own needs and feedback through interaction with financial institutions, so as to encourage financial institutions to better meet the needs of users and continuously optimize and improve products and services. In summary, the present invention can comprehensively consider multiple factors such as the industry to which the financial product belongs, the economic environment, the risk resistance of the enterprise, the influence of the product spokesperson, and the credit status of the investor, and comprehensively evaluate and grade the risks of the financial product. Traditional evaluation methods usually only consider a single dimension, and it is difficult to fully understand the risks of financial products. The method based on data mining can comprehensively evaluate the risks of financial products from different angles by comprehensively considering multiple factors, and improve the accuracy and reliability of the evaluation. Traditional evaluation methods often rely on static models and are difficult to adapt to the dynamic changes of the market environment. The method based on data mining can update the evaluation results in time and better adapt to market changes by real-time monitoring of changes in economic indicator data, product spokesperson influence and other factors. The method based on data mining can perform dynamic evaluation as the data is continuously updated and changed. By real-time monitoring of data, combined with machine learning and model updates, the risk level labeling can be adjusted and updated in time to adapt to changes in the market environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0013] Figure 1 A schematic flow chart of the steps of a financial risk level labeling method based on data mining according to an embodiment is shown.

[0014] Figure 2 A detailed flow chart of step S2 of an embodiment is shown.

[0015] Figure 3 A detailed flow chart of step S25 of an embodiment is shown. DETAILED DESCRIPTION

[0016] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0017] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0018] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0019] To achieve this, please refer to Figures 1 to 3 The present invention provides a financial risk level labeling method based on data mining, the method comprising the following steps:

[0020] Step S1: Obtain target financial product data; determine the business field of the target financial product based on the target financial product data, thereby obtaining the industry data to which the product belongs;

[0021] Specifically, for example, the historical price, trading volume and other relevant data of the stock investment product can be obtained from the API or database of the financial data provider. The obtained financial product data is analyzed, including statistical characteristics, price trend analysis, etc., to understand the basic situation of the product. Based on the obtained financial product data, the business field of the product is determined, for example, whether the product belongs to the financial services industry.

[0022] Step S2: Acquire data of the target city for data collection; collect economic indicator data of the target city based on the data of the target city for data collection, thereby obtaining an economic indicator data set; mine the economic indicator data set for implicit economic operation laws, thereby obtaining economic operation law data; perform preliminary risk factor labeling on the target financial product based on the economic operation law data, thereby obtaining preliminary risk factor data;

[0023] Specifically, for example, the name of the target city for data collection can be obtained as the target city for data collection. The acquired economic indicator data is cleaned and sorted to ensure the accuracy and completeness of the data. The economic indicator data is analyzed and mined using statistical analysis methods, time series analysis, or machine learning algorithms to discover the implicit economic operation laws. Based on the mined economic operation law data, it is applied to the risk labeling of the target financial product. For example, the fluctuation of a certain economic indicator can be used as a preliminary risk factor to label the risk level of the financial product.

[0024] Step S3: Perform asset association network analysis on the enterprise to which the target financial product belongs, thereby obtaining asset association network data; perform technology innovation capability assessment on the enterprise to which the target financial product belongs, thereby obtaining technology innovation capability assessment data; perform risk resistance assessment on the enterprise to which the target financial product belongs based on the asset association network data and technology innovation capability assessment data, thereby obtaining comprehensive risk resistance data of the enterprise;

[0025] Specifically, for example, the target company's financial statements, annual reports and other public information can be collected. Data mining and relationship analysis techniques, such as association rule mining algorithms, are used to analyze the company's asset structure, investment relationships, related subsidiaries and other information. Graph theory algorithms and network analysis methods, such as graph traversal algorithms and community discovery algorithms, are used to build asset association networks to clarify the asset relationships and mutual influences within and outside the company. Collect the target company's financial statements, annual reports and other public information. Data mining and relationship analysis techniques, such as association rule mining algorithms, are used to analyze the company's asset structure, investment relationships, related subsidiaries and other information. Graph theory algorithms and network analysis methods, such as graph traversal algorithms and community discovery algorithms, are used to build asset association networks to clarify the asset relationships and mutual influences within and outside the company. Comprehensively consider factors such as the company's asset structure, related relationships, and technological innovation capabilities to build a risk resistance assessment model or indicator system. Statistical analysis, risk models or machine learning methods, such as regression analysis, Monte Carlo simulation, etc., are used to quantitatively evaluate the company's risk resistance. The assessment results may include comprehensive indicators such as the company's risk tolerance, stress resistance, and crisis response capabilities.

[0026] Step S4: Collect enterprise operation data of the target financial product to obtain an enterprise operation data set; perform feature extraction and dimension reduction on the enterprise operation data set to obtain product operating risk feature data; perform product potential operating risk rating on the product operating risk feature data to obtain product potential risk data;

[0027] Specifically, for example, the enterprise operation data, including financial statements, sales data, production data, etc., can be obtained from the internal database of the financial company. Relevant economic indicator data, such as industry output value, industry growth rate, etc., can be obtained from public data sources. Data crawler technology is used to collect external information such as news reports and industry analysis reports of the manufacturing company from the Internet. Data cleaning and data preprocessing are performed on the collected enterprise operation data to remove noise and outliers. Feature engineering technology is applied to extract features related to product operating risks, such as financial indicators (profit margin, debt-to-asset ratio, etc.), market indicators (market share, industry competitiveness, etc.). Dimensionality reduction methods (such as principal component analysis) are used to convert high-dimensional feature data into low-dimensional representations to reduce the dimension and complexity of the data. To establish an operating risk assessment model, supervised learning algorithms (such as decision trees, random forests, support vector machines, etc.) or unsupervised learning algorithms (such as clustering algorithms) can be used. Use labeled historical data as a training set to train an operating risk assessment model to predict the potential operating risks of products. According to the model prediction results, the products are rated, such as high risk, medium risk, and low risk levels.

[0028] Step S5: construct a comprehensive risk factor network for the preliminary risk factor data and the product potential risk data, thereby obtaining product risk factor network architecture data; label the target financial product with a final risk level according to the product risk factor network architecture data, thereby obtaining final risk level data; perform multi-dimensional visualization processing on the final risk level data, thereby obtaining a product risk level chart;

[0029] Specifically, for example, preliminary risk factor data can be comprehensively analyzed with product potential risk data to construct a product risk factor network architecture. By considering the interrelationships and weights between various factors, a comprehensive risk factor network can be established. Based on the product risk factor network architecture data, a pre-defined assessment model or algorithm is used to label the target financial product with a final risk level. This may involve setting risk level classification standards and scoring systems, and evaluating products based on the weights and impact of different factors. The final risk level data is visualized in multiple dimensions to generate a product risk level chart. This can include using charts, graphs, and indicators to show the distribution of different risk levels, as well as comparisons with other products or market indicators.

[0030] Step S6: Collect relevant records of the target financial product account of the investor to obtain the investor account record data; analyze the personal asset distribution of the investor to obtain the personal asset distribution data; conduct credit risk assessment on the investor based on the investor account record data and personal asset distribution data to obtain the investor credit risk assessment data; send the investor credit risk assessment data to the enterprise to which the target financial product belongs.

[0031] Specifically, for example, account record data related to investors can be collected, including investment transaction records, fund flows, etc. This data can be obtained from the trading platform used by the investor or the relevant financial institution. Analyze the investor's personal assets, including fund allocation, portfolio distribution, etc. This can be obtained by statistically analyzing the account record data. Based on the investor's account record data and personal asset distribution data, use a credit assessment model or algorithm to assess the credit risk of the investor. This may involve analysis and comprehensive assessment of factors such as credit history, repayment ability, debt situation, and asset stability. Based on the results of the credit risk assessment, generate the investor's credit risk assessment data. This can be a score or grade used to indicate the investor's credit risk level. Send the investor's credit risk assessment data to the investment company to which the target financial product belongs. This can be transmitted through an encrypted and secure communication channel to ensure the confidentiality and integrity of the data.

[0032] The present invention can obtain the data of the industry to which the product belongs by acquiring the data of the target financial product and determining its business field. The value of this step is that it provides basic information for subsequent risk assessment, because the risk factors faced by financial products in different industries may be different. By acquiring the data of the industry to which the product belongs, the financial risk level can be more accurately evaluated and marked. By acquiring the economic indicator data of the target city for data collection and mining the implicit economic operation law, the economic operation law data can be obtained. These law data can be used to mark the preliminary risk factors of the target financial product, thereby providing more comprehensive information for risk assessment. By analyzing the economic operation law, the risk level of the financial product can be more accurately assessed. By performing asset association network analysis and technology innovation capability assessment on the enterprise to which the target financial product belongs, asset association network data and technology innovation capability assessment data can be obtained. These data can be used to assess the risk resistance of the enterprise and obtain the comprehensive risk resistance data of the enterprise. Considering that the risk of financial products is often related to the risk of the enterprise to which they belong, the value of this step is that it provides important indicators for assessing the risk of financial products. By collecting, extracting features and reducing the dimension of the operation data of the enterprise to which the target financial product belongs, the product operating risk feature data can be obtained. These data can be used to rate the potential operating risks of products, thereby obtaining the potential risk data of products. By constructing a comprehensive risk factor network for the preliminary risk factor data and the potential risk data of products, the correlation between different factors can be taken into account in the risk assessment. This can more comprehensively assess the risk level of the target financial product and improve the accuracy and reliability of the assessment. By marking the final risk level of the target financial product according to the product risk factor network architecture data, the comprehensive risk factors can be incorporated into the assessment results, making the assessment results more operational. By marking the risk factors of financial products, assessing the risk resistance of enterprises and analyzing the potential risks of products, the risk level of financial products can be comprehensively assessed from multiple perspectives. This helps financial institutions to understand the potential risks of products more accurately and adopt corresponding risk management strategies and measures. By collecting relevant records of investor accounts and analyzing the distribution of personal assets, the personal situation and asset status of investors can be understood. By conducting credit risk assessment on investors based on these data, the credit risk level of investors in financial products can be assessed. Sending the investor credit risk assessment data to the enterprise to which the target financial product belongs can allow the enterprise to have a more comprehensive understanding of the credit status of investors and better manage risks. The present invention realizes the level labeling of financial risks, and realizes two-way information transmission and interaction between investors and enterprises through credit risk analysis and intelligent labeling of investors. In this two-way information transmission and interaction process, financial institutions can not only better understand users, but also provide risk management and financial management advice to users, helping users make more intelligent financial decisions.At the same time, users can also provide their own needs and feedback through interaction with financial institutions, so as to encourage financial institutions to better meet the needs of users and continuously optimize and improve products and services. In summary, the present invention can comprehensively consider multiple factors such as the industry to which the financial product belongs, the economic environment, the risk resistance of the enterprise, the influence of the product spokesperson, and the credit status of the investor, and comprehensively evaluate and grade the risks of the financial product. Traditional evaluation methods usually only consider a single dimension, and it is difficult to fully understand the risks of financial products. The method based on data mining can comprehensively evaluate the risks of financial products from different angles by comprehensively considering multiple factors, and improve the accuracy and reliability of the evaluation. Traditional evaluation methods often rely on static models and are difficult to adapt to the dynamic changes of the market environment. The method based on data mining can update the evaluation results in time and better adapt to market changes by real-time monitoring of changes in economic indicator data, product spokesperson influence and other factors. The method based on data mining can perform dynamic evaluation as the data is continuously updated and changed. By real-time monitoring of data, combined with machine learning and model updates, the risk level labeling can be adjusted and updated in time to adapt to changes in the market environment.

[0033] Preferably, step S2 comprises the following steps:

[0034] Step S21: Acquire data collection target city data;

[0035] Specifically, for example, the target city for data collection may be Beijing, Shanghai, etc.

[0036] Step S22: formulating a data source strategy according to the data collection target city data, thereby obtaining data source strategy data;

[0037] Specifically, for example, the characteristics and requirements of the data collection target city data can be analyzed to determine the goal of formulating the data source strategy. The data source strategy is formulated by considering factors such as data reliability, coverage, update frequency, and data format.

[0038] Step S23: collecting economic indicator data of the target city according to the data source strategy data, thereby obtaining an economic indicator data set, wherein the economic indicator data set includes economic indicator data collected from multiple different data sources;

[0039] Specifically, for example, economic indicator data can be obtained from corresponding data sources according to the formulated data source strategy. Data from different data sources can be cleaned and integrated to ensure data accuracy and consistency. Economic indicator data from different data sources can be integrated into a comprehensive economic indicator data set. Economic indicator data can be stored and managed by using tools such as databases or data warehouses for data storage and indexing.

[0040] Step S24: Calculate the credibility of each data source in the economic indicator data set, thereby obtaining the credibility data corresponding to each data source, and remove the economic indicator data collected by the data source corresponding to the credibility data lower than the preset credibility threshold, thereby obtaining a credible economic indicator data set;

[0041] Specifically, for example, the following method can be used for credibility calculation and data cleaning: evaluate each data source, consider factors such as the authority of the data source, the credibility of the publishing agency, the accuracy of the data, and assign a credibility score to each data source. Or the data source credibility calculation formula performs credibility calculation on each data source in the economic indicator data set to obtain the credibility data of each data source. Set a preset credibility threshold, for example, set it to 80 points (out of 100 points). For each data source in the economic indicator data set, check its corresponding credibility data. If the credibility is lower than the preset threshold (for example, lower than 80 points), the economic indicator data corresponding to the data source is eliminated.

[0042] Step S25: Preliminary risk factor labeling of the target financial product is performed based on the credible economic indicator data set, thereby obtaining preliminary risk factor data.

[0043] Specifically, for example, the risk factors of the target financial products can be defined, and the economic indicators related to the target financial products can be determined based on the needs and domain knowledge. For example, for bond products, factors such as interest rates, economic growth rates, and inflation rates may be included. Based on the credible economic indicator data set, economic indicator data related to the risk factors of the target financial products are extracted. The extracted economic indicator data are preliminarily labeled and classified as high-risk factors, medium-risk factors, or low-risk factors based on historical data or professional judgment. The results of the preliminary labeling are sorted to obtain preliminary risk factor data, which contains each economic indicator and its corresponding risk level.

[0044] The present invention can provide basic data for subsequent financial risk assessment by acquiring the target city for data collection as the object of data collection in the subsequent steps. By formulating the data source strategy, the specific method and source of data collection can be determined to ensure that the collected data is reliable and effective. The data source strategy data provides guidance for managing and controlling the data collection process, which can improve the efficiency and accuracy of data collection, thereby providing a reliable data basis for subsequent risk assessment. By collecting economic indicator data for the target city, various indicator data reflecting the economic status of the city can be obtained. These economic indicator data are one of the important bases for assessing financial risks, and can be used to analyze information on the city's economic activities, employment, industrial structure, etc., thereby providing a comprehensive market environment background for financial risk assessment. By calculating the credibility of each data source in the economic indicator data set, the credibility of each data source can be evaluated and measured. This helps to screen out data sources with higher credibility, thereby improving the quality and reliability of the data. By eliminating the economic indicator data collected from data sources with a credibility lower than a preset threshold, the interference of unreliable data on financial risk assessment can be reduced, and the accuracy and credibility of the assessment results can be improved. By analyzing and marking the credible economic indicator data set, factors related to financial risk can be preliminarily determined. These preliminary risk factor data are one of the important bases for assessing the risk level of financial products. By marking the risk factors, basic information can be provided for subsequent risk assessments, and the risk level of financial products can be further analyzed and assessed.

[0045] Preferably, step S24 calculates the credibility of each data source in the economic indicator data set by using a data source credibility calculation formula, wherein the data source credibility calculation formula is as follows:

[0046]

[0047] In the formula, C is the credibility of the data source, n is the number of economic indicator data collected by the data source, i is the sequence number of the economic indicator data collected by the data source, and d is the i is the deviation between the i-th economic indicator data and the true value, π is the circumference of a circle, σ i is the standard deviation of the ith economic indicator data, t i is the collection time of the ith economic indicator data, e is the base of the natural logarithm, and x is any positive real number.

[0048] The present invention constructs a data source credibility calculation formula, in which represents the logarithmic deviation of the i-th economic indicator data collected by the data source, where d i Represents the deviation between the data and the true value, σ iRepresents the standard deviation of the data, and π represents the ratio of pi. This item can measure the accuracy and stability of the data. The smaller the deviation of the data, the smaller the standard deviation, the larger the logarithmic deviation, and the more reliable the data; conversely, the larger the deviation of the data, the larger the standard deviation, the smaller the logarithmic deviation, and the less reliable the data. Using the logarithmic function can make the ratio of the deviation and standard deviation of the data smoother and more reasonable, avoiding extreme values ​​that are too large or too small. represents the integral of the collection time of the i-th economic indicator data collected by the data source, where t i represents the data collection time, e represents the base of the natural logarithm, and x represents an arbitrary positive real number. This item can measure the timeliness of the data. The shorter the data collection time, the smaller the integral, which means the data is fresher; conversely, the longer the data collection time, the larger the integral, which means the data is older. Using the integral function can make the impact of the data collection time smoother and more reasonable, avoiding extreme values ​​that are too large or too small. Represents an infinite limit, where x represents an arbitrary positive real number. This term can be used as a correction factor to make the influence of other terms in the formula more significant and not affected by this term. The value of this term tends to 1, indicating that the credibility of the data source is not affected by this term, but mainly depends on the deviation, standard deviation and collection time of the data. The use of limit functions can make the formula more complete and rigorous, avoiding undefined or meaningless situations. This formula can comprehensively reflect the quality and reliability of economic indicator data collected by the data source, thereby evaluating the credibility of the data source and providing an effective basis and reference for data mining and risk level labeling. Multiple influencing factors such as data deviation, standard deviation, and collection time can be considered to make the credibility of the data source more objective and reasonable, avoiding the abnormal or accidental influence of individual data. Commonly used mathematical constants such as pi and the base of natural logarithms can be used to make the credibility of the data source more universal and standard, avoiding the influence of different units or systems.

[0049] Preferably, step S25 comprises the following steps:

[0050] Step S251: collecting basic enterprise information of the target city according to the data collection target city data, thereby obtaining an enterprise basic information data set, wherein the enterprise basic information data set includes a plurality of enterprise basic information data; monitoring the enterprise business conditions of the target city, thereby obtaining an enterprise business data set, wherein the enterprise business data set includes a plurality of enterprise business data; wherein the enterprise basic information data in the enterprise basic information data set and the enterprise business data in the enterprise business data set have a one-to-one correspondence;

[0051] Specifically, for example, you can obtain basic information about an enterprise, such as company name, registered capital, legal representative, etc., by visiting a website or using the API of an industrial and commercial data service provider. Monitor announcements and financial statements of stock exchanges to obtain operating data such as financial data, revenue, and profits of listed companies. By accessing enterprise databases or using the API of a commercial data provider, you can obtain more extensive enterprise data, including financial data and operating conditions of non-listed companies. It should be noted that the enterprise basic information data in the enterprise basic information data set and the enterprise operating data in the enterprise operating data set should correspond one to one, that is, they can be matched through a common identifier (such as the enterprise registration number).

[0052] Step S252: dividing the corresponding enterprises into industry ecology according to the enterprise basic information data set, thereby obtaining industrial ecological network data;

[0053] Specifically, for example, for each enterprise in the enterprise basic information data set, the industry information to which it belongs can be extracted. Enterprises are classified according to their industries using industry classification standards (such as national economic industry classification standards). An industrial ecological network is constructed based on the classification results, in which enterprises are nodes and the relationship between industries is the edge in the network. Graph theory algorithms (such as graph traversal, shortest path algorithm, etc.) can be used to analyze the structure and characteristics of the industrial ecological network, such as node degree centrality, betweenness centrality and other indicators. Betweenness centrality is an indicator used to measure the importance of nodes in a network. It measures the importance of a node in connecting other nodes in the network. Specifically, betweenness centrality indicates the frequency of passing through the node in the shortest path in the network, that is, the intermediary role played by the node in the shortest path connecting other nodes. The calculation method of betweenness centrality is to calculate the number of shortest paths between each pair of nodes in the network and calculate the proportion of the number of shortest paths passing through a node to the number of all shortest paths. A node with high betweenness centrality means that it has greater control or intermediary role in the propagation of information flow, resource transfer or influence in the network. In the above embodiment, by calculating the betweenness centrality of each node (i.e., enterprise) in the industrial ecology network, the importance and degree of intermediary role of the enterprise in the entire industrial ecology can be evaluated. Enterprises with high betweenness centrality may play a key role in the industrial ecology and have an important impact on information dissemination, resource flow, etc. This indicator can help understand and analyze the structural characteristics of the industrial ecology network and identify enterprises that have a significant influence on the entire network.

[0054] Step S253: constructing an enterprise ecological network based on the industrial ecological network data, thereby obtaining enterprise ecological network architecture data;

[0055] Specifically, for example, based on the industrial ecological network data, enterprises can be used as nodes to build an enterprise ecological network. According to the relationship between enterprises (such as cooperative relationship, competitive relationship, etc.), the connection edges between enterprises are established. Social network analysis methods (such as node degree centrality, clustering coefficient, etc.) can be used to analyze the topological structure and key nodes of the enterprise ecological network.

[0056] Step S254: mining the implicit economic operation law of the credible economic indicator data set according to the enterprise operation data set and the enterprise ecological network architecture data, thereby obtaining economic operation law data;

[0057] Specifically, for example, enterprises can be associated to form an enterprise network based on the cooperative relationship, supply chain relationship, etc. in the enterprise ecological network architecture data. Analyze the association law between enterprise business data and economic indicator data. The association rule mining method can be used to discover the implicit law and correlation between enterprise business data and economic indicator data. For example, the Apriori algorithm can be used to mine frequent item sets to find the associated item sets between business data and economic indicator data. Association rule evaluation indicators such as support, confidence, and lift can be used to determine the strength and importance of association rules. Perform cluster analysis. Cluster enterprises according to the similarity of business data and economic indicator data to reveal the economic operation mode and law between different enterprise groups. K-means clustering algorithm or hierarchical clustering algorithm can be used to cluster enterprises to form different economic operation modes. The characteristics and behaviors of different enterprise groups can be analyzed based on the clustering results to infer the economic operation law. Result analysis and interpretation. According to the mining results, the interaction, dependency and relationship between enterprises and economic indicators are analyzed to reveal the law and trend of economic operation.

[0058] Step S255: Preliminary risk factor labeling of the target financial product is performed according to the economic operation law data, thereby obtaining preliminary risk factor data.

[0059] Specifically, for example, economic indicators related to risk factors can be selected as features according to the characteristics of the target financial product and the purpose of the research. For example, GDP growth rate, inflation rate and interest rate level are selected as representatives of risk factors. Based on the selected features, the historical performance of the target financial product is labeled. For example, according to the changes in GDP growth rate, inflation rate and interest rate level, the risk factors of each period are labeled as high risk, medium risk or low risk. This can be automatically labeled by setting thresholds or using machine learning algorithms. Data analysis and statistics are performed based on the labeled risk factor data. Statistical methods (such as mean, variance, correlation, etc.) can be used to analyze the relationship between different risk factors and make a preliminary assessment of the risk characteristics of the target financial product. Based on the data analysis results, the preliminary risk factors of the target financial product are interpreted. For example, if it is found that a high inflation rate is highly correlated with the risk level of the fund, a preliminary conclusion can be drawn that the inflation rate is one of the important risk factors of the fund.

[0060] The present invention can obtain a data set including basic information data and business data of enterprises by collecting basic information and monitoring the business conditions of enterprises in the target city. These data are crucial for understanding the enterprise structure, scale, nature and business conditions of the target city. The basic information data of the enterprise can provide basic information about the name of the enterprise, registered capital, legal representative, etc., while the business data of the enterprise can provide data about the financial status, sales, profitability, etc. of the enterprise. Such data helps stakeholders such as financial institutions and investors to conduct a comprehensive assessment of the economic activities of the target city, so as to formulate corresponding investment decisions and risk management measures. By analyzing and processing the basic information data set of the enterprise, the corresponding enterprises can be divided into industry ecology, so as to obtain industrial ecological network data. These data can reveal the cooperative relationship, supply chain relationship, competitive relationship, etc. between different enterprises, and then build a network structure about the industrial ecology of the target city. Industrial ecological network data is of great significance for understanding the industrial layout of the target city, the interaction between key enterprises and the development trend of the industry. Such data helps financial institutions and decision makers to formulate industries, market forecasts and risk assessments, promote industrial development and promote economic growth. By processing and analyzing the industrial ecological network data, an enterprise ecological network can be constructed, so as to obtain enterprise ecological network architecture data. Enterprise ecological network architecture data can reveal the associations, cooperation models, dependencies, etc. between different enterprises. Such data is of great value for understanding the structure and operation of the enterprise ecosystem of the target city. For financial institutions and investors, enterprise ecological network architecture data can provide a more comprehensive perspective on enterprise relationships, helping them understand the stability, vulnerability and potential risks of the enterprise ecosystem of the target city, so as to formulate corresponding risk management and decision-making measures. Through the comprehensive analysis of enterprise operation data sets and enterprise ecological network architecture data, it is possible to mine implicit economic operation laws and obtain economic operation law data. These data can reveal the operation laws, trends and key factors of the target city's economy. By mining economic operation law data, it can help financial institutions and investors better understand the economic development of the target city, predict economic trends, and discover potential risks and opportunities. Such data is of great significance for formulating investment strategies and conducting risk management. By using economic operation law data to preliminarily label the target financial products with risk factors, preliminary risk factor data can be obtained. Preliminary risk factor data can reveal the potential risks and uncertainties faced by the target financial products. Such data can provide a more comprehensive risk assessment and decision-making basis for financial institutions and investors. By understanding preliminary risk factor data, financial institutions can take appropriate risk management measures and investors can make more informed investment decisions, thereby reducing risks and improving investment returns.

[0061] Preferably, step S254 includes the following steps:

[0062] Step S2541: performing endangered enterprise analysis on the enterprise operation data set, thereby obtaining endangered enterprise data;

[0063] Specifically, for example, the enterprise operation data set can be analyzed based on the definition and indicators of endangered enterprises, such as financial indicators (profit margin, debt repayment ability, etc.), market indicators (market share, sales growth rate, etc.), and competition indicators (industry status, market prospects, etc.). Use appropriate methods and algorithms, such as machine learning-based classification algorithms (such as decision trees, support vector machines, etc.) or rule engines to classify enterprises as endangered enterprises. Based on the classification results, determine the endangered enterprise set and obtain endangered enterprise data.

[0064] Step S2542: marking the corresponding enterprises in the enterprise ecological network architecture data as endangered enterprises according to the endangered enterprise data, thereby obtaining high-level enterprise ecological network architecture data;

[0065] Specifically, for example, the endangered enterprise data can be matched with the enterprise ecological network architecture data to find out the position of the endangered enterprise in the ecological network. Enterprise ecological network architecture data, which describes the cooperative relationship and supply chain relationship between enterprises. For endangered enterprises and their affiliated enterprises, endangered enterprise labeling can be achieved using professional software or algorithms, such as network analysis algorithms or community discovery algorithms. After labeling is completed, high-level enterprise ecological network architecture data is obtained, which contains the labeling information of endangered enterprises and can be used for subsequent analysis and mining.

[0066] Step S2543: analyzing the business model of the corresponding enterprise according to the enterprise business data set, thereby obtaining the enterprise business model data set;

[0067] Specifically, for example, the business data set can be used to analyze the business model of each enterprise. The business model can include product portfolio, market positioning, sales channels, operation strategy and other aspects. Use appropriate methods and techniques, such as cluster analysis and association rule mining, to analyze and mine the business data of enterprises to find the differences and commonalities in the business models of different enterprises. According to the analysis results, corresponding business model data is generated for each enterprise, which can be in the form of classification labels, association rule sets, etc., for subsequent economic operation law mining.

[0068] Step S2544: Mining the implicit economic operation laws of the trusted economic indicator data set based on the enterprise business model data set and the advanced enterprise ecological network architecture data, thereby obtaining economic operation law data.

[0069] Specifically, for example, a comprehensive data set can be constructed by combining the enterprise business model data set and the advanced enterprise ecological network architecture data, which includes enterprise business model data, enterprise relationship data and economic indicator data. Use appropriate data mining techniques and algorithms, such as association analysis, time series analysis, machine learning, etc., to analyze and mine the comprehensive data set to reveal the laws of economic operation. Based on the analysis results, extract and summarize the data of economic operation laws, which can be in the form of statistical indicators, trend patterns, association rules, etc., for understanding and predicting economic operation.

[0070] The present invention can identify endangered enterprises and obtain corresponding endangered enterprise data by performing endangered enterprise analysis on enterprise operation data sets. These data are helpful to understand the operating conditions, financial health, and possible risks and challenges of endangered enterprises. Such analysis can help relevant stakeholders such as financial institutions and investors to take appropriate measures, such as providing loan assistance, risk management, or rescue measures, to maintain financial stability and promote economic development. By combining endangered enterprise data with enterprise ecological network architecture data, endangered enterprises in the enterprise ecological network can be marked, thereby obtaining advanced enterprise ecological network architecture data. Such data helps to reveal the associations, cooperation models, and dependencies between enterprises. For stakeholders such as financial institutions and investors, these data can provide a more comprehensive perspective of the enterprise ecological network, helping them to better understand the impact and potential transmission effects of endangered enterprises on the entire economic system, and to take corresponding risk management and decision-making measures. By performing enterprise business model analysis on the enterprise operation data set, an enterprise business model data set can be obtained. These data can reveal key elements such as the enterprise's business strategy, market positioning, profit model, and business model. For financial institutions, investors and decision makers, understanding the business model of an enterprise can help them evaluate the competitiveness, profitability and sustainability of the enterprise, so as to make more accurate investment decisions and risk assessments. By combining the enterprise business model data set and the advanced enterprise ecological network architecture data, the implicit economic operation law mining of the trusted economic indicator data set can obtain economic operation law data. These data can help reveal the potential laws, correlations and trends in the economic system. For financial institutions and decision makers, understanding the economic operation law can provide deeper insights, help them formulate economic forecasts and manage risks, thereby promoting economic development and maintaining financial stability.

[0071] Preferably, step S2544 includes the following steps:

[0072] Step S25441: performing balance sheet analysis on each enterprise business model data in the enterprise business model data set, thereby obtaining an enterprise financial structure data set, wherein the enterprise financial structure data set includes enterprise financial structure data corresponding to each enterprise;

[0073] Specifically, for example, the balance sheet data of each enterprise can be extracted from the enterprise operation data set, including items such as assets, liabilities and owner's equity. Based on the balance sheet data, a series of financial ratios are calculated, such as current ratio, quick ratio, debt ratio, capital structure ratio, etc. These ratios reflect the financial status and structure of the enterprise. The calculated financial ratio data are cleaned to remove outliers and missing values ​​to ensure the accuracy and completeness of the data. The financial ratio data are normalized to have the same scale and range to avoid different influences of different ratios on the analysis results. The financial ratio data corresponding to each enterprise are organized into an enterprise financial structure data set, where each enterprise contains a set of financial ratio data to describe its financial structure.

[0074] Step S25442: performing enterprise profitability analysis on the corresponding enterprise according to the enterprise financial structure data set, thereby obtaining the enterprise profitability data set;

[0075] Specifically, for example, a series of profitability indicators, such as gross profit margin, net profit margin, return on assets, etc., can be calculated based on the enterprise financial structure data set. These indicators reflect the profitability and efficiency of the enterprise. The profitability indicator data corresponding to each enterprise is organized into an enterprise profitability data set, where each enterprise contains a set of profitability indicator data to describe its profitability.

[0076] Step S25443: Performing enterprise intelligent risk assessment on the corresponding enterprise according to the enterprise business model data set and the enterprise profitability data set, thereby obtaining an enterprise intelligent risk feature data set; topologically analyzing the high-level enterprise ecological network architecture data using the enterprise intelligent risk feature data set, thereby obtaining enterprise ecological network topological structure data;

[0077] Specifically, for example, based on the enterprise business model data set and the enterprise profitability data set, a machine learning algorithm (such as decision tree, random forest, etc.) can be used to establish an intelligent risk assessment model. Use the established model to perform intelligent risk assessment on each enterprise to obtain risk assessment results. Organize the intelligent risk assessment results of each enterprise into an enterprise intelligent risk feature data set, where each enterprise contains its risk assessment results and describes its intelligent risk characteristics. Enterprise ecological network architecture data, including information such as the association relationship and connection mode between enterprises. Take the enterprise as a node and build a network graph based on the association relationship between them. You can use a graph theory library (such as NetworkX) to create and operate a network graph. Apply topological methods and algorithms to parse and analyze the enterprise ecological network to reveal the structure and characteristics of the network. You can calculate indicators such as degree centrality, betweenness centrality, and closeness of nodes to evaluate the importance and influence of nodes in the network. You can detect community structure, identify subnetworks with close connections, and reveal the association and interaction between different enterprises. According to the results of topological analysis, extract key topological structure data, such as node centrality indicators, community division results, etc. These data are organized into a dataset of the topological structure of the enterprise ecological network, where each enterprise contains its location and characteristic information in the network. It should be noted that the specific methods and algorithms for topological analysis can be selected according to actual needs, such as using algorithms based on graph theory (such as the shortest path algorithm, clustering algorithm, etc.) or complex network analysis methods (such as small-world networks, scale-free networks, etc.). The selection of appropriate tools and techniques to achieve topological analysis, as well as the specific form and structure of the dataset, also requires decision-making and adjustment based on specific circumstances.

[0078] Step S25444: determining the target economic indicator for the credible economic indicator data set, thereby obtaining the target economic indicator data;

[0079] Specifically, for example, indicators related to economic operation can be selected from the credible economic indicator data set, and the calculation methods and data sources of these indicators can be determined to ensure their accuracy and reliability. The selected economic indicator data are cleaned to remove outliers and missing values ​​to ensure the accuracy and completeness of the data. The data of different indicators are organized into an economic indicator data set, with each indicator corresponding to a column of data and each row representing data at a time point.

[0080] Step S25445: classify and label the credible economic indicator data set according to the target economic indicator data, thereby constructing an economic indicator library;

[0081] Specifically, for example, each data point in the credible economic indicator data set can be classified and labeled based on economic indicator data, such as economic growth period, economic recession period, stable period, etc. Supervised learning algorithms (such as decision trees, support vector machines, etc.) or time series analysis methods (such as ARIMA models, exponential smoothing methods, etc.) can be used for labeling. The credible economic indicator data set after classification and labeling is organized into an economic indicator library, and each data point contains an economic indicator value and a corresponding classification label.

[0082] Step S25446: extracting the implicit dependency structure of economic operation from the economic indicator database, thereby obtaining an implicit dependency structure table of economic operation;

[0083] Specifically, for example, the implicit dependencies between different economic indicators can be extracted based on economic indicator data using correlation analysis or causal analysis. Statistical methods (such as correlation coefficient, grey correlation, etc.) or machine learning methods (such as causal graph model, Bayesian network, etc.) can be used for analysis. The extracted implicit dependencies are organized into an economic operation implicit dependency structure table to record the dependencies and strengths between different economic indicators.

[0084] Step S25447: Based on the enterprise ecological network topology data and the economic operation implicit dependency structure table, the economic operation law mining is carried out on the trusted economic indicator data set to obtain the economic operation law data.

[0085] Specifically, for example, data mining techniques (such as association rule mining, time series analysis, etc.) can be used to mine economic operation law data based on the enterprise ecological network topology data and the implicit dependency structure table of economic operation. The association law between different economic indicators can be discovered through the association rule mining method, or the trend and change of future economic indicators can be predicted through the time series analysis method. The economic operation laws mined are organized into data sets or reports to show the relationship and trend between different economic indicators, helping decision makers make corresponding economic policy adjustments and decisions.

[0086] The present invention can obtain an enterprise financial structure data set by performing balance sheet analysis on enterprise business model data. Enterprise financial structure data is an important indicator of the enterprise's financial status, including data on assets, liabilities and owner's equity. These data are of great significance for understanding the financial health, asset allocation and liability structure of the enterprise. Financial institutions and investors can use these data to evaluate the financial risk, capital operation ability and debt repayment ability of the enterprise, so as to formulate corresponding risk management and decision-making measures. By analyzing the enterprise financial structure data set, enterprise profitability analysis can be performed to obtain an enterprise profitability data set. Enterprise profitability data can reveal the profit level, profit structure and profitability indicators of the enterprise. These data are of great significance for evaluating the business status, profitability and competitiveness of the enterprise. Financial institutions and investors can use these data to evaluate the profit potential, development trend and competitive position of the enterprise, so as to make corresponding investment decisions. Through comprehensive analysis of the enterprise business model data set and the enterprise profitability data set, enterprise intelligent risk assessment can be performed to obtain an enterprise intelligent risk feature data set. Enterprise intelligent risk feature data can reveal the risk type, risk level and risk characteristics faced by the enterprise. These data can provide more comprehensive risk assessment and decision-making basis for financial institutions and investors. By analyzing the enterprise intelligent risk feature data set, the topological structure data of the enterprise ecological network can be constructed, which reveals the association, cooperation mode and dependency between different enterprises. Such data is of great significance for understanding the stability, vulnerability and potential risks of the enterprise ecosystem, and helps to formulate corresponding risk management and decision-making measures. By analyzing the trusted economic indicator data set, the target economic indicators can be determined, thereby obtaining the target economic indicator data. Target economic indicators are important indicators for measuring economic development and economic performance. Determining target economic indicators can help stakeholders such as financial institutions and research institutions to clearly measure the important aspects and goals of economic development, and provide quantitative evaluation and comparison basis. It is of great significance for formulating economic policies, monitoring economic conditions and evaluating economic development trends. By classifying and marking the trusted economic indicator data set according to the target economic indicators, an economic indicator library can be constructed. The economic indicator library is a database that integrates various economic indicator data, organized and stored according to different categories, dimensions and time. Building an economic indicator library can facilitate economic analysis, formulation and decision support. Stakeholders can obtain various economic indicator data through the economic indicator library, conduct data analysis, comparison and research, so as to better understand the development status, trends and influencing factors of the economy, and provide a scientific basis for decision-making and planning. By analyzing and processing the economic indicator database, the implicit dependency structure of economic operation can be extracted, thereby obtaining the implicit dependency structure table of economic operation. The implicit dependency structure table of economic operation reveals the mutual relationship, influence mechanism and degree of dependence between different economic indicators.This information is of great significance for understanding the complexity of the economic system, identifying key indicators and predicting economic trends. The extraction of the implicit dependency structure of economic operation can help stakeholders better understand the mechanism and laws of economic operation, and provide a scientific basis for risk management and decision support. By analyzing and processing the economic indicator library, the implicit dependency structure of economic operation can be extracted, thereby obtaining the implicit dependency structure table of economic operation. The implicit dependency structure table of economic operation reveals the mutual relationship, influence mechanism and degree of dependence between different economic indicators. This information is of great significance for understanding the complexity of the economic system, identifying key indicators and predicting economic trends. The extraction of the implicit dependency structure of economic operation can help stakeholders better understand the mechanism and laws of economic operation, and provide a scientific basis for risk management and decision support.

[0087] Preferably, step S3 comprises the following steps:

[0088] Step S31: Collect corporate governance data of the enterprise to which the target financial product belongs, thereby obtaining product corporate governance data;

[0089] Specifically, for example, reliable data sources such as annual reports, financial statements, and public disclosure documents can be identified for the collection of corporate governance data. Data related to corporate governance, such as board member information, equity structure, and internal control system, can be extracted from the selected data sources. Data can be collected and organized manually, or obtained from public data platforms using automated tools and crawler technology. The collected corporate governance data can be converted into data formats and named uniformly for subsequent analysis and use.

[0090] Step S32: performing asset association network analysis on the enterprise to which the target financial product belongs based on the product enterprise governance data, thereby obtaining asset association network data;

[0091] Specifically, for example, the information provided in the product enterprise governance data can be used to collect asset data of the enterprise to which the target financial product belongs, such as subsidiaries, controlling shares, investment projects, etc. The corresponding asset association information can be obtained through channels such as corporate annual reports, financial statements, and operating data. Based on the collected asset association data, an asset association network of the enterprise to which the target financial product belongs is constructed. This can be represented by establishing a relationship between nodes and edges, where nodes represent enterprises or assets and edges represent the association between them. The constructed asset association network is analyzed to explore the asset relationships and connection patterns within the enterprise. Graph theory algorithms (such as social network analysis, connectivity analysis, etc.) can be used to reveal network characteristics such as the degree of association between assets and key nodes.

[0092] Step S33: Analyze the composition of the core technical talent team of the enterprise to which the target financial product belongs, so as to obtain talent composition data;

[0093] Specifically, for example, the talent information of the company to which the target financial product belongs can be collected through the company's internal human resources department or related databases. This includes key information such as the title, education, and work experience of core technical personnel. Use the collected talent data to analyze the composition of the core technical talent team. The number of people with different titles, education, or work experience can be counted to understand the talent structure and distribution of the company. Explore the correlation between core technical talents, such as cooperation between personnel, sharing of experience, etc. Social network analysis methods (such as co-occurrence analysis, relationship strength analysis, etc.) can be used to reveal the degree of correlation between talents and key figures.

[0094] Step S34: Conduct in-depth mining of patent data of the enterprise to which the target financial product belongs, thereby obtaining patent technology data;

[0095] Specifically, for example, patent databases (such as patent search systems, patent database platforms, etc.) can be used to search and obtain patent information of companies to which target financial products belong. Searches can be conducted based on company names, keywords, technical fields, and other conditions to obtain patent data sets related to the company. The acquired patent data is cleaned and sorted to remove duplicate, invalid, or irrelevant patent information. Patents can be screened and classified using technical means such as patent classification and keyword extraction to ensure the accuracy and availability of the data. Technical analysis is performed on the cleaned and sorted patent data to explore the company's patent technology fields, innovation priorities, and so on. Technical methods such as text mining and natural language processing can be used to extract key information from patent texts, such as technical keywords, patent citation relationships, and so on.

[0096] Step S35: Evaluate the technological innovation capability of the enterprise to which the target financial product belongs based on the talent composition data and patent technology data, thereby obtaining technological innovation capability evaluation data;

[0097] Specifically, for example, a series of indicators reflecting technological innovation capabilities can be defined based on talent composition data and patent technology data. Indicators may include the number of patents, patent quality, number and structure of technical personnel, technical cooperation relationships, etc. Integrate talent composition data and patent technology data to establish a comprehensive technological innovation capability evaluation data set. Data fusion and data processing techniques can be used to uniformly weight or normalize indicators from different data sources. Based on the integrated data set, an evaluation model or algorithm is used to evaluate the technological innovation capabilities of the company to which the target financial product belongs. Multi-dimensional evaluation methods, such as indicator weighting method and hierarchical analysis method, can be used to comprehensively consider the importance and weight of each indicator.

[0098] Step S36: Evaluate the risk resistance of the enterprise to which the target financial product belongs based on the asset association network data and the technological innovation capability assessment data, thereby obtaining the enterprise's comprehensive risk resistance data.

[0099] Specifically, for example, a series of indicators reflecting the risk resistance of enterprises can be defined based on asset-related network data and technology innovation capability assessment data. Indicators may include asset diversity, the relationship between technology innovation capability and risk resistance, and the stability of key talents. Integrate asset-related network data and technology innovation capability assessment data to establish a comprehensive risk resistance assessment data set. Data fusion and data processing techniques can be used to uniformly weight or normalize indicators from different data sources. Based on the integrated data set, an assessment model or algorithm is used to assess the risk resistance of the enterprise to which the target financial product belongs. Comprehensive evaluation methods, such as indicator weighting method and hierarchical analysis method, can be used to comprehensively consider the importance and weight of various indicators.

[0100] By collecting product enterprise governance data, the present invention can understand the governance structure, decision-making mechanism, internal control and other situations of the enterprise to which the target financial product belongs, and provide an important basis for evaluating the stability and sustainability of the enterprise. This helps investors, regulators and other stakeholders to judge the governance risks and potential problems of the enterprise, and then make corresponding decisions. Asset association network analysis can reveal the asset structure, association relationship and risk transmission path of the enterprise. By analyzing the asset association network of the enterprise to which the target financial product belongs, the core assets, dependencies and risk exposure of the enterprise can be identified. This helps to understand the asset quality, liquidity risk and systemic risk of the enterprise, and provides a reference for investors and regulators to evaluate the risk tolerance of the enterprise. Core technical talents are an important driving factor for enterprise innovation and competitiveness. By analyzing the core technical talent composition of the enterprise to which the target financial product belongs, the technical strength, innovation ability and talent reserve of the enterprise can be understood. This helps to evaluate the technical competitiveness, long-term development potential and innovation ability of the enterprise, and provides an important reference for investment decisions and strategic planning. Patents are an important manifestation of the innovation achievements of enterprises, and are of great significance to the technical strength and competitive advantage of enterprises. By deeply mining the patent data of the enterprise to which the target financial product belongs, the technical field, technical layout and innovation activities of the enterprise can be understood. This helps to evaluate the technical strength, technical barriers and technical innovation capabilities of enterprises, and provides investors and decision makers with important information about the technical value and development potential of enterprises. By comprehensively evaluating the talent composition data and patent technology data, the technical innovation capabilities of the enterprises to which the target financial products belong can be evaluated to understand the innovation strength, technology R&D investment and innovation results of the enterprises. This helps to evaluate the leading position of enterprises in the field of technology, the sustainability of innovation capabilities and future growth potential, and provides key judgment indicators for investors and decision makers. By comprehensively analyzing the asset association network data and the technical innovation capability evaluation data, the comprehensive risk resistance of the enterprises to which the target financial products belong can be evaluated. This helps to understand the risk tolerance of enterprises, the degree of systemic risk exposure and the strength of risk resistance. The evaluation results can provide important references for investors, regulators and other stakeholders to help them make risk management and decision-making.

[0101] Preferably, step S4 comprises the following steps:

[0102] Step S41: collecting enterprise operation data of the target financial product, thereby obtaining an enterprise operation data set;

[0103] Specifically, for example, you can determine which data sources to collect enterprise operation data from, such as financial statements, market data, industry reports, etc. You can obtain data from public data sources, financial institution data, or internal databases. Use appropriate methods to capture enterprise operation data, which can be through API interfaces, crawlers, or manual collection. Ensure the accuracy and completeness of the data.

[0104] Step S42: Structuring the enterprise operation data set to obtain a structured enterprise operation data set;

[0105] Specifically, for example, unstructured or semi-structured enterprise operation data sets can be converted into structured data, such as text mining and natural language processing for text data, image recognition and feature extraction for image data, etc. The processed data is organized in a certain format for subsequent analysis and modeling. For example, the data is organized into a table format, with each row representing a sample and each column representing a feature.

[0106] Step S43: inputting the structured enterprise operation data set into a preset convolutional autoencoder deep learning model, and automatically extracting high-dimensional features of the structured enterprise operation data set through an unsupervised feature learning algorithm, thereby obtaining product operation risk feature data;

[0107] Specifically, for example, a suitable convolutional autoencoder deep learning model can be selected for feature learning and feature extraction. Existing model architectures can be selected, such as Convolutional and Autoencoder. Use the structured enterprise operation dataset as input data to ensure that the data format matches the model requirements. Use the structured enterprise operation dataset to train the convolutional autoencoder deep learning model, learn data features through unsupervised learning, and extract high-dimensional feature representations. Use the trained convolutional autoencoder model to extract features from the structured enterprise operation dataset to obtain product operating risk feature data. These features can be the output of the hidden layer in the model or the output of the encoder.

[0108] Step S44: construct a GBDT gradient boosting decision tree model, input the product operation risk feature vector as an input sample feature into the GBDT gradient boosting decision tree model for training, thereby obtaining a product operation risk rating model;

[0109] Specifically, for example, a training data set including a product operating risk feature vector and a corresponding product operating risk rating label can be prepared. Ensure the accuracy of the data and the consistency of the annotations. Set the parameters of the GBDT model, such as the number of trees, the depth of the tree, the learning rate, etc. These parameters can be tuned according to the specific problem and data set. Use the product operating risk feature vector as the input sample feature, the product operating risk rating label as the output label, and use the training data set to train the GBDT model. During the training process, the model gradually fits the training data in an iterative manner to continuously improve the prediction effect. Use the validation data set to evaluate the trained GBDT model, and calculate evaluation indicators (such as accuracy, recall, F1 value, etc.) to measure the performance of the model. Save the trained GBDT model for subsequent use and prediction.

[0110] Step S45: Input the product operating risk feature vector into the product operating risk rating model for feature mapping and operating risk prediction, thereby obtaining product potential risk data.

[0111] Specifically, for example, you can prepare the product business risk feature vector data to be predicted. Input the product business risk feature vector to be predicted into the trained GBDT model, perform feature mapping through the model, and obtain the corresponding potential risk features. Input the mapped features into the rating model, and use the model to predict business risks. The model will give the corresponding risk rating results based on the input features. Based on the business risk prediction results, the potential risk data of the product is obtained. It can be a specific value of the risk rating or a risk level classification.

[0112] The present invention can obtain an enterprise operation data set by collecting the operation data of the enterprise associated with the target financial product. Enterprise operation data refers to various data generated by the enterprise in its business activities, including financial data, sales data, production data, etc. These data are of great significance for evaluating the business status, performance and risk level of the enterprise. By collecting enterprise operation data, a comprehensive and accurate data foundation can be established to provide support for subsequent data processing and analysis. By performing structured processing on the enterprise operation data set, the original data can be converted into a structured data form, so that the data has better organization, readability and analyzability. The structured enterprise operation data set can provide a clearer and easier to understand data expression method, which is convenient for subsequent data analysis and modeling. The structured data set can also perform data cleaning, noise removal and outlier processing to improve the quality and accuracy of the data. By using the convolutional autoencoder deep learning model, the structured enterprise operation data set can be automatically extracted for high-dimensional features. This unsupervised feature learning algorithm can learn from the data and discover potential important features without manually defining feature engineering. By extracting product operating risk feature data, the potential risk factors in the operation of the enterprise can be revealed, providing a basis for subsequent risk rating and risk prediction. By constructing the GBDT gradient boosting decision tree model, the product operating risk feature vector can be used to evaluate and predict the business risk of the enterprise. The GBDT model is a powerful machine learning algorithm that can effectively capture the nonlinear relationship and interaction between features. By training the product operating risk rating model, enterprises can be classified and rated according to the product operating risk feature vector, helping financial institutions, investors and decision makers to better understand and evaluate the business risks of enterprises. By inputting the product operating risk feature vector into the product operating risk rating model, feature mapping and business risk prediction can be performed. In this way, potential risk data based on the product operating risk characteristics can be obtained, that is, the prediction results of the business risk of the enterprise. These data can provide information about the potential risk level of the product and the challenges it may face. By predicting and evaluating the operating risks of products, financial institutions, investors and decision makers can better understand the risk status of products and take corresponding measures to manage risks, formulate strategies and make decisions.

[0113] Preferably, step S5 comprises the following steps:

[0114] Step S51: Perform risk sensitivity analysis on product potential risk data to obtain network key node factor data;

[0115] Specifically, for example, based on the product potential risk data, the sensitivity of different risk factors can be analyzed using statistical analysis methods or machine learning algorithms. The contribution of risk factors to product risks can be measured using correlation coefficients, variance analysis, regression analysis and other methods. Based on the results of the risk sensitivity analysis, risk factors with greater influence are determined and used as network key node factors. Ranking or threshold methods can be used to determine network key node factors.

[0116] Step S52: constructing a comprehensive risk factor network based on the preliminary risk factor data and the network key node factor data, thereby obtaining product risk factor network architecture data;

[0117] Specifically, for example, the preliminary risk factor data can be integrated with the network key node factor data to construct a product risk factor network. Graph theory algorithms and network analysis tools (such as Gephi) can be used to construct a risk factor network, where nodes represent risk factors and edges represent the relationship between factors. Based on the construction results of the comprehensive risk factor network, the architecture data of the product risk factor network is obtained, including node attributes, edge weights, etc.

[0118] Step S53: Obtaining a market volatility data set;

[0119] Specifically, for example, market-related data can be obtained from data sources such as financial data suppliers and exchanges. Stock index data, futures price data, foreign exchange data, etc. are collected, which reflect market volatility. API interfaces or data subscription services can be used to obtain real-time or historical market volatility data.

[0120] Step S54: performing risk propagation simulation on the product risk factor network architecture data according to the market volatility data set, thereby obtaining risk propagation path data;

[0121] Specifically, for example, appropriate risk propagation simulation methods can be selected, such as Monte Carlo simulation, system dynamics model, etc. Based on the market volatility data set, the propagation relationship between different risk factors is simulated. The market volatility data is used as input to simulate the risk propagation of the product risk factor network architecture data. Random sampling, simulation and other techniques can be used to simulate the risk propagation process, considering the impact of market volatility on different risk factors. During the risk propagation simulation process, the risk propagation path data is recorded and obtained, including the state change of each node, the time series of the risk propagation path and other information.

[0122] Step S55: performing risk communication influence analysis on the risk communication path data, thereby obtaining risk communication influence data;

[0123] Specifically, for example, appropriate risk propagation influence analysis methods can be selected, such as node centrality analysis, network connectivity analysis, etc. These methods can help determine which nodes in the risk propagation path have a greater impact on the risk propagation of the entire system. Based on the risk propagation path data, the propagation influence of each node is calculated and evaluated. Graph theory algorithms, network analysis tools, etc. can be used for analysis to determine key nodes and propagation paths.

[0124] Step S56: Using the risk propagation influence data, the target financial product is labeled with a final risk level, thereby obtaining final risk level data;

[0125] Specifically, for example, an appropriate final risk level labeling method can be selected based on business needs and risk assessment criteria. The final risk level can be determined based on risk communication influence data and pre-set assessment rules. Based on risk communication influence data and assessment rules, the final risk level of the target financial product is labeled. Classification algorithms, decision trees and other methods can be used for labeling to divide products into different risk levels.

[0126] Step S57: Perform multi-dimensional visualization processing on the final risk level data to obtain a product risk level chart.

[0127] Specifically, for example, you can select appropriate data visualization tools, such as Python's Matplotlib, Tableau, PowerBI and other tools. Based on the final risk level data, use the selected data visualization tool to generate a product risk level chart. You can draw bar charts, pie charts, radar charts and other forms to visualize products of different risk levels. You can add labels, color codes and other methods to enhance the readability of the chart and the effect of information transmission.

[0128] The present invention can identify the sensitive factors of potential product risks, that is, factors that have an important impact on product risks, through risk sensitivity analysis. These key node factor data can help understand the main sources and key factors of product risks, and provide an important basis for subsequent risk assessment and control. Through the construction of a comprehensive risk factor network, the preliminary risk factor data can be combined with the key node factor data to construct a comprehensive network architecture of product risk factors. These data can help understand the relationship and propagation path between different risk factors, so as to more comprehensively evaluate the risk level and potential risk of the product. The comprehensive risk factor network architecture data provides an important basis for subsequent risk propagation simulation and risk level labeling. By obtaining a market volatility data set, the impact of market dynamics on financial product risks can be understood. These data can help understand the changing trend of market risks and the path of risk propagation, and provide an important basis for subsequent risk propagation simulation and influence analysis. Through risk propagation simulation, the propagation path of product risks in the market can be simulated and predicted. These data can help understand the propagation effect and potential impact of product risks in the market, and provide an important basis for subsequent risk propagation influence analysis and risk level labeling. Through risk propagation influence analysis, the degree of influence of different propagation paths on product risks can be evaluated. These data can help understand the transmission effect and potential impact of different transmission paths on product risks, and provide an important basis for subsequent risk level labeling. Risk transmission influence data can quantify the influence of product risk transmission, providing a more accurate and comprehensive evaluation basis for risk level labeling. By comprehensively considering the influence of risk transmission, the risk level of financial products can be more accurately assessed. Risk transmission influence data provides quantitative information on risk transmission paths and transmission effects, which helps to determine the accuracy and reliability of risk levels. By marking risk levels, the impact of different risk factors on financial products can be quantified. This helps decision makers better understand the importance and potential impact of different risk factors, and provides a basis for formulating corresponding risk management strategies and decisions. By labeling different financial products with risk levels, the risk levels of different products can be compared horizontally. This helps decision makers understand the risk characteristics and risk exposure of different products, and then conduct effective risk management and select appropriate investment portfolios. The risk level labeling results can provide important reference information for decision makers, helping them to formulate risk control strategies and management measures. Based on accurate risk level assessment, decision makers can better understand the risk characteristics and potential risks of products, so as to make wise decisions and take appropriate risk management measures. By performing multi-dimensional visualization on the final risk level data, the risk level charts of the products can be presented intuitively. These charts can help decision makers better understand and analyze the risk situation of the products, so as to make corresponding decisions and measures. Multi-dimensional visualization can provide more intuitive and easy-to-understand product risk level information, and improve the accuracy and efficiency of decision-making.

[0129] Preferably, step S6 comprises the following steps:

[0130] Step S61: collecting relevant records of the target financial product account of the investor, thereby obtaining the investor account record data;

[0131] Specifically, for example, the source of investor account record data can be determined, such as financial institutions such as banks, securities companies, and fund companies. Determine whether it is necessary to establish a data interface with relevant financial institutions or obtain data licenses. Obtain investor account record data through data interfaces with financial institutions or other legal means. Collect data related to target financial products, including account transaction records, fund flows, and position information.

[0132] Step S62: Perform high-frequency trading detection on the investor account record data, thereby obtaining the investor's high-frequency trading data;

[0133] Specifically, for example, the definition of high-frequency trading can be determined based on business needs and regulatory requirements, such as the number of transactions per unit time or the frequency of transactions. Appropriate high-frequency trading detection methods can be selected, such as statistical analysis based on transaction frequency, algorithm models, etc. High-frequency trading detection can be performed using data mining, machine learning and other technologies. High-frequency trading detection can be performed on investor account record data to screen out transaction records that meet the definition of high-frequency trading. Time windows, statistical indicators and other methods can be used to detect high-frequency trading and identify high-frequency trading behaviors.

[0134] Step S63: statistically summarize the asset data of the investor, thereby obtaining the asset structure data of the investor; and analyze the individual asset distribution of the investor according to the asset structure data of the investor, thereby obtaining the individual asset distribution data;

[0135] Specifically, for example, the assets of investors can be counted and summarized based on the fund flow and position information in the investor account record data. The market value or fund proportion of different types of assets (such as stocks, bonds, funds, etc.) is counted. Based on the asset structure data, the distribution of investors' personal assets is analyzed. The proportion of different types of assets in total assets can be calculated, and pie charts or bar charts can be drawn to show the distribution of personal assets. According to the results of the personal asset distribution analysis, personal asset distribution data is obtained. Personal assets can be classified according to different asset types, different market value ranges, etc. to obtain corresponding data.

[0136] Step S64: performing transaction behavior pattern recognition on the investor's high-frequency transaction data and personal asset distribution data, thereby obtaining the investor's transaction behavior pattern data;

[0137] Specifically, for example, appropriate trading behavior pattern recognition methods can be selected, such as machine learning algorithms, time series analysis, etc. According to specific needs and data characteristics, a suitable model or algorithm is selected to identify trading behavior patterns. The selected method is used to analyze and model investors' high-frequency trading data and personal asset distribution data to identify investors' trading behavior patterns. Patterns in terms of investors' trading frequency, trading size, asset allocation, etc. can be explored.

[0138] Step S65: conducting credit risk assessment on the investor based on the investor's transaction behavior pattern data, thereby obtaining the investor's credit risk assessment data;

[0139] Specifically, for example, appropriate credit risk assessment indicators such as transaction frequency, transaction size, position concentration, etc. can be selected according to business needs and risk management standards. Based on the selected credit risk assessment indicators, an assessment model can be established, and statistical analysis, machine learning and other methods can be used. Use past investor data as historical data for model training and verification to optimize the model's predictive ability and stability. Historical data can include investors' trading records, position information, asset distribution, trading frequency and other relevant data. A credit risk assessment indicator system for investors can be formulated in combination with industry experience and expert opinions. Credit risk assessment of investors is performed based on the investor's trading behavior pattern data and the constructed credit risk assessment model. The assessment result can be a quantitative score or grade reflecting the investor's credit risk level.

[0140] Step S66: perform credit risk measurement and analysis on the investor's credit risk assessment data, thereby obtaining the investor's credit risk assessment data; use the investor's credit risk assessment data to intelligently label the credit files of the corresponding investors, thereby obtaining the investor's credit file label data; and send the investor's credit file label data to the enterprise to which the target financial product belongs.

[0141] Specifically, for example, the credit risk assessment data of investors can be measured and analyzed, and indicators such as average risk level and risk distribution can be calculated. The performance and risk exposure of investors of different risk levels in the market can be further analyzed. According to the credit risk assessment data of investors, the corresponding credit file labels are intelligently labeled for each investor. The labels can be risk levels, risk categories, etc., which are used to quickly identify and classify the credit risk levels of investors. The credit file label data of investors is sent to the enterprise to which the target financial product belongs. The label data can be provided to the enterprise through data interfaces, file transfers, etc. to support risk management and decision-making processes.

[0142] The present invention can obtain the transaction records and account information of investors on target financial products by collecting investor account record data. These data can provide information on the investor's trading behavior, preferences and risk tolerance, and provide basic data for risk assessment and personalized recommendation in subsequent steps. By performing high-frequency trading detection on the investor account record data, the investor's high-frequency trading behavior can be identified. These high-frequency trading data can help understand the investor's trading activity and trading preferences, and further analyze the investor's risk preferences and investment behavior. By statistics and analysis of the investor's asset data, the investor's asset structure and asset allocation can be understood. These personal asset distribution data can help evaluate the investor's financial status and risk tolerance, and provide a basis for subsequent credit risk assessment and personalized recommendation. By performing trading behavior pattern recognition on the investor's high-frequency trading data and personal asset distribution data, the investor's trading behavior pattern and preference can be understood. These trading behavior pattern data can help understand the investor's investment strategy and risk preference, and provide a basis for subsequent credit risk assessment and personalized recommendation. By performing credit risk assessment on the investor's trading behavior pattern data, the investor's credit status and default risk can be assessed. These credit risk assessment data can help financial institutions understand the credit risk level of investors and provide a basis for risk management and decision-making. By measuring and analyzing the credit risk assessment data of investors, we can comprehensively assess the credit status and risk level of investors. These credit risk assessment data and credit file label data can help financial institutions better understand the credit status and risk characteristics of investors, provide a basis for risk management and decision-making, and also provide personalized services and product recommendations for target financial products.

[0143] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0144] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A financial risk level labeling method based on data mining, characterized in that: The following steps are involved: Step S1: Obtain target financial product data; determine the business field of the target financial product based on the target financial product data, thereby obtaining the industry data to which the product belongs; Step S2: Acquire data of a data collection target city; collect economic indicator data of the target city according to the data collection target city data, thereby obtaining an economic indicator data set; mine the economic indicator data set for implicit economic operation laws, thereby obtaining economic operation law data; Conduct preliminary risk factor labeling of target financial products based on economic operation law data to obtain preliminary risk factor data; Step S3: Perform asset association network analysis on the enterprise to which the target financial product belongs, thereby obtaining asset association network data; Conduct technological innovation capability assessment on the target financial product’s enterprise, thereby obtaining technological innovation capability assessment data; Evaluate the risk resistance of the target financial product’s enterprise based on asset-related network data and technology innovation capability assessment data, thereby obtaining the enterprise’s comprehensive risk resistance data; Step S4: Collect enterprise operation data of the target financial product, thereby obtaining an enterprise operation data set; Perform feature extraction and dimensionality reduction on enterprise operation data sets to obtain product operation risk feature data; Rating the product's potential operating risk based on the product's operating risk characteristic data, thereby obtaining the product's potential risk data; Step S4 includes the following steps: Step S41: collecting enterprise operation data of the target financial product, thereby obtaining an enterprise operation data set; Step S42: Structuring the enterprise operation data set to obtain a structured enterprise operation data set; Step S43: inputting the structured enterprise operation data set into a preset convolutional autoencoder deep learning model, and automatically extracting high-dimensional features of the structured enterprise operation data set through an unsupervised feature learning algorithm, thereby obtaining product operation risk feature data; Step S44: construct a GBDT gradient boosting decision tree model, input the product operation risk feature vector as an input sample feature into the GBDT gradient boosting decision tree model for training, thereby obtaining a product operation risk rating model; Step S45: inputting the product operating risk feature vector into the product operating risk rating model for feature mapping and operating risk prediction, thereby obtaining product potential risk data; Step S5: construct a comprehensive risk factor network for the preliminary risk factor data and the product potential risk data, thereby obtaining product risk factor network architecture data; label the target financial product with a final risk level according to the product risk factor network architecture data, thereby obtaining final risk level data; perform multi-dimensional visualization processing on the final risk level data, thereby obtaining a product risk level chart; Step S6: collecting relevant records of the target financial product account of the investor, thereby obtaining the investor account record data; analyzing the personal asset distribution of the investor, thereby obtaining the personal asset distribution data; conducting a credit risk assessment on the investor based on the investor account record data and the personal asset distribution data, thereby obtaining the investor credit risk assessment data; sending the investor credit risk assessment data to the enterprise to which the target financial product belongs; Step S6 includes the following steps: Step S61: collecting relevant records of the target financial product account of the investor, thereby obtaining the investor account record data; Step S62: Perform high-frequency trading detection on the investor account record data, thereby obtaining the investor's high-frequency trading data; Step S63: statistically summarize the asset data of the investor, thereby obtaining the asset structure data of the investor; and analyze the individual asset distribution of the investor according to the asset structure data of the investor, thereby obtaining the individual asset distribution data; Step S64: performing transaction behavior pattern recognition on the investor's high-frequency transaction data and personal asset distribution data, thereby obtaining the investor's transaction behavior pattern data, including exploring the investor's transaction frequency, transaction scale, and asset allocation patterns; Step S65: conducting credit risk assessment on the investor based on the investor's transaction behavior pattern data, thereby obtaining the investor's credit risk assessment data; Step S66: perform credit risk measurement and analysis on the investor's credit risk assessment data, thereby obtaining the investor's credit risk assessment data; use the investor's credit risk assessment data to intelligently label the credit files of the corresponding investors, thereby obtaining the investor's credit file label data; and send the investor's credit file label data to the enterprise to which the target financial product belongs.

2. The financial risk level labeling method based on data mining according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: Acquire data collection target city data; Step S22: formulating a data source strategy according to the data collection target city data, thereby obtaining data source strategy data; Step S23: collecting economic indicator data of the target city according to the data source strategy data, thereby obtaining an economic indicator data set, wherein the economic indicator data set includes economic indicator data collected from multiple different data sources; Step S24: Calculate the credibility of each data source in the economic indicator data set, thereby obtaining the credibility data corresponding to each data source, and remove the economic indicator data collected by the data source corresponding to the credibility data lower than the preset credibility threshold, thereby obtaining a credible economic indicator data set; Step S25: Preliminary risk factor labeling of the target financial product is performed based on the credible economic indicator data set, thereby obtaining preliminary risk factor data.

3. The financial risk level labeling method based on data mining according to claim 2 is characterized in that: Step S24 calculates the credibility of each data source in the economic indicator data set using a data source credibility calculation formula, where the data source credibility calculation formula is as follows: ； In the formula, The credibility of the data source, is the number of economic indicator data collected by the data source, It is the serial number of the economic indicator data collected by the data source. For the The deviation of economic indicator data from the true value, is the circumference of a circle, For the The standard deviation of economic indicators, For the The collection time of economic indicator data, is the base of natural logarithms, is any positive real number.

4. The financial risk level labeling method based on data mining according to claim 2 is characterized in that: Step S25 includes the following steps: Step S251: collecting basic enterprise information of the target city according to the data collection target city data, thereby obtaining an enterprise basic information data set, wherein the enterprise basic information data set includes a plurality of enterprise basic information data; monitoring the enterprise business conditions of the target city, thereby obtaining an enterprise business data set, wherein the enterprise business data set includes a plurality of enterprise business data; wherein the enterprise basic information data in the enterprise basic information data set and the enterprise business data in the enterprise business data set have a one-to-one correspondence; Step S252: dividing the corresponding enterprises into industry ecology according to the enterprise basic information data set, thereby obtaining industrial ecological network data; Step S253: constructing an enterprise ecological network based on the industrial ecological network data, thereby obtaining enterprise ecological network architecture data; Step S254: mining the implicit economic operation law of the credible economic indicator data set according to the enterprise operation data set and the enterprise ecological network architecture data, thereby obtaining economic operation law data; Step S255: Preliminary risk factor labeling of the target financial product is performed according to the economic operation law data, thereby obtaining preliminary risk factor data.

5. The financial risk level labeling method based on data mining according to claim 4 is characterized in that: Step S254 includes the following steps: Step S2541: performing endangered enterprise analysis on the enterprise operation data set, thereby obtaining endangered enterprise data; Step S2542: marking the corresponding enterprises in the enterprise ecological network architecture data as endangered enterprises according to the endangered enterprise data, thereby obtaining high-level enterprise ecological network architecture data; Step S2543: analyzing the business model of the corresponding enterprise according to the enterprise business data set, thereby obtaining the enterprise business model data set; Step S2544: Mining the implicit economic operation laws of the trusted economic indicator data set based on the enterprise business model data set and the advanced enterprise ecological network architecture data, thereby obtaining economic operation law data.

6. The financial risk level labeling method based on data mining according to claim 5 is characterized in that: Step S2544 includes the following steps: Step S25441: performing balance sheet analysis on each enterprise business model data in the enterprise business model data set, thereby obtaining an enterprise financial structure data set, wherein the enterprise financial structure data set includes enterprise financial structure data corresponding to each enterprise; Step S25442: performing enterprise profitability analysis on the corresponding enterprise according to the enterprise financial structure data set, thereby obtaining the enterprise profitability data set; Step S25443: Performing enterprise intelligent risk assessment on the corresponding enterprise according to the enterprise business model data set and the enterprise profitability data set, thereby obtaining an enterprise intelligent risk feature data set; topologically analyzing the high-level enterprise ecological network architecture data using the enterprise intelligent risk feature data set, thereby obtaining enterprise ecological network topological structure data; Step S25444: determining the target economic indicator for the credible economic indicator data set, thereby obtaining the target economic indicator data; Step S25445: classify and label the credible economic indicator data set according to the target economic indicator data, thereby constructing an economic indicator library; Step S25446: extracting the implicit dependency structure of economic operation from the economic indicator database, thereby obtaining an implicit dependency structure table of economic operation; Step S25447: Based on the enterprise ecological network topology data and the economic operation implicit dependency structure table, the economic operation law mining is carried out on the trusted economic indicator data set to obtain the economic operation law data.

7. The financial risk level labeling method based on data mining according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31: Collect corporate governance data of the enterprise to which the target financial product belongs, thereby obtaining product corporate governance data; Step S32: performing asset association network analysis on the enterprise to which the target financial product belongs based on the product enterprise governance data, thereby obtaining asset association network data; Step S33: Analyze the composition of the core technical talent team of the enterprise to which the target financial product belongs, so as to obtain talent composition data; Step S34: Conduct in-depth mining of patent data of the enterprise to which the target financial product belongs, thereby obtaining patent technology data; Step S35: Evaluate the technological innovation capability of the enterprise to which the target financial product belongs based on the talent composition data and patent technology data, thereby obtaining technological innovation capability evaluation data; Step S36: Evaluate the risk resistance of the enterprise to which the target financial product belongs based on the asset association network data and the technological innovation capability assessment data, thereby obtaining the enterprise's comprehensive risk resistance data.

8. The financial risk level labeling method based on data mining according to claim 1 is characterized in that: Step S5 includes the following steps: Step S51: Perform risk sensitivity analysis on product potential risk data to obtain network key node factor data; Step S52: constructing a comprehensive risk factor network based on the preliminary risk factor data and the network key node factor data, thereby obtaining product risk factor network architecture data; Step S53: Obtaining a market volatility data set; Step S54: performing risk propagation simulation on the product risk factor network architecture data according to the market volatility data set, thereby obtaining risk propagation path data; Step S55: performing risk communication influence analysis on the risk communication path data, thereby obtaining risk communication influence data; Step S56: Using the risk propagation influence data, the target financial product is labeled with a final risk level, thereby obtaining final risk level data; Step S57: Perform multi-dimensional visualization processing on the final risk level data to obtain a product risk level chart.

Images