Enterprise financial data intelligent processing method and system

Abstract

The invention provides an enterprise financial data intelligent processing method and system, and relates to the technical field of enterprise financial management. The method comprises the steps that to-be-analyzed financial data and historical financial data of an enterprise are acquired, and target financial feature vectors of multiple departments are constructed based on the to-be-analyzed financial data; performing clustering analysis on the plurality of departments based on the historical financial data to determine a plurality of department sets; extracting a deviation value of each department in the department set about each financial index parameter; constructing a global financial deviation vector of each department set, and calculating a global financial deviation value of each department set; and collecting financial planning data of each department, extracting a financial planning feature vector and a financial planning coefficient of each department, generating a first financial planning deviation vector and a second financial planning deviation vector of each department, and obtaining a financial planning deviation analysis result of the enterprise about the plurality of departments. According to the invention, the precision of enterprise financial data analysis is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise financial management, and in particular to a method and system for intelligently processing enterprise financial data. Background Art

[0002] As the scale of enterprises continues to expand, the diversity and complexity of corporate financial data are increasing. Traditional financial analysis methods are more based on the separate analysis of financial data of each department, lacking a global and systematic analysis of the overall financial model. Specifically, some methods use static financial ratios or indicators to analyze the financial health of a single department. This approach ignores the temporal differences and pattern similarities in financial performance between departments, fails to effectively capture and utilize the commonalities and regularities in financial data performance between departments, and fails to effectively consider the long-term trend of historical data, resulting in limited accuracy of analysis results and reliability of predictions. How to conduct multi-dimensional and dynamic analysis of financial data based on historical financial data and departmental characteristics is an important direction to improve the accuracy of financial data analysis and decision-making quality. Summary of the invention

[0003] In order to solve the above technical problems, the present invention proposes a method and system for intelligent processing of enterprise financial data. On the basis of independent analysis of the financial data of the enterprise, it combines the changing trends of historical financial data and the financial behavior patterns among departments, and conducts multi-dimensional intelligent analysis of the financial data. It can provide more accurate financial analysis results and improve the financial management efficiency of the enterprise.

[0004] To achieve the above-mentioned purpose, the present invention provides a first aspect of a method for intelligently processing enterprise financial data, comprising: Obtain the company's financial data to be analyzed and historical financial data, extract multiple financial indicator parameters of each department of the company from the financial data to be analyzed, and construct target financial feature vectors of multiple departments; Based on the historical financial data, the historical financial feature vector of each department and the parameter time series feature vector of each financial indicator parameter of each department are extracted, and a cluster analysis is performed on multiple departments according to the historical financial feature vectors and the parameter time series feature vectors to determine multiple department sets; Determine a reference financial feature vector for each department set, analyze multiple target financial feature vectors in each department set based on the reference financial feature vector, and extract the deviation value of each financial indicator parameter of each department in the department set; Calculate the deviation mean of each financial indicator parameter and construct a global financial deviation vector for each department set, and calculate the global financial deviation value of each department set based on the global financial deviation vector; For a set of departments whose global financial deviation value is less than a preset global deviation threshold, the financial planning data of each department is collected, the financial planning feature vector and the financial planning coefficient of each department are extracted from the financial planning data, and the first financial planning deviation vector of each department is generated according to the financial planning feature vector and the target financial feature vector; Feature optimization is performed on multiple target financial feature vectors in the department set to generate a second financial planning deviation vector for each department, and financial planning deviation analysis results for multiple departments of the enterprise are generated based on the first financial planning deviation vector, the second financial planning deviation vector and the financial planning coefficient.

[0005] Preferably, cluster analysis is performed on multiple departments according to the historical financial feature vectors and the parameter time series feature vectors to determine multiple department sets, including: Calculate the time series similarity of each financial indicator parameter between any two departments according to the parameter time series feature vector, and determine the time series distance parameter between any two departments according to the parameter time series feature vector and the time series similarity; The method includes processing multiple pairs of parameter time series feature vectors of any two departments based on the DTW algorithm, calculating the time series similarity between any two departments with respect to each financial indicator parameter, determining the weight of each financial indicator parameter between any two departments based on multiple time series similarities, calculating the weighted Euclidean distance between any two departments based on the parameter time series feature vector according to the weights of multiple financial indicator parameters, and obtaining the time series distance parameter between any two departments; The temporal distance matrix corresponding to multiple departments is constructed according to multiple temporal distance parameters, and the DBSCAN algorithm is used to cluster multiple departments based on the temporal distance matrix to generate multiple department sets.

[0006] Preferably, calculating the global financial deviation value of each department set according to the global financial deviation vector includes: Calculate the Euclidean norm of the global financial deviation vector of each department set to obtain the global financial deviation value of each department set; For the global financial deviation vector, after determining the reference financial feature vector of each department set, determine the first element value of each financial indicator parameter in the department set according to the reference financial feature vector, and determine multiple second element values ​​of each financial indicator parameter in the department set according to multiple target financial feature vectors; Calculate the difference between the second element values ​​and the first element value under each financial indicator parameter, and obtain the deviation value of each financial indicator parameter for each department in the department set; The average of multiple deviation values ​​under each financial indicator parameter is calculated to obtain the deviation mean of each financial indicator parameter, and the global financial deviation vector of each department set is constructed based on the deviation mean of multiple financial indicator parameters.

[0007] Preferably, feature optimization is performed on multiple target financial feature vectors in the department set to generate a second financial planning deviation vector for each department, including: Generate a target financial mean vector for each department set based on multiple target financial feature vectors of the department set, and extract a mean deviation vector for each department in the department set based on the target financial mean vector, including calculating the difference between the target financial feature vector of the department and the target financial mean vector of the department set with respect to the element value corresponding to each financial indicator parameter, and construct a mean deviation vector for each department; The mean deviation vector and the target financial mean vector of each department are merged, including summing the element values ​​of the mean deviation vector and the target financial mean vector corresponding to each financial indicator parameter to generate a modified financial feature vector for each department, and a second financial planning deviation vector for each department is generated based on the financial planning feature vector and the modified financial feature vector.

[0008] Preferably, generating the financial planning deviation analysis results of the enterprise for multiple departments based on the first financial planning deviation vector, the second financial planning deviation vector and the financial planning coefficient includes: Determine a first characteristic value of each financial indicator parameter of each department according to the first financial planning deviation vector, and mark multiple financial indicator parameters whose first characteristic values ​​are greater than a preset planning threshold as normal parameters; For multiple financial indicator parameters whose first eigenvalues ​​are not greater than the preset planning threshold, determine the second eigenvalues ​​respectively corresponding to the multiple financial indicator parameters according to the second financial planning deviation vector, mark the multiple financial indicator parameters whose second eigenvalues ​​are greater than the preset planning threshold as normal parameters, and mark the remaining multiple financial indicator parameters as abnormal parameters, so as to obtain the local deviation analysis results of the enterprise on multiple departments; The target characteristic value of each financial indicator parameter is extracted from the first financial planning deviation vector and the second financial planning deviation vector, and the target financial deviation vector of each department is constructed; the target growth coefficient of each department in the financial data to be analyzed is determined based on the target financial deviation vector; the global deviation analysis results of the enterprise for multiple departments are generated according to the target growth coefficient and the financial planning coefficient; the local deviation analysis results and the global deviation analysis results of the enterprise for multiple departments are summarized to generate the financial planning deviation analysis results of the enterprise for multiple departments.

[0009] Preferably, extracting the target characteristic value of each financial indicator parameter from the first financial planning deviation vector and the second financial planning deviation vector and constructing a target financial deviation vector for each department includes: If the first eigenvalue of the financial indicator parameter is greater than the preset planning threshold, the first eigenvalue of the financial indicator parameter is recorded as the target eigenvalue of the financial indicator parameter. Otherwise, it is determined whether the second eigenvalue of the financial indicator parameter is greater than the preset planning threshold. If so, the second eigenvalue is recorded as the target eigenvalue of the financial indicator parameter. For the financial indicator parameter whose first eigenvalue and second eigenvalue are not greater than the preset planning threshold, the eigenvalue with the smallest deviation from the preset planning threshold is selected from the first eigenvalue and the second eigenvalue of the financial indicator parameter and is recorded as the target eigenvalue. The target financial deviation vector of each department is constructed according to the target eigenvalues ​​of multiple financial indicator parameters.

[0010] A second aspect of the present invention provides a system for intelligently processing enterprise financial data, the system being used to implement the above-mentioned method for intelligently processing enterprise financial data, comprising: The data acquisition module is used to obtain the company's financial data to be analyzed and historical financial data, extract multiple financial indicator parameters of each department of the company from the financial data to be analyzed, and construct target financial feature vectors of multiple departments; The historical data analysis module is used to extract the historical financial feature vector of each department and the parameter time series feature vector of each financial indicator parameter of each department based on the historical financial data, and to perform cluster analysis on multiple departments according to the historical financial feature vectors and the parameter time series feature vectors to determine multiple department sets; The financial indicator deviation analysis module is used to determine the reference financial feature vector of each department set, analyze multiple target financial feature vectors in each department set according to the reference financial feature vector, extract the deviation value of each financial indicator parameter of each department in the department set, calculate the deviation mean of each financial indicator parameter and construct a global financial deviation vector for each department set, and calculate the global financial deviation value of each department set according to the global financial deviation vector; A local deviation analysis module is used to collect the financial planning data of each department for a set of departments whose global financial deviation value is less than a preset global deviation threshold, extract the financial planning feature vector and financial planning coefficient of each department from the financial planning data, and generate a first financial planning deviation vector for each department according to the financial planning feature vector and the target financial feature vector; A global deviation analysis module is used to optimize the characteristics of multiple target financial feature vectors in the department set and generate a second financial planning deviation vector for each department; The financial planning deviation analysis module is used to generate the financial planning deviation analysis results of the enterprise regarding multiple departments based on the first financial planning deviation vector, the second financial planning deviation vector and the financial planning coefficient.

[0011] Preferably, the financial planning deviation analysis module generates financial planning deviation analysis results of the enterprise for multiple departments based on the first financial planning deviation vector, the second financial planning deviation vector and the financial planning coefficient, including: Determine a first characteristic value of each financial indicator parameter of each department according to the first financial planning deviation vector, and mark multiple financial indicator parameters whose first characteristic values ​​are greater than a preset planning threshold as normal parameters; For multiple financial indicator parameters whose first eigenvalues ​​are not greater than the preset planning threshold, determine the second eigenvalues ​​respectively corresponding to the multiple financial indicator parameters according to the second financial planning deviation vector, mark the multiple financial indicator parameters whose second eigenvalues ​​are greater than the preset planning threshold as normal parameters, and mark the remaining multiple financial indicator parameters as abnormal parameters, so as to obtain the local deviation analysis results of the enterprise on multiple departments; The target characteristic value of each financial indicator parameter is extracted from the first financial planning deviation vector and the second financial planning deviation vector, and the target financial deviation vector of each department is constructed; the target growth coefficient of each department in the financial data to be analyzed is determined based on the target financial deviation vector; the global deviation analysis results of the enterprise for multiple departments are generated according to the target growth coefficient and the financial planning coefficient; the local deviation analysis results and the global deviation analysis results of the enterprise for multiple departments are summarized to generate the financial planning deviation analysis results of the enterprise for multiple departments.

[0012] The present invention has the following beneficial effects: The present invention conducts department clustering analysis on enterprises based on time series characteristics through historical financial data, identifies similar financial patterns in multiple departments of the enterprise, conducts multi-dimensional analysis on the financial data to be analyzed according to the financial planning feature vectors of each department, and introduces the first financial planning deviation vector and the second financial planning deviation vector to deeply explore the financial status of each department with respect to different financial indicator parameters from both individual and global perspectives, conducts deviation analysis on the financial data, and accurately determines whether each financial indicator reaches the expected target, thereby identifying potential anomalies and risks, taking into account the similarities between departments while also taking into account individual differences, thereby improving the accuracy of enterprise financial data analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 A flowchart of a method for intelligent processing of enterprise financial data provided by an embodiment of the present invention.

[0014] Figure 2 A schematic diagram of the structure of an enterprise financial data intelligent processing system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0015] In order to make those skilled in the art better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0016] See also Figure 1 , which shows a flow chart of a method for intelligent processing of enterprise financial data provided by an embodiment of the present invention. The method can be specifically applied to Figure 2 A system for intelligently processing enterprise financial data is shown. A method for intelligently processing enterprise financial data specifically includes the following steps: Step S1: Obtain the financial data to be analyzed and the historical financial data of the enterprise, extract multiple financial indicator parameters of each department of the enterprise from the financial data to be analyzed, and construct target financial feature vectors of multiple departments.

[0017] Specifically, the financial data to be analyzed and the historical financial data usually include multi-dimensional data such as the company's asset and liability information, profit information, cash flow information, etc. for each department, covering multiple financial indicators such as revenue, cost, profit, liability, asset, cash flow, etc. The financial data to be analyzed can be preprocessed and cleaned to ensure the accuracy and consistency of the data. For the financial data to be analyzed, multiple financial indicator parameters of each department are extracted from it, such as revenue, gross profit margin, net profit, return on assets, debt ratio, etc., and the financial feature vector of each department is constructed based on these financial indicator parameters. The financial feature vector containing multiple financial indicator parameters can comprehensively and intuitively describe the financial status of each department of the enterprise, which can serve as the basis for subsequent analysis and clustering.

[0018] Step S2: extract the historical financial feature vector of each department and the parameter time series feature vector of each department regarding each financial indicator parameter based on the historical financial data, and perform cluster analysis on multiple departments according to the historical financial feature vectors and the parameter time series feature vectors to determine multiple department sets.

[0019] Specifically, for historical financial data, based on the method of extracting financial feature vectors, historical financial feature vectors of each department are extracted to reflect the financial performance of each department in the enterprise in a specific time period in the past, such as one month, one quarter, etc., and considering the time series characteristics of different financial indicator parameters in historical financial data, the parameter time series feature vectors of each financial indicator parameter of each department are extracted to reveal the trends and cyclical characteristics of different financial indicator parameters changing over time in different departments.

[0020] Based on the historical financial feature vectors and parameter time series feature vectors, cluster analysis methods are used to cluster multiple departments. For example, common clustering algorithms include K-means, DBSCAN, hierarchical clustering, etc. The goal of cluster analysis is to group departments with similar performance into the same set, that is, to identify the same financial patterns that may exist in multiple targets. For example, some departments may appear to be financially sound in historical data, while other departments may appear to be high-risk or highly volatile. Clustering can identify parts with similar financial behavior patterns, thereby constructing multiple department sets. In the process of department clustering, on the basis of conventional clustering, the time series change relationship of financial indicator parameters is also combined to achieve clustering of each department based on the fusion of time series features, so as to obtain more accurate department clustering results.

[0021] Step S3: determine the reference financial feature vector of each department set, analyze multiple target financial feature vectors in each department set according to the reference financial feature vector, and extract the deviation value of each financial indicator parameter of each department in the department set.

[0022] Specifically, after cluster analysis, multiple department sets can be determined based on multiple clusters generated by clustering, and for each department set, a reference financial feature vector representing the overall financial characteristics of each department set can be determined based on the center of the cluster to represent the overall level of financial characteristics of all departments in the set, which can reflect the overall financial behavior pattern of each department in the set. The reference financial feature vector can be obtained by calculating the mean or weighted mean of the financial feature vectors of each department in the set. This is not specifically limited in this embodiment. Then, for multiple departments in each department set, the reference financial feature vector and multiple target financial feature vectors are analyzed respectively. Specifically, the difference between the two element values ​​under each financial indicator parameter in the reference financial feature vector and the target financial feature vector is obtained, so as to obtain the deviation value of each financial indicator parameter of each department in the department set. A financial indicator parameter with a large deviation value may mean that a department has a large deviation from the historical data under a certain financial indicator parameter.

[0023] Step S4, calculating the deviation mean of each financial indicator parameter and constructing a global financial deviation vector for each department set, and calculating the global financial deviation value of each department set based on the global financial deviation vector.

[0024] Specifically, the deviation values ​​of multiple financial indicator parameters in each department set are statistically analyzed to calculate the deviation mean of each financial indicator parameter. The deviation mean can reflect the overall performance difference of each financial indicator within the department set. For example, if the deviation of a certain financial indicator parameter in all departments is small, it means that most departments have similar performance on this financial indicator; conversely, a large deviation means that the financial status is quite different. Based on the deviation mean of each financial indicator parameter, a global financial deviation vector is constructed for each department set. The global financial deviation vector will contain the deviation mean of each financial indicator and comprehensively reflect the financial performance of the department set. Then, the global financial deviation value of each department set is calculated based on the global financial deviation vector as a quantitative indicator of the deviation between the current overall financial status of the department set and the historical law.

[0025] Step S5: For the set of departments whose global financial deviation value is less than the preset global deviation threshold, collect the financial planning data of each department, extract the financial planning feature vector and financial planning coefficient of each department from the financial planning data, and generate the first financial planning deviation vector of each department according to the financial planning feature vector and the target financial feature vector.

[0026] Specifically, for the set of departments whose global financial deviation value is less than the preset global deviation threshold, it means that the current financial performance of these departments is relatively stable and close to the financial model reflected by the historical data. For these stable departments, relevant information will be extracted from the financial planning data. Financial planning data usually contains the budget, expected income, expenditure, profit target and other contents of each department, reflecting the company's planning for the future financial performance of the department. Based on the financial planning data of each department, the financial planning feature vector and financial planning coefficient of the department are extracted. The financial planning feature vector is a multi-dimensional representation of the content of the department's financial planning, which is specifically used to reflect some of the expected expectations about the parameters of each financial indicator, that is, in the financial planning data, it is expected that the department can achieve the financial level in the established plan, and the financial planning coefficient is used to characterize the specific increase in the financial level, such as the overall improvement compared with the historical data, such as the increase in percentage points. The financial planning feature vector and the financial planning coefficient can be obtained by reasonable analysis based on the financial planning data of each part, and are not specifically limited here. After determining the financial planning characteristic vector of each department, the target financial characteristic vector of the department is analyzed according to the financial planning characteristic vector to determine the deviation. The first financial planning deviation vector of each department can be calculated, that is, the difference between the actual level and the expected level of each financial indicator parameter. The deviation vector reflects the difference between the actual financial situation of the department and the financial plan as a whole, and is used to quantify the degree of deviation from the department's financial plan.

[0027] Step S6: perform feature optimization on multiple target financial feature vectors in the department set to generate a second financial planning deviation vector for each department, and generate the enterprise's financial planning deviation analysis results for multiple departments based on the first financial planning deviation vector, the second financial planning deviation vector and the financial planning coefficient.

[0028] Specifically, for each department set, multiple target financial feature vectors are optimized to improve the accuracy of financial analysis results. The feature optimization process specifically includes considering the similarity of the financial status of multiple departments, conducting an overall analysis based on the actual financial status of each department, and obtaining a feature vector that integrates the common financial characteristics of each department. Then, based on the financial planning feature vector, the feature vector that integrates the common financial characteristics of each department is analyzed to determine the deviation, and the second financial planning deviation vector of each department is calculated, which integrates the common characteristics of the actual financial status of multiple departments, and obtains the second financial planning deviation vector of the actual level of each department under the common characteristics of the actual financial status of multiple departments. The deviation vector reflects the difference between the actual financial status of each department and the financial plan under the influence of the common characteristics of the actual financial status of multiple departments, and is used to characterize the degree of deviation of the department's financial plan under the joint influence of the external environment, such as market fluctuations. Analyzing the first financial planning deviation vector and the second financial planning deviation vector can measure whether the financial planning of different departments meets expectations from the perspective of personalized and global analysis. On this basis, the actual financial growth of each department of the enterprise can be further determined, and the financial status of each department of the enterprise can be measured as a whole by combining the financial planning coefficient. Compared with the traditional method of analyzing a single department, the above solution can provide more accurate and stable financial analysis results. It helps enterprises grasp the financial status of each department from a macro perspective, discover potential systemic problems, and provide enterprises with more operational decision-making support, thereby improving the efficiency and flexibility of financial decision-making.

[0029] As an optional implementation scheme, in step S2, cluster analysis is performed on multiple departments according to the historical financial feature vectors and the parameter time series feature vectors to determine multiple department sets, specifically including: Calculate the time series similarity of each financial indicator parameter between any two departments according to the parameter time series feature vector, and determine the time series distance parameter between any two departments according to the parameter time series feature vector and the time series similarity; Specifically, for any two departments, their time series data on each financial indicator are extracted to generate the corresponding parameter time series feature vector. For example, the values ​​of a certain financial indicator of department A, such as revenue indicator, in the past 6 months constitute a time series vector, and the same financial indicator of department B can also obtain a time series vector. Then, the DTW algorithm is used to process a pair of parameter time series feature vectors under each financial indicator parameter of the two departments to obtain the time series similarity of the two departments under each financial indicator parameter. After processing multiple departments in this way, the time series similarity of each financial indicator parameter between any two departments can be obtained.

[0030] Then, the weight of each financial indicator parameter between any two departments is determined based on multiple time series similarities. According to the weights of multiple financial indicator parameters, the weight represents the importance of the financial indicator parameter in the overall similarity calculation. For example, some financial indicator parameters such as revenue and net profit show relatively similar time series correlations in the historical data of the two departments. In the process of measuring the overall similarity of the two departments, the financial indicator parameter can be assigned a higher weight. Then, based on the weight of each financial indicator parameter, the weighted Euclidean distance between any two departments is calculated based on the parameter time series feature vector to obtain the time series distance parameter between any two departments.

[0031] Among them, for the weighted Euclidean distance, the calculation formula is as follows: ; In the formula, represents the weighted Euclidean distance between department A and department B, , Represents the first Financial indicator parameters, For the The weight of each financial indicator parameter, is the number of financial indicators.

[0032] By calculating the weighted Euclidean distance, the time series distance parameters between each pair of departments can be obtained. The time series distance matrix corresponding to multiple departments can be constructed based on the calculated multiple time series distance parameters. The rows and columns of the matrix represent different departments, and each element in the matrix represents the time series distance between different departments. Finally, DBSCAN (density clustering algorithm) is used to cluster the time series distance matrix. The DBSCAN algorithm does not require the number of clusters to be specified in advance, but clusters according to the density of the data. It can automatically identify dense areas in the data and divide these dense areas into the same cluster. The DBSCAN algorithm is used to cluster similar departments in the same cluster, and finally multiple department sets are generated. The departments in each department set have high similarity in financial performance, and the time series correlation of financial indicator parameters is integrated, which can effectively reflect the similar financial behavior patterns between departments.

[0033] As an optional implementation scheme, in step S3, the global financial deviation value of each department set is calculated according to the global financial deviation vector, specifically including: Calculate the Euclidean norm of the global financial deviation vector of each department set to obtain the global financial deviation value of each department set; Specifically, by calculating the Euclidean norm corresponding to the global financial deviation vector of each department set and recording it as the global financial deviation value of the department set, the degree of financial deviation of multiple departments in the department set is quantified.

[0034] For the global financial deviation vector, after determining the reference financial feature vector of each department set, the reference financial feature vector of each department set is first analyzed to determine the deviation value of each department in the department set with respect to each financial indicator parameter. Specifically, the first element value of each financial indicator parameter in the department set is determined according to the reference financial feature vector, and the multiple second element values ​​of each financial indicator parameter in the department set are determined according to the multiple target financial feature vectors. Then, the difference between the multiple second element values ​​and the first element value under each financial indicator parameter is calculated, that is, the difference between the reference financial feature vector and the target financial feature vector with respect to the value of the financial indicator parameter under each financial indicator parameter is calculated, so as to obtain the deviation value of each department in the department set with respect to each financial indicator parameter. After averaging the multiple deviation values ​​under each financial indicator parameter, the deviation mean of the department set with respect to each financial indicator parameter can be obtained, and finally the global financial deviation vector of each department set is constructed according to the deviation mean of multiple financial indicator parameters.

[0035] The above steps of calculating the global financial deviation vector for each department set take into account the difference between each department and the reference vector, and integrate the financial performance between departments, providing a comprehensive financial deviation assessment for different department sets.

[0036] As an optional implementation scheme, in step S6, multiple target financial feature vectors in the department set are optimized to generate a second financial planning deviation vector for each department, specifically including: Generate a target financial mean vector for each department set based on multiple target financial feature vectors of the department set, and extract a mean deviation vector for each department in the department set based on the target financial mean vector; Specifically, the mean of each financial indicator parameter in multiple target financial feature vectors is obtained, and the target financial mean vector of the department set is constructed according to the mean of each financial indicator parameter. Then, the target financial mean vector of each department is analyzed based on the target financial mean vector of the department set, including calculating the difference between the target financial feature vector of the department and the target financial mean vector of the department set with respect to the element value corresponding to each financial indicator parameter, and constructing the mean deviation vector of each department according to the difference between the element value corresponding to each financial indicator parameter.

[0037] The mean deviation vector of each department and the target financial mean vector are merged to generate the modified financial feature vector of each department. Specifically, the element values ​​corresponding to each financial indicator parameter of the mean deviation vector and the target financial mean vector are summed, and the second financial planning deviation vector of each department is constructed based on the value obtained by summing the element values ​​corresponding to each financial indicator parameter. In this way, the original financial data to be analyzed is corrected according to the overall level of the department set, so that the financial data can better reflect the relationship between the department and the whole.

[0038] As an optional implementation scheme, in step S6, the financial planning deviation analysis results of the enterprise for multiple departments are generated based on the first financial planning deviation vector, the second financial planning deviation vector and the financial planning coefficient, including: A first characteristic value of each financial indicator parameter of each department is determined according to the first financial planning deviation vector, and a plurality of financial indicator parameters whose first characteristic values ​​are greater than a preset planning threshold are marked as normal parameters.

[0039] Specifically, the first eigenvalue of each financial indicator parameter represents the deviation of each financial indicator parameter when comparing the financial planning eigenvector with the target financial eigenvector, or it can be understood as the excess of the true value over the expected value, that is, the degree to which the true value of the enterprise's financial indicator parameter exceeds the expected value when analyzing a certain enterprise alone without considering the common characteristics of the enterprise. For example, the difference between the actual income of a department and the income under the expected plan. The larger the difference, the greater the excess of the actual value of the financial indicator parameter over the expected value. The degree of excess is measured by the preset planning threshold. If the first eigenvalue of the financial indicator parameter is greater than the preset planning threshold, it means that the true data of these financial indicator parameters have not deviated from the expected financial plan.

[0040] For multiple financial indicator parameters whose first eigenvalues ​​are not greater than the preset planning threshold, the second eigenvalues ​​respectively corresponding to the multiple financial indicator parameters are determined according to the second financial planning deviation vector, and the multiple financial indicator parameters whose second eigenvalues ​​are greater than the preset planning threshold are marked as normal parameters, and the remaining multiple financial indicator parameters are marked as abnormal parameters, so as to obtain the local deviation analysis results of the enterprise for multiple departments.

[0041] Specifically, for multiple financial indicator parameters whose first eigenvalues ​​are not greater than the preset planning threshold, the conventional scheme may simply consider that they have not reached the expected target. However, this application further analyzes these financial indicator parameters in combination with the second financial planning deviation vector. Specifically, the second eigenvalues ​​corresponding to the multiple financial indicator parameters are determined according to the second financial planning deviation vector, which represents the excess of each financial indicator parameter about the expected value under the comparison of the financial planning eigenvector and the target financial eigenvector. The generation of the second financial planning deviation vector takes into account the overall association between departments with similar financial models, such as the overall impact of the market environment impact. In this case, individual differences at the overall level are considered, and the second eigenvalue represents the excess of the financial indicator parameter about the expected value. If the second eigenvalue is greater than the preset planning threshold in this case, it is also marked as a normal parameter, that is, although the financial indicator parameter does not meet expectations from the perspective of individual data alone, it is still in line with expectations from the overall situation, but it may be affected by market fluctuations, resulting in certain differences in individual data. For the remaining multiple financial indicator parameters marked as abnormal parameters, it is considered that the expected financial level has not been reached. In this way, combined with the relationship between departments, the financial status of each department is analyzed from the individual level, that is, each department has a separate financial analysis of each financial indicator parameter, so as to determine the financial level of each financial indicator parameter within the department and obtain accurate individual financial analysis results for each department.

[0042] On this basis, the target eigenvalue of each financial indicator parameter is extracted from the first financial planning deviation vector and the second financial planning deviation vector and a target financial deviation vector is constructed. The construction of the target financial deviation vector is specifically combined with the above-mentioned analysis results. First, if the first eigenvalue of the financial indicator parameter meets expectations, the first eigenvalue is recorded as the target eigenvalue of the financial indicator parameter. Otherwise, it is determined whether the second eigenvalue of the financial indicator parameter meets expectations. If it does, the second eigenvalue is recorded as the target eigenvalue of the financial indicator parameter. For financial indicator parameters whose first eigenvalue and second eigenvalue do not meet expectations, the eigenvalue with the smallest deviation from the preset planning threshold can be selected from the first eigenvalue and the second eigenvalue of the financial indicator parameter and recorded as the target eigenvalue, thereby constructing a target financial deviation vector that characterizes the overall financial level of the department. The target growth coefficient of the financial data to be analyzed for each department can be determined by the target financial deviation vector, that is, the sub-growth coefficient of each financial indicator parameter relative to the initial data of the financial data to be analyzed, and then the average of multiple sub-growth coefficients is obtained to obtain the target growth coefficient of the financial data to be analyzed for each department, which is used to characterize the overall growth level of the financial data to be analyzed. By comparing it with the financial planning coefficient corresponding to the department, the financial status of each department at the overall level can be obtained to judge whether it meets the expected level as a whole, and the global deviation analysis results of each department of the enterprise representing the group financial level are obtained. After summarizing the local deviation analysis results and global deviation analysis results of the enterprise for multiple departments, the financial planning deviation analysis results of the enterprise for multiple departments are generated. The local deviation analysis reveals the abnormal and normal conditions of the specific financial indicators of each department, while the global deviation analysis provides an assessment of the degree of achievement of the overall financial goals of each department. Through the combination of the two, the performance of each department of the enterprise in the implementation of financial planning can be more comprehensively understood, problems can be identified and financial strategies can be optimized. This analysis method can help enterprises accurately identify deviations in the implementation of financial goals, and make timely adjustments and optimization decisions through multi-dimensional analysis to ensure that the implementation of financial planning is more in line with the planning goals of the enterprise.

[0043] See also Figure 2 Based on the same concept as the above-mentioned enterprise financial data intelligent processing method, a system for intelligent processing of enterprise financial data is also provided, which specifically includes: The data acquisition module is used to obtain the company's financial data to be analyzed and historical financial data, extract multiple financial indicator parameters of each department of the company from the financial data to be analyzed, and construct target financial feature vectors of multiple departments; The historical data analysis module is used to extract the historical financial feature vector of each department and the parameter time series feature vector of each financial indicator parameter of each department based on the historical financial data, and to perform cluster analysis on multiple departments according to the historical financial feature vectors and the parameter time series feature vectors to determine multiple department sets; Specifically, cluster analysis is performed on multiple departments based on historical financial feature vectors and parameter time series feature vectors to determine multiple department sets, including: Calculate the time series similarity of each financial indicator parameter between any two departments according to the parameter time series feature vector, and determine the time series distance parameter between any two departments according to the parameter time series feature vector and the time series similarity; The method includes processing multiple pairs of parameter time series feature vectors of any two departments based on the DTW algorithm, calculating the time series similarity between any two departments with respect to each financial indicator parameter, determining the weight of each financial indicator parameter between any two departments based on multiple time series similarities, calculating the weighted Euclidean distance between any two departments based on the parameter time series feature vector according to the weights of multiple financial indicator parameters, and obtaining the time series distance parameter between any two departments; The temporal distance matrix corresponding to multiple departments is constructed according to multiple temporal distance parameters, and the DBSCAN algorithm is used to cluster multiple departments based on the temporal distance matrix to generate multiple department sets.

[0044] The financial indicator deviation analysis module is used to determine the reference financial feature vector of each department set, analyze multiple target financial feature vectors in each department set according to the reference financial feature vector, extract the deviation value of each financial indicator parameter of each department in the department set, calculate the deviation mean of each financial indicator parameter and construct a global financial deviation vector for each department set, and calculate the global financial deviation value of each department set according to the global financial deviation vector;

[0045] A local deviation analysis module is used to collect the financial planning data of each department for a set of departments whose global financial deviation value is less than a preset global deviation threshold, extract the financial planning feature vector and financial planning coefficient of each department from the financial planning data, and generate a first financial planning deviation vector for each department according to the financial planning feature vector and the target financial feature vector; A global deviation analysis module is used to optimize the characteristics of multiple target financial feature vectors in the department set and generate a second financial planning deviation vector for each department; The financial planning deviation analysis module is used to generate the financial planning deviation analysis results of the enterprise regarding multiple departments based on the first financial planning deviation vector, the second financial planning deviation vector and the financial planning coefficient.

[0046] Specifically, based on the first financial planning deviation vector, the second financial planning deviation vector and the financial planning coefficient, the financial planning deviation analysis results of the enterprise for multiple departments are generated, including: Determine a first characteristic value of each financial indicator parameter of each department according to the first financial planning deviation vector, and mark multiple financial indicator parameters whose first characteristic values ​​are greater than a preset planning threshold as normal parameters; For multiple financial indicator parameters whose first eigenvalues ​​are not greater than the preset planning threshold, determine the second eigenvalues ​​respectively corresponding to the multiple financial indicator parameters according to the second financial planning deviation vector, mark the multiple financial indicator parameters whose second eigenvalues ​​are greater than the preset planning threshold as normal parameters, and mark the remaining multiple financial indicator parameters as abnormal parameters, so as to obtain the local deviation analysis results of the enterprise on multiple departments; The target characteristic value of each financial indicator parameter is extracted from the first financial planning deviation vector and the second financial planning deviation vector, and the target financial deviation vector of each department is constructed; the target growth coefficient of each department in the financial data to be analyzed is determined based on the target financial deviation vector; the global deviation analysis results of the enterprise for multiple departments are generated according to the target growth coefficient and the financial planning coefficient; the local deviation analysis results and the global deviation analysis results of the enterprise for multiple departments are summarized to generate the financial planning deviation analysis results of the enterprise for multiple departments.

[0047] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. Parts not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. A method for intelligent processing of enterprise financial data, characterized in that: include: Obtain the company's financial data to be analyzed and historical financial data, extract multiple financial indicator parameters of each department of the company from the financial data to be analyzed, and construct target financial feature vectors of multiple departments; Based on the historical financial data, the historical financial feature vector of each department and the parameter time series feature vector of each financial indicator parameter of each department are extracted, and a cluster analysis is performed on multiple departments according to the historical financial feature vectors and the parameter time series feature vectors to determine multiple department sets; Determine a reference financial feature vector for each department set, analyze multiple target financial feature vectors in each department set based on the reference financial feature vector, and extract the deviation value of each financial indicator parameter of each department in the department set; Calculate the deviation mean of each financial indicator parameter and construct a global financial deviation vector for each department set, and calculate the global financial deviation value of each department set based on the global financial deviation vector; For a set of departments whose global financial deviation value is less than a preset global deviation threshold, the financial planning data of each department is collected, the financial planning feature vector and the financial planning coefficient of each department are extracted from the financial planning data, and the first financial planning deviation vector of each department is generated according to the financial planning feature vector and the target financial feature vector; Feature optimization is performed on multiple target financial feature vectors in the department set to generate a second financial planning deviation vector for each department, and financial planning deviation analysis results for multiple departments of the enterprise are generated based on the first financial planning deviation vector, the second financial planning deviation vector and the financial planning coefficient.

2. The method for intelligent processing of enterprise financial data according to claim 1, characterized in that: Based on the historical financial feature vectors and parameter time series feature vectors, cluster analysis is performed on multiple departments to determine multiple department sets, including: Calculate the time series similarity of each financial indicator parameter between any two departments according to the parameter time series feature vector, and determine the time series distance parameter between any two departments according to the parameter time series feature vector and the time series similarity; The method includes processing multiple pairs of parameter time series feature vectors of any two departments based on the DTW algorithm, calculating the time series similarity between any two departments with respect to each financial indicator parameter, determining the weight of each financial indicator parameter between any two departments based on multiple time series similarities, calculating the weighted Euclidean distance between any two departments based on the parameter time series feature vector according to the weights of multiple financial indicator parameters, and obtaining the time series distance parameter between any two departments; The temporal distance matrix corresponding to multiple departments is constructed according to multiple temporal distance parameters, and the DBSCAN algorithm is used to cluster multiple departments based on the temporal distance matrix to generate multiple department sets.

3. The method for intelligent processing of enterprise financial data according to claim 1, characterized in that: The global financial deviation value of each department set is calculated based on the global financial deviation vector, including: Calculate the Euclidean norm of the global financial deviation vector of each department set to obtain the global financial deviation value of each department set; For the global financial deviation vector, after determining the reference financial feature vector of each department set, determine the first element value of each financial indicator parameter in the department set according to the reference financial feature vector, and determine multiple second element values ​​of each financial indicator parameter in the department set according to multiple target financial feature vectors; Calculate the difference between the second element values ​​and the first element value under each financial indicator parameter, and obtain the deviation value of each financial indicator parameter for each department in the department set; The average of multiple deviation values ​​under each financial indicator parameter is calculated to obtain the deviation mean of each financial indicator parameter, and the global financial deviation vector of each department set is constructed based on the deviation mean of multiple financial indicator parameters.

4. The method for intelligent processing of enterprise financial data according to claim 3, characterized in that: Perform feature optimization on multiple target financial feature vectors in the department set to generate the second financial planning deviation vector for each department, including: Generate a target financial mean vector for each department set based on multiple target financial feature vectors of the department set, and extract a mean deviation vector for each department in the department set based on the target financial mean vector, including calculating the difference between the target financial feature vector of the department and the target financial mean vector of the department set with respect to the element value corresponding to each financial indicator parameter, and construct a mean deviation vector for each department; The mean deviation vector and the target financial mean vector of each department are merged, including summing the element values ​​of the mean deviation vector and the target financial mean vector corresponding to each financial indicator parameter to generate a modified financial feature vector for each department, and a second financial planning deviation vector for each department is generated based on the financial planning feature vector and the modified financial feature vector.

5. The method for intelligent processing of enterprise financial data according to claim 1, characterized in that: The financial planning deviation analysis results of the enterprise for multiple departments are generated based on the first financial planning deviation vector, the second financial planning deviation vector and the financial planning coefficient, including: Determine a first characteristic value of each financial indicator parameter of each department according to the first financial planning deviation vector, and mark multiple financial indicator parameters whose first characteristic values ​​are greater than a preset planning threshold as normal parameters; For multiple financial indicator parameters whose first eigenvalues ​​are not greater than the preset planning threshold, determine the second eigenvalues ​​respectively corresponding to the multiple financial indicator parameters according to the second financial planning deviation vector, mark the multiple financial indicator parameters whose second eigenvalues ​​are greater than the preset planning threshold as normal parameters, and mark the remaining multiple financial indicator parameters as abnormal parameters, so as to obtain the local deviation analysis results of the enterprise on multiple departments; The target characteristic value of each financial indicator parameter is extracted from the first financial planning deviation vector and the second financial planning deviation vector, and the target financial deviation vector of each department is constructed; the target growth coefficient of each department in the financial data to be analyzed is determined based on the target financial deviation vector; the global deviation analysis results of the enterprise for multiple departments are generated according to the target growth coefficient and the financial planning coefficient; the local deviation analysis results and the global deviation analysis results of the enterprise for multiple departments are summarized to generate the financial planning deviation analysis results of the enterprise for multiple departments.

6. The method for intelligent processing of enterprise financial data according to claim 5, characterized in that: The target characteristic value of each financial indicator parameter is extracted from the first financial planning deviation vector and the second financial planning deviation vector, and the target financial deviation vector of each department is constructed, including: If the first eigenvalue of the financial indicator parameter is greater than the preset planning threshold, the first eigenvalue of the financial indicator parameter is recorded as the target eigenvalue of the financial indicator parameter. Otherwise, it is determined whether the second eigenvalue of the financial indicator parameter is greater than the preset planning threshold. If so, the second eigenvalue is recorded as the target eigenvalue of the financial indicator parameter. For the financial indicator parameter whose first eigenvalue and second eigenvalue are not greater than the preset planning threshold, the eigenvalue with the smallest deviation from the preset planning threshold is selected from the first eigenvalue and the second eigenvalue of the financial indicator parameter and is recorded as the target eigenvalue. The target financial deviation vector of each department is constructed according to the target eigenvalues ​​of multiple financial indicator parameters.

7. An intelligent processing system for enterprise financial data, characterized in that: The system is used to implement the enterprise financial data intelligent processing method described in any one of claims 1 to 6, including: The data acquisition module is used to obtain the company's financial data to be analyzed and historical financial data, extract multiple financial indicator parameters of each department of the company from the financial data to be analyzed, and construct target financial feature vectors of multiple departments; The historical data analysis module is used to extract the historical financial feature vector of each department and the parameter time series feature vector of each financial indicator parameter of each department based on the historical financial data, and to perform cluster analysis on multiple departments according to the historical financial feature vectors and the parameter time series feature vectors to determine multiple department sets; The financial indicator deviation analysis module is used to determine the reference financial feature vector of each department set, analyze multiple target financial feature vectors in each department set according to the reference financial feature vector, extract the deviation value of each financial indicator parameter of each department in the department set, calculate the deviation mean of each financial indicator parameter and construct a global financial deviation vector for each department set, and calculate the global financial deviation value of each department set according to the global financial deviation vector; A local deviation analysis module is used to collect the financial planning data of each department for a set of departments whose global financial deviation value is less than a preset global deviation threshold, extract the financial planning feature vector and financial planning coefficient of each department from the financial planning data, and generate a first financial planning deviation vector for each department according to the financial planning feature vector and the target financial feature vector; A global deviation analysis module is used to optimize the characteristics of multiple target financial feature vectors in the department set and generate a second financial planning deviation vector for each department; The financial planning deviation analysis module is used to generate the financial planning deviation analysis results of the enterprise regarding multiple departments based on the first financial planning deviation vector, the second financial planning deviation vector and the financial planning coefficient.

8. The enterprise financial data intelligent processing system according to claim 7, characterized in that: For the financial planning deviation analysis module, the financial planning deviation analysis results of the enterprise for multiple departments are generated based on the first financial planning deviation vector, the second financial planning deviation vector and the financial planning coefficient, including: Determine a first characteristic value of each financial indicator parameter of each department according to the first financial planning deviation vector, and mark multiple financial indicator parameters whose first characteristic values ​​are greater than a preset planning threshold as normal parameters; For multiple financial indicator parameters whose first eigenvalues ​​are not greater than the preset planning threshold, determine the second eigenvalues ​​respectively corresponding to the multiple financial indicator parameters according to the second financial planning deviation vector, mark the multiple financial indicator parameters whose second eigenvalues ​​are greater than the preset planning threshold as normal parameters, and mark the remaining multiple financial indicator parameters as abnormal parameters, so as to obtain the local deviation analysis results of the enterprise on multiple departments; The target characteristic value of each financial indicator parameter is extracted from the first financial planning deviation vector and the second financial planning deviation vector, and the target financial deviation vector of each department is constructed; the target growth coefficient of each department in the financial data to be analyzed is determined based on the target financial deviation vector; the global deviation analysis results of the enterprise for multiple departments are generated according to the target growth coefficient and the financial planning coefficient; the local deviation analysis results and the global deviation analysis results of the enterprise for multiple departments are summarized to generate the financial planning deviation analysis results of the enterprise for multiple departments.

Images