An enterprise financial process management system based on big data

Abstract

The present invention relates to the technical field of enterprise financial data management, and in particular to an enterprise financial process management system based on big data. The present invention obtains enterprise financial data, wherein the enterprise financial data includes financial statements, business data and tax data; performs dimensionality reduction processing on the enterprise financial data to obtain a sample set after dimensionality reduction processing; performs clustering processing on the enterprise financial data to complete clustering of new data, wherein the new data is the latest added enterprise financial data; performs security risk assessment based on the clustered enterprise financial data to obtain an assessment result, and issues an early warning to the enterprise based on the assessment result. It can improve the efficiency of enterprise financial process management, reduce management costs, and reduce the risk of enterprise financial process management.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise financial data management, and in particular to an enterprise financial process management system based on big data. Background Art

[0002] With the rapid development and widespread application of big data technology, corporate financial management faces unprecedented challenges and opportunities. Traditional financial management models often rely on manual operations and limited data processing capabilities, resulting in low work efficiency, insufficient data accuracy, and difficulty in coping with complex and changing market environments. Therefore, the emergence of enterprise financial process management systems based on big data technology has become a key way to solve these problems.

[0003] Big data technology has the advantages of rich data types, large volume, and fast processing speed. Through intelligent analysis and data screening, it can extract data information with practical application value. With the support of big data technology, enterprises can comprehensively and quickly obtain, process and analyze various financial data, thereby improving the efficiency and accuracy of financial management.

[0004] The traditional corporate financial management model has the following major challenges:

[0005] Insufficient data processing capabilities: Faced with massive amounts of financial data, traditional systems find it difficult to process them efficiently, resulting in long data processing cycles and low accuracy.

[0006] The phenomenon of information islands is serious: It is difficult to effectively share financial data among departments within the enterprise, resulting in serious information islands and affecting decision-making efficiency.

[0007] Limited risk management capabilities: Traditional financial management models make it difficult to provide timely warnings and effective management of potential risks, increasing the financial risks of the company.

[0008] Therefore, there is an urgent need for an enterprise financial process management system based on big data to solve the above problems. Summary of the invention

[0009] The purpose of the present invention is to provide an enterprise financial process management system based on big data: to solve the technical problems that the traditional enterprise financial management model of the existing solutions mainly relies on manual data processing, resulting in low efficiency, high cost and high risk of enterprise financial process management.

[0010] The purpose of the present invention can be achieved through the following technical solutions:

[0011] An enterprise financial process management system based on big data, the system includes a financial data acquisition module, a financial data dimensionality reduction processing module, a financial data clustering processing module and a financial data security risk assessment module;

[0012] The financial data acquisition module is used to acquire enterprise financial data, wherein the enterprise financial data includes financial statements, business data and tax data;

[0013] The financial data dimensionality reduction processing module is used to perform dimensionality reduction processing on corporate financial data: construct a sample matrix X based on corporate financial data, perform a mean operation on the matrix X, and construct a mean matrix M; distribute the mean-processed data evenly to each local node, and construct a local matrix of each node , calculate the three variable information of the square matrix H, linear sum vector B, and square sum vector F of each local node data; calculate the correlation coefficient of each local node data based on the three variable information, construct the correlation coefficient matrix R based on the correlation coefficient, solve the eigenvalue λ of the correlation coefficient matrix R, arrange the eigenvalue λ in descending order, and determine the first k eigenvalues ​​by the preset contribution rate , based on the k eigenvalues, calculate the corresponding eigenvector , k eigenvectors are used to construct a projection matrix, and the sample matrix X is mapped onto the projection matrix to obtain the sample set after dimensionality reduction;

[0014] The financial data clustering processing module is used to perform clustering processing on the enterprise financial data and complete the clustering of the newly arrived data, wherein the newly arrived data is the latest added enterprise financial data;

[0015] The financial data security risk assessment module is used to conduct security risk assessment based on clustered enterprise financial data, obtain assessment results, and issue early warnings to enterprises based on the assessment results.

[0016] Furthermore, a sample matrix X is constructed based on the enterprise financial data, the matrix X is averaged, and the averaged matrix M is constructed. Specifically, the following process is included:

[0017] Based on the enterprise financial data, n data samples with d-dimensional attribute vectors are expressed as , and each sample is represented as , then the sample matrix X is expressed as follows:

[0018] ;

[0019] The sample matrix X is averaged as follows:

[0020] ;

[0021] in, , ;

[0022] The resulting mean matrix M is:

[0023] .

[0024] Furthermore, the calculation of the three variable information of the square matrix H, the linear sum vector B, and the square sum vector F of each local node data specifically includes the following process:

[0025] Construct a formula based on a square matrix:

[0026] ;

[0027] Then the square matrix H is:

[0028] ;

[0029] Compute the linear sum of the matrix M and the vector B:

[0030] ,in, , ;

[0031] Compute the square of the matrix M and the vector F:

[0032] ,in, , .

[0033] Furthermore, the correlation coefficients of the local node data are calculated based on the three variable information, and the correlation coefficient matrix R is constructed based on the correlation coefficients, which specifically includes the following process:

[0034] Calculate the correlation coefficient of each local node data based on three variable information , the calculation formula is as follows:

[0035] ;

[0036] Based on the correlation coefficient Construct the correlation coefficient matrix R:

[0037] .

[0038] Furthermore, the eigenvalue λ of the correlation coefficient matrix R is solved, the eigenvalue λ is arranged in descending order, and the first k eigenvalues ​​are determined by the preset contribution rate , based on the k eigenvalues, calculate the corresponding eigenvector , k eigenvectors to construct the projection matrix specifically include the following process:

[0039] Use contribution rate To determine the number of principal components:

[0040] ;

[0041] The first k eigenvalues The corresponding eigenvector are the k eigenvectors determined, where each eigenvector , ;

[0042] k eigenvectors construct the projection matrix V:

[0043] .

[0044] Furthermore, the enterprise financial data is clustered, and the clustering of the new data specifically includes the following processes:

[0045] Using the k-means algorithm to cluster the initially read preset financial data to obtain a number of initial micro-clusters, and combining the initial micro-clusters into an initial core micro-cluster set, the initial core micro-cluster set including a number of core micro-clusters;

[0046] Get the newly arrived data and determine whether the newly arrived data can be absorbed by an existing core micro-cluster. If so, add the data to this micro-cluster.

[0047] If not, calculate the newly arrived data The local outlier factor LDOF is:

[0048] ;in, is the distance to the nearest neighbor data, Represents the distance to the newly arrived data The set of the most recent z data, Core micro-cluster center With the new data Euclidean distance;

[0049] ;in, is the internal distance between z's nearest neighbor data;

[0050] ;

[0051] judge Does it exceed the preset threshold? If so, the new data If it is abnormal data, add it to the buffer micro-cluster set for later observation; if not, the newly arrived data is normal data, then it is new data Create a new core micro-cluster and add it to the core micro-cluster collection;

[0052] Among them, the later observation includes judging whether the newly arrived data is seed data or outlier data based on the condition. The condition is whether the newly arrived data attracts a preset number of data to form a new cluster. If so, the newly arrived data is seed data, and if not, the newly arrived data is outlier data.

[0053] Furthermore, security risk assessment is performed based on the clustered enterprise financial data, and the assessment results are obtained by the following process:

[0054] Based on the clustered enterprise financial data, the data completeness CWZ on the financial statements is obtained: the ratio of the actual data items on the financial statements to the data items that should be on the financial statements is calculated, and the ratio is used as the quantitative data completeness CWZ on the financial statements;

[0055] Based on the clustered enterprise financial data, the data completeness YWZ of the business data is obtained: the ratio of the actual data items on the business data to the data items on the business data should be calculated, and the ratio is used as the data completeness YWZ of the business data;

[0056] Based on the clustered enterprise financial data, the data completeness SWZ of the tax data is obtained: the ratio of the actual data items on the tax data to the data items that should be on the tax data is calculated, and the ratio is used as the data completeness SWZ of the tax data;

[0057] Substituting CWZ, YWZ and SWZ into the safety risk assessment index calculation formula, the safety risk assessment index LMS is obtained. The calculation formula is as follows:

[0058] ;

[0059] in, , is the weight coefficient, which is set by the system;

[0060] The safety risk assessment index LMS is recorded as the assessment result.

[0061] Furthermore, when the safety risk assessment index LMS is greater than a preset safety risk assessment index threshold, a warning message is generated and fed back to the enterprise, wherein the preset safety risk assessment index threshold is stored in the system and set by the system.

[0062] Compared with the existing solutions, the present invention achieves the following beneficial effects:

[0063] The present invention obtains enterprise financial data, wherein the enterprise financial data includes financial statements, business data and tax data; performs dimensionality reduction processing on the enterprise financial data to obtain a sample set after dimensionality reduction processing; performs clustering processing on the enterprise financial data to complete clustering of new data, wherein the new data is the latest added enterprise financial data; performs security risk assessment based on the clustered enterprise financial data to obtain an assessment result, and issues an early warning to the enterprise based on the assessment result. The present invention can improve the processing efficiency of financial data, further improve the efficiency of enterprise financial process management, reduce management costs, and reduce the risk of enterprise financial process management. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0065] Figure 1 This is a system block diagram of an enterprise financial process management system based on big data according to an embodiment of the present invention;

[0066] Figure 2 It is a workflow diagram of an enterprise financial process management system based on big data according to an embodiment of the present invention;

[0067] Figure 3 This is another workflow diagram of an enterprise financial process management system based on big data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0068] 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.

[0069] In addition, the described features, structures or characteristics may be combined in one or more example embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the example embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced while omitting one or more of the specific details, or other methods, components, steps, etc. may be adopted. In other cases, well-known structures, methods, implementations or operations are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0070] This embodiment provides an enterprise financial process management system based on big data. Figure 1 is a system block diagram of an enterprise financial process management system based on big data according to an embodiment of the present invention. Figure 1 As shown, the system includes a financial data acquisition module, a financial data dimensionality reduction processing module, a financial data clustering processing module and a financial data security risk assessment module;

[0071] The financial data acquisition module is used to acquire enterprise financial data, wherein the enterprise financial data includes financial statements, business data and tax data;

[0072] The financial data dimensionality reduction processing module is used to perform dimensionality reduction processing on corporate financial data: construct a sample matrix X based on corporate financial data, perform a mean operation on matrix X, and construct a mean matrix M; distribute the mean-processed data evenly to each local node. It is worth noting that in the process of data dimensionality reduction, local nodes are the local structure and relationship of data in the original high-dimensional space. These local nodes refer to a small group of neighboring data points around each data point in the data set. These neighboring points together constitute the local environment or "neighborhood" of the data point. In dimensionality reduction technology, local nodes help to maintain the original structure and characteristics of the data as much as possible while reducing the data dimension. Local nodes can be focused on based on Laplacian Eigenmaps (LE) to construct local matrices for each node. , calculate the three variable information of the square matrix H, linear sum vector B, and square sum vector F of each local node data; calculate the correlation coefficient of each local node data based on the three variable information, construct the correlation coefficient matrix R based on the correlation coefficient, solve the eigenvalue λ of the correlation coefficient matrix R, arrange the eigenvalue λ in descending order, and determine the first k eigenvalues ​​by the preset contribution rate , based on the k eigenvalues, calculate the corresponding eigenvector , k eigenvectors are used to construct a projection matrix, and the sample matrix X is mapped onto the projection matrix to obtain the sample set after dimensionality reduction;

[0073] The financial data clustering processing module is used to perform clustering processing on the enterprise financial data and complete the clustering of the newly arrived data, wherein the newly arrived data is the latest added enterprise financial data;

[0074] The financial data security risk assessment module is used to conduct security risk assessment based on clustered enterprise financial data, obtain assessment results, and issue early warnings to enterprises based on the assessment results.

[0075] In summary, the present invention obtains enterprise financial data, wherein the enterprise financial data includes financial statements, business data and tax data; performs dimensionality reduction processing on the enterprise financial data to obtain a sample set after dimensionality reduction processing; performs clustering processing on the enterprise financial data to complete clustering of new data, wherein the new data is the latest added enterprise financial data; performs security risk assessment based on the clustered enterprise financial data to obtain an assessment result, and issues an early warning to the enterprise based on the assessment result. The present invention can improve the processing efficiency of financial data, further improve the efficiency of enterprise financial process management, reduce management costs, and reduce the risk of enterprise financial process management.

[0076] In some embodiments, constructing a sample matrix X based on enterprise financial data, performing a mean operation on the matrix X, and constructing a mean matrix M specifically includes the following process:

[0077] Based on the enterprise financial data, n data samples with d-dimensional attribute vectors are expressed as , and each sample is represented as ,in, It is expressed as one or more of financial statements, business data and tax data. Financial statements include the amount and composition of assets, liabilities and owner's equity (or shareholder's equity), as well as the company's income, expenses, and gains and losses that should be recorded in the current profit. Business data includes but is not limited to data on sales, production, procurement, inventory, finance, human resources and other links. Tax data includes but is not limited to basic tax information, declaration and payment information, etc. The sample matrix X is expressed as follows:

[0078] ;

[0079] The sample matrix X is averaged as follows:

[0080] ;

[0081] in, , ;

[0082] The resulting mean matrix M is:

[0083] .

[0084] In some embodiments, calculating the three variable information of the square matrix H, the linear sum vector B, and the square sum vector F of each local node data specifically includes the following process:

[0085] Construct a formula based on a square matrix:

[0086] ;

[0087] Then the square matrix H is:

[0088] ;

[0089] Compute the linear sum of the matrix M and the vector B:

[0090] ,in, , ;

[0091] Compute the square of the matrix M and the vector F:

[0092] ,in, , .

[0093] Furthermore, the correlation coefficients of the local node data are calculated based on the three variable information, and the correlation coefficient matrix R is constructed based on the correlation coefficients, which specifically includes the following process:

[0094] Calculate the correlation coefficient of each local node data based on three variable information , the calculation formula is as follows:

[0095] ;

[0096] Based on the correlation coefficient Construct the correlation coefficient matrix R:

[0097] .

[0098] It is worth noting that the three variable information of the square matrix H, linear sum vector B, and square sum vector F of each local node data is calculated to construct the correlation coefficient , quickly get the correlation coefficient based on the three variable information , which can further improve the efficiency of dimensionality reduction processing.

[0099] It is worth noting that based on the correlation coefficient The eigenvalue λ of the correlation coefficient matrix R can be solved, and the dimension reduction processing of the enterprise financial data can be realized more quickly, and further, the efficiency of the security risk assessment of the enterprise financial data can be improved.

[0100] In some embodiments, the eigenvalue λ of the correlation coefficient matrix R is solved, the eigenvalue λ is arranged in descending order, and the first k eigenvalues ​​are determined by a preset contribution rate. , based on the k eigenvalues, calculate the corresponding eigenvector , k eigenvectors to construct the projection matrix specifically include the following process:

[0101] Use contribution rate To determine the number of principal components:

[0102] ;

[0103] The first k eigenvalues The corresponding eigenvector are the k eigenvectors determined, where each eigenvector , ;

[0104] k eigenvectors construct the projection matrix V:

[0105] .

[0106] In some embodiments, Figure 2 is a workflow diagram of an enterprise financial process management system based on big data according to an embodiment of the present invention, such as Figure 2 As shown, clustering of enterprise financial data is performed to complete the clustering of new data, which specifically includes the following steps:

[0107] Step S201: clustering the initially read preset financial data using a k-means algorithm to obtain a number of initial micro-clusters, and combining the initial micro-clusters into an initial core micro-cluster set, the initial core micro-cluster set including a number of core micro-clusters;

[0108] Step S202: Acquire new data and determine whether the new data can be absorbed by an existing core micro-cluster. If so, add the data to the micro-cluster.

[0109] If not, calculate the newly arrived data The local outlier factor LDOF is:

[0110] ;in, is the distance to the nearest neighbor data, Represents the distance to the newly arrived data The set of the most recent z data, Core micro-cluster center With the new data Euclidean distance;

[0111] ;in, is the internal distance between z's nearest neighbor data;

[0112] ;

[0113] Step S203: Determination Does it exceed the preset threshold? If so, the new data If it is abnormal data, add it to the buffer micro-cluster set for later observation; if not, the newly arrived data is normal data, then it is new data Create a new core micro-cluster and add it to the core micro-cluster collection;

[0114] Among them, the later observation includes judging whether the newly arrived data is seed data or outlier data based on the condition. The condition is whether the newly arrived data attracts a preset number of data to form a new cluster. If so, the newly arrived data is seed data, and if not, the newly arrived data is outlier data.

[0115] Further, Figure 3 is another workflow diagram of an enterprise financial process management system based on big data according to an embodiment of the present invention. Figure 3 As shown, security risk assessment is performed based on the clustered enterprise financial data, and the assessment results are obtained by the following steps:

[0116] Step S301: based on the clustered enterprise financial data, obtain the data completeness CWZ on the financial statement: calculate the ratio of the actual data items on the financial statement to the expected data items on the financial statement, and use the ratio as the quantitative data completeness CWZ on the financial statement;

[0117] Step S302: based on the clustered enterprise financial data, obtain the data integrity YWZ of the business data: calculate the ratio of the actual data items in the business data to the expected data items in the business data, and use the ratio as the data integrity YWZ of the business data;

[0118] Step S303: based on the clustered enterprise financial data, obtain the data completeness SWZ of the tax data: calculate the ratio of the actual data items on the tax data to the expected data items on the tax data, and use the ratio as the data completeness SWZ of the tax data;

[0119] Step S304: Substitute CWZ, YWZ and SWZ into the safety risk assessment index calculation formula to obtain the safety risk assessment index LMS, which is calculated as follows:

[0120] ;

[0121] in, , is the weight coefficient, which is set by the system;

[0122] Step S305: Record the security risk assessment index LMS as the assessment result.

[0123] When the safety risk assessment index LMS is greater than a preset safety risk assessment index threshold, a warning message is generated and fed back to the enterprise, wherein the preset safety risk assessment index threshold is stored in the system and set by the system.

[0124] The entire system of this application is an enterprise financial process management system based on big data. The technical problem to be solved is: in the traditional enterprise financial management model, it mainly relies on manual data processing, which leads to low efficiency, high cost and high risk of enterprise financial process management. After obtaining the financial data of the enterprise, the present application enters the financial data into the system. Among them, the financial data includes data such as tables and texts, and there will be some useless information data. The system of the present application first performs dimensionality reduction processing on the data. The matrix and vector are nouns in dimensionality reduction. For example, taking the financial statement as an example, each financial statement data is an element of the matrix X, which can be realized through a specific identifier. For example, for the data of paying a tax of 10,000 yuan, the present application can pre-set the amount as the main feature, convert it into an element in the matrix X through binary conversion, and then perform dimensionality reduction processing on the matrix X, including: the data of paying a tax of 10,000 yuan has n data samples of a d-dimensional attribute vector, and the d dimension includes a time dimension or a monthly dimension, and then perform averaging processing. The averaging processing is data normalization. After the averaging processing, the square matrix H, linear sum vector B, and square sum vector F of each local node (the local node data is the data corresponding to each enterprise, and each enterprise corresponds to a node. The system of the present application processes multiple enterprises or multiple data at the same time to increase the overall work efficiency) data are calculated. The calculation process is According to the technical solution of the present application, the present application involves mathematical matrix calculation, solving the eigenvalue of the correlation coefficient matrix R, using the system computer to automatically calculate according to the calculation formula in the present application, and then clustering the reduced-dimensional data. Clustering is to cluster the financial data of an enterprise together, especially useful information, such as information about a tax payment of 10,000 yuan, which belongs to a certain category, that is, financial information. Because the enterprise also includes other data, such as business transaction data, it can achieve rapid classification and improve data processing efficiency. Finally, the system of the present application also includes risk warning for the clustered data, checking whether the data has security risks, and obtaining the data integrity CWZ on the financial statements based on the clustered enterprise financial data: calculating the ratio of the actual data items on the financial statements to the data items that should be on the financial statements, and using the ratio as the quantitative data integrity CWZ on the financial statements; obtaining the data integrity YWZ on the business data based on the clustered enterprise financial data: calculating the ratio of the actual data items on the business data to the data items that should be on the business data, and using the ratio as the data integrity YWZ on the business data;

[0125] Based on the clustered enterprise financial data, the data completeness SWZ of the tax data is obtained: the ratio of the actual data items on the tax data to the data items on the tax data is calculated, and the ratio is used as the data completeness SWZ of the tax data; then the security risk assessment index LMS is obtained. If the security risk assessment index LMS is less than a value, such as 60, it means that the risk is very high and an alarm will be issued. After the enterprise's financial data is uploaded to the system of this application, if the uploaded data is incomplete, the system will issue an alarm without manually comparing them one by one and then issuing a message that it cannot be passed.

[0126] When new financial data is added to the system of the present application, the newly arrived data is obtained, and it is determined whether the newly arrived data can be absorbed by an existing core micro-cluster. If so, the data is added to the micro-cluster; if not, the local outlier factor of the newly arrived data is calculated to determine whether it exceeds a preset threshold. If so, the newly arrived data is abnormal data and is added to the buffer micro-cluster set for later observation; if not, the newly arrived data is normal data, and a new core micro-cluster is created for the newly arrived data and added to the core micro-cluster set; wherein the later observation includes determining whether the newly arrived data is seed data or outlier data based on the condition, and the condition is whether the newly arrived data attracts a preset number of data to form a new cluster. If so, the newly arrived data is seed data, and if not, the newly arrived data is outlier data. The data is defined as useless outlier data.

[0127] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0128] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0129] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0130] In the several embodiments provided in 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 only schematic. For example, the division of the units is only some logical function divisions. There may be other division methods in actual implementation, such as 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 mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0131] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0132] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An enterprise financial process management system based on big data, characterized in that: The system includes a financial data acquisition module, a financial data dimensionality reduction processing module, a financial data clustering processing module and a financial data security risk assessment module; The financial data acquisition module is used to acquire enterprise financial data, wherein the enterprise financial data includes financial statements, business data and tax data; The financial data dimensionality reduction processing module is used to perform dimensionality reduction processing on corporate financial data: construct a sample matrix X based on corporate financial data, perform averaging operation on matrix X, and construct an average matrix M; distribute the averaged data evenly to each local node, and construct the local matrix M of each node i , calculate the three variable information of the square matrix H, linear sum vector B, and square sum vector F of each local node data; calculate the correlation coefficient of each local node data based on the three variable information, construct the correlation coefficient matrix R based on the correlation coefficient, solve the eigenvalue λ of the correlation coefficient matrix R, arrange the eigenvalue λ in descending order, and determine the first k eigenvalues ​​λ1, λ2...λ by the preset contribution rate k , based on the k eigenvalues, calculate the corresponding eigenvectors v1, v2...v k , k eigenvectors are used to construct a projection matrix, and the sample matrix X is mapped onto the projection matrix to obtain the sample set after dimensionality reduction; The financial data clustering processing module is used to perform clustering processing on the enterprise financial data and complete the clustering of the newly arrived data, wherein the newly arrived data is the latest added enterprise financial data; The financial data security risk assessment module is used to conduct security risk assessment based on the clustered enterprise financial data, obtain the assessment results, and issue early warnings to the enterprise based on the assessment results; The calculation of the three variable information of the square matrix H, linear sum vector B, and square sum vector F of each local node data specifically includes the following process: Construct a formula based on a square matrix: Then the square matrix H is: Compute the linear sum of the matrix M and the vector B: B=[b 1 , b 2 ...b d ],in, Compute the square of the matrix M and the vector F: F=[f 1 , f 2 ...f d ],in, Calculating the correlation coefficient of each local node data based on the three variable information and constructing the correlation coefficient matrix R based on the correlation coefficient specifically includes the following process: Calculate the correlation coefficient of each local node data based on three variable information The calculation formula is as follows: Based on the correlation coefficient Construct the correlation coefficient matrix R: Based on the clustered enterprise financial data, security risk assessment is performed to obtain the assessment results, which specifically includes the following process: Based on the clustered enterprise financial data, the data completeness CWZ on the financial statements is obtained: the ratio of the actual data items on the financial statements to the data items that should be on the financial statements is calculated, and the ratio is used as the quantitative data completeness CWZ on the financial statements; Based on the clustered enterprise financial data, the data completeness YWZ of the business data is obtained: the ratio of the actual data items on the business data to the data items on the business data should be calculated, and the ratio is used as the data completeness YWZ of the business data; Based on the clustered enterprise financial data, the data completeness SWZ of the tax data is obtained: the ratio of the actual data items on the tax data to the data items that should be on the tax data is calculated, and the ratio is used as the data completeness SWZ of the tax data; Substituting CWZ, YWZ and SWZ into the safety risk assessment index calculation formula, the safety risk assessment index LMS is obtained. The calculation formula is as follows: Among them, α, β, and θ are weight coefficients, which are set by the system; The safety risk assessment index LMS is recorded as the assessment result.

2. According to the big data-based enterprise financial process management system of claim 1, it is characterized in that: Constructing a sample matrix X based on corporate financial data, performing a mean operation on the matrix X, and constructing a mean matrix M specifically includes the following processes: Based on the enterprise financial data, n data samples with d-dimensional attribute vectors are obtained and expressed as X = {x1, x2...x n }, and each sample is represented as Then the sample matrix X is expressed as follows: The sample matrix X is averaged as follows: Where, i = 1, 2...n, j = 1, 2...d; The resulting mean matrix M is:

3. The enterprise financial process management system based on big data according to claim 1 is characterized in that: Solve the eigenvalue λ of the correlation coefficient matrix R, arrange the eigenvalue λ in descending order, and determine the first k eigenvalues ​​λ1, λ2...λ by the preset contribution rate k , based on the k eigenvalues, calculate the corresponding eigenvectors v1, v2...v k , k eigenvectors to construct the projection matrix specifically include the following process: Use the contribution rate δ to determine the number of principal components: The first k eigenvalues ​​λ1, λ2, ...λ k The corresponding eigenvectors v1, v2...v k are the k eigenvectors determined, where each eigenvector k eigenvectors construct the projection matrix V:

4. The enterprise financial process management system based on big data according to claim 1 is characterized in that: Clustering of enterprise financial data to complete the clustering of new data includes the following processes: Using the k-means algorithm to cluster the initially read preset financial data to obtain a number of initial micro-clusters, and combining the initial micro-clusters into an initial core micro-cluster set, the initial core micro-cluster set including a number of core micro-clusters; Get the newly arrived data and determine whether the newly arrived data can be absorbed by an existing core micro-cluster. If so, add the data to this micro-cluster. If not, calculate the new data x p The local outlier factor LDOF is: Among them, d p is the distance to the nearest data z, and O represents the distance to the newly arrived data x p The set of the most recent z data, dist(x i , x p ) represents the core micro-cluster center x i With the new data x p Euclidean distance; Among them, D p is the internal distance between z's nearest neighbor data; Determine whether LDOF exceeds the preset threshold. If so, the new data x p If it is abnormal data, add it to the buffer micro-cluster set for later observation; if not, the new data x p is normal data, then it is the new data x p Create a new core micro-cluster and add it to the core micro-cluster collection; Among them, the later observation includes judging whether the newly arrived data is seed data or outlier data based on the condition. The condition is whether the newly arrived data attracts a preset number of data to form a new cluster. If so, the newly arrived data is seed data, and if not, the newly arrived data is outlier data.

5. The enterprise financial process management system based on big data according to claim 1 is characterized in that: When the safety risk assessment index LMS is greater than a preset safety risk assessment index threshold, a warning message is generated and fed back to the enterprise, wherein the preset safety risk assessment index threshold is stored in the system and set by the system.

Images