AI algorithm-based agent bookkeeping intelligent auditing system and method

Abstract

The invention discloses an agency bookkeeping intelligent auditing system and method based on an AI algorithm, and belongs to the technical field of agency bookkeeping intelligent auditing. The agency bookkeeping intelligent auditing system based on the AI algorithm comprises a data acquisition module, a risk assessment module, a resource allocation module and a self-adaptive adjustment module; the data acquisition module is used for collecting enterprise financial historical data, obtaining a historical trend chart of financial indexes, analyzing a historical trend of financial data and determining a financial auditing period; the risk assessment module is used for assessing the risk of the financial data, calculating a deviation index of a financial index, and performing risk grade division on the financial data according to the calculated deviation index of the financial index; the resource allocation module is used for allocating auditing resources; according to the agency bookkeeping auditing system and method, the agency bookkeeping auditing process is analyzed through the AI algorithm, and the problem that the auditing resources are distributed unevenly is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent auditing of agency accounting, and specifically to an intelligent auditing system and method for agency accounting based on an AI algorithm. Background Art

[0002] As enterprises' requirements for refined financial management continue to increase, agency accounting services are becoming more and more widely used in various small and medium-sized enterprises. As a core link in financial management, the accuracy, efficiency and security of agency accounting directly affect the financial health of enterprises and the effectiveness of business decisions.

[0003] However, traditional agency accounting audit methods often face the following problems when dealing with complex financial data processing, multi-dimensional data analysis and security reviews: First, the audit efficiency is low, the data volume is huge, and manual processing is time-consuming and labor-intensive. Enterprises generate a large amount of financial vouchers, invoices and contract data in daily operations. Traditional agency accounting relies on manual audits. When faced with massive data, it is easy to have problems with untimely audits, especially during business peak periods. Manual audits are not only slow in processing, but also prone to omissions, resulting in a decline in the quality of financial data audits, which may seriously affect the company's financial decisions and compliance; secondly, the audit process lacks intelligent means and it is difficult to discover complex financial problems. Traditional financial data audits usually check financial vouchers and bills one by one, relying mainly on manual experience for judgment, and are unable to quickly identify potential financial risks or abnormal data. With the complexity of the company's financial structure and business model, financial data often contains cross-information in multiple dimensions. Traditional methods are unable to discover hidden risks from the multi-dimensional correlation of data. Especially when dealing with financial fraud and duplicate accounting issues, manual audit methods are limited in effectiveness and it is difficult to detect potential problems in a timely manner. In addition, financial data processing lacks global review capabilities and does not comprehensively analyze the changing trends and mutual influence of multiple financial dimensions such as income, expenditure, and assets and liabilities. In response to the above problems, although the existing agency accounting audit system has certain automation functions, it still has great limitations, especially when dealing with massive and multi-dimensional financial data. It lacks intelligent risk assessment and dynamic adjustment capabilities and cannot meet the growing intelligent needs of enterprises. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent audit system and method for agency accounting based on AI algorithm to solve the problems raised in the above background technology.

[0005] To achieve the above purpose, the present invention provides the following technical solution: an intelligent audit method for agency accounting based on AI algorithm, comprising the following steps:

[0006] S1: Collect historical financial data of the enterprise, obtain historical trend charts of financial indicators, analyze historical trends of financial data, and determine the financial audit cycle;

[0007] S2: Obtain the financial data of the enterprise, calculate the deviation of financial indicators, and classify the risk level of the financial data according to the deviation index of the calculated financial indicators;

[0008] S3: Obtain the historical audit records of each audit module, and evaluate the audit workload indicators of each audit module based on these historical audit data and corresponding financial data;

[0009] S4: Calculate the historical audit service quality score of each audit module based on the historical audit data of each audit module, and judge the rationality of audit resource allocation in combination with the audit workload indicator of each audit module;

[0010] S5: For audit modules with unreasonable resource allocation, reallocate audit resources and make dynamic adjustments based on real-time data.

[0011] In step S1: said step S1 comprises the following steps:

[0012] Step S1-1: Obtain historical data of financial indicators of the enterprise's income, expenditure and debt-to-asset ratio, sort the acquired historical data of the financial indicators in chronological order, generate continuous historical financial data points, and mark each financial data point in a plane rectangular coordinate system, connect each financial data point adjacently, and thus draw a historical trend chart of each financial indicator, where the horizontal axis represents the time series and the vertical axis represents the corresponding financial data value;

[0013] Step S1-2: In the historical trend chart, extract the continuous time intervals corresponding to the maximum and minimum values ​​of the financial data, calculate the fluctuation amplitude, and then sort the continuous time intervals from large to small according to the fluctuation amplitude, and select the continuous time interval with the smallest fluctuation amplitude as the selected audit period T, where the fluctuation amplitude represents the difference between the maximum and minimum values ​​of the extracted financial data within a specific time area.

[0014] In step S2: said step S2 comprises the following steps:

[0015] Step S2-1: Obtain historical data of the financial indicators of the enterprise within the selected audit period T, and calculate the average value U of each financial indicator;

[0016] Step S2-2: Calculate the deviation index R between each financial indicator and its average value i , the specific calculation formula is as follows:

[0017]

[0018] Among them, M i represents the actual value of the ith financial indicator, Ui represents the average value of the i-th indicator in the audit period T, MAX i represents the maximum value of the i-th financial indicator;

[0019] Step S2-3: Obtain a set of deviation indices R and normalize them: When the deviation index R i If it is greater than or equal to 0, then the deviation index R i Divide by the maximum positive value of the deviation index in the financial index set to calculate the updated deviation index. For the deviation index R i Less than 0, then the deviation index R i Divided by the minimum negative value of the deviation index in the financial index set, the normalized deviation index is R i ';

[0020] Step S2-4: According to the normalized deviation index Ri', the financial data is divided into high risk, medium risk and low risk, and further audited for data of different risk levels, wherein the range of the normalized deviation index Ri' for high risk level is greater than or equal to -1 and less than a; the range of the normalized deviation index Ri' for medium risk level is greater than or equal to a and less than b; the range of the normalized deviation index Ri' for low risk level is greater than or equal to b and less than 1; parameters a and b are determined by statistical analysis or industry standards;

[0021] Step S2-5: The risk level result is passed to the corresponding financial audit module, including the income audit, expenditure audit and asset audit modules. Each module prepares to audit the high-risk, medium-risk and low-risk financial data according to the passed risk level.

[0022] In step S3: said step S3 comprises the following steps:

[0023] Step S3-1: For each financial audit module, obtain the corresponding audit history data within the selected period T. Each financial audit module includes income audit, expenditure audit and asset and liability audit. Obtain the historical audit data of each module within the selected period T, including audit duration, audit complexity and audit task volume information;

[0024] Step S3-2: Combine the audit task volume and complexity in the historical audit data and the fluctuation of financial data to calculate the workload index Q of each audit module. The specific calculation formula is as follows:

[0025]

[0026] Where k represents the total number of tasks to be reviewed within the selected period T; W jThe workload of the jth audit task includes the number of audited documents and the amount of data processed; S represents the standard workload of the audit task; C j The complexity of the j-th audit task includes the types of financial data and the complexity of the vouchers involved in the audit task; A represents the complexity of the standard audit task; α and β represent the influencing factors of workload and complexity respectively, and their weight values ​​are determined based on historical data or industry standards.

[0027] In step S4: said step S4 comprises the following steps:

[0028] Step S4-1: Calculate the historical audit service quality score F of the audit module based on the audit efficiency and audit duration in the historical audit data of each financial audit module. The specific calculation formula is as follows:

[0029]

[0030] Where C represents the number of audit tasks completed by the financial audit module within the selected period T; V represents the processing speed of audit tasks; t represents the total time spent by the module audit task processing all tasks within the selected period T;

[0031] Step S4-2: Combine the audit task volume index Q of the financial audit module and the actual data volume s of the financial audit module to calculate the judgment index P to evaluate whether the audit resource allocation is reasonable. The specific calculation formula is as follows:

[0032]

[0033] Among them, Q represents the workload indicator of the financial audit module; s represents the actual data volume of the financial audit module; F j represents the historical service quality score of the jth auditor; d is the correction coefficient, which is used to balance the relationship between the actual module data volume s and the workload indicator Q in the audit resource allocation evaluation;

[0034] Step S4-3: Compare the calculated judgment index P with the judgment threshold P0 to determine whether the audit resource allocation is reasonable: when the judgment index P is equal to the judgment threshold P0, it means that the audit resource allocation of the audit module is reasonable; when the judgment index P is not equal to the judgment threshold P0, it means that there is a problem with the resource allocation of the audit module, and the audit module resource allocation problem is sent to the abnormal monitoring unit for further processing by the abnormal monitoring unit.

[0035] In step S5: said step S5 comprises the following steps:

[0036] Step S5-1: The abnormal monitoring unit obtains the number of the financial audit module whose judgment result is unreasonable audit resource allocation and the corresponding judgment index P, and divides the financial audit module number according to the size relationship between the judgment index P and the judgment threshold P0: the numbers of the financial audit modules whose judgment index P is greater than the judgment threshold P0 are aggregated and marked as modules with excessive audit resource allocation; the numbers of the modules whose judgment index P is less than the judgment threshold P0 are aggregated and marked as modules with insufficient audit resource allocation;

[0037] Step S5-2: For each financial audit module where audit resources are over-allocated, reduce the computing resources or audit personnel's work data allocated to the audit task, and recalculate the judgment index P ’ , traverse all audit task resources until the judgment index P' is equal to the judgment threshold P0, and record the deleted audit task or resource number in this process;

[0038] Step S5-3: For each financial audit module with insufficient audit resource allocation, increase the computing resources or auditor's work data allocated for the audit task, and recalculate the judgment index P ’ , traverse all audit task resources until the judgment indicator P ’ Equal to the judgment threshold P0, during which the deleted audit task or resource number recorded in step S5-2 is used for increase.

[0039] The invention discloses an intelligent audit system for agency accounting based on AI algorithm. The system includes: a data acquisition module, a risk assessment module, a resource allocation module and an adaptive adjustment module. The data acquisition module is used to collect historical financial data of the enterprise, obtain historical trend charts of financial indicators, analyze historical trends of financial data, and determine financial audit cycles; the risk assessment module is used to assess the risks of financial data, calculate the deviation index of financial indicators, and classify the financial data into risk levels according to the calculated deviation index of financial indicators; the resource allocation module is used to allocate audit resources; the adaptive adjustment module is used to adjust the allocation of audit resources based on real-time data and risk changes. The output end of the data acquisition module is connected to the input end of the risk assessment module, the output end of the risk assessment module is connected to the input end of the resource allocation module, and the output end of the resource allocation module is connected to the input end of the adaptive adjustment module.

[0040] The data acquisition module includes a financial data acquisition unit and a data preprocessing unit. The financial data acquisition unit is used to extract financial data, including invoices, bills and reports, from the financial system and documents of the enterprise. The data preprocessing unit deletes duplicate invoices, vouchers or contracts by comparing the unique identifier of the data; the risk assessment module includes a deviation index calculation unit and a risk level classification unit. The deviation index calculation unit is used to calculate the deviation index of the financial indicator based on the financial data of the enterprise. The risk level classification unit is used to classify the financial data into different risk levels according to the calculation result of the deviation index and a preset risk threshold.

[0041] The resource allocation module includes a workload assessment unit, a historical audit service quality scoring unit and an audit resource allocation judgment unit. The workload assessment unit is used to assess the workload of each financial audit module, and generate a workload indicator for each audit module in combination with historical audit records, current task volume and audit complexity; the historical audit service quality scoring unit is used to calculate the historical audit service quality score of the audit module based on the audit efficiency and audit duration in the historical audit data of each financial audit module; the audit resource allocation judgment unit is used to comprehensively consider the workload indicator and historical audit service quality score of the audit module, calculate a rationality judgment indicator for audit resource allocation, and compare the calculated judgment indicator with the judgment threshold to determine whether the audit resource allocation is reasonable.

[0042] The adaptive adjustment module includes an anomaly monitoring unit and an adjustment unit. The anomaly monitoring unit receives the judgment result of the audit resource allocation unit to identify the financial audit modules with over-use or insufficient resources. The adjustment unit is used to adjust the computing resources allocated to the audit task or the work data of the auditor by reducing the resources in the modules with over-use of resources and supplementing the modules with insufficient resources.

[0043] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0044] 1. By dynamically analyzing the changing trends of financial data, combining multiple financial dimensions such as income, expenditure, assets and liabilities, and calculating the deviation index, potential financial risks can be identified. Unlike the single indicator audit in the prior art, the present invention can capture the complex coupling relationship between financial data and realize multi-dimensional comprehensive risk assessment, thereby effectively improving the ability to identify financial risks.

[0045] 2. The present invention introduces an intelligent resource allocation module, which combines workload assessment and historical audit service quality scores to dynamically allocate audit resources and make adjustments based on real-time data. Unlike the static resource configuration in the prior art, the present invention can automatically adjust the allocation of auditors and computing resources based on changes in financial data and audit load to ensure efficient and timely completion of audit tasks. At the same time, the adaptive adjustment module can automatically optimize resource allocation based on real-time monitoring results to avoid waste or shortage of resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a structural diagram of the intelligent audit system for agency accounting based on AI algorithm of the present invention;

[0047] Figure 2 It is a flow chart of the intelligent audit method for agency accounting based on AI algorithm of the present invention. DETAILED DESCRIPTION

[0048] 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.

[0049] like Figure 1-Figure 2 As shown, the present invention provides a technical solution: an intelligent audit method for agency accounting based on AI algorithm, comprising the following steps:

[0050] S1: Collect historical financial data of the enterprise, obtain historical trend charts of financial indicators, analyze historical trends of financial data, and determine the financial audit cycle;

[0051] S2: Obtain the financial data of the enterprise, calculate the deviation of financial indicators, and classify the risk level of the financial data according to the deviation index of the calculated financial indicators;

[0052] S3: Obtain the historical audit records of each audit module, and evaluate the audit workload indicators of each audit module based on these historical audit data and corresponding financial data;

[0053] S4: Calculate the historical audit service quality score of each audit module based on the historical audit data of each audit module, and judge the rationality of audit resource allocation in combination with the audit workload indicator of each audit module;

[0054] S5: For audit modules with unreasonable resource allocation, reallocate audit resources and make dynamic adjustments based on real-time data.

[0055] In step S1: said step S1 comprises the following steps:

[0056] Step S1-1: Obtain historical data of financial indicators of the enterprise's income, expenditure and debt-to-asset ratio, sort the acquired historical data of the financial indicators in chronological order, generate continuous historical financial data points, and mark each financial data point in a plane rectangular coordinate system, connect each financial data point adjacently, and thus draw a historical trend chart of each financial indicator, where the horizontal axis represents the time series and the vertical axis represents the corresponding financial data value;

[0057] Step S1-2: In the historical trend chart, extract the continuous time intervals corresponding to the maximum and minimum values ​​of the financial data, calculate the fluctuation amplitude, and then sort the continuous time intervals from large to small according to the fluctuation amplitude, and select the continuous time interval with the smallest fluctuation amplitude as the selected audit period T, where the fluctuation amplitude represents the difference between the maximum and minimum values ​​of the extracted financial data within a specific time area.

[0058] In step S2: said step S2 comprises the following steps:

[0059] Step S2-1: Obtain historical data of the financial indicators of the enterprise within the selected audit period T, and calculate the average value U of each financial indicator;

[0060] Step S2-2: Calculate the deviation index R between each financial indicator and its average value i , the specific calculation formula is as follows:

[0061]

[0062] Among them, M i represents the actual value of the ith financial indicator, U i represents the average value of the i-th indicator in the audit period T, MAX i represents the maximum value of the i-th financial indicator;

[0063] Step S2-3: Obtain a set of deviation indices R and normalize them: When the deviation index R i If it is greater than or equal to 0, then the deviation index R i Divide by the maximum positive value of the deviation index in the financial index set to calculate the updated deviation index. For the deviation index R i Less than 0, then the deviation index R i Divided by the minimum negative value of the deviation index in the financial index set, the normalized deviation index is R i ';

[0064] Step S2-4: According to the normalized deviation index Ri', the financial data is divided into high risk, medium risk and low risk, and further audited for data of different risk levels, wherein the range of the normalized deviation index Ri' for high risk level is greater than or equal to -1 and less than a; the range of the normalized deviation index Ri' for medium risk level is greater than or equal to a and less than b; the range of the normalized deviation index Ri' for low risk level is greater than or equal to b and less than 1; parameters a and b are determined by statistical analysis or industry standards;

[0065] Step S2-5: The risk level result is passed to the corresponding financial audit module, including the income audit, expenditure audit and asset audit modules. Each module prepares to audit the high-risk, medium-risk and low-risk financial data according to the passed risk level.

[0066] In step S3: said step S3 comprises the following steps:

[0067] Step S3-1: For each financial audit module, obtain the corresponding audit history data within the selected period T. Each financial audit module includes income audit, expenditure audit and asset and liability audit. Obtain the historical audit data of each module within the selected period T, including audit duration, audit complexity and audit task volume information;

[0068] Step S3-2: Combine the audit task volume and complexity in the historical audit data and the fluctuation of financial data to calculate the workload index Q of each audit module. The specific calculation formula is as follows:

[0069]

[0070] Where k represents the total number of tasks to be reviewed within the selected period T; W j The workload of the jth audit task includes the number of audited documents and the amount of data processed; S represents the standard workload of the audit task; C j The complexity of the j-th audit task includes the types of financial data and the complexity of the vouchers involved in the audit task; A represents the complexity of the standard audit task; α and β represent the influencing factors of workload and complexity respectively, and their weight values ​​are determined based on historical data or industry standards.

[0071] In step S4: said step S4 comprises the following steps:

[0072] Step S4-1: Calculate the historical audit service quality score F of the audit module based on the audit efficiency and audit duration in the historical audit data of each financial audit module. The specific calculation formula is as follows:

[0073]

[0074] Where C represents the number of audit tasks completed by the financial audit module within the selected period T; V represents the processing speed of audit tasks; t represents the total time spent by the module audit task processing all tasks within the selected period T;

[0075] Step S4-2: Combine the audit task volume index Q of the financial audit module and the actual data volume s of the financial audit module to calculate the judgment index P to evaluate whether the audit resource allocation is reasonable. The specific calculation formula is as follows:

[0076]

[0077] Among them, Q represents the workload indicator of the financial audit module; s represents the actual data volume of the financial audit module; F j represents the historical service quality score of the jth auditor; d is the correction coefficient, which is used to balance the relationship between the actual module data volume s and the workload indicator Q in the audit resource allocation evaluation;

[0078] Step S4-3: Compare the calculated judgment index P with the judgment threshold P0 to determine whether the audit resource allocation is reasonable: when the judgment index P is equal to the judgment threshold P0, it means that the audit resource allocation of the audit module is reasonable; when the judgment index P is not equal to the judgment threshold P0, it means that there is a problem with the resource allocation of the audit module, and the audit module resource allocation problem is sent to the abnormal monitoring unit for further processing by the abnormal monitoring unit.

[0079] In step S5: said step S5 comprises the following steps:

[0080] Step S5-1: The abnormal monitoring unit obtains the number of the financial audit module whose judgment result is unreasonable audit resource allocation and the corresponding judgment index P, and divides the financial audit module number according to the size relationship between the judgment index P and the judgment threshold P0: the numbers of the financial audit modules whose judgment index P is greater than the judgment threshold P0 are aggregated and marked as modules with excessive audit resource allocation; the numbers of the modules whose judgment index P is less than the judgment threshold P0 are aggregated and marked as modules with insufficient audit resource allocation;

[0081] Step S5-2: For each financial audit module where audit resources are over-allocated, reduce the computing resources or audit personnel's work data allocated to the audit task, and recalculate the judgment index P ’ , traverse all audit task resources until the judgment index P' is equal to the judgment threshold P0, and record the deleted audit task or resource number in this process;

[0082] Step S5-3: For each financial audit module with insufficient audit resource allocation, increase the computing resources or auditor's work data allocated for the audit task, and recalculate the judgment index P’ , traverse all audit task resources until the judgment indicator P ’ Equal to the judgment threshold P0, during which the deleted audit task or resource number recorded in step S5-2 is used for increase.

[0083] The invention discloses an intelligent audit system for agency accounting based on AI algorithm. The system includes: a data acquisition module, a risk assessment module, a resource allocation module and an adaptive adjustment module. The data acquisition module is used to collect historical financial data of the enterprise, obtain historical trend charts of financial indicators, analyze historical trends of financial data, and determine financial audit cycles; the risk assessment module is used to assess the risks of financial data, calculate the deviation index of financial indicators, and classify the financial data into risk levels according to the calculated deviation index of financial indicators; the resource allocation module is used to allocate audit resources; the adaptive adjustment module is used to adjust the allocation of audit resources based on real-time data and risk changes. The output end of the data acquisition module is connected to the input end of the risk assessment module, the output end of the risk assessment module is connected to the input end of the resource allocation module, and the output end of the resource allocation module is connected to the input end of the adaptive adjustment module.

[0084] The data acquisition module includes a financial data acquisition unit and a data preprocessing unit. The financial data acquisition unit is used to extract financial data, including invoices, bills and reports, from the financial system and documents of the enterprise. The data preprocessing unit deletes duplicate invoices, vouchers or contracts by comparing the unique identifier of the data; the risk assessment module includes a deviation index calculation unit and a risk level classification unit. The deviation index calculation unit is used to calculate the deviation index of the financial indicator based on the financial data of the enterprise. The risk level classification unit is used to classify the financial data into different risk levels according to the calculation result of the deviation index and a preset risk threshold.

[0085] The resource allocation module includes a workload assessment unit, a historical audit service quality scoring unit and an audit resource allocation judgment unit. The workload assessment unit is used to assess the workload of each financial audit module, and generate a workload indicator for each audit module in combination with historical audit records, current task volume and audit complexity; the historical audit service quality scoring unit is used to calculate the historical audit service quality score of the audit module based on the audit efficiency and audit duration in the historical audit data of each financial audit module; the audit resource allocation judgment unit is used to comprehensively consider the workload indicator and historical audit service quality score of the audit module, calculate a rationality judgment indicator for audit resource allocation, and compare the calculated judgment indicator with the judgment threshold to determine whether the audit resource allocation is reasonable.

[0086] The adaptive adjustment module includes an anomaly monitoring unit and an adjustment unit. The anomaly monitoring unit receives the judgment result of the audit resource allocation unit to identify the financial audit modules with over-use or insufficient resources. The adjustment unit is used to adjust the computing resources allocated to the audit task or the work data of the auditor by reducing the resources in the modules with over-use of resources and supplementing the modules with insufficient resources.

[0087] In the embodiment: the intelligent audit system for agency accounting based on AI algorithm obtains the historical data of the financial indicators of the company's income, expenditure and debt-to-asset ratio in the past year through the data acquisition module, and performs standardized processing through the preprocessing module to generate continuous financial data points, and obtains the company's income data, which is 5 million yuan for quarter 1, 6 million yuan for quarter 2, 5.5 million yuan for quarter 3, and 7 million yuan for quarter 4; the expenditure data is 3 million yuan for quarter 1, 3.5 million yuan for quarter 2, 3.3 million yuan for quarter 3, and 4 million yuan for quarter 4; the debt-to-asset ratio data is 40% for quarter 1, 45% for quarter 2, 43% for quarter 3, and 50% for quarter 4, and the average value of each financial indicator is calculated. , the average income is 5.875 million yuan, the average expenditure is 3.45 million yuan, and the average debt-to-asset ratio is 44.5%. The deviation index between each financial indicator and its average value is calculated, and the deviation index of income is 0.16, the deviation index of expenditure is -0.043, and the deviation index of debt-to-asset ratio is 0.11. The set of deviation indexes R is obtained and normalized. The normalized income deviation index is 1, the normalized expenditure deviation index is 1, and the normalized debt-to-asset ratio deviation index is 0.6875. According to statistical analysis or industry standards, a=0.5 and b=0.8 are determined. Risk classification is performed according to the normalized deviation index. Input data and expenditure The data belongs to low-risk data, and the normalized debt-to-asset ratio deviation index is 0.6875, which is greater than or equal to a and less than b. The debt-to-asset ratio data is judged to be medium-risk data, and the debt-to-asset ratio data is further audited to check whether there is potential financial risk. In the selected period, financial audit is used to process and audit a total of k = 10 tasks. The workload of each task in the selected period is as follows: [120, 150, 130, 110, 90, 160, 140, 170, 125, 135] The annotation workload of each task is set to 100, and the complexity of each task is as follows: [1.1, 1.3, 1.2, 1.0, 0.9, 1.4, 1.2, 1.5, 1.1, 1.2], the complexity of the standard audit task is A = 1.0, the influencing factors α = 0.6 and β = 0.4, the workload indicator calculation formula of each audit module is 13.08, the number of audit tasks completed per hour is 5, the total time spent on processing all tasks is 50 hours, the service quality score is 1 according to the service quality score calculation formula, the actual data volume processed by the financial audit module is 1000, the correction coefficient d is 1, and the judgment indicator calculation formula is 0.001308. Compared with the set judgment threshold of 0.0015, it is judged that the audit resources are over-allocated, and the computing resources allocated to the audit tasks or the work data of the auditors are reduced.

[0088] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. The intelligent audit method for agency accounting based on AI algorithm is characterized by: The following steps are involved: S1: Collect historical financial data of the enterprise, obtain historical trend charts of financial indicators, analyze historical trends of financial data, and determine the financial audit cycle; S2: Obtain the financial data of the enterprise, calculate the deviation of financial indicators, and classify the risk level of the financial data according to the deviation index of the calculated financial indicators; S3: Obtain the historical audit records of each audit module, and evaluate the audit workload indicators of each audit module based on these historical audit data and corresponding financial data; S4: Calculate the historical audit service quality score of each audit module based on the historical audit data of each audit module, and judge the rationality of audit resource allocation in combination with the audit workload indicator of each audit module; S5: For audit modules with unreasonable resource allocation, reallocate audit resources and make dynamic adjustments based on real-time data.

2. The intelligent audit method for agency accounting based on AI algorithm according to claim 1 is characterized by: In step S1: said step S1 comprises the following steps: Step S1-1: Obtain historical data of financial indicators of the enterprise's income, expenditure and debt-to-asset ratio, sort the acquired historical data of the financial indicators in chronological order, generate continuous historical financial data points, and mark each financial data point in a plane rectangular coordinate system, connect each financial data point adjacently, and thus draw a historical trend chart of each financial indicator, where the horizontal axis represents the time series and the vertical axis represents the corresponding financial data value; Step S1-2: In the historical trend chart, extract the continuous time intervals corresponding to the maximum and minimum values ​​of the financial data, calculate the fluctuation amplitude, and then sort the continuous time intervals from large to small according to the fluctuation amplitude, and select the continuous time interval with the smallest fluctuation amplitude as the selected audit period T, where the fluctuation amplitude represents the difference between the maximum and minimum values ​​of the extracted financial data within a specific time area.

3. The intelligent audit method for agency accounting based on AI algorithm according to claim 1 is characterized in that: In step S2: said step S2 comprises the following steps: Step S2-1: Obtain historical data of the financial indicators of the enterprise within the selected audit period T, and calculate the average value U of each financial indicator; Step S2-2: Calculate the deviation index R between each financial indicator and its average value i , the specific calculation formula is as follows: Among them, M i represents the actual value of the ith financial indicator, U i represents the average value of the i-th indicator in the audit period T, MAX i represents the maximum value of the i-th financial indicator; Step S2-3: Obtain a set of deviation indices R and normalize them: When the deviation index R i If it is greater than or equal to 0, then the deviation index R i Divide by the maximum positive value of the deviation index in the financial index set to calculate the updated deviation index. For the deviation index R i Less than 0, then the deviation index R i Divided by the minimum negative value of the deviation index in the financial index set, the normalized deviation index is R i '; Step S2-4: According to the normalized deviation index Ri', the financial data is divided into high risk, medium risk and low risk, and further audited for data of different risk levels, wherein the range of the normalized deviation index Ri' for high risk level is greater than or equal to -1 and less than a; the range of the normalized deviation index Ri' for medium risk level is greater than or equal to a and less than b; the range of the normalized deviation index Ri' for low risk level is greater than or equal to b and less than 1; parameters a and b are determined by statistical analysis or industry standards; Step S2-5: The risk level result is passed to the corresponding financial audit module, including the income audit, expenditure audit and asset audit modules. Each module prepares to audit the high-risk, medium-risk and low-risk financial data according to the passed risk level.

4. The intelligent audit method for agency accounting based on AI algorithm according to claim 1 is characterized in that: In step S3: said step S3 comprises the following steps: Step S3-1: For each financial audit module, obtain the corresponding audit history data within the selected period T. Each financial audit module includes income audit, expenditure audit and asset and liability audit. Obtain the historical audit data of each module within the selected period T, including audit duration, audit complexity and audit task volume information; Step S3-2: Combine the audit task volume and complexity in the historical audit data and the fluctuation of financial data to calculate the workload index Q of each audit module. The specific calculation formula is as follows: Where k represents the total number of tasks to be reviewed within the selected period T; W j The workload of the jth audit task includes the number of audited documents and the amount of data processing; S represents the standard workload of the audit task; C j The complexity of the j-th audit task includes the types of financial data and the complexity of the vouchers involved in the audit task; A represents the complexity of the standard audit task; α and β represent the influencing factors of workload and complexity respectively, and their weight values ​​are determined based on historical data or industry standards.

5. The intelligent audit method for agency accounting based on AI algorithm according to claim 1 is characterized in that: In step S4: said step S4 comprises the following steps: Step S4-1: Calculate the historical audit service quality score F of the audit module based on the audit efficiency and audit duration in the historical audit data of each financial audit module. The specific calculation formula is as follows: Where C represents the number of audit tasks completed by the financial audit module within the selected period T; V represents the processing speed of audit tasks; t represents the total time spent by the module audit task processing all tasks within the selected period T; Step S4-2: Combine the audit task volume index Q of the financial audit module and the actual data volume s of the financial audit module to calculate the judgment index P to evaluate whether the audit resource allocation is reasonable. The specific calculation formula is as follows: Among them, Q represents the workload indicator of the financial audit module; s represents the actual data volume of the financial audit module; F j represents the historical service quality score of the jth auditor; d is the correction coefficient, which is used to balance the relationship between the actual module data volume s and the workload indicator Q in the audit resource allocation evaluation; Step S4-3: Compare the calculated judgment index P with the judgment threshold P0 to determine whether the audit resource allocation is reasonable: when the judgment index P is equal to the judgment threshold P0, it means that the audit resource allocation of the audit module is reasonable; when the judgment index P is not equal to the judgment threshold P0, it means that there is a problem with the resource allocation of the audit module, and the audit module resource allocation problem is sent to the abnormal monitoring unit for further processing by the abnormal monitoring unit.

6. The intelligent audit method for agency accounting based on AI algorithm according to claim 1 is characterized in that: In step S5: said step S5 comprises the following steps: Step S5-1: The abnormal monitoring unit obtains the number of the financial audit module whose judgment result is unreasonable audit resource allocation and the corresponding judgment index P, and divides the financial audit module number according to the size relationship between the judgment index P and the judgment threshold P0: the numbers of the financial audit modules whose judgment index P is greater than the judgment threshold P0 are aggregated and marked as modules with excessive audit resource allocation; the numbers of the modules whose judgment index P is less than the judgment threshold P0 are aggregated and marked as modules with insufficient audit resource allocation; Step S5-2: For each financial audit module where audit resources are over-allocated, reduce the computing resources or audit personnel's work data allocated to the audit task, and recalculate the judgment index P ’ , traverse all audit task resources until the judgment index P' is equal to the judgment threshold P0, and record the deleted audit task or resource number in this process; Step S5-3: For each financial audit module with insufficient audit resource allocation, increase the computing resources or auditor's work data allocated for the audit task, and recalculate the judgment index P ’ , traverse all audit task resources until the judgment indicator P ’ Equal to the judgment threshold P0, during which the deleted audit task or resource number recorded in step S5-2 is used for increase.

7. The intelligent audit system for agency accounting based on AI algorithm is characterized by: The system shown includes: a data acquisition module, a risk assessment module, a resource allocation module and an adaptive adjustment module. The data acquisition module is used to collect historical financial data of the enterprise, obtain historical trend charts of financial indicators, analyze historical trends of financial data, and determine financial audit cycles; the risk assessment module is used to assess the risks of financial data, calculate the deviation index of financial indicators, and classify the financial data into risk levels according to the calculated deviation index of financial indicators; the resource allocation module is used to allocate audit resources; the adaptive adjustment module is used to adjust the allocation of audit resources based on real-time data and risk changes, the output end of the data acquisition module is connected to the input end of the risk assessment module, the output end of the risk assessment module is connected to the input end of the resource allocation module, and the output end of the resource allocation module is connected to the input end of the adaptive adjustment module.

8. The AI ​​algorithm-based intelligent auditing system for agency accounting according to claim 7 is characterized by: The data acquisition module includes a financial data acquisition unit and a data preprocessing unit. The financial data acquisition unit is used to extract financial data, including invoices, bills and reports, from the financial system and documents of the enterprise. The data preprocessing unit deletes duplicate invoices, vouchers or contracts by comparing the unique identifier of the data; the risk assessment module includes a deviation index calculation unit and a risk level classification unit. The deviation index calculation unit is used to calculate the deviation index of the financial indicator based on the financial data of the enterprise. The risk level classification unit is used to classify the financial data into different risk levels according to the calculation result of the deviation index and a preset risk threshold.

9. The intelligent audit system for agency accounting based on AI algorithm according to claim 7 is characterized by: The resource allocation module includes a workload assessment unit, a historical audit service quality scoring unit and an audit resource allocation judgment unit. The workload assessment unit is used to assess the workload of each financial audit module, and generate a workload indicator for each audit module in combination with historical audit records, current task volume and audit complexity; the historical audit service quality scoring unit is used to calculate the historical audit service quality score of the audit module based on the audit efficiency and audit duration in the historical audit data of each financial audit module; the audit resource allocation judgment unit is used to comprehensively consider the workload indicator and historical audit service quality score of the audit module, calculate a rationality judgment indicator for audit resource allocation, and compare the calculated judgment indicator with the judgment threshold to determine whether the audit resource allocation is reasonable.

10. The intelligent auditing system for agency accounting based on AI algorithm according to claim 7 is characterized in that: The adaptive adjustment module includes an anomaly monitoring unit and an adjustment unit. The anomaly monitoring unit receives the judgment result of the audit resource allocation unit to identify the financial audit modules with over-use or insufficient resources. The adjustment unit is used to adjust the computing resources allocated to the audit task or the work data of the auditor by reducing the resources in the modules with over-use of resources and supplementing the modules with insufficient resources.

Images