A risk warning method and system for information innovation based on big data

Abstract

The present invention discloses a risk warning method and system for financial credible innovation based on big data. First, the multi-source credible innovation original data of the target financial credible innovation project is obtained, and the data dimension is reduced to form multi-source credible innovation dimension reduction data; then, the credible innovation risk features are extracted from the dimension reduction data based on feature engineering to obtain the multi-dimensional risk features of the target project; then, a financial credible innovation risk warning model is constructed based on big data technology; finally, the real-time data of the target project is monitored using the model, risk warning information is identified and generated, and risk warning operations are performed according to the warning information to form a risk warning plan. The present invention can effectively monitor and warn risks of financial credible innovation projects in real time, and improve the security and stability of financial credible innovation projects.

Description

Technical Field

[0001] The present invention relates to the field of risk warning technology, and in particular to a method and system for early warning of information technology risks based on big data. Background Art

[0002] With the rapid development of information technology and financial technology, financial information innovation projects have been widely used in the financial industry. However, these projects also face numerous risks during their innovation and development, such as fraud in financial transactions, risks arising from market dynamics, and potential anomalies in user behavior data. If these risks are not promptly identified and warned of, they could cause significant losses to financial institutions and even trigger systemic risks.

[0003] Existing risk warning methods mostly rely on traditional risk management models. These models typically assess risk based on historical data and expert experience, resulting in poor timeliness and adaptability. Meanwhile, the development of big data technologies has provided a rich source of data on finance, market dynamics, and user behavior for risk warning. However, effectively utilizing this multi-source data to improve the accuracy and real-time nature of risk warnings remains a pressing challenge.

[0004] To address these challenges, the present invention proposes a big data-based risk warning method and system for financial credible innovation projects. By acquiring and processing multi-source data, applying feature engineering, and combining big data technologies, the present invention constructs a system capable of real-time monitoring and early warning of financial credible innovation project risks. Through comprehensive analysis and processing of financial transaction data, market dynamics data, and user behavior data, the present invention can more accurately identify potential risks and issue early warning information in a timely manner, thereby helping financial institutions adopt effective risk management measures and ensure the security and stability of financial credible innovation projects. Summary of the Invention

[0005] In order to solve at least one of the above technical problems, the present invention proposes a risk warning method and system for information technology innovation based on big data.

[0006] The first aspect of the present invention provides a risk warning method for information innovation based on big data, comprising:

[0007] Obtain multi-source original data of a target financial trust innovation project, wherein the multi-source original data includes financial transaction data, market dynamic data, and user behavior data, and perform data dimensionality reduction on the multi-source original data to form multi-source trust innovation reduced dimensionality data;

[0008] Extract risk features of the multi-source credential creation dimensionality reduction data based on feature engineering to obtain multi-dimensional risk features of the target financial credential creation project;

[0009] Constructing a financial information innovation risk warning model based on big data technology, and importing the multi-dimensional risk characteristics into the financial information innovation risk warning model for training;

[0010] Based on the financial credential creation risk warning model, real-time data of the target financial credential creation project is monitored to identify the credential creation risks of the target financial credential creation project and generate risk warning information;

[0011] Based on the risk warning information, risk warning operations are performed on the target financial information innovation project to form a risk warning plan.

[0012] In this solution, the multi-source original data of the target financial trust creation project is obtained, and the multi-source original data includes financial transaction data, market dynamic data, and user behavior data. The multi-source original data is subjected to data dimensionality reduction to form multi-source trust creation dimensionality reduction data, specifically:

[0013] Building a data acquisition system for the target financial trust innovation project based on distributed data acquisition technology, and obtaining multi-source trust innovation original data of the target financial trust innovation project according to the data acquisition system, wherein the multi-source trust innovation original data includes financial transaction data, market dynamic data, and user behavior data;

[0014] Classify the multi-source credential creation original data according to data categories, assign a category label to each multi-source credential creation original data, and obtain multi-source credential creation classified data;

[0015] Introducing the LDA algorithm, calculating the mean vector of each category of data in the multi-source information creation classification data according to the LDA algorithm, and calculating the intra-class scatter matrix of each category of data according to the mean vector;

[0016] Calculate the overall mean vector of each category of data based on the multi-source information creation classification data, and calculate the inter-class scatter matrix of the multi-source information creation classification data based on the overall mean vector;

[0017] Calculate the inverse matrix of the intra-class scatter matrix, multiply the inverse matrix by the inter-class scatter matrix to obtain a generalized eigenvalue problem matrix, solve the eigenvalues ​​of the generalized eigenvalue problem matrix to obtain the eigenvalues ​​and corresponding eigenvectors of the generalized eigenvalue problem matrix, sort the eigenvalues ​​of the generalized eigenvalue problem matrix in descending order, and select the eigenvectors corresponding to the first N largest eigenvalues ​​to construct a data dimensionality reduction matrix of the multi-source information creation original data;

[0018] The multi-source information creation original data is linearly transformed according to the data dimensionality reduction matrix to form multi-source information creation dimensionality reduction data.

[0019] In this solution, the risk features of the multi-source credential creation dimensionality reduction data are extracted based on feature engineering to obtain the multi-dimensional risk features of the target financial credential creation project, specifically:

[0020] Performing time series data smoothing processing on the multi-source credential creation dimensionality reduction data, and performing feature enhancement processing on the multi-source credential creation dimensionality reduction data through a feature enhancement operator to obtain feature-enhanced multi-source credential creation data;

[0021] Performing a data normalization operation on the feature-enhanced multi-source trust-innovation data, calculating the variance of each feature data in the feature-enhanced multi-source trust-innovation data, removing feature data with variances lower than a preset threshold, and obtaining a first screening risk feature of the target financial trust-innovation project;

[0022] Obtain historical multi-source original data of the target financial credential creation project, obtain risk labels of each feature data in the target financial credential creation project in different numerical ranges based on the historical multi-source original data, construct target variables with the risk labels of each feature data in different numerical ranges, perform a chi-square test on each data feature in the feature-enhanced multi-source credential creation data based on the target variable, and select features with chi-square statistics ranked before a preset ranking as the second screening risk features;

[0023] Combine the first screening risk characteristics and the second screening risk characteristics, and deduplicate the combined risk characteristics to obtain the multi-dimensional risk characteristics of the target financial information innovation project.

[0024] In this solution, the financial information innovation risk warning model is constructed based on big data technology, and the multi-dimensional risk characteristics are introduced into the financial information innovation risk warning model for training, specifically:

[0025] Extracting original data of each risk feature in the multi-dimensional risk features of the target financial credential creation project based on the multi-source credential creation original data, and visualizing the original data of each risk feature through a box plot to obtain a box plot of each risk feature;

[0026] Performing a distribution hypothesis test on the box plot to determine the distribution type of the original data of each risk feature, and estimating the distribution parameters of each risk feature based on the distribution type;

[0027] Calculating a risk threshold based on the distribution parameter, determining a risk range for each risk feature based on the risk threshold, and determining a risk level for each risk range;

[0028] Backtest the risk range based on historical multi-source information innovation data, evaluate the accuracy of the risk interval, and calculate the frequency and severity of events within different risk level intervals;

[0029] A financial information innovation risk warning model is constructed based on a convolutional neural network. The risk range of each risk feature, the risk level of each risk range, the frequency and severity of occurrence within different risk level intervals are imported into the financial information innovation risk warning model for training to obtain a complete financial information innovation risk warning model.

[0030] In this solution, the real-time data of the target financial credential creation project is monitored based on the financial credential creation risk warning model, and the credential creation risk of the target financial credential creation project is identified to generate risk warning information, specifically:

[0031] Acquire real-time data of each risk feature of the target financial information innovation project according to the data acquisition system, import the real-time data of each risk feature into the financial information innovation risk early warning model for real-time risk monitoring, output the risk probability value of each risk feature of the target financial information innovation project, and obtain the probability data of risk event occurrence;

[0032] Calculate the risk score of the target financial information innovation project based on the probability data of the risk event, and divide the risk level according to the risk score;

[0033] The risk event occurrence probability data, risk score, and risk level are used to generate risk warning information.

[0034] In this solution, the risk warning operation is performed on the target financial information innovation project based on the risk warning information to form a risk warning plan, specifically:

[0035] Determine a risk event based on the risk warning information, determine the impact scope of the risk event based on expert experience, and determine the warning scope of the risk warning information based on the impact scope;

[0036] Classify the risk warning information according to the risk level to obtain risk level warning information;

[0037] Based on the risk level warning information, when the risk level is high, the high-risk project is isolated; when the risk level is medium, the project management personnel are arranged to manually review the risk event, determine the nature and severity of the risk based on the manual review, and formulate emergency response measures based on the review results; when the risk level is low, the risk event is monitored in real time, and the real-time risk changes and risk diffusion changes of the low-risk event are monitored. The risk level is adjusted according to the real-time risk changes and risk diffusion changes to obtain a risk warning plan;

[0038] The risk warning plan is subjected to risk warning operations according to the warning scope.

[0039] The second aspect of the present invention further provides a credible innovation risk warning system based on big data, the system comprising: a memory, a processor, the memory including a credible innovation risk warning method program based on big data, and when the credible innovation risk warning method program based on big data is executed by the processor, the following steps are implemented:

[0040] Obtain multi-source original data of a target financial trust innovation project, wherein the multi-source original data includes financial transaction data, market dynamic data, and user behavior data, and perform data dimensionality reduction on the multi-source original data to form multi-source trust innovation reduced dimensionality data;

[0041] Extract risk features of the multi-source credential creation dimensionality reduction data based on feature engineering to obtain multi-dimensional risk features of the target financial credential creation project;

[0042] Constructing a financial information innovation risk warning model based on big data technology, and importing the multi-dimensional risk characteristics into the financial information innovation risk warning model for training;

[0043] Based on the financial credential creation risk warning model, real-time data of the target financial credential creation project is monitored to identify the credential creation risks of the target financial credential creation project and generate risk warning information;

[0044] Based on the risk warning information, risk warning operations are performed on the target financial information innovation project to form a risk warning plan.

[0045] In this solution, the financial information innovation risk warning model is constructed based on big data technology, and the multi-dimensional risk characteristics are introduced into the financial information innovation risk warning model for training, specifically:

[0046] Extracting original data of each risk feature in the multi-dimensional risk features of the target financial credential creation project based on the multi-source credential creation original data, and visualizing the original data of each risk feature through a box plot to obtain a box plot of each risk feature;

[0047] Performing a distribution hypothesis test on the box plot to determine the distribution type of the original data of each risk feature, and estimating the distribution parameters of each risk feature based on the distribution type;

[0048] Calculating a risk threshold based on the distribution parameter, determining a risk range for each risk feature based on the risk threshold, and determining a risk level for each risk range;

[0049] Backtest the risk range based on historical multi-source information innovation data, evaluate the accuracy of the risk interval, and calculate the frequency and severity of events within different risk level intervals;

[0050] A financial information innovation risk warning model is constructed based on a convolutional neural network. The risk range of each risk feature, the risk level of each risk range, the frequency and severity of occurrence within different risk level intervals are imported into the financial information innovation risk warning model for training to obtain a complete financial information innovation risk warning model.

[0051] In this solution, the real-time data of the target financial credential creation project is monitored based on the financial credential creation risk warning model, and the credential creation risk of the target financial credential creation project is identified to generate risk warning information, specifically:

[0052] Acquire real-time data of each risk feature of the target financial information innovation project according to the data acquisition system, import the real-time data of each risk feature into the financial information innovation risk early warning model for real-time risk monitoring, output the risk probability value of each risk feature of the target financial information innovation project, and obtain the probability data of risk event occurrence;

[0053] Calculate the risk score of the target financial information innovation project based on the probability data of the risk event, and divide the risk level according to the risk score;

[0054] The risk event occurrence probability data, risk score, and risk level are used to generate risk warning information.

[0055] In this solution, the risk warning operation is performed on the target financial information innovation project based on the risk warning information to form a risk warning plan, specifically:

[0056] Determine a risk event based on the risk warning information, determine the impact scope of the risk event based on expert experience, and determine the warning scope of the risk warning information based on the impact scope;

[0057] Classify the risk warning information according to the risk level to obtain risk level warning information;

[0058] Based on the risk level warning information, when the risk level is high, the high-risk project is isolated; when the risk level is medium, the project management personnel are arranged to manually review the risk event, determine the nature and severity of the risk based on the manual review, and formulate emergency response measures based on the review results; when the risk level is low, the risk event is monitored in real time, and the real-time risk changes and risk diffusion changes of the low-risk event are monitored. The risk level is adjusted according to the real-time risk changes and risk diffusion changes to obtain a risk warning plan;

[0059] The risk warning plan is subjected to risk warning operations according to the warning scope.

[0060] The present invention discloses a risk warning method and system for financial credible innovation based on big data. First, the multi-source credible innovation original data of the target financial credible innovation project is obtained, and the data dimension is reduced to form multi-source credible innovation dimension reduction data; then, the credible innovation risk features are extracted from the dimension reduction data based on feature engineering to obtain the multi-dimensional risk features of the target project; then, a financial credible innovation risk warning model is constructed based on big data technology; finally, the real-time data of the target project is monitored using the model, risk warning information is identified and generated, and risk warning operations are performed according to the warning information to form a risk warning plan. The present invention can effectively monitor and warn risks of financial credible innovation projects in real time, and improve the security and stability of financial credible innovation projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 The present invention shows a flowchart of a risk warning method for information innovation based on big data;

[0062] Figure 2 The flowchart of the present invention for generating risk warning information is shown;

[0063] Figure 3 A flow chart showing a risk warning scheme formed according to the present invention is shown;

[0064] Figure 4 A block diagram of an information innovation risk warning system based on big data is shown in the present invention. DETAILED DESCRIPTION

[0065] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the absence of conflict, the embodiments of the present application and the features therein can be combined with each other.

[0066] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0067] Figure 1 A flow chart of a method for early warning of information innovation risks based on big data is shown in the present invention.

[0068] like Figure 1 As shown, the first aspect of the present invention provides a risk warning method for information innovation based on big data, comprising:

[0069] S102, obtaining multi-source original data of a target financial trust-based innovation project, wherein the multi-source original data includes financial transaction data, market dynamic data, and user behavior data, and performing data dimensionality reduction on the multi-source original data to form multi-source trust-based innovation dimensionality reduction data;

[0070] S104, extracting risk features of the multi-source credential creation dimensionality reduction data based on feature engineering to obtain multi-dimensional risk features of the target financial credential creation project;

[0071] S106, constructing a financial information innovation risk warning model based on big data technology, and importing the multi-dimensional risk characteristics into the financial information innovation risk warning model for training;

[0072] S108: Monitoring the real-time data of the target financial credential creation project based on the financial credential creation risk warning model, identifying the credential creation risks of the target financial credential creation project and generating risk warning information;

[0073] S110, perform risk warning operations on the target financial information innovation project based on the risk warning information to form a risk warning plan.

[0074] It should be noted that by obtaining multi-source original data including financial transaction data, market dynamics data and user behavior data, all aspects of the target financial trust innovation project can be fully covered, which is helpful to identify the full picture of potential risks. Data dimensionality reduction improves the efficiency of data processing by reducing redundant information in the data set, reduces computational complexity and storage requirements; feature engineering can effectively extract and transform the reduced dimensionality data to make it more representative and discriminative, extract the risk characteristics of the target financial trust innovation project, and help understand which factors have a greater impact on the risk of the target trust innovation project. Through the extraction and analysis of multi-dimensional risk characteristics, risks can be comprehensively assessed from multiple angles to improve risk management. The comprehensiveness and accuracy of risk identification; the financial information innovation risk warning model built based on big data technology can process massive data and complex feature relationships, significantly improving the accuracy and real-time nature of risk prediction; using the risk warning model to monitor the real-time data of the target financial information innovation project can quickly identify potential risks and generate risk warning information in a timely manner to avoid the spread and proliferation of risks; through dynamic monitoring and analysis of real-time data, it can accurately identify abnormal behaviors and abnormal patterns, thereby achieving efficient risk identification and warning; based on the generated risk warning information, risk warning operations can be taken in a timely manner to proactively respond to potential risks and avoid or mitigate the negative impact of risks.

[0075] According to an embodiment of the present invention, the multi-source original data of the target financial trust creation project is obtained, and the multi-source original data includes financial transaction data, market dynamic data, and user behavior data. The multi-source original data is subjected to data dimensionality reduction to form multi-source trust creation dimensionality reduction data, specifically:

[0076] Building a data acquisition system for the target financial trust innovation project based on distributed data acquisition technology, and obtaining multi-source trust innovation original data of the target financial trust innovation project according to the data acquisition system, wherein the multi-source trust innovation original data includes financial transaction data, market dynamic data, and user behavior data;

[0077] Classify the multi-source credential creation original data according to data categories, assign a category label to each multi-source credential creation original data, and obtain multi-source credential creation classified data;

[0078] Introducing the LDA algorithm, calculating the mean vector of each category of data in the multi-source information creation classification data according to the LDA algorithm, and calculating the intra-class scatter matrix of each category of data according to the mean vector;

[0079] Calculate the overall mean vector of each category of data based on the multi-source information creation classification data, and calculate the inter-class scatter matrix of the multi-source information creation classification data based on the overall mean vector;

[0080] Calculate the inverse matrix of the intra-class scatter matrix, multiply the inverse matrix by the inter-class scatter matrix to obtain a generalized eigenvalue problem matrix, solve the eigenvalues ​​of the generalized eigenvalue problem matrix to obtain the eigenvalues ​​and corresponding eigenvectors of the generalized eigenvalue problem matrix, sort the eigenvalues ​​of the generalized eigenvalue problem matrix in descending order, and select the eigenvectors corresponding to the first N largest eigenvalues ​​to construct a data dimensionality reduction matrix of the multi-source information creation original data;

[0081] The multi-source information creation original data is linearly transformed according to the data dimensionality reduction matrix to form multi-source information creation dimensionality reduction data.

[0082] It should be noted that the data collection system of the target financial information innovation project is constructed through distributed data collection technology to obtain the financial transaction data, market dynamics data, and user behavior data of the target financial information innovation project. The financial transaction data include transaction date, transaction amount, transaction type (such as buy, sell, transfer, payment, etc.), transaction object identity information, transaction subject (such as bought or sold stocks, bonds, foreign exchange, futures, options, etc.); the market dynamics data include securities market data (such as stock price, trading volume, etc.), foreign exchange market data (such as exchange rate, trading volume, interest rate data); the user behavior data includes browsing behavior, search behavior, and trading behavior of the target financial information innovation project system; the multi-source information innovation original data is classified according to data categories to define target categories based on business needs and data characteristics. For example, financial transaction data is classified by transaction type (such as payment, transfer, investment); market dynamics data is classified by industry (such as technology, medicine, energy); user behavior data is classified by behavior type (such as browsing, purchasing, evaluation); the target financial information technology innovation project refers to information technology innovation (information technology innovation) projects in the financial field, including digital banks, financial management systems, and anti-money laundering systems; these projects usually use modern information technology (such as big data, cloud computing, blockchain, artificial intelligence, etc.) to improve the efficiency, security and innovation of financial services; by calculating the mean vector and intra-class scatter matrix of each category of data, the distribution characteristics of the data within each category and the relationship between the data points can be described. For each category of data, its mean vector (that is, the average value of each feature) is calculated, and the difference and similarity between each data point and the mean vector are measured by the intra-class scatter matrix, which is one of the basic steps for data feature extraction and dimensionality reduction in the LDA algorithm; the overall mean vector is obtained by weighted averaging the mean vectors of all categories of data, which is used to measure the central trend of the overall data distribution. The inter-class scatter matrix reflects the degree of data separation between different categories by calculating the difference between the mean vector of each category and the overall mean vector; first, the inverse matrix of the intra-class scatter matrix is ​​calculated. This step usually requires that the intra-class scatter matrix is ​​non-singular to ensure the existence of the inverse matrix. The inverse matrix is ​​then multiplied by the inter-class scatter matrix to obtain the generalized eigenvalue problem matrix. Then, by solving the generalized eigenvalue problem matrix, its eigenvalues ​​and corresponding eigenvectors are obtained. The properties of these eigenvalues ​​and eigenvectors can reveal the main direction of change and important features of the data: the eigenvalue descending sorting is based on the size of the eigenvalue, and the main eigenvectors are selected to retain the most significant information and change patterns in the data. The construction of the data dimensionality reduction matrix uses these eigenvectors to perform linear transformations to map the high-dimensional multi-source information creation original data to a low-dimensional space to form multi-source information creation dimensionality reduction data.This process can effectively reduce the dimension of the data and retain important data structure features; the LDA (Linear Discriminant Analysis) is a supervised learning algorithm, and the purpose of LDA is to project the data into a low dimension.

[0083] According to an embodiment of the present invention, the risk features of the multi-source credential creation dimensionality reduction data are extracted based on feature engineering to obtain the multi-dimensional risk features of the target financial credential creation project, specifically:

[0084] Performing time series data smoothing processing on the multi-source credential creation dimensionality reduction data, and performing feature enhancement processing on the multi-source credential creation dimensionality reduction data through a feature enhancement operator to obtain feature-enhanced multi-source credential creation data;

[0085] Performing a data normalization operation on the feature-enhanced multi-source trust-innovation data, calculating the variance of each feature data in the feature-enhanced multi-source trust-innovation data, removing feature data with variances lower than a preset threshold, and obtaining a first screening risk feature of the target financial trust-innovation project;

[0086] Obtain historical multi-source original data of the target financial credential creation project, obtain risk labels of each feature data in the target financial credential creation project in different numerical ranges based on the historical multi-source original data, construct target variables with the risk labels of each feature data in different numerical ranges, perform a chi-square test on each data feature in the feature-enhanced multi-source credential creation data based on the target variable, and select features with chi-square statistics ranked before a preset ranking as the second screening risk features;

[0087] Combine the first screening risk characteristics and the second screening risk characteristics, and deduplicate the combined risk characteristics to obtain the multi-dimensional risk characteristics of the target financial information innovation project.

[0088] It should be noted that through time series data smoothing, the noise and fluctuation of the data can be reduced, making the data more stable and predictable. Feature enhancement processing can further extract and enhance the hidden feature information in the data, enhance the expressiveness and discrimination of the data; data normalization operation can eliminate the dimensional influence between different features, so that each feature can be compared and analyzed within the same numerical range. By variance filtering, the feature data with low variance can be removed, which can reduce the complexity and computational cost of model training, while retaining features that contribute more to risk assessment. For example, features with too low variance of transaction amount in financial transaction data, features with too low variance of transaction frequency, and features with low variance of transaction frequency can be removed to obtain risk features after the first feature screening; using the chi-square test to evaluate the features, it can be determined which features have a significant impact on the target variable in different numerical ranges, thereby selecting the features with the most discrimination and risk prediction capabilities, which helps to accurately capture and assess the potential risks in financial information innovation projects; the target variable is the risk label of each feature data converted into a numerical value within a different numerical range. The word represents, 0 means no risk, 1 means risk, for example, if the transaction frequency in the financial transaction data has no risk within a certain range, this range will be marked as 0, and 0 is the target variable of this range; the feature data is each data item of the target financial information innovation project in the multi-source information innovation dimensionality reduction data, such as transaction date, transaction amount, and transaction type data; the feature engineering includes removing feature data with a variance lower than a preset threshold and selecting features with a chi-square statistic before a preset ranking as the second screening risk feature; by removing feature data with a variance lower than a preset threshold and selecting features with a chi-square statistic before a preset ranking as the second screening risk feature, double screening of risk features is performed, thereby improving the accuracy of risk feature determination.

[0089] According to an embodiment of the present invention, the financial information innovation risk warning model is constructed based on big data technology, and the multi-dimensional risk characteristics are introduced into the financial information innovation risk warning model for training, specifically:

[0090] Extracting original data of each risk feature in the multi-dimensional risk features of the target financial credential creation project based on the multi-source credential creation original data, and visualizing the original data of each risk feature through a box plot to obtain a box plot of each risk feature;

[0091] Performing a distribution hypothesis test on the box plot to determine the distribution type of the original data of each risk feature, and estimating the distribution parameters of each risk feature based on the distribution type;

[0092] Calculating a risk threshold based on the distribution parameter, determining a risk range for each risk feature based on the risk threshold, and determining a risk level for each risk range;

[0093] Backtest the risk range based on historical multi-source information innovation data, evaluate the accuracy of the risk interval, and calculate the frequency and severity of events within different risk level intervals;

[0094] A financial information innovation risk warning model is constructed based on a convolutional neural network. The risk range of each risk feature, the risk level of each risk range, the frequency and severity of occurrence within different risk level intervals are imported into the financial information innovation risk warning model for training to obtain a complete financial information innovation risk warning model.

[0095] It should be noted that visualizing the original data distribution of each risk feature through box plots and determining the distribution type and parameters of the data in combination with distribution hypothesis tests can help analysts intuitively understand the data distribution characteristics of risk features; the risk threshold calculated based on the distribution parameters determines the risk range and corresponding risk level of each risk feature, and the risk threshold and risk level provide quantitative indicators for risk assessment; the determined risk range is backtested and evaluated using historical multi-source information technology data to verify the accuracy and actual effect of the risk interval, ensuring the reliability and effectiveness of the risk warning model, and further optimizing the model's parameter settings and risk analysis strategies by statistically analyzing the frequency and severity of events within different risk level intervals; using the relevant information extracted for each risk feature, including the risk range, risk level, and statistical analysis results of historical data, a financial information technology risk warning model is constructed, which not only integrates the information of multi-dimensional risk features, but also considers its performance within different risk ranges, thereby improving the ability to predict and respond to risk events.

[0096] Figure 2 A flow chart of generating risk warning information according to the present invention is shown.

[0097] According to an embodiment of the present invention, the real-time data of the target financial credential creation project is monitored based on the financial credential creation risk warning model, and the credential creation risk of the target financial credential creation project is identified to generate risk warning information, specifically:

[0098] S202, acquiring real-time data of each risk feature of the target financial information innovation project according to the data acquisition system, importing the real-time data of each risk feature into the financial information innovation risk early warning model for real-time risk monitoring, outputting the risk probability value of each risk feature of the target financial information innovation project, and obtaining risk event occurrence probability data;

[0099] S204, calculating a risk score for the target financial information innovation project based on the risk event occurrence probability data, and classifying the risk level based on the risk score;

[0100] S206: Generate risk warning information based on the risk event occurrence probability data, risk score, and risk level.

[0101] It should be noted that the data collection system acquires real-time data on each risk characteristic of the target financial information innovation project and imports this data into the financial information innovation risk warning model for monitoring. Based on the real-time monitoring data, the financial information innovation risk warning model outputs a risk probability value for each risk characteristic. These probability values ​​reflect the likelihood of a risk event occurring at the current moment for each risk characteristic. The overall risk score of the target financial information innovation project is calculated based on the probability data of risk events. The higher the probability, the higher the risk score. Based on the calculated risk score, the risk is divided into different levels, including high risk, medium risk, and low risk.

[0102] Figure 3 A flow chart of forming a risk early warning scheme according to the present invention is shown.

[0103] According to an embodiment of the present invention, the risk warning operation is performed on the target financial information innovation project based on the risk warning information to form a risk warning plan, specifically:

[0104] S302, determining a risk event based on the risk warning information, determining an impact range of the risk event based on expert experience, and determining a warning range of the risk warning information based on the impact range;

[0105] S304, classifying the risk warning information according to risk levels to obtain risk level warning information;

[0106] S306: Based on the risk level warning information, if the risk level is high, the high-risk project is isolated. If the risk level is medium, the project management personnel are arranged to manually review the risk event. The nature and severity of the risk are determined based on the manual review. Emergency response measures are formulated based on the review results. If the risk level is low, the risk event is monitored in real time to monitor the real-time risk changes and risk diffusion changes of the low-risk event. The risk level is adjusted based on the real-time risk changes and risk diffusion changes to obtain a risk warning plan.

[0107] S308: Perform risk warning operations on the risk warning plan according to the warning range.

[0108] It should be noted that specific risk events are identified and determined based on risk warning information, and the impact that each risk event may have on the target financial information innovation project is clarified. Based on expert experience and system analysis, the scope of impact of each risk event is determined, including the business processes, assets or markets it may affect; risks are classified according to the risk scores and levels in the risk warning information, which helps to effectively manage and handle risk events according to their severity and urgency. For projects or areas identified as high-risk, the system can automatically perform isolation operations, which include suspending high-risk event business and restricting user transactions; project managers are arranged to conduct detailed manual reviews of medium-risk events to determine the specific nature and severity of the risks and formulate corresponding emergency response measures, which include suspending transactions, reconfiguring investment portfolios, and issuing notifications to internal teams and key decision makers; for low-risk events, the system conducts real-time monitoring to promptly detect any changes or spread of risks, and adjusts risk levels and response strategies based on real-time monitoring results; through systematic risk warning operations, financial information innovation projects can be helped to promptly identify, assess and respond to potential risks to ensure the stable operation and sustainable development of the projects.

[0109] According to an embodiment of the present invention, the further embodiment includes:

[0110] Acquire historical user behavior data of users operating in the target financial trust innovation project system within a preset time period according to the data collection system of the target financial trust innovation project, based on the historical user behavior data, the historical user behavior data includes user login information, transaction records, page access behavior, and operation timestamps;

[0111] Sorting the historical user behavior data according to the process sequence to obtain historical user behavior time series data, analyzing the historical user behavior time series data based on the sequence pattern mining method, identifying the user's transaction path and operation process in the target financial trust innovation project system, and obtaining the user's operation path data;

[0112] The operation path data is used to collect statistics on the user's transactions and operation behaviors in the target financial trust innovation project system, identify high-frequency trading behaviors and operation behaviors, and obtain the user's operation habit data in the target financial trust innovation project system;

[0113] Based on the K-means clustering algorithm, cluster the user's operating habit data, classify users with similar operating habits into one category, and label each user in one category with an operation preference label to obtain user classification data of the target financial information innovation project system;

[0114] Analyze the acceptance of each category of users for the financial products in the target financial trust innovation project system based on the preference labels of the user classification data;

[0115] Obtain the user's browsing time for each financial product in the target financial trust innovation project system and the completeness of the browsing of the financial product introduction data based on the operation path data, and evaluate the user's understanding of each financial product in the target financial trust innovation project system based on the browsing time of the financial product and the completeness of the browsing of the financial product introduction data;

[0116] Obtaining information on the impact of the acceptance of the financial products and the degree of understanding of each financial product on the transaction risk of the user's financial product transactions;

[0117] Acquire real-time user behavior data, update the user's acceptance and understanding of financial products based on the real-time behavior data, and when the user trades a target financial product, perform a risk score based on the transaction risk impact weight information and the user's acceptance and understanding of the target financial product to obtain a transaction risk score result;

[0118] According to the transaction risk scoring result, when the transaction risk score is greater than a preset value, the user is restricted from trading the target financial product.

[0119] It should be noted that users can trade financial products in the target financial credible innovation project system. In real life, if users do not understand a type of financial product or do not have sufficient understanding of the financial product before trading, it is easy for the users to suffer property losses and create transaction risks. Therefore, by identifying the transaction path and operation process of the user in the target financial credible innovation project system, the operation habit data is determined. The operation habit data includes the operation habit data of what type of financial products the user trades and browses. The user's operation habit data is clustered based on the K-means clustering algorithm, and users with similar operation habits are classified into one category. In this way, preference labels can be assigned to users in the target financial credible innovation project system for batch operations, without the need to repeatedly analyze each user to determine the preference label, thereby improving data analysis efficiency. Preference labels are assigned to each type of user, and the preference labels include financial product type investment preference, investment preference, etc. The user's investment acceptance for each type of financial product can be understood based on the preference labels. The user's understanding of financial products is also a decisive indicator of the user's investment risk. Therefore, the user's understanding of financial products is assessed by the user's browsing time and the completeness of the browsing of the financial product introduction data. Finally, the risk of the user's financial product transaction is scored based on the financial product acceptance and the understanding of each financial product. When the transaction risk score is greater than the preset value, it can be considered that the user's understanding of the target financial product is too low and the investment risk is relatively high. The trading behavior is restricted to prevent the user from blindly investing in a certain financial product and causing financial losses. At the same time, it prevents the user from blindly listening to other people and investing in non-high-quality financial products, ensuring the high-quality investment of the user and improving the quality of investment decisions.

[0120] Figure 4 A block diagram of an information innovation risk warning system based on big data is shown in the present invention.

[0121] The second aspect of the present invention further provides a big data-based credential creation risk warning system 4, which includes: a memory 41 and a processor 42. The memory includes a big data-based credential creation risk warning method program. When the big data-based credential creation risk warning method program is executed by the processor, the following steps are implemented:

[0122] Obtain multi-source original data of a target financial trust innovation project, wherein the multi-source original data includes financial transaction data, market dynamic data, and user behavior data, and perform data dimensionality reduction on the multi-source original data to form multi-source trust innovation reduced dimensionality data;

[0123] Extract risk features of the multi-source credential creation dimensionality reduction data based on feature engineering to obtain multi-dimensional risk features of the target financial credential creation project;

[0124] Constructing a financial information innovation risk warning model based on big data technology, and importing the multi-dimensional risk characteristics into the financial information innovation risk warning model for training;

[0125] Based on the financial credential creation risk warning model, real-time data of the target financial credential creation project is monitored to identify the credential creation risks of the target financial credential creation project and generate risk warning information;

[0126] Based on the risk warning information, risk warning operations are performed on the target financial information innovation project to form a risk warning plan.

[0127] The present invention discloses a risk warning method and system for financial credible innovation based on big data. First, the multi-source credible innovation original data of the target financial credible innovation project is obtained, and the data dimension is reduced to form multi-source credible innovation dimension reduction data; then, the credible innovation risk features are extracted from the dimension reduction data based on feature engineering to obtain the multi-dimensional risk features of the target project; then, a financial credible innovation risk warning model is constructed based on big data technology; finally, the real-time data of the target project is monitored using the model, risk warning information is identified and generated, and risk warning operations are performed according to the warning information to form a risk warning plan. The present invention can effectively monitor and warn risks of financial credible innovation projects in real time, and improve the security and stability of financial credible innovation projects.

[0128] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0129] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed across multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0130] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0131] Those skilled in the art will appreciate that all or part of the steps of the above-mentioned method embodiments may be implemented by hardware associated with program instructions, and the aforementioned program may be stored in a computer-readable storage medium. When the program is executed, the program executes the steps of the above-mentioned method embodiments. The aforementioned storage medium includes various media that can store program codes, such as mobile storage devices, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0132] Alternatively, if the integrated units described above are implemented as software modules and sold or used as standalone products, they can also be stored on a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present invention, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product, stored on a storage medium, includes instructions for enabling a computer device (such as a personal computer, server, or network device) to execute all or part of the methods described in various embodiments of the present invention. The aforementioned storage media include various media capable of storing program code, such as removable storage devices, ROM, RAM, magnetic disks, or optical disks.

[0133] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A risk warning method for information innovation based on big data, characterized in that: The following steps are involved: Obtain multi-source original data of a target financial trust innovation project, wherein the multi-source original data includes financial transaction data, market dynamic data, and user behavior data, and perform data dimensionality reduction on the multi-source original data to form multi-source trust innovation reduced dimensionality data; The financial transaction data includes transaction date, transaction amount, transaction type, transaction counterparty identity information, and transaction subject; the user behavior data includes browsing behavior, search behavior, and transaction behavior on the target financial information innovation project system; Extract risk features of the multi-source credential creation dimensionality reduction data based on feature engineering to obtain multi-dimensional risk features of the target financial credential creation project; Constructing a financial information innovation risk warning model based on big data technology, and importing the multi-dimensional risk characteristics into the financial information innovation risk warning model for training; Based on the financial credential creation risk warning model, real-time data of the target financial credential creation project is monitored to identify the credential creation risks of the target financial credential creation project and generate risk warning information; Conduct risk warning operations on the target financial information innovation project based on the risk warning information to form a risk warning plan; The financial information innovation risk warning model is constructed based on big data technology, and the multi-dimensional risk characteristics are introduced into the financial information innovation risk warning model for training, specifically: Extracting original data of each risk feature in the multi-dimensional risk features of the target financial credential creation project based on the multi-source credential creation original data, and visualizing the original data of each risk feature through a box plot to obtain a box plot of each risk feature; Performing a distribution hypothesis test on the box plot to determine the distribution type of the original data of each risk feature, and estimating the distribution parameters of each risk feature based on the distribution type; Calculating a risk threshold based on the distribution parameter, determining a risk range for each risk feature based on the risk threshold, and determining a risk level for each risk range; Backtest the risk range based on historical multi-source information and innovation data, evaluate the accuracy of the risk interval, and calculate the frequency and severity of events within different risk level intervals; A financial information innovation risk warning model is constructed based on a convolutional neural network. The risk range of each risk feature, the risk level of each risk range, and the frequency and severity of events within different risk level intervals are imported into the financial information innovation risk warning model for training to obtain a complete financial information innovation risk warning model. Among them, also include: Acquire historical user behavior data of users operating in the target financial trust innovation project system within a preset time period according to the data collection system of the target financial trust innovation project, based on the historical user behavior data, the historical user behavior data includes user login information, transaction records, page access behavior, and operation timestamps; Sorting the historical user behavior data according to the process sequence to obtain historical user behavior time series data, analyzing the historical user behavior time series data based on the sequence pattern mining method, identifying the user's transaction path and operation process in the target financial trust innovation project system, and obtaining the user's operation path data; Collect statistics on the user's transactions and operation behaviors in the target financial trust innovation project system based on the operation path data, identify high-frequency trading behaviors and operation behaviors, and obtain the user's operation habit data in the target financial trust innovation project system; Based on the K-means clustering algorithm, cluster the user's operating habit data, classify users with similar operating habits into one category, and label each user in one category with an operation preference label to obtain user classification data of the target financial information innovation project system; Analyze the acceptance of each category of users for the financial products in the target financial trust innovation project system based on the preference labels of the user classification data; Obtain the user's browsing time for each financial product in the target financial trust innovation project system and the completeness of the browsing of the financial product introduction data based on the operation path data, and evaluate the user's understanding of each financial product in the target financial trust innovation project system based on the browsing time of the financial product and the completeness of the browsing of the financial product introduction data; Obtaining information on the impact of the acceptance of the financial products and the degree of understanding of each financial product on the transaction risk of the user's financial product transactions; Acquire real-time user behavior data, update the user's acceptance and understanding of financial products based on the real-time behavior data, and when the user trades a target financial product, perform a risk score based on the transaction risk impact weight information and the user's acceptance and understanding of the target financial product to obtain a transaction risk score result; According to the transaction risk scoring result, when the transaction risk score is greater than a preset value, the user is restricted from trading the target financial product.

2. The method for early warning of information innovation risk based on big data according to claim 1 is characterized in that: The multi-source original data of the target financial trust creation project is obtained, and the multi-source original data includes financial transaction data, market dynamic data, and user behavior data. The multi-source original data is subjected to data dimensionality reduction to form multi-source trust creation dimensionality reduction data, specifically: Building a data acquisition system for the target financial trust innovation project based on distributed data acquisition technology, and obtaining multi-source trust innovation original data of the target financial trust innovation project according to the data acquisition system, wherein the multi-source trust innovation original data includes financial transaction data, market dynamic data, and user behavior data; Classify the multi-source credential creation original data according to data categories, assign a category label to each multi-source credential creation original data, and obtain multi-source credential creation classified data; Introducing the LDA algorithm, calculating the mean vector of each category of data in the multi-source information creation classification data according to the LDA algorithm, and calculating the intra-class scatter matrix of each category of data according to the mean vector; Calculate the overall mean vector of each category of data based on the multi-source information creation classification data, and calculate the inter-class scatter matrix of the multi-source information creation classification data based on the overall mean vector; Calculate the inverse matrix of the intra-class scatter matrix, multiply the inverse matrix by the inter-class scatter matrix to obtain a generalized eigenvalue problem matrix, solve the eigenvalues ​​of the generalized eigenvalue problem matrix to obtain the eigenvalues ​​and corresponding eigenvectors of the generalized eigenvalue problem matrix, sort the eigenvalues ​​of the generalized eigenvalue problem matrix in descending order, and select the eigenvectors corresponding to the first N largest eigenvalues ​​to construct a data dimensionality reduction matrix of the multi-source information creation original data; The multi-source information creation original data is linearly transformed according to the data dimensionality reduction matrix to form multi-source information creation dimensionality reduction data.

3. The method for early warning of information innovation risk based on big data according to claim 1 is characterized in that: The feature engineering-based extraction of risk features of the multi-source credential creation dimensionality reduction data is performed to obtain the multi-dimensional risk features of the target financial credential creation project, specifically: Performing time series data smoothing processing on the multi-source credential creation dimensionality reduction data, and performing feature enhancement processing on the multi-source credential creation dimensionality reduction data through a feature enhancement operator to obtain feature-enhanced multi-source credential creation data; Performing a data normalization operation on the feature-enhanced multi-source trust-innovation data, calculating the variance of each feature data in the feature-enhanced multi-source trust-innovation data, removing feature data with variances lower than a preset threshold, and obtaining a first screening risk feature of the target financial trust-innovation project; Obtain historical multi-source original data of the target financial credential creation project, obtain risk labels of each feature data in the target financial credential creation project in different numerical ranges based on the historical multi-source original data, construct target variables with the risk labels of each feature data in different numerical ranges, perform a chi-square test on each data feature in the feature-enhanced multi-source credential creation data based on the target variable, and select features with chi-square statistics ranked before a preset ranking as the second screening risk features; Combine the first screening risk characteristics and the second screening risk characteristics, and deduplicate the combined risk characteristics to obtain the multi-dimensional risk characteristics of the target financial information innovation project.

4. The method for early warning of information innovation risk based on big data according to claim 1 is characterized in that: The real-time data of the target financial credential creation project is monitored based on the financial credential creation risk warning model, and the credential creation risk of the target financial credential creation project is identified to generate risk warning information, specifically: Acquire real-time data of each risk feature of the target financial information innovation project according to the data acquisition system, import the real-time data of each risk feature into the financial information innovation risk early warning model for real-time risk monitoring, output the risk probability value of each risk feature of the target financial information innovation project, and obtain the probability data of risk event occurrence; Calculate the risk score of the target financial information innovation project based on the probability data of the risk event, and divide the risk level according to the risk score; The risk event occurrence probability data, risk score, and risk level are used to generate risk warning information.

5. The method for early warning of information innovation risk based on big data according to claim 1 is characterized in that: The risk warning operation is performed on the target financial information innovation project according to the risk warning information to form a risk warning plan, specifically: Determine a risk event based on the risk warning information, determine the impact scope of the risk event based on expert experience, and determine the warning scope of the risk warning information based on the impact scope; Classify the risk warning information according to the risk level to obtain risk level warning information; Based on the risk level warning information, when the risk level is high, the high-risk project is isolated; when the risk level is medium, the project management personnel are arranged to manually review the risk event, determine the nature and severity of the risk based on the manual review, and formulate emergency response measures based on the review results; when the risk level is low, the risk event is monitored in real time, and the real-time risk changes and risk diffusion changes of the low-risk event are monitored. The risk level is adjusted according to the real-time risk changes and risk diffusion changes to obtain a risk warning plan; The risk warning plan is subjected to risk warning operations according to the warning scope.

6. A risk warning system for information innovation based on big data, characterized by: The big data-based credible innovation risk warning system includes a storage device and a processor. The storage device includes a big data-based credible innovation risk warning method program. When the big data-based credible innovation risk warning method program is executed by the processor, the following steps are implemented: Obtain multi-source original data of a target financial trust innovation project, wherein the multi-source original data includes financial transaction data, market dynamic data, and user behavior data, and perform data dimensionality reduction on the multi-source original data to form multi-source trust innovation reduced dimensionality data; The financial transaction data includes transaction date, transaction amount, transaction type, transaction counterparty identity information, and transaction subject; the user behavior data includes browsing behavior, search behavior, and transaction behavior on the target financial information innovation project system; Extract risk features of the multi-source credential creation dimensionality reduction data based on feature engineering to obtain multi-dimensional risk features of the target financial credential creation project; Constructing a financial information innovation risk warning model based on big data technology, and importing the multi-dimensional risk characteristics into the financial information innovation risk warning model for training; Based on the financial credential creation risk warning model, real-time data of the target financial credential creation project is monitored to identify the credential creation risks of the target financial credential creation project and generate risk warning information; Conduct risk warning operations on the target financial information innovation project based on the risk warning information to form a risk warning plan; The financial information innovation risk warning model is constructed based on big data technology, and the multi-dimensional risk characteristics are introduced into the financial information innovation risk warning model for training, specifically: Extracting original data of each risk feature in the multi-dimensional risk features of the target financial credential creation project based on the multi-source credential creation original data, and visualizing the original data of each risk feature through a box plot to obtain a box plot of each risk feature; Performing a distribution hypothesis test on the box plot to determine the distribution type of the original data of each risk feature, and estimating the distribution parameters of each risk feature based on the distribution type; Calculating a risk threshold based on the distribution parameter, determining a risk range for each risk feature based on the risk threshold, and determining a risk level for each risk range; Backtest the risk range based on historical multi-source information and innovation data, evaluate the accuracy of the risk interval, and calculate the frequency and severity of events within different risk level intervals; A financial information innovation risk warning model is constructed based on a convolutional neural network. The risk range of each risk feature, the risk level of each risk range, and the frequency and severity of events within different risk level intervals are imported into the financial information innovation risk warning model for training to obtain a complete financial information innovation risk warning model. Among them, also include: Acquire historical user behavior data of users operating in the target financial trust innovation project system within a preset time period according to the data collection system of the target financial trust innovation project, based on the historical user behavior data, the historical user behavior data includes user login information, transaction records, page access behavior, and operation timestamps; Sorting the historical user behavior data according to the process sequence to obtain historical user behavior time series data, analyzing the historical user behavior time series data based on the sequence pattern mining method, identifying the user's transaction path and operation process in the target financial trust innovation project system, and obtaining the user's operation path data; Collect statistics on the user's transactions and operation behaviors in the target financial trust innovation project system based on the operation path data, identify high-frequency trading behaviors and operation behaviors, and obtain the user's operation habit data in the target financial trust innovation project system; Based on the K-means clustering algorithm, cluster the user's operating habit data, classify users with similar operating habits into one category, and label each user in one category with an operation preference label to obtain user classification data of the target financial information innovation project system; Analyze the acceptance of each category of users for the financial products in the target financial trust innovation project system based on the preference labels of the user classification data; Obtain the user's browsing time for each financial product in the target financial trust innovation project system and the completeness of the browsing of the financial product introduction data based on the operation path data, and evaluate the user's understanding of each financial product in the target financial trust innovation project system based on the browsing time of the financial product and the completeness of the browsing of the financial product introduction data; Obtaining information on the impact of the acceptance of the financial products and the degree of understanding of each financial product on the transaction risk of the user's financial product transactions; Acquire real-time user behavior data, update the user's acceptance and understanding of financial products based on the real-time behavior data, and when the user trades a target financial product, perform a risk score based on the transaction risk impact weight information and the user's acceptance and understanding of the target financial product to obtain a transaction risk score result; According to the transaction risk scoring result, when the transaction risk score is greater than a preset value, the user is restricted from trading the target financial product.

7. The big data-based risk early warning system for information innovation according to claim 6 is characterized in that: The real-time data of the target financial credential creation project is monitored based on the financial credential creation risk warning model, and the credential creation risk of the target financial credential creation project is identified to generate risk warning information, specifically: Acquire real-time data of each risk feature of the target financial information innovation project according to the data acquisition system, import the real-time data of each risk feature into the financial information innovation risk early warning model for real-time risk monitoring, output the risk probability value of each risk feature of the target financial information innovation project, and obtain the probability data of risk event occurrence; Calculate the risk score of the target financial information innovation project based on the probability data of the risk event, and divide the risk level according to the risk score; The risk event occurrence probability data, risk score, and risk level are used to generate risk warning information.

8. The big data-based early warning system for information innovation risks according to claim 6 is characterized in that: The risk warning operation is performed on the target financial information innovation project according to the risk warning information to form a risk warning plan, specifically: Determine a risk event based on the risk warning information, determine the impact scope of the risk event based on expert experience, and determine the warning scope of the risk warning information based on the impact scope; Classify the risk warning information according to the risk level to obtain risk level warning information; Based on the risk level warning information, when the risk level is high, the high-risk project is isolated; when the risk level is medium, the project management personnel are arranged to manually review the risk event, determine the nature and severity of the risk based on the manual review, and formulate emergency response measures based on the review results; when the risk level is low, the risk event is monitored in real time, and the real-time risk changes and risk diffusion changes of the low-risk event are monitored. The risk level is adjusted according to the real-time risk changes and risk diffusion changes to obtain a risk warning plan; The risk warning plan is subjected to risk warning operations according to the warning scope.

Images