A comprehensive financial audit system based on big data

Abstract

The present invention discloses a comprehensive financial audit system based on big data, which relates to the technical field of financial auditing and includes a data collection module, a financial audit module, and a computing power regulation module. The data collection module is used to continuously collect financial data uploaded by the audit object and its related business pre-event information, and perform classification and time feature analysis. The financial audit module performs automatic audit tasks based on the uploaded financial data, and identifies its periodic or event-triggered characteristics in combination with the data type. The computing power regulation module is used to dynamically adjust the allocation strategy of computing resources according to the financial data upload mode and business event prediction results, so as to achieve efficient utilization and intelligent preheating of computing power. The data collection module includes a data upload record module, a data classification module, an interface authority acquisition module, a business event monitoring module, and a data storage module. The present invention has the characteristics of prediction rules.

Description

Technical Field

[0001] The present invention relates to the technical field of financial auditing, and in particular to a comprehensive financial auditing system based on big data. Background Art

[0002] A comprehensive financial audit system refers to an integrated audit platform that integrates multiple audit objects, technical modules, and data sources. It utilizes an audit algorithm model based on big data and AI, with the goal of achieving automated, intelligent, and closed-loop audit management from data access, rule checking, risk identification to result feedback.

[0003] In existing technology, audit subjects often upload large quantities of financial data to the audit system at fixed times, such as at the end of each month, the end of each quarter, and during tax filing periods. This places an excessive burden on the system's computing servers during these times. To cope with these peaks and troughs in computing power demand, powerful servers are required, which is prohibitively expensive. While uploading financial data immediately after it is generated will avoid significant peaks and troughs in computing power demand, the unpredictable patterns of financial data generation will result in redundant computing power. Therefore, it is necessary to design a comprehensive financial audit system based on big data that can predict these patterns. Summary of the Invention

[0004] The purpose of the present invention is to provide a comprehensive financial audit system based on big data to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: a comprehensive financial audit system based on big data, including a data collection module, a financial audit module, and a computing power regulation module. The data collection module is used to continuously collect financial data uploaded by the audit object and its related business pre-event information, and perform classification and time feature analysis. The financial audit module performs automatic audit tasks based on the uploaded financial data, and identifies its periodic or event-triggered characteristics in combination with the data type. The computing power regulation module is used to dynamically adjust the allocation strategy of computing resources according to the financial data upload pattern and business event prediction results, so as to achieve efficient utilization and intelligent preheating of computing power.

[0006] According to the above technical solution, the data collection module includes a data upload recording module, a data classification module, an interface authority acquisition module, a business event monitoring module, and a data storage module. The data upload recording module is electrically connected to the data classification module, and the interface authority acquisition module is electrically connected to the business event monitoring module. The data upload recording module is used to record the upload time of the financial data of each audit object, and the data classification module is used to divide the collected financial data into two categories: fixed periodicity and non-fixed periodicity. The interface authority acquisition module is used to apply for and manage access rights to the business system of each audit object, call relevant interfaces to obtain financial data and business pre-event information, and the business event monitoring module is used to record the business pre-events that trigger the upload of financial data in real time. The data storage module is used to store financial data and pre-business data;

[0007] The financial audit module includes an upload threshold judgment module, a cycle pattern recognition module, a delay probability algorithm model, and an automatic audit module. The upload threshold judgment module is electrically connected to the cycle pattern recognition module and the business event monitoring module, the data upload recording module is electrically connected to the automatic audit module, and the delay probability algorithm model is electrically connected to the business event monitoring module. The upload threshold judgment module is used to judge whether the current amount of uploaded financial data reaches the batch threshold. The cycle pattern recognition module is used to identify the cycle of financial data upload of each audit object. The delay probability algorithm model is used to construct a delay probability model of business pre-events and data capacity that are highly correlated with the generation of batch financial data. The automatic audit module audits financial data using an audit algorithm based on big data and AI;

[0008] The computing power regulation module includes a peak prediction module, a computing resource division module, a server computing power allocation module, and a delay trigger module. The peak prediction module and the delay trigger module are electrically connected to the computing resource division module. The computing resource division module is electrically connected to the server computing power allocation module. The peak prediction module is used to predict the peaks generated by the comprehensive financial data of each audit object. The computing resource division module is used to divide the computing resources of the computing server. The server computing power allocation module is used to allocate redundant server computing power. The delay trigger module is used to delay the allocation of server computing power after a business pre-event that is highly relevant to the generation of batch financial data occurs.

[0009] According to the above technical solution, the working method of the system is:

[0010] S1. Divide the computing server's resources and match each divided resource with each audit object. A single data processing unit corresponds to multiple audit objects. When each audit object uploads financial data, the upload time and data volume of each audit object are collected. Access rights are requested from the audit object's business system, and business pre-event information highly relevant to the financial data upload is collected in real time.

[0011] S2. Determine whether the financial data capacity has reached the batch financial data threshold, perform time series analysis on fixed-periodic data, extract the periodic characteristics of data upload, and predict the peak financial data upload time for each audit object in a specific future time period;

[0012] S3. Based on the collected business pre-event information, a delay probability model is constructed between the event and the generation of batch financial data to predict the timing of uploading non-fixed periodic data. The prediction result is provided to the delay trigger module as a trigger signal.

[0013] S4. Pre-allocate computing power to computing servers based on predicted periodic peaks and non-periodic batch financial data upload events, using more computing resources for processing financial data and allocating redundant computing resources to other data processing units during non-peak periods.

[0014] S5. Predict the total financial data capacity received by the subsequent computing server and allocate redundant computing resources to other entities that require computing power of the data processing unit.

[0015] According to the above technical solution, S2 specifically includes:

[0016] S2-1, let the statistical period be ,exist An audit object uploaded Financial data, in The data capacity for uploading financial data is , the uploaded time series is , let the batch financial data threshold be , when the data capacity of a certain upload When the upload is complete, it is recorded as the upload of batch financial data;

[0017] S2-2, analyze the periodicity of the upload behavior of batch financial data. When the data capacity of multiple uploads is close, that is, the data capacity error is Within, and in When the data exhibits periodic characteristics, it is determined to be fixed-periodic financial data, and the subsequent upload time points and data capacity of all financial data exhibiting periodic characteristics are predicted.

[0018] According to the above technical solution, in S3, the method for constructing the delay probability model is:

[0019] S3-1. Filter all batch financial data, exclude the financial data with periodic characteristics, and record the remaining financial data with non-periodic characteristics. Let the time series of a business pre-event be , and the data capacity error between them is Financial data within constitutes a data capacity interval and belongs to a certain data capacity The time series of financial data upload with non-periodic characteristics of interval is ;

[0020] S3-2, determine the correlation between the two sequences based on the time difference between the two sequences, as long as the time interval after the business pre-event occurs Within a certain non-periodic feature of financial data upload time series, we can find a one-to-one corresponding time point, that is, to judge whether this business pre-event will trigger the subsequent The data capacity in the time is The financial data upload behavior in the specified interval is used to find all batch financial data corresponding to all business pre-event information, and to predict the capacity and time of subsequent corresponding batch financial data uploads.

[0021] According to the above technical solution, in S4, the specific method for pre-allocating the computing power of the computing server is:

[0022] S4-1, let the total computing power of the computing server be ,common Data processing units, the computing power that a single data processing unit can carry is , a data processing unit receives data with a capacity of When the financial data is collected, the computing power consumed by the AI ​​audit algorithm model increases rapidly with the data capacity. The computing power required to maintain the normal financial audit process is , ,in is the conversion factor, is the growth coefficient;

[0023] S4-2, each data processing unit corresponds to Audit objects, follow-up If the current data processing unit does not predict this If the batch financial data of audit objects is uploaded, the data processing unit retains the minimum computing power. , The minimum computing power coefficient is used for itself, and the remaining computing power is used to undertake financial data from other data processing units;

[0024] S4-3, when the follow-up When the current data processing unit predicts that financial data will be uploaded within the time period, the total data capacity will be calculated based on the current forecast. , reduce the financial data taken over from other data processing units, and distribute part of the financial data of the corresponding audit objects to other data processing units.

[0025] According to the above technical solution, in S4-3, the specific allocation method is: the computing power currently allocated by this data processing unit to other data processing units ,when When the calculation result is less than 0, the financial data of other data processing units will not be accepted. When the predicted time point is reached, the financial data of the data capacity that cannot be accepted will be uploaded to the The smallest other data processing unit.

[0026] According to the above technical solution, in S5, the redundant computing resources are allocated to other entities that need the computing power of the data processing unit. Specifically, it is predicted that the computing server will The total amount of data received during this time is ,in The number of audit objects, assigned The computing resources of computing power are provided to other entities that need computing power of data processing units. To allocate the coefficient for redundancy, it is evenly distributed to each data processing unit. The computing power allocated to other entities needs to be deducted in advance in S4-3.

[0027] Compared with the existing technology, the present invention has the following beneficial effects: when each audit object uploads financial data in normal times, the present invention collects statistics on the upload time of the financial data, and after big data analysis, divides the financial data into two situations: fixed-cycle generation and irregular-cycle generation. After summarizing the regularity of the batch financial data generated in the fixed cycle, the comprehensive financial data of each audit object is calculated to generate peaks, and the future peak time periods are predicted. During these time periods, the computing server is allocated more computing power, and the redundant computing power is allocated to other systems at other times, so that a large amount of computing power will not be idle.

[0028] For batch financial data generated at irregular intervals, we apply to the audit object to monitor business pre-events that are highly relevant to the generation of batch financial data, build a delay probability model for events and data, and provide it to the system backend as an input signal. Based on these event signals, we dynamically adjust computing power resources and implement intelligent pre-adjustment before the peak of financial data upload. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0030] Figure 1 It is a schematic diagram of the overall module structure of the present invention. DETAILED DESCRIPTION

[0031] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0032] See also Figure 1 The present invention provides a technical solution: a comprehensive financial audit system based on big data, including a data collection module, a financial audit module, and a computing power control module. The data collection module is used to continuously collect financial data uploaded by the audit object and its related business pre-event information, and perform classification and time feature analysis. The financial audit module performs automatic audit tasks based on the uploaded financial data, and identifies its periodic or event-triggered characteristics based on the data type. The computing power control module is used to dynamically adjust the allocation strategy of computing resources according to the financial data upload pattern and business event prediction results, thereby achieving efficient utilization and intelligent preheating of computing power.

[0033] The data collection module includes a data upload recording module, a data classification module, an interface authority acquisition module, a business event monitoring module, and a data storage module. The data upload recording module is electrically connected to the data classification module, and the interface authority acquisition module is electrically connected to the business event monitoring module. The data upload recording module is used to record the upload time of the financial data of each audit object. The data classification module is used to divide the collected financial data into two categories: fixed periodicity and non-fixed periodicity. The interface authority acquisition module is used to apply for and manage access rights to the business system of each audit object, call relevant interfaces to obtain financial data and business pre-event information, the business event monitoring module is used to record the business pre-events that trigger the upload of financial data in real time, and the data storage module is used to store financial data and pre-business data;

[0034] The financial audit module includes an upload threshold judgment module, a cycle pattern recognition module, a delay probability algorithm model, and an automatic audit module. The upload threshold judgment module is electrically connected to the cycle pattern recognition module and the business event monitoring module, the data upload record module is electrically connected to the automatic audit module, and the delay probability algorithm model is electrically connected to the business event monitoring module. The upload threshold judgment module is used to judge whether the current amount of uploaded financial data reaches the batch threshold. The cycle pattern recognition module is used to identify the cycle of financial data upload of each audit object. The delay probability algorithm model is used to construct a delay probability model of business pre-events and data capacity that are highly correlated with the generation of batch financial data. The automatic audit module uses an audit algorithm based on big data and AI to audit financial data;

[0035] The computing power control module includes a peak prediction module, a computing resource division module, a server computing power allocation module, and a delay trigger module. The peak prediction module and the delay trigger module are electrically connected to the computing resource division module, and the computing resource division module is electrically connected to the server computing power allocation module. The peak prediction module is used to predict the peaks generated by the comprehensive financial data of each audit object. The computing resource division module is used to divide the computing resources of the computing server. The server computing power allocation module is used to allocate the redundant server computing power. The delay trigger module is used to delay the allocation of server computing power after a business pre-event that is highly relevant to the generation of batch financial data occurs.

[0036] The system works as follows:

[0037] S1. Divide the computing server's resources and match each divided resource with each audit object. A single data processing unit corresponds to multiple audit objects. When each audit object uploads financial data, the upload time and data volume of each audit object are collected. Access rights are requested from the audit object's business system, and business pre-event information highly relevant to the financial data upload is collected in real time.

[0038] S2. Determine whether the financial data capacity has reached the batch financial data threshold, perform time series analysis on fixed-periodic data, extract the periodic characteristics of data upload, and predict the peak financial data upload time for each audit object in a specific future time period;

[0039] S3. Based on the collected business pre-event information, a delay probability model is constructed between the event and the generation of batch financial data to predict the timing of uploading non-fixed periodic data. The prediction result is provided to the delay trigger module as a trigger signal.

[0040] Pre-business events refer to highly relevant, system-monitored business operations or management behaviors that precede the bulk upload of financial data. These typically fall into the following categories: settlement events (such as project settlement, supplier reconciliation, and sales commission settlement), reimbursement events (such as centralized travel reimbursement and monthly expense reimbursement summary), tax events (such as VAT invoice authentication and tax return initiation), approval events (such as financial review approval and budget approval completion), system operation events (such as financial system export, bulk voucher generation, and period-end closing operations), and business milestone events (such as project completion, contract fulfillment, and large-value purchase receipt). These events often serve as the precursors to the bulk generation and bulk upload of financial data and can serve as signal inputs for delayed probability models.

[0041] S4. Pre-allocate computing power to computing servers based on predicted periodic peaks and non-periodic batch financial data upload events, using more computing resources for processing financial data and allocating redundant computing resources to other data processing units during non-peak periods.

[0042] S5. Predict the total financial data capacity received by the subsequent computing server and allocate redundant computing resources to other entities that require computing power from the data processing unit;

[0043] S2 specifically includes:

[0044] S2-1, let the statistical period be ,exist An audit object uploaded Financial data, in The data capacity for uploading financial data is , the uploaded time series is , let the batch financial data threshold be , when the data capacity of a certain upload When the upload is complete, it is recorded as the upload of batch financial data;

[0045] S2-2, analyze the periodicity of the upload behavior of batch financial data. When the data capacity of multiple uploads is close, that is, the data capacity error is Within, and in When the data exhibits periodic characteristics, it is determined to be fixed periodic financial data, and the subsequent upload time and data capacity of all financial data exhibiting periodic characteristics are predicted;

[0046] In S3, the delay probability model is constructed as follows:

[0047] S3-1. Filter all batch financial data, exclude the financial data with periodic characteristics, and record the remaining financial data with non-periodic characteristics. Let the time series of a business pre-event be , and the data capacity error between them is Financial data within constitutes a data capacity interval and belongs to a certain data capacity The time series of financial data upload with non-periodic characteristics of interval is ;

[0048] S3-2, determine the correlation between the two sequences based on the time difference between the two sequences, as long as the time interval after the business pre-event occurs Within a certain non-periodic feature of financial data upload time series, we can find a one-to-one corresponding time point, that is, to judge whether this business pre-event will trigger the subsequent The data capacity in the time is The financial data upload behavior in the interval is used to find all batch financial data corresponding to all business precursor event information, and predict the capacity and time of subsequent corresponding batch financial data uploads; this method searches for corresponding patterns between the uploaded financial data and the business precursor events. Compared with directly matching the predecessor business time with the financial data, which will directly cause the direct leakage of the one-to-one correspondence between business matters and financial data, it better protects the privacy of the audit subject and only establishes a statistical-level probability model for the event category and data upload pattern. It does not track the specific content of the financial data, does not store it, and is not bound to specific original business details or uploaded content.

[0049] By predicting the peaks of periodic and non-periodic financial data uploads, computing servers can prepare and allocate computing power in advance to avoid an influx of excessive financial data within a certain period of time.

[0050] In S4, the specific method for pre-allocating the computing power of the computing server is as follows:

[0051] S4-1, let the total computing power of the computing server be ,common Data processing units, the computing power that a single data processing unit can carry is , a data processing unit receives data with a capacity of When the financial data is collected, the computing power consumed by the AI ​​audit algorithm model increases rapidly with the data capacity. The computing power required to maintain the normal financial audit process is , ,in is the conversion factor, is the growth coefficient;

[0052] S4-2, each data processing unit corresponds to Audit objects, follow-up If the current data processing unit does not predict this If the batch financial data of audit objects is uploaded, the data processing unit retains the minimum computing power. , The minimum computing power coefficient is used for itself, and the remaining computing power is used to undertake financial data of other data processing units;

[0053] S4-3, when the follow-up When the current data processing unit predicts that financial data will be uploaded within the time period, the total data capacity will be calculated based on the current forecast. , reduce the financial data taken over from other data processing units, and distribute part of the financial data of the corresponding audit objects to other data processing units;

[0054] In S4-3, the specific allocation method is: the computing power currently allocated by this data processing unit to other data processing units ,when When the calculation result is less than 0, the financial data of other data processing units will not be accepted. When the predicted time point is reached, the financial data of the data capacity that cannot be accepted will be uploaded to the The smallest other data processing unit;

[0055] In S5, the redundant computing resources are allocated to other entities that need the computing power of the data processing unit. Specifically, it is predicted that the computing server will The total amount of data received during this time is ,in The number of audit objects, assigned The computing resources of computing power are provided to other entities that need computing power of data processing units. To allocate the coefficient for redundancy, it is evenly distributed to each data processing unit. The computing power allocated to other entities needs to be deducted in advance in S4-3.

[0056] The present invention collects statistics on the upload time of financial data when each audit object uploads it. After big data analysis, the financial data is divided into two types: those generated in fixed periods and those generated in irregular periods. After summarizing the regularity of the batch financial data generated in fixed periods, the present invention calculates the peaks of the comprehensive financial data of each audit object and predicts the future peak time periods. During these time periods, the computing server is allocated more computing power, and the redundant computing power is allocated to other systems at other times, so that a large amount of computing power will not be idle.

[0057] For batch financial data generated at irregular intervals, we apply to the audit object to monitor business pre-events that are highly relevant to the generation of batch financial data, build a delay probability model for events and data, and provide it to the system backend as an input signal. Based on these event signals, we dynamically adjust computing power resources and implement intelligent pre-adjustment before the peak of financial data upload.

[0058] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0059] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A comprehensive financial audit system based on big data, characterized by: It includes a data collection module, a financial audit module, and a computing power regulation module. The data collection module is used to continuously collect financial data uploaded by the audit object and its related business pre-event information, and perform classification and time feature analysis. The financial audit module performs automatic audit tasks based on the uploaded financial data and identifies its periodic or event-triggered characteristics based on the data type. The computing power regulation module is used to dynamically adjust the allocation strategy of computing resources based on the financial data upload pattern and business event prediction results, thereby achieving efficient utilization and intelligent preheating of computing power. The data collection module includes a data upload recording module, a data classification module, an interface authority acquisition module, a business event monitoring module, and a data storage module. The data upload recording module is electrically connected to the data classification module, and the interface authority acquisition module is electrically connected to the business event monitoring module. The data upload recording module is used to record the upload time of the financial data of each audit object. The data classification module is used to divide the collected financial data into two categories: fixed periodicity and non-fixed periodicity. The interface authority acquisition module is used to apply for and manage access rights to the business system of each audit object, call relevant interfaces to obtain financial data and business pre-event information, the business event monitoring module is used to record the business pre-events that trigger the upload of financial data in real time, and the data storage module is used to store financial data and pre-business data; The financial audit module includes an upload threshold judgment module, a cycle pattern recognition module, a delay probability algorithm model, and an automatic audit module. The upload threshold judgment module is electrically connected to the cycle pattern recognition module and the business event monitoring module, the data upload recording module is electrically connected to the automatic audit module, and the delay probability algorithm model is electrically connected to the business event monitoring module. The upload threshold judgment module is used to judge whether the current amount of uploaded financial data reaches the batch threshold. The cycle pattern recognition module is used to identify the cycle of financial data upload of each audit object. The delay probability algorithm model is used to construct a delay probability model of business pre-events and data capacity that are highly correlated with the generation of batch financial data. The automatic audit module audits financial data using an audit algorithm based on big data and AI; The computing power control module includes a peak prediction module, a computing resource division module, a server computing power allocation module, and a delay trigger module. The peak prediction module and the delay trigger module are electrically connected to the computing resource division module, and the computing resource division module is electrically connected to the server computing power allocation module. The peak prediction module is used to predict the peak of the comprehensive financial data of each audit object. The computing resource division module is used to divide the computing resources of the computing server. The server computing power allocation module is used to allocate redundant server computing power. The delay trigger module is used to delay the allocation of server computing power after a business pre-event that is highly relevant to the generation of batch financial data occurs; The system works as follows: S1. Divide the computing server's resources and match each divided resource with each audit object. When each audit object uploads financial data, collect the upload time and volume of the financial data, apply for access rights from the audit object's business system, and collect business event information that is highly relevant to the financial data upload in real time. S2. Determine whether the financial data capacity has reached the batch financial data threshold, perform time series analysis on fixed-periodic data, extract the periodic characteristics of data upload, and predict the peak financial data upload time for each audit object in a specific future time period; S3. Based on the collected business pre-event information, a delay probability model is constructed between the event and the generation of batch financial data to predict the timing of uploading non-fixed periodic data. The prediction result is provided to the delay trigger module as a trigger signal. S4. Pre-allocate computing power to the computing servers based on predicted periodic peaks and non-periodic batch financial data upload events, and allocate redundant computing resources to other data processing units during non-peak periods. S5. Predict the total financial data capacity received by the subsequent computing server and allocate redundant computing resources to other entities that require computing power from the data processing unit; In S3, the delay probability model is constructed as follows: S3-1. Record the remaining non-periodic financial data and let the time series of a business antecedent event be , and the data capacity error between them is Financial data within a data capacity interval constitutes a data capacity interval and belongs to a certain data capacity The time series of financial data upload with non-periodic characteristics of interval is ; S3-2, determine the correlation between the two sequences based on the time difference between the two sequences, as long as the time interval after the business pre-event occurs Within a certain non-periodic feature of financial data upload time series, we can find a one-to-one corresponding time point, that is, to judge whether this business pre-event will trigger the subsequent The data capacity in the time is Financial data upload behavior in the current interval, find all batches of financial data corresponding to all business pre-event information, and predict the capacity and time of subsequent corresponding batch financial data uploads; In S4, the specific method for allocating the computing power of the computing server in advance is: S4-1, let the total computing power of the computing server be ,common Data processing units, the computing power that a single data processing unit can carry is , a data processing unit receives data with a capacity of When the financial data is collected, the computing power required to maintain the normal financial audit process is , ,in is the conversion factor, is the growth coefficient; S4-2, each data processing unit corresponds to Audit objects, follow-up If the current data processing unit does not predict this If the batch financial data of audit objects is uploaded, the data processing unit retains the minimum computing power. , The minimum computing power coefficient is used for itself, and the remaining computing power is used to undertake financial data of other data processing units; S4-3, when the follow-up When the current data processing unit predicts that financial data will be uploaded within the time period, the total data capacity will be calculated based on the current forecast. , reduce the financial data that needs to be taken over from other data processing units; The business pre-events are highly relevant business operations and management behaviors that can be monitored before the batch uploading of financial data, including settlement events, reimbursement events, tax events, approval events, system operation events and business node events.

2. The comprehensive financial audit system based on big data according to claim 1, characterized in that: Said S2 specifically includes: S2-1, let the statistical period be ,exist An audit object uploaded Financial data, in The data capacity for uploading financial data is , the uploaded time series is , let the batch financial data threshold be , when the data capacity of a certain upload When the upload is complete, it is recorded as the upload of batch financial data; S2-2, analyze the periodicity of the upload behavior of batch financial data. When the data capacity of multiple uploads is close, that is, the data capacity error is Within, and in When the data exhibits periodic characteristics, it is determined to be fixed-periodic financial data, and the subsequent upload time points and data capacity of all financial data exhibiting periodic characteristics are predicted.

3. The comprehensive financial audit system based on big data according to claim 2, characterized in that: In S4-3, the specific allocation method is: the computing power currently allocated by this data processing unit to other data processing units ,when When the calculation result is less than 0, the financial data of other data processing units will not be accepted. When the predicted time point is reached, the financial data of the data capacity that cannot be accepted will be uploaded to the The smallest other data processing unit.

4. The comprehensive financial audit system based on big data according to claim 3, characterized in that: In S5, the redundant computing resources are allocated to other entities that need the computing power of the data processing unit. Specifically, it is predicted that the computing server will The total amount of data received during this time is ,in The number of audit objects, assigned The computing resources of computing power are provided to other entities that need computing power of data processing units. is the redundancy allocation coefficient.

Images