Enterprise financial management risk assessment method and system based on data analysis

Abstract

The invention discloses an enterprise financial management risk assessment method and system based on data analysis, and relates to the technical field of financial management.The method comprises the steps that a fluctuation coefficient of a financial index is constructed, a target index is screened out through the fluctuation coefficient, trend analysis is conducted on the target index, a corresponding trend degree is obtained, and if the obtained trend degree exceeds a trend degree threshold value, a risk assessment result is obtained; sending a prediction instruction to the outside; the method comprises the following steps: acquiring prediction data and constructing risk degrees, after the risk degrees of all financial indexes are acquired, constructing financial risk values according to the risk degrees, and if the financial risk values exceed a risk threshold value, sending out an early warning instruction to the outside; and carrying out feature identification on the prediction data to obtain a plurality of risk features, and giving a corresponding financial risk optimization scheme through the financial risk optimization knowledge graph according to the correspondence between the risk features and the financial risk optimization scheme. The current management and operation of the enterprise are optimized, and the subsequent financial risk is reduced on the basis of evaluating the financial risk.

Description

Technical Field

[0001] The present invention relates to the technical field of financial management, and specifically to a method and system for evaluating financial management risks of enterprises based on data analysis. Background Art

[0002] Financial risk refers to the opportunity and possibility of economic losses formed due to the deviation between the actual income and the expected goal of an enterprise in various financial activities under the action of internal and external environments and various unpredictable and uncontrollable factors. Financial risk is one of the inevitable risks in the operation process of an enterprise. Therefore, an enterprise needs to carry out effective risk management to ensure financial security and stability. There are various types of financial risks, mainly including the following: Financing risk: It refers to the uncertainty brought about by changes in the capital supply and demand market and the macroeconomic environment when an enterprise raises funds, which may affect the financial results of the enterprise. Investment risk: It refers to the risk that the final income deviates from the expected income after the enterprise invests funds due to changes in market demand or other factors. Operating risk: It refers to the stagnation of the enterprise's capital movement and the change in value caused by the influence of uncertain factors in various links such as supply, production, and sales during the production and operation process of the enterprise.

[0003] In the Chinese invention patent with the application publication number CN117593142A, a financial risk assessment and management method and system include the following steps: Based on a financial data set, using a decision tree classification algorithm, conduct subdivision and identification of historical risk events, analyze the mutual correlation between risk events through association rule learning, reveal potential risk patterns, and generate a risk pattern analysis result. By using the decision tree classification algorithm and association rule learning, it is possible to process a large financial data set more quickly and effectively, effectively handle the randomness and uncertainty of financial data, and improve the flexibility and adaptability of short-term and long-term financial behavior prediction.

[0004] Combined with the above application and the content in the prior art; when an enterprise is in an operating state, due to the changeable market economic state, economic indicators often fluctuate to a large extent. If the enterprise's fund management or operation is poor, the enterprise will have a relatively large operating risk, and financial risk will also follow. At this time, if the financial risk is not handled in time, it may lead to the difficulty in the continuous operation of the enterprise; in the existing financial risk management, by monitoring some financial parameters and sending out early warning information to the outside when they are abnormal, and then financial management personnel give treatment methods based on the early warning information. However, these financial indicators that play an early warning role are usually randomly selected and have relatively poor representativeness. Further, if the management personnel give treatment methods after receiving the early warning instruction, it may not necessarily achieve the expected effect. For example, the management personnel may not notice the early warning instruction, or it has been a long time after noticing the early warning instruction, and it is difficult to make effective treatment.

[0005] To this end, the present invention provides a method and system for evaluating the financial management risk of an enterprise based on data analysis. Summary of the Invention

[0006] (I) Technical problems to be solved In view of the deficiencies of the prior art, the present invention provides a method and system for evaluating the financial management risk of an enterprise based on data analysis. By performing trend analysis on target indicators and obtaining corresponding trend degrees, if the obtained trend degree exceeds the trend degree threshold, a prediction instruction is sent to the outside; by obtaining prediction data and constructing a risk degree, after obtaining the risk degrees of each financial indicator, a financial risk value is constructed from the risk degrees. If the financial risk value exceeds the risk threshold, a warning instruction is sent to the outside; the prediction data is subjected to feature recognition to obtain a number of risk features, and according to the correspondence between the risk features and the financial risk optimization scheme, the financial risk optimization knowledge graph gives the corresponding financial risk optimization scheme. Optimize the current management and operation of the enterprise, reduce the subsequent financial risks, and thus solve the technical problems raised in the background art.

[0007] (II) Technical solutions To achieve the above objectives, the present invention is implemented through the following technical solutions: A method for evaluating the financial management risk of an enterprise based on data analysis, including: Collecting enterprise financial data and constructing a number of data categories, performing quality screening on the data categories, optimizing the screened low-quality data categories, and replacing outliers; Performing quality analysis on the financial indicator columns and constructing a fluctuation coefficient of the financial indicators From the fluctuation coefficient screening out target indicators, performing trend analysis on the target indicators and obtaining corresponding trend degrees If the obtained trend degree exceeds the trend degree threshold, sending a prediction instruction to the outside; wherein, the construction method of the fluctuation coefficient is to perform linear normalization on the data values of the financial indicators according to the following method:

[0008] wherein, n, n is the number of time nodes, is the financial indicator at the i-th time node, is the corresponding mean value, and the weight coefficient: and ; Obtaining prediction data and constructing a risk degree After obtaining the risk degrees of each financial indicator, from the risk degree Construct a financial risk value , if the financial risk value exceeds the risk threshold, send a warning instruction to the outside; Perform feature recognition on the predicted data to obtain a number of risk features, and according to the correspondence between the risk features and the financial risk optimization plan, the financial risk optimization knowledge graph gives the corresponding financial risk optimization plan.

[0009] Furthermore, within the management cycle, collect the financial data within the enterprise, summarize the collected financial data, and after preprocessing the data, construct an enterprise financial data set; classify the data within the enterprise financial data set to obtain a number of data classes, and perform quality analysis on the data within the data class to obtain the relative range and the quality difference index .

[0010] Furthermore, construct the quality degree of each data class , if the quality degree is lower than the quality threshold, determine the corresponding data class as a low-quality data class, screen out the outliers from the low-quality data class, and replace the outliers by interpolation method to obtain the optimized data class; the acquisition method of the quality degree is as follows: after normalizing the relative range and the quality difference index , where:

[0011] Weight coefficient: , .

[0012] Furthermore, construct the financial indicators required to evaluate the financial risk from the financial data in the enterprise financial data set, summarize the obtained financial indicators to construct a financial indicator set, sort the financial indicators according to the generation time to obtain the corresponding financial indicator column; perform quality analysis on the financial indicator column and construct the fluctuation coefficient of the financial indicator, and select the financial indicator with the highest fluctuation coefficient as the target indicator.

[0013] Furthermore, perform function fitting on the financial indicator column corresponding to the target indicator to construct the trend degree of obtaining the target indicator, and obtain the slope value of the target indicator between two time nodes, according to the following method:

[0014] Among them, m, m is the number of slope values, is the mean value of the slope values, is the i-th slope value; if the obtained trend degree When exceeding the trend degree threshold, send a prediction instruction to the outside.

[0015] Furthermore, after receiving the prediction instruction, obtain a financial indicator prediction model through training with sample data. After setting several prediction nodes, the financial indicators at each prediction node are output by the financial indicator prediction model; sort the nodes before and after the optimization of the target enterprise according to the time axis and correspond them one by one to construct the risk degree F .

[0016] Furthermore, construct the risk degree F in the following way:

[0017] Among them, is the intermediate value of the risk degree at the i-th node, is its mean value, i is the node serial number, , n is the number of nodes, and are the trend degrees at the i-th node before and after optimization respectively, and are the corresponding mean values.

[0018] Furthermore, after obtaining the risk degrees of each financial indicator in the same way , construct a financial risk value from the risk degrees. If the obtained financial risk value exceeds the risk threshold, send a warning instruction to the outside. If it does not exceed, no processing is done; among them, the construction method of the financial risk value is as follows:

[0019] Among them, p, p is the number of financial indicators, is the risk degree of the -th financial indicator, is the corresponding mean value, weight coefficient: , , and .

[0020] Furthermore, after receiving the warning instruction, obtain the financial indicators and corresponding prediction data at each prediction node. According to the management expectation of the enterprise's financial data, pre-set the corresponding risk criteria, identify the features of the prediction data, and obtain several risk features; use the enterprise financial risk control as the target word, after deep retrieval and constructing entity relationships, construct a financial risk optimization knowledge graph; according to the correspondence between the risk features and the financial risk optimization plan, the financial risk optimization knowledge graph gives the corresponding financial risk optimization plan.

[0021] An enterprise financial management risk assessment system based on data analysis, comprising: A data optimization unit, which collects enterprise financial data and constructs several data classes, performs quality screening on the data classes, optimizes the screened low-quality data classes, and replaces outliers; A trend analysis unit, which performs quality analysis on financial indicator columns and constructs a fluctuation coefficient of financial indicators , and filters out target indicators based on the fluctuation coefficient , performs trend analysis on the target indicators and obtains corresponding trend degrees , if the obtained trend degree exceeds the trend degree threshold, sends a prediction instruction to the outside; A risk prediction unit, which obtains prediction data and constructs a risk degree , after obtaining the risk degrees of each financial indicator, constructs a financial risk value based on the risk degree ; if the financial risk value exceeds the risk threshold, sends a warning instruction to the outside; A risk processing unit, which performs feature recognition on the prediction data, obtains several risk features, and based on the correspondence between the risk features and the financial risk optimization plan, gives a corresponding financial risk optimization plan by the financial risk optimization knowledge graph.

[0022] (III) Beneficial effects The present invention provides an enterprise financial management risk assessment method and system based on data analysis, having the following beneficial effects: 1. Evaluate and judge the data quality of each data class according to the quality degree. If the data quality is low, optimize it after preprocessing, optimize and process the outliers in it, so as to ensure the data quality. By generating a fluctuation coefficient filter out the indicators with large fluctuations from several different target indicators, and when using it as the target indicator, by using it as a representative parameter, the amount of data analysis can be reduced and the efficiency of data analysis can be improved; 2. By constructing a financial indicator prediction model, predict the financial indicators at each node, and construct a risk degree from the prediction data , and based on the risk degree can judge whether there are abnormalities in the corresponding financial indicators. If there are abnormalities, targeted processing can be carried out to reduce the overall financial risk. 3. By constructing a financial risk value from the risk degrees of each financial indicator to make an overall judgment on the current financial risk of the enterprise. If the current financial risk of the enterprise is low, the current business status can be maintained. On the contrary, if the risk is high, the current business management strategy of the enterprise needs to be adjusted.​​

[0023] 4. By obtaining the corresponding risk characteristics after feature identification, the corresponding financial risk optimization plan is given based on the risk characteristics by the financial risk optimization knowledge map. After evaluating and judging the current financial risks, the optimization plan can be given quickly. From the perspective of financial data, the current management and operation of the enterprise can be optimized, and the subsequent financial risks can be reduced based on the assessment of financial risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a flow chart of the enterprise financial management risk assessment method of the present invention; Figure 2 This is a schematic diagram of the structure of the enterprise financial management risk assessment system of the present invention. The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] See also Figure 1 The present invention provides a method for assessing enterprise financial management risks based on data analysis, comprising: Step 1: Collect enterprise financial data and construct several data classes. After quality screening of the data classes, optimize the low-quality data classes and replace outliers. The step 1 includes the following contents: Step 101: Set up an enterprise financial management cycle, collect the enterprise's internal financial data, such as balance sheet, income statement, cash flow statement, etc., and external data, such as market data, industry data, macroeconomic data, etc., during the management cycle; summarize the collected financial data, pre-process the data, and construct an enterprise financial data set; Step 102: Classify the data in the enterprise financial data set according to the data type, obtain several data classes, and perform quality analysis on the data in the data class to obtain the relative range of each data class. Heterogeneity Index , and after normalization, construct the quality of each data class , to judge the data quality in the following way:

[0026] Weight coefficient: , ; Step 103: Preset a quality threshold based on historical data and the management expectations for the quality of each data category; if the quality degree is lower than the quality threshold, determine the corresponding data category as a low-quality data category, and screen out outliers from the low-quality data category. Among them, a conventional threshold is pre-constructed, and the specific method is as follows:

[0027] Among them, is the first quartile, is the third quartile, is the interquartile range, is the data mean, is the data maximum value, is the data minimum value; if the data of the data category is not within the conventional threshold , then determine it as an outlier, and replace the outlier by interpolation method to obtain an optimized data category; when in use, combine the content in Steps 101 to 103: When the enterprise is in a continuous operation state, collect and classify the enterprise's business data and financial data, conduct quality analysis on each data category, and on this basis, construct the quality degree of each data category. Evaluate and judge the data quality of each data category according to the quality degree. If the data quality is low, optimize it after preprocessing, and optimize and process the outliers among them, so as to ensure the data quality.

[0028] Combine the above application and the content in the prior art; When the enterprise is in an operating state, due to the changeable market economic state, economic indicators often fluctuate to a large extent. If the enterprise's capital management or operation is poor, the enterprise will have a large degree of business risks, and financial risks will also follow. At this time, if the financial risks are not processed in time, it may lead to the difficulty of the enterprise's continuous operation; in the existing financial risk management, it is usually to monitor some financial parameters, and when they are abnormal, send out early warning information to the outside, and then according to the early warning information, the financial management personnel give treatment methods. However, the financial indicators used for early warning are usually randomly selected and have relatively poor representativeness. Further, if the treatment method is given by the management personnel after receiving the early warning instruction, it may not be able to achieve the expected effect. For example, the management personnel may not notice the early warning instruction, or it has been a long time after noticing the early warning instruction, and it is difficult to make an effective treatment.

[0029] Step Two: Conduct quality analysis on the financial indicator columns and construct the fluctuation coefficient of the financial indicators , from the fluctuation coefficient Screen out the target indicators, conduct trend analysis on the target indicators and obtain the corresponding trend degree , if the obtained trend degree exceeds the trend degree threshold, send a prediction instruction to the outside; The second step includes the following contents: Step 201, construct financial indicators required for evaluating financial risks from the financial data in the enterprise financial data set, including: financial leverage ratio, liquidity ratio, profitability indicators, cash flow indicators, etc., summarize the obtained financial indicators to construct a financial indicator set, sort the financial indicators according to the generation time, and obtain the corresponding financial indicator column; Step 202, conduct quality analysis on the financial indicator column and construct the fluctuation coefficient of the financial indicator , among which, perform linear normalization processing on the data values of the financial indicators according to the following method:

[0030] Among them, n, n is the number of time nodes, refers to the financial indicator at the i-th time node, is the corresponding mean value, weight coefficient: , , and ; The weight coefficient is obtained by referring to the analytic hierarchy process; Select the financial indicator with the largest degree of fluctuation, that is, the financial indicator with the highest fluctuation coefficient, as the target indicator; when used, after constructing the financial indicator column, screen out the indicators with larger fluctuations from several different target indicators by generating the fluctuation coefficient When using it as the target indicator, by using it as a representative parameter, the amount of data analysis can be reduced and the efficiency of data analysis can be improved; Step 203, perform function fitting on the financial indicator column corresponding to the target indicator, and after passing the K-S verification, judge the change trend of the target indicator by the fitting function and construct and obtain the trend degree of the target indicator. Among them, obtain the slope value of the target indicator between two time nodes , according to the following method:

[0031] Among them, m, m is the number of slope values, is the mean value of the slope values, is the i-th slope value; According to historical data and management expectations for financial indicators, preset the trend degree threshold; if the obtained trend degree exceeds the trend degree threshold, that is, the trend degree exceeds the expectation, indicating that there will be a large change in the corresponding target indicator. At this time, send a prediction instruction to the outside; When in use, combine the contents of steps 201 to 203: after selecting the target indicator, after fitting the function to the target indicator, the changes in the target indicator can be described and visualized through the fitting function, which can be more intuitive when managing the enterprise. At the same time, a trend degree is constructed according to the degree of change of the fitting function, which can describe the changing trend of the target indicator. Therefore, when an abnormality occurs in the target indicator, it can be quickly discovered and processed.

[0032] Step 3: Obtain forecast data and construct risk After obtaining the risk level of each financial indicator, the risk level Constructing Financial Value at Risk , if the financial risk value If the risk threshold is exceeded, an early warning instruction will be issued to the outside world; The step three includes the following contents: Step 301: after receiving the prediction instruction, economic indicator data, enterprise financial data, business status data and other related data are aggregated to construct an enterprise status data set; the data in the enterprise status set is used as sample data; an initial model is constructed by a neural convolutional network, and a financial indicator prediction model is obtained by training the sample data; after setting a number of prediction nodes, the financial indicator prediction model outputs the financial indicators on each prediction node; Step 302: Construct risk level from prediction data , sort the nodes before and after the optimization of the target enterprise according to the time axis, and correspond one by one, and construct the risk degree F according to the following method :

[0033] in, is the median risk value on the i-th node, is its mean, i is the node number, , n is the number of nodes, and are the trend degrees on the i-th node before and after optimization, and is the corresponding mean; when used, the financial indicator prediction model is constructed to predict the financial indicators at each node, and the risk degree is constructed from the predicted data. , according to the risk It can determine whether there are any abnormalities in the corresponding financial indicators. If there are any abnormalities, targeted processing can be carried out to reduce the overall financial risk.

[0034] Step 303: Obtain the risk level of each financial indicator in the same way Then, the financial risk value is constructed by the risk degree. , to evaluate its financial risks in the following ways:

[0035] Among them, p, where p is the number of financial indicators, is the risk degree of the th financial indicator, is the corresponding average value, and the weight coefficient: , , and ; the value of the weight coefficient is the same as the previous value; Preset a risk threshold. If the obtained financial risk value exceeds the risk threshold, it indicates that there may be financial risks in the subsequent operations, and it is necessary to correct the current enterprise management and send out a warning instruction to the outside. If it does not exceed, no processing is required; When in use, combine the content in steps 301 and 302: Construct a financial risk value through the risk degrees of each financial indicator to evaluate the financial risks and make an overall judgment on the current financial risks of the enterprise. If the current financial risks of the enterprise are relatively low, the current business status can be maintained. On the contrary, if the risks are relatively high, it is necessary to adjust the current business management strategy.

[0036] Step Four: Identify the characteristics of the predicted data to obtain several risk characteristics, and based on the correspondence between the risk characteristics and the financial risk optimization plan, the financial risk optimization knowledge graph gives the corresponding financial risk optimization plan; The said Step Four includes the following content: Step 401: After receiving the warning instruction, obtain the financial indicators and the corresponding predicted data on each prediction node, preset the corresponding risk standards according to the management expectations of the enterprise's financial data, identify the characteristics of the predicted data, and obtain several risk characteristics; Step 402: Using the enterprise financial risk control as the target word, after deep retrieval and constructing entity relationships, construct a financial risk optimization knowledge graph; use the trained matching model, and based on the correspondence between the risk characteristics and the financial risk optimization plan, the financial risk optimization knowledge graph gives the corresponding financial risk optimization plan; for example, carry out debt restructuring, optimize asset allocation, strengthen internal control, etc.; regularly monitor the changes in risk indicators and the market environment to adjust risk management measures in a timely manner.

[0037] When in use, combine the content in steps 401 and 402: When the current financial risk of an enterprise is relatively high, by identifying the characteristics of the current financial data, including analyzing the change range, trend, and predicted value of the financial data, after obtaining the corresponding risk characteristics through feature recognition, according to the risk characteristics, the financial risk optimization knowledge graph gives the corresponding financial risk optimization plan. After evaluating and judging that there is a financial risk currently, it can quickly give an optimization plan, and optimize the current management and operation of the enterprise from the perspective of financial data. On the basis of evaluating the financial risk, the subsequent financial risk can be reduced. Please refer to Figure 2 , the present invention provides an enterprise financial management risk assessment system based on data analysis, including: A data optimization unit, which collects enterprise financial data and constructs several data classes, performs quality screening on the data classes, optimizes the screened low-quality data classes, and replaces outliers; A trend analysis unit, which performs quality analysis on the financial indicator columns and constructs the fluctuation coefficient of the financial indicators , and screens out target indicators from the fluctuation coefficient , performs trend analysis on the target indicators and obtains the corresponding trend degree , if the obtained trend degree exceeds the trend degree threshold, issues a prediction instruction to the outside; A risk prediction unit, which obtains prediction data and constructs a risk degree , after obtaining the risk degrees of each financial indicator, constructs a financial risk value from the risk degree ; if the financial risk value exceeds the risk threshold, issues a warning instruction to the outside; A risk processing unit, which performs feature recognition on the prediction data, obtains several risk characteristics, and according to the correspondence between the risk characteristics and the financial risk optimization plan, the financial risk optimization knowledge graph gives the corresponding financial risk optimization plan. It should be noted that:

[0038] The construction method of the financial risk optimization knowledge graph can be mainly divided into the following steps: Define financial risks and optimization goals: First, it is necessary to clarify the definition and scope of financial risks, as well as the specific optimization goals. This includes identifying and understanding various financial risks, such as market risks, credit risks, liquidity risks, etc., and determining the specific indicators and expected results of risk optimization. Data collection and collation: This is the basic step in constructing the knowledge graph. It is necessary to collect various data related to financial risks, including the financial statements of the enterprise, market environment data, policies and regulations, etc. These data should be as comprehensive, accurate, and timely as possible to ensure the reliability and effectiveness of the knowledge graph.

[0039]

[0040] Knowledge Representation and Modeling: Operations such as cleaning, classifying, and correlation analysis are performed on the collected data to convert unstructured data into structured knowledge representation. This includes entity recognition, attribute extraction, relationship definition, etc. for financial risks to form the ontology or schema layer in the field of financial risks.

[0041] Constructing a Knowledge Graph: Based on the results of knowledge representation and modeling, a knowledge graph for optimizing financial risks is constructed in a graph structure. In the graph, nodes can represent financial risks, optimization strategies, related entities, etc., and edges represent the relationships between these elements.

[0042] Improvement and Optimization of the Knowledge Graph: As new data and knowledge are continuously acquired, the knowledge graph needs to be updated and optimized regularly to maintain its timeliness and accuracy. At the same time, the knowledge graph can also be used for predicting and warning of financial risks, providing decision-making support for the enterprise's risk management and optimization.

[0043] In the specific implementation process, technical means such as natural language processing and machine learning can be used to improve the efficiency and accuracy of data processing. At the same time, attention also needs to be paid to the visual display and interaction methods of the knowledge graph to better understand and utilize the information in the knowledge graph.

[0044] Generally speaking, the construction of a knowledge graph for optimizing financial risks is a complex but valuable process. It can help enterprises better understand and cope with financial risks, improve the level of risk management, and optimize financial decisions.

[0045] The Analytic Hierarchy Process (AHP) is a decision-making analysis method that decomposes decision-related elements into levels such as goals, criteria, and solutions, and then conducts qualitative and quantitative analyses on this basis. The core of this method lies in decomposing complex decision-making problems into a series of hierarchical factors, establishing a hierarchical structure, and transforming human judgment into the comparison of the importance between pairs of several factors, thus transforming qualitative judgments that are difficult to quantify into comparable importance that can be operated.

[0046] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated here.

[0047] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only for some logical function divisions. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of the devices or units can be in electrical, mechanical, or other forms. As described above, this is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0048] As described above, this is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application.

Claims

1. An enterprise financial management risk assessment method based on data analysis, characterized in that: including, collecting enterprise financial data and constructing several data classes, performing quality screening on the data classes, optimizing the screened low-quality data classes, and replacing outliers; Conduct a quality analysis on the financial indicators column and construct the fluctuation coefficient of the financial indicators , and screen out the target indicators from the fluctuation coefficient , conduct a trend analysis on the target indicators and obtain the corresponding trend degree . If the obtained trend degree exceeds the trend degree threshold, send a prediction instruction to the outside; among them, the construction method of the fluctuation coefficient is to perform linear normalization on the data values of the financial indicators in the following manner: Among them, n , n is the number of time nodes, represents the financial index at the i-th time node, is the corresponding mean value, weight coefficient: , and ; By obtaining prediction data and constructing a risk level , after obtaining the risk levels of various financial indicators, construct a financial risk value from the risk levels . If the financial risk value exceeds the risk threshold, send a warning instruction to the outside . performing feature recognition on the prediction data to obtain several risk features, and based on the correspondence between the risk features and the financial risk optimization solutions, the financial risk optimization knowledge graph gives the corresponding financial risk optimization solutions.

2. The method for evaluating enterprise financial management risks based on data analysis according to claim 1, wherein: collecting the financial data within the enterprise during the management cycle, summarizing the collected financial data, and constructing an enterprise financial data set after preprocessing the data; Classify the data in the enterprise financial data set to obtain several data classes, and perform quality analysis on the data within the data classes to obtain the relative range of each data class and the quality difference index .

3. The method for evaluating enterprise financial management risks based on data analysis according to claim 2, wherein: Construct the quality degree of each data class , if the quality degree is lower than the quality threshold, determine the corresponding data class as a low-quality data class, screen outliers from the low-quality data class, replace the outliers by interpolation method, and obtain the optimized data class; the method for obtaining the quality degree is as follows. After normalizing the relative range and the quality difference index , where: Weight coefficient: , .

4. The method for evaluating enterprise financial management risks based on data analysis according to claim 1, wherein: Construct financial indicators required for evaluating financial risks from the financial data within the enterprise financial data set, summarize the obtained financial indicators to construct a financial indicator set, sort the financial indicators according to the generation time, and obtain the corresponding financial indicator column; conduct a quality analysis on the financial indicator column and construct the fluctuation coefficient of the financial indicators , and select the financial indicator with the highest fluctuation coefficient as the target indicator.

5. The method for evaluating enterprise financial management risks based on data analysis according to claim 1, wherein: Perform a functional fit on the financial indicator column corresponding to the target indicator, construct the trend degree of the target indicator, and obtain the slope value of the target indicator between two time nodes , according to the following method: Among them, m , m is the number of slope values, is the average value of the output slope values, is the i th slope value; if the obtained trend degree exceeds the trend degree threshold, a prediction instruction is sent to the outside.

6. The method for evaluating enterprise financial management risks based on data analysis according to claim 5, wherein: After receiving the prediction instruction, a financial index prediction model is obtained through training with sample data. After setting several prediction nodes, the financial index prediction model outputs the financial indexes at each prediction node. The nodes before and after the optimization of the target enterprise are sorted according to the time axis and corresponding one by one to construct the risk degree F 。 7. The method for evaluating enterprise financial management risks based on data analysis according to claim 6, wherein: Build the risk level F The method is as follows: Among them, is the median risk degree of the i th node, is its mean value, i is the node number, , n is the number of nodes, and are the trend degrees of the i th node before and after optimization respectively, and are the corresponding mean values.

8. The method for evaluating enterprise financial management risks based on data analysis according to claim 7, wherein: Obtain the risk degrees of each financial indicator in the same way After that, construct a financial risk value from the risk degrees , if the obtained financial risk value exceeds the risk threshold, issue a warning instruction to the outside; if it does not exceed, no processing is performed; Financial risk value is constructed as follows: Among them, p , p is the number of financial indicators, is the risk degree of the th financial indicator, is the corresponding mean value, weight coefficient: , and .

9. The method for evaluating enterprise financial management risks based on data analysis according to claim 8, wherein: after receiving the early warning instruction, obtaining the financial indicators and corresponding prediction data on each prediction node, presetting the corresponding risk standards according to the management expectations of the enterprise financial data, performing feature recognition on the prediction data, and obtaining several risk features; using enterprise financial risk control as the target word, constructing a financial risk optimization knowledge graph after in-depth retrieval and constructing entity relationships; based on the correspondence between the risk features and the financial risk optimization solutions, the financial risk optimization knowledge graph gives the corresponding financial risk optimization solutions.

10. Enterprise financial management risk assessment system based on data analysis, characterized in that: including: a data optimization unit that collects enterprise financial data and constructs several data classes, performs quality screening on the data classes, optimizes the screened low-quality data classes, and replaces outliers; Trend analysis unit, which conducts quality analysis on a column of financial indicators and constructs a volatility coefficient of the financial indicators , and filters out target indicators based on the volatility coefficient , conducts trend analysis on the target indicators and obtains corresponding trend degrees . If the obtained trend degree exceeds the trend degree threshold, a prediction instruction is sent externally; Risk prediction unit, which obtains prediction data and constructs a risk level , after obtaining the risk levels of each financial indicator, constructs a financial risk value from the risk levels , if the financial risk value exceeds the risk threshold, issues a warning instruction to the outside; a risk processing unit that performs feature recognition on the prediction data to obtain several risk features, and based on the correspondence between the risk features and the financial risk optimization solutions, the financial risk optimization knowledge graph gives the corresponding financial risk optimization solutions.

Images