Enterprise tax potential risk analysis method based on cloud computing and data mining

Abstract

The invention discloses an enterprise tax potential risk analysis method based on cloud computing and data mining, and relates to the financial field, and the method comprises the steps: obtaining an accounting index of a tax accounting table and an accounting mode of the accounting index; obtaining a sample value of the accounting index; obtaining a to-be-analyzed numerical value of the accounting index; forming a transverse deviation coefficient of the numerical value to be analyzed, and forming a judgment critical value of the transverse deviation coefficient; taking tax disclosure shortage information as a risk point; obtaining an informatization weak point of the to-be-analyzed enterprise; obtaining a risk coefficient of a manual accounting work step; and forming a critical value for risk coefficient judgment, and taking the informatization weak points corresponding to the artificial accounting working steps of which the risk coefficients are greater than the critical value as risk points. By forming the transverse deviation coefficient and the judgment critical value, tax disclosure shortage information is obtained, the risk coefficient is obtained, and the formation critical value is obtained, so that the risk point can be determined, and the risk point is refined to the project, so that the self-inspection efficiency can be improved.

Description

Technical Field

[0001] The present invention relates to the financial field, and specifically to an enterprise tax potential risk analysis method based on cloud computing and data mining. Background Art

[0002] Tax risk is an inevitable part of enterprise operation, mainly involving the standardization of tax management. Non-compliance of tax payment behaviors with the provisions of tax laws and regulations, such as failure to pay tax or underpayment of tax, may lead to risks such as tax supplement, fine, late payment surcharge, criminal penalty, and damage to reputation for the enterprise. Such risks mainly stem from the enterprise's lack of understanding of tax laws and regulations or intentional tax evasion.

[0003] Since the calculation of various tax burdens is a well-known and perfect technology, there is no tax risk for the enterprise's operation data when sufficient information is provided. However, due to insufficient understanding of laws and regulations, insufficient data is provided during accounting, or due to insufficient data digitization, manual accounting is required for the non-digitized part. As a result, there will be risks of incorrect tax accounting, and the specific reasons for the potential tax risks caused by these situations are difficult to determine, which will further increase the difficulty of the enterprise itself. Summary of the Invention

[0004] To solve the above technical problems, an enterprise tax potential risk analysis method based on cloud computing and data mining is provided, and this technical solution solves the problems raised in the above background art.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] An enterprise tax potential risk analysis method based on cloud computing and data mining, comprising:

[0007] Obtain at least one sample enterprise in the region where the enterprise to be analyzed is located, obtain the sample tax accounting form of the sample enterprise, obtain the tax accounting form to be analyzed of the enterprise to be analyzed, and collectively refer to the sample tax accounting form and the tax accounting form to be analyzed as the tax accounting form, and the tax accounting form contains the operation data disclosed by the enterprise;

[0008] Based on big data, obtain at least one accounting index of the tax accounting form and the accounting method of the accounting index;

[0009] Using the accounting method of the accounting index, calculate the sample value of the accounting index in the sample tax accounting form;

[0010] Using the accounting method of the accounting index, calculate the value to be analyzed of the accounting index in the tax accounting form to be analyzed;

[0011] Form a horizontal deviation coefficient of the value to be analyzed and form a judgment critical value of the horizontal deviation coefficient;

[0012] When the horizontal deviation coefficients are all less than the corresponding judgment critical values, it is determined that there is no potential tax risk for the enterprise to be analyzed; otherwise, it is determined that there is potential tax risk for the enterprise to be analyzed.

[0013] When there is potential tax risk for the enterprise to be analyzed, analyze the information disclosure situation of the enterprise to be analyzed, obtain the lacking tax disclosure information of the enterprise to be analyzed, and take the lacking tax disclosure information as the risk point.

[0014] Evaluate the informatization level of the enterprise to be analyzed to obtain the weak points of the enterprise's informatization.

[0015] Form manual accounting work steps for the weak points of informatization, and calculate the risk coefficient of the manual accounting work steps.

[0016] Form a critical value for risk coefficient judgment, and take the weak points of informatization corresponding to the manual accounting work steps with a risk coefficient greater than the critical value as risk points.

[0017] Preferably, the obtaining of at least one sample enterprise in the region where the enterprise to be analyzed is located includes the following steps:

[0018] Obtain at least one characteristic enterprise in the region where the enterprise to be analyzed is located, obtain at least one business item, asset scale, operating income, and number of employees of the enterprise to be analyzed, and obtain at least one business item, asset scale, operating income, and number of employees of the characteristic enterprise.

[0019] Based on empirical data, form influence factors for business items, asset scale, operating income, and number of employees respectively, where the sum of the influence factors for business items, asset scale, operating income, and number of employees is equal to 1.

[0020] Use the correlation formula to calculate the correlation coefficient between the characteristic enterprise and the enterprise to be analyzed.

[0021] When the difference between the correlation coefficient and 1 is less than the preset difference, the enterprise with a qualified historical tax inspection among the characteristic enterprises corresponding to the correlation coefficient is taken as the sample enterprise.

[0022] The formation of the preset difference is as follows:

[0023] Obtain at least one reference enterprise of the same type and scale as the enterprise to be analyzed, calculate the correlation coefficient between the enterprise to be analyzed and the reference enterprise, and take the maximum value of the difference between the correlation coefficient between the enterprise to be analyzed and the reference enterprise and 1 as the preset difference.

[0024] The correlation formula is as follows:

[0025]

[0026] Among them, A is the correlation coefficient between the characteristic enterprise and the enterprise to be analyzed, α is the influence factor of the business project, β is the influence factor of the asset scale, γ is the influence factor of the operating income, δ is the influence factor of the number of employees, a is the number of overlapping business projects between the enterprise to be analyzed and the characteristic enterprise, b is the asset scale of the characteristic enterprise, c is the operating income of the characteristic enterprise, d is the number of employees of the characteristic enterprise, e is the number of business projects of the enterprise to be analyzed, f is the asset scale of the enterprise to be analyzed, g is the operating income of the enterprise to be analyzed, and h is the number of employees of the enterprise to be analyzed.

[0027] Preferably, the method for obtaining at least one accounting index of the tax accounting form and the accounting method of the accounting index based on big data includes the following steps:

[0028] Obtain the tax items that the enterprise should report based on big data, and use the tax items as accounting indexes;

[0029] Obtain the data items required for the accounting of the accounting index, obtain the positions of the data items in the tax accounting form, obtain the usage steps of the data items, and use the data items and the usage steps of the data items as the accounting method of the accounting index.

[0030] Preferably, the steps for forming the horizontal deviation coefficient of the value to be analyzed include the following:

[0031] Use the accounting index corresponding to the value to be analyzed as the target accounting index;

[0032] Subtract the sample value of the target accounting index from the value to be analyzed to obtain a deviation value;

[0033] Take the mean of the results of taking the absolute value of at least one deviation value to obtain the horizontal deviation coefficient of the value to be analyzed.

[0034] Preferably, the steps for forming the judgment critical value of the horizontal deviation coefficient include the following:

[0035] Take the mean of the sample values of the target accounting index to obtain the sample average value;

[0036] Use one of the sample values of the target accounting index as the target sample value;

[0037] Subtract the sample average value from the target sample value and take the absolute value to obtain the allowable deviation value;

[0038] When the target sample value traverses all the sample values of the target accounting index, obtain at least one allowable deviation value, and take the mean of at least one allowable deviation value to obtain the judgment critical value of the horizontal deviation coefficient of the value to be analyzed corresponding to the target accounting index.

[0039] Preferably, analyzing the information disclosure situation of the enterprise to be analyzed to obtain the tax disclosure lack information of the enterprise to be analyzed includes the following steps:

[0040] In the sample tax accounting table, obtain at least one business data disclosed by the sample enterprise, summarize the business data belonging to the business items of the sample enterprise to obtain a sample business data set, and correspond the business items to the sample business data set;

[0041] Classify the business data in the sample business data set to obtain sample revenue data and at least one sample non-revenue data;

[0042] In the tax accounting table to be analyzed, obtain at least one business data disclosed by the enterprise to be analyzed, summarize the business data belonging to the business items of the enterprise to be analyzed to obtain an analyzed business data set, and correspond the business items to the analyzed business data set;

[0043] Classify the business data in the analyzed business data set to obtain analyzed revenue data and at least one analyzed non-revenue data;

[0044] Correspond the sample business data set and the analyzed business data set with the same corresponding business items;

[0045] Calculate the ratio of the analyzed revenue data in the analyzed business data set to the sample revenue data in the corresponding sample business data set to obtain a display coefficient;

[0046] Take the sample business data set with the display coefficient closest to 1 as the target business data set, and take the sample revenue data and sample non-revenue data in the target business data set as the target revenue data and target non-revenue data respectively;

[0047] Calculate the ratio of the analyzed non-revenue data with the same attribute to the target non-revenue data to obtain a test coefficient;

[0048] When the gap between the test coefficient and the display coefficient exceeds a preset value, then take the analyzed non-revenue data corresponding to the test coefficient as the disclosure lack data, and the preset value is set based on empirical data;

[0049] Summarize all the disclosure lack data to obtain the tax disclosure lack information of the enterprise to be analyzed.

[0050] Preferably, evaluating the informatization level of the enterprise to be analyzed to obtain the informatization weak points of the enterprise to be analyzed includes the following steps:

[0051] Judge the business data in the analyzed business data set. When the business data is not electronic, then take the business data as the informatization weak point of the enterprise to be analyzed.

[0052] Preferably, the steps for forming manual accounting for information technology weak points include the following steps:

[0053] Take the data items consistent with the information technology weak points as characteristic data items;

[0054] Take the accounting indicators for using the characteristic data items during accounting as characteristic accounting indicators;

[0055] Take the usage steps of the characteristic data items involved during the accounting of the characteristic accounting indicators as characteristic usage steps;

[0056] Summarize all the characteristic usage steps to obtain the manual accounting work steps.

[0057] Preferably, the steps for calculating the risk coefficient of the manual accounting work steps include the following steps:

[0058] Set a benchmark accounting step, which is composed of the usage steps of any data items;

[0059] Obtain the first difficulty characteristic of the benchmark accounting step, where the first difficulty characteristic is equal to the result of multiplying the occupied space of the data processed by the benchmark accounting step by the number of arithmetic operations of the processed data;

[0060] Obtain the second difficulty characteristic of the manual accounting work steps, where the second difficulty characteristic is equal to the result of multiplying the occupied space of the data processed by the manual accounting work steps by the number of arithmetic operations of the processed data;

[0061] Based on big data, obtain the average error rate of the benchmark accounting step under manual accounting conditions;

[0062] Use the risk calculation formula to calculate the risk coefficient of the manual accounting work steps;

[0063] The risk calculation formula is as follows:

[0064]

[0065] Among them, B is the risk coefficient, C is the second difficulty characteristic, D is the first difficulty characteristic, and E is the average error rate.

[0066] Preferably, the steps for forming the critical value for risk coefficient judgment include the following steps:

[0067] Obtain the upper limit of the enterprise compensation ability of the enterprise to be analyzed;

[0068] Based on historical data, obtain the value range of the accounting errors of the characteristic accounting indicators;

[0069] Based on the fine grading standard, divide the value range of the accounting error to obtain at least one identification interval, such that when the accounting result of the characteristic accounting index belongs to the identification interval, the fine amount is consistent;

[0070] Based on the fine grading standard, obtain the fine amount of the identification interval;

[0071] Use the comprehensive fine formula to calculate the comprehensive fine value of the characteristic accounting index;

[0072] Accumulate the comprehensive fine values of all characteristic accounting indexes to obtain the total fine value;

[0073] Divide the upper limit of the enterprise's compensation ability by the total fine value to obtain the critical value for risk coefficient judgment;

[0074] The comprehensive fine formula is as follows:

[0075]

[0076] Wherein, F is the comprehensive fine value of the characteristic accounting index, i is the subscript, n is the number of identification intervals, G i is the proportion of the i-th identification interval in the value range of the accounting error, and H i is the fine amount of the i-th identification interval.

[0077] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0078] By forming the horizontal deviation coefficient and judgment critical value of the value to be analyzed, obtaining the tax disclosure lack information of the enterprise to be analyzed, calculating the risk coefficient of the manual accounting work steps, and forming the critical value for risk coefficient judgment, thus, it is possible to determine whether there is a potential tax risk in the enterprise by means of horizontal comparison, and through horizontal comparison, estimate the possible tax risk of the enterprise from the perspectives of the enterprise's data disclosure situation and the risk of manual calculation, and thereby determine the risk points. Since the risk points are refined to the projects, when the enterprise conducts self-inspection, it is not necessary to check all projects, so the efficiency of its self-inspection can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 is a schematic flowchart of the enterprise tax potential risk analysis method based on cloud computing and data mining of the present invention;

[0080] Figure 2 is a schematic flowchart of obtaining at least one sample enterprise in the area where the enterprise to be analyzed is located according to the present invention;

[0081] Figure 3 is a schematic flowchart of obtaining at least one accounting index and the accounting method of the accounting index of the tax accounting form based on big data according to the present invention;

[0082] Figure 4 Schematic flow chart of forming the horizontal deviation coefficient of the value to be analyzed in the present invention;

[0083] Figure 5 Schematic flow chart of forming the judgment critical value of the horizontal deviation coefficient in the present invention;

[0084] Figure 6 Schematic flow chart of analyzing the information disclosure situation of the enterprise to be analyzed in the present invention to obtain the tax disclosure lack information of the enterprise to be analyzed;

[0085] Figure 7 Schematic flow chart of forming the manual accounting work steps for the weak points of informatization in the present invention;

[0086] Figure 8 Schematic flow chart of calculating the risk coefficient of the manual accounting work steps in the present invention;

[0087] Figure 9 Schematic flow chart of forming the critical value for risk coefficient judgment in the present invention. Detailed implementation manners

[0088] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments in the following description are only examples, and those skilled in the art can think of other obvious variations.

[0089] Referring to Figure 1 As shown, a method for analyzing potential enterprise tax risks based on cloud computing and data mining includes:

[0090] Obtain at least one sample enterprise in the area where the enterprise to be analyzed is located, obtain the sample tax accounting form of the sample enterprise, obtain the to-be-analyzed tax accounting form of the enterprise to be analyzed, and collectively refer to the sample tax accounting form and the to-be-analyzed tax accounting form as the tax accounting form, and the tax accounting form contains the business data disclosed by the enterprise;

[0091] Based on big data, obtain at least one accounting index of the tax accounting form and the accounting method of the accounting index;

[0092] Use the accounting method of the accounting index to calculate the sample value of the accounting index in the sample tax accounting form;

[0093] Use the accounting method of the accounting index to calculate the to-be-analyzed value of the accounting index in the to-be-analyzed tax accounting form;

[0094] Form the horizontal deviation coefficient of the to-be-analyzed value and form the judgment critical value of the horizontal deviation coefficient;

[0095] When the horizontal deviation coefficients are all less than the corresponding judgment critical values, it is determined that there is no potential tax risk for the enterprise to be analyzed; otherwise, it is determined that there is potential tax risk for the enterprise to be analyzed. Since there are multiple values to be analyzed, the horizontal deviation coefficient of each value to be analyzed needs to be less than the corresponding judgment critical value to determine that there is no potential tax risk.

[0096] When there is potential tax risk for the enterprise to be analyzed, analyze the information disclosure situation of the enterprise to be analyzed, obtain the lacking information on tax disclosure of the enterprise to be analyzed, and take the lacking information on tax disclosure as a risk point.

[0097] Evaluate the informatization level of the enterprise to be analyzed to obtain the weak points of informatization of the enterprise to be analyzed.

[0098] Form manual accounting work steps for the weak points of informatization, and calculate the risk coefficient of the manual accounting work steps.

[0099] Form a critical value for risk coefficient judgment, and take the weak points of informatization corresponding to the manual accounting work steps with a risk coefficient greater than the critical value as risk points.

[0100] It is easy to know that in the existing technology, the electronicization of taxes is quite sufficient. When an enterprise provides the required data in accordance with the law and all these data are electronic data, the tax burden calculation system can automatically process these data, and moreover, the calculated tax burden is risk-free. However, in actual situations, the data provided by the enterprise may be insufficient. On the other hand, the data provided by the enterprise may not be all electronic data, and some are paper data in the form of account books. For this part of the data, it cannot be automatically calculated and manual accounting must be used. However, there must be a certain error rate in manual calculation. Thus, it will lead to potential tax risks. But when conducting a review, in order to discover potential tax risks, it is necessary to re-review the provision of all data in the accounts and recalculate. However, since the review is carried out manually, a non-discriminatory comprehensive review will consume a lot of time. In this solution, corresponding algorithms are formed. By means of horizontal comparison, it is determined whether there are potential tax risks for the enterprise and the risk points causing potential tax risks. Thus, the time for self-inspection can be reduced.

[0101] Refer to Figure 2 As shown, the steps to obtain at least one sample enterprise in the region where the enterprise to be analyzed is located include the following:

[0102] Obtain at least one characteristic enterprise in the region where the enterprise to be analyzed is located, obtain at least one business item, asset scale, operating income, and number of employees of the enterprise to be analyzed, and obtain at least one business item, asset scale, operating income, and number of employees of the characteristic enterprise.

[0103] Based on empirical data, impact factors for business items, asset size, operating income, and number of employees are respectively formed, where the sum of the impact factors for business items, asset size, operating income, and number of employees is equal to 1;

[0104] Use the correlation formula to calculate the correlation coefficient between the characteristic enterprise and the enterprise to be analyzed;

[0105] When the difference between the correlation coefficient and 1 is less than the preset difference, the enterprises with qualified historical tax inspections among the characteristic enterprises corresponding to the correlation coefficient are used as sample enterprises;

[0106] The formation of the preset difference is as follows:

[0107] Obtain at least one reference enterprise of the same type and scale as the enterprise to be analyzed, calculate the correlation coefficient between the enterprise to be analyzed and the reference enterprise, and take the maximum value of the difference between the correlation coefficient of the enterprise to be analyzed and the reference enterprise and 1 as the preset difference;

[0108] The correlation formula is as follows:

[0109]

[0110] Among them, A is the correlation coefficient between the characteristic enterprise and the enterprise to be analyzed, α is the impact factor of business items, β is the impact factor of asset size, γ is the impact factor of operating income, δ is the impact factor of number of employees, a is the number of overlapping business items between the enterprise to be analyzed and the characteristic enterprise, b is the asset size of the characteristic enterprise, c is the operating income of the characteristic enterprise, d is the number of employees of the characteristic enterprise, e is the number of business items of the enterprise to be analyzed, f is the asset size of the enterprise to be analyzed, g is the operating income of the enterprise to be analyzed, and h is the number of employees of the enterprise to be analyzed.

[0111] Sample enterprises are enterprises with the same attributes and similar businesses as the enterprise to be analyzed. At the same time, their scales are similar. Therefore, the taxes they involve are similar. Since sample enterprises are all enterprises that have passed the inspection and there is no tax risk for them, the tax situation of sample enterprises can be used for horizontal comparison with the situation of the enterprise to be analyzed.

[0112] Refer to Figure 3 As shown, based on big data, obtaining at least one accounting index of the tax accounting form and the accounting method of the accounting index includes the following steps:

[0113] Based on big data, obtain the tax items that the enterprise should report, and use the tax items as accounting indexes;

[0114] Obtain the data items required for accounting the accounting index, obtain the positions of the data items in the tax accounting form, obtain the usage steps of the data items, and use the data items and the usage steps of the data items as the accounting method of the accounting index.

[0115] The calculation of various tax burdens is existing knowledge, and for each tax burden calculation, a corresponding formula can be found, which includes the required data and the way of data operation.

[0116] Refer to Figure 4 As shown, the steps to form the horizontal deviation coefficient of the value to be analyzed are as follows:

[0117] Take the accounting index corresponding to the value to be analyzed as the target accounting index;

[0118] Subtract the sample value of the target accounting index from the value to be analyzed to obtain the deviation value;

[0119] Take the mean of the results after taking the absolute value of at least one deviation value to obtain the horizontal deviation coefficient of the value to be analyzed.

[0120] Since the enterprises corresponding to the value to be analyzed and the sample values are similar enterprises, the deviation between the value to be analyzed and the sample values will inevitably fluctuate within a very small range, that is, the judgment critical value. The judgment critical value is actually obtained from the fluctuation of the sample values. When the value to be analyzed has no risk, it is actually the sample value. Therefore, the horizontal deviation coefficient of the value to be analyzed will inevitably not exceed the judgment critical value. When it does not meet this condition, it indicates that the value to be analyzed is likely to be abnormal, and thus tax risk investigation is required.

[0121] Refer to Figure 5 As shown, the steps to form the judgment critical value of the horizontal deviation coefficient are as follows:

[0122] Take the mean of the sample values of the target accounting index to obtain the sample average value;

[0123] Take one of the sample values of the target accounting index as the target sample value;

[0124] Subtract the sample average value from the target sample value, take the absolute value, to obtain the allowable deviation value;

[0125] When the target sample value traverses all the sample values of the target accounting index, at least one allowable deviation value is obtained. Take the mean of at least one allowable deviation value to obtain the judgment critical value of the horizontal deviation coefficient of the value to be analyzed corresponding to the target accounting index.

[0126] Refer to Figure 6 As shown, the steps to analyze the information disclosure situation of the enterprise to be analyzed and obtain the tax disclosure lack information of the enterprise to be analyzed are as follows:

[0127] In a sample tax accounting form, obtain at least one operating data disclosed by a sample enterprise, summarize the operating data belonging to the operating items of the sample enterprise to obtain a sample operating data set, and correspond the operating items to the sample operating data set;

[0128] Classify the operating data in the sample operating data set to obtain sample revenue data and at least one sample non-revenue data;

[0129] In a tax accounting form to be analyzed, obtain at least one operating data disclosed by an enterprise to be analyzed, summarize the operating data belonging to the operating items of the enterprise to be analyzed to obtain an operating data set to be analyzed, and correspond the operating items to the operating data set to be analyzed;

[0130] Classify the operating data in the operating data set to be analyzed to obtain revenue data to be analyzed and at least one non-revenue data to be analyzed;

[0131] Correspond the sample operating data set and the operating data set to be analyzed with the same corresponding operating items;

[0132] Calculate the ratio of the revenue data to be analyzed in the operating data set to be analyzed to the sample revenue data in the corresponding sample operating data set to obtain a display coefficient;

[0133] Take the sample operating data set with the display coefficient closest to 1 as the target operating data set, and take the sample revenue data and sample non-revenue data in the target operating data set as the target revenue data and target non-revenue data respectively;

[0134] Calculate the ratio of the non-revenue data to be analyzed with the same attribute to the target non-revenue data to obtain a test coefficient;

[0135] When the gap between the test coefficient and the display coefficient exceeds a preset value, take the non-revenue data to be analyzed corresponding to the test coefficient as the lacking disclosure data, and the preset value is set based on empirical data;

[0136] Summarize all the lacking disclosure data to obtain the tax lacking disclosure information of the enterprise to be analyzed.

[0137] When calculating the tax burden, the common situation is underreporting or omitting, which will not cause the data of a certain item to completely disappear, but the data of this item is smaller than the actual value. Therefore, it is necessary to find the item where this situation occurs as the tax lacking disclosure information by forming an algorithm;

[0138] Here, the sample business data set with the display coefficient closest to 1 is taken as the target business data set, and the data of the target business data set is used for horizontal comparison. According to the setting method of the display coefficient, the sample enterprise corresponding to the target business data set is the enterprise closest to the enterprise to be analyzed. When conducting the analysis, it is necessary to consider its revenue situation, because the larger the revenue, the corresponding non-revenue data will also increase proportionally. Since the sample enterprise corresponding to the target business data set and the enterprise to be analyzed are different enterprises, they cannot be compared equally. That is, when the non-revenue data to be analyzed with the same attributes is different from the target non-revenue data, it cannot be determined that there is data concealment or other situations in the non-revenue data to be analyzed. It is necessary to compare the test coefficient with the display coefficient to obtain that there is data concealment or other situations in the non-revenue data to be analyzed. Here, the display coefficient is the ratio of the revenue data to be analyzed in the business data set to be analyzed to the sample revenue data in the corresponding sample business data set, and the test coefficient is the ratio of the non-revenue data to be analyzed with the same attributes to the target non-revenue data. The test coefficient and the display coefficient eliminate the revenue gap between the sample enterprise and the enterprise to be analyzed. Therefore, the test coefficient and the display coefficient can be directly compared. When the gap is large, it indicates that there is an abnormality in the non-revenue data to be analyzed;

[0139] The preset value can be calculated by calculating the display coefficient and the test coefficient of two sample enterprises, that is, replacing the business data set to be analyzed in the above calculation process of the display coefficient and the test coefficient with another sample business data set. Thus, take the absolute value of the difference between the display coefficient and the test coefficient and obtain the value range of this result. Because there are many combinations of two sample enterprises, the maximum value of the value range of this result can be used as the preset value.

[0140] To evaluate the informatization level of the enterprise to be analyzed and obtain the informatization weak points of the enterprise to be analyzed, the following steps are included:

[0141] Judge the business data in the business data set to be analyzed. When the business data is not digitized, the business data is taken as the informatization weak point of the enterprise to be analyzed.

[0142] Refer to Figure 7 As shown, the steps for forming manual accounting for the informatization weak points include the following steps:

[0143] The data items consistent with the informatization weak points are taken as the characteristic data items;

[0144] The accounting indicators for using the characteristic data items during accounting are taken as the characteristic accounting indicators;

[0145] When calculating the characteristic accounting indicators, the usage steps involving the characteristic data items are taken as the characteristic usage steps;

[0146] Summarize all the steps of using features to obtain the steps of manual accounting work.

[0147] For data that has not been digitized, only manual methods can be used for accounting.

[0148] Refer to Figure 8 As shown, calculating the risk coefficient of the manual accounting work steps includes the following steps:

[0149] Set a benchmark accounting step, which is composed of the usage steps of any data item;

[0150] Obtain the first difficulty feature of the benchmark accounting step. The first difficulty feature is equal to the result of multiplying the occupied space of the data processed by the benchmark accounting step by the number of arithmetic operations of the processed data;

[0151] Obtain the second difficulty feature of the manual accounting work steps. The second difficulty feature is equal to the result of multiplying the occupied space of the data processed by the manual accounting work steps by the number of arithmetic operations of the processed data;

[0152] Based on big data, obtain the average error rate of the benchmark accounting step under manual accounting conditions;

[0153] Use the risk calculation formula to calculate the risk coefficient of the manual accounting work steps;

[0154] The risk calculation formula is as follows:

[0155]

[0156] Among them, B is the risk coefficient, C is the second difficulty feature, D is the first difficulty feature, and E is the average error rate.

[0157] During accounting, its risk is related to the length of the data and the number of arithmetic operations in the calculation process. The longer the data, the greater the possibility of input errors. The more arithmetic operations, the more input times, and the greater the possibility of errors. For the convenience of calculation, a benchmark accounting step is set. When calculating later, it is compared with the benchmark accounting step, and according to the average error rate of the benchmark accounting step, the risk coefficient of the manual accounting work steps is obtained.

[0158] Refer to Figure 9 As shown, forming the critical value for risk coefficient judgment includes the following steps:

[0159] Obtain the upper limit of the enterprise's compensation ability of the enterprise to be analyzed;

[0160] Based on historical data, obtain the value range of accounting errors of the feature accounting index;

[0161] Based on the fine grading standard, divide the value range of the accounting error to obtain at least one identification interval, such that when the accounting result of the characteristic accounting index belongs to the identification interval, the fine amount is consistent;

[0162] Based on the fine grading standard, obtain the fine amount for the identification interval;

[0163] Use the comprehensive fine formula to calculate the comprehensive fine value of the characteristic accounting index;

[0164] Accumulate the comprehensive fine values of all characteristic accounting indexes to obtain the total fine value;

[0165] Divide the upper limit of the enterprise's compensation ability by the total fine value to obtain the critical value for risk coefficient judgment;

[0166] The comprehensive fine formula is as follows:

[0167]

[0168] Among them, F is the comprehensive fine value of the characteristic accounting index, i is the subscript, n is the number of identification intervals, G i is the proportion of the i-th identification interval in the value range of the accounting error, and H i is the fine amount for the i-th identification interval.

[0169] Since errors and risks are inevitable, not all risks need to be investigated, and investigation also incurs costs. Therefore, a critical value for the risk coefficient is formed based on the upper limit of the enterprise's compensation ability that the enterprise can afford. As long as the fine caused by the risk is acceptable to the enterprise, there is no need to conduct a risk investigation. The total fine value is fined only when an error occurs. Therefore, the actual fine is the total fine value and the risk coefficient, and it must not exceed the upper limit of the enterprise's compensation ability. Thus, the critical value of the risk coefficient can be calculated.

[0170] Furthermore, this solution also proposes a storage medium on which a computer-readable program is stored. When the computer-readable program is called, it executes the above-mentioned enterprise tax potential risk analysis method based on cloud computing and data mining.

[0171] It can be understood that the storage medium can be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid-state disk (SSD).

[0172] In summary, the advantages of the present invention are as follows: by forming the horizontal deviation coefficient and the judgment threshold value of the value to be analyzed, obtaining the lacking tax disclosure information of the enterprise to be analyzed, calculating the risk coefficient of the manual accounting work steps, and forming the threshold value for risk coefficient judgment. Thus, it is possible to determine whether there is a potential tax risk in the enterprise by means of horizontal comparison, and through horizontal comparison, estimate the possible tax risks of the enterprise from the perspectives of the enterprise's data disclosure situation and the risk of manual calculation, and thereby determine the risk points. Since the risk points are refined to projects, when the enterprise conducts self-inspection, it is not necessary to check all projects, so the efficiency of its self-inspection can be improved.

[0173] The foregoing has shown and described the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and the principles described in the specification are only the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection required by the present invention is defined by the appended claims and their equivalents.

Claims

1. An enterprise tax potential risk analysis method based on cloud computing and data mining, characterized in that, Including: Obtain at least one sample enterprise in the region where the enterprise to be analyzed is located, obtain the sample tax accounting table of the sample enterprise, obtain the tax accounting table to be analyzed of the enterprise to be analyzed, and collectively refer to the sample tax accounting table and the tax accounting table to be analyzed as the tax accounting table. The tax accounting table contains the business data disclosed by the enterprise; Based on big data, obtain at least one accounting indicator of the tax accounting table and the accounting method of the accounting indicator; Using the accounting method of the accounting indicator, calculate the sample value of the accounting indicator in the sample tax accounting table; Using the accounting method of the accounting indicator, calculate the value to be analyzed of the accounting indicator in the tax accounting table to be analyzed; Form the horizontal deviation coefficient of the value to be analyzed and form the judgment critical value of the horizontal deviation coefficient; When the horizontal deviation coefficients are all less than the corresponding judgment critical values, it is determined that there is no potential tax risk for the enterprise to be analyzed. Otherwise, it is determined that the enterprise to be analyzed has potential tax risk; When there is potential tax risk for the enterprise to be analyzed, analyze the information disclosure situation of the enterprise to be analyzed to obtain the lacking tax disclosure information of the enterprise to be analyzed, and use the lacking tax disclosure information as the risk point; Evaluate the informatization level of the enterprise to be analyzed to obtain the weak points of the informatization of the enterprise to be analyzed; Form the manual accounting work steps for the weak points of informatization and calculate the risk coefficient of the manual accounting work steps; Form the critical value for risk coefficient judgment, and use the weak points of informatization corresponding to the manual accounting work steps with the risk coefficient greater than the critical value as the risk points.

2. The enterprise tax potential risk analysis method based on cloud computing and data mining according to claim 1, wherein, The obtaining of at least one sample enterprise in the region where the enterprise to be analyzed is located includes the following steps: Obtain at least one characteristic enterprise in the region where the enterprise to be analyzed is located, obtain at least one business item, asset scale, operating income, and number of employees of the enterprise to be analyzed, and obtain at least one business item, asset scale, operating income, and number of employees of the characteristic enterprise; Based on empirical data, form the influencing factors of business item, asset scale, operating income, and number of employees respectively, where the sum of the influencing factors of business item, asset scale, operating income, and number of employees is equal to 1; Use the correlation formula to calculate the correlation coefficient between the characteristic enterprise and the enterprise to be analyzed; When the difference between the correlation coefficient and 1 is less than the preset difference, the enterprise with qualified historical tax inspections among the characteristic enterprises corresponding to the correlation coefficient is used as the sample enterprise; The formation of the preset difference is as follows: Obtain at least one reference enterprise of the same type and scale as the enterprise to be analyzed, calculate the correlation coefficient between the enterprise to be analyzed and the reference enterprise, and use the maximum value of the difference between the correlation coefficient between the enterprise to be analyzed and the reference enterprise and 1 as the preset difference; The correlation formula is as follows: Among them, A is the correlation coefficient between the characteristic enterprise and the enterprise to be analyzed, α is the influencing factor of the operating project, β is the influencing factor of the asset scale, γ is the influencing factor of the operating income, δ is the influencing factor of the number of employees, a is the number of overlapping operating projects between the enterprise to be analyzed and the characteristic enterprise, b is the asset scale of the characteristic enterprise, c is the operating income of the characteristic enterprise, d is the number of employees of the characteristic enterprise, e is the number of operating projects of the enterprise to be analyzed, f is the asset scale of the enterprise to be analyzed, g is the operating income of the enterprise to be analyzed, and h is the number of employees of the enterprise to be analyzed.

3. A method for analyzing potential risks of enterprise tax based on cloud computing and data mining according to claim 2, characterized in that, The method of obtaining at least one accounting indicator and the calculation method of the accounting indicator of the tax accounting table based on big data includes the following steps: Obtain tax items that enterprises should report based on big data, and use tax items as accounting indicators; Obtain the data items required for calculating the accounting indicators, obtain the location of the data items in the tax accounting table, obtain the steps for using the data items, and use the data items and the steps for using the data items as the calculation method for the accounting indicators.

4. The enterprise tax potential risk analysis method based on cloud computing and data mining according to claim 3, characterized in that The forming of the lateral deviation coefficient of the value to be analyzed comprises the following steps: The accounting indicator corresponding to the value to be analyzed is used as the target accounting indicator; The sample value of the target accounting indicator is subtracted from the value to be analyzed to obtain the deviation value; The results of taking the absolute value of at least one deviation value are averaged to obtain the lateral deviation coefficient of the value to be analyzed.

5. A method for analyzing potential risks of enterprise tax based on cloud computing and data mining according to claim 4, characterized in that, The forming of the critical value of the lateral deviation coefficient comprises the following steps: Take the average of the sample values ​​of the target accounting indicator to obtain the sample average value; Use one of the sample values ​​of the target accounting indicator as the target sample value; The target sample value is subtracted from the sample average value and the absolute value is taken to obtain the allowable deviation value; When the target sample value traverses all sample values ​​of the target accounting indicator, at least one allowable deviation value is obtained, and the average of the at least one allowable deviation value is taken to obtain the judgment critical value of the lateral deviation coefficient of the value to be analyzed corresponding to the target accounting indicator.

6. A method for analyzing potential risks of enterprise taxation based on cloud computing and data mining according to claim 5, characterized in that, Analyzing the information disclosure of the enterprise to be analyzed and obtaining the tax disclosure missing information of the enterprise to be analyzed includes the following steps: In the sample tax accounting table, at least one operating data disclosed by the sample enterprise is obtained, the operating data of the operating items belonging to the sample enterprise are aggregated to obtain a sample operating data set, and the operating items are matched with the sample operating data set; Classifying the operating data in the sample operating data set to obtain sample revenue data and at least one sample non-revenue data; In the tax accounting table to be analyzed, at least one operating data disclosed by the enterprise to be analyzed is obtained, the operating data of the operating items belonging to the enterprise to be analyzed are aggregated to obtain the operating data set to be analyzed, and the operating items are matched with the operating data set to be analyzed; Classify the operating data in the operating data set to be analyzed to obtain revenue data to be analyzed and at least one non-revenue data to be analyzed; Match the sample business data set and the business data set to be analyzed that correspond to the same business project; Calculate the ratio of the revenue data to be analyzed of the operating data set to be analyzed to the sample revenue data of the corresponding sample operating data set to obtain a display coefficient; The sample business data set with the display coefficient closest to 1 is used as the target business data set, and the sample revenue data and sample non-revenue data in the target business data set are used as the target revenue data and target non-revenue data respectively; Calculate the ratio of the non-revenue data to be analyzed with the same attributes to the target non-revenue data to obtain a test coefficient; When the gap between the test coefficient and the display coefficient exceeds a preset value, the non-revenue data to be analyzed corresponding to the test coefficient is used as the data lacking in disclosure, and the preset value is set based on empirical data; Summarize all the data lacking in disclosure to obtain the tax disclosure lack information of the enterprise to be analyzed.

7. A method for analyzing potential risks of enterprise tax based on cloud computing and data mining according to claim 6, characterized in that, Evaluating the informatization level of the enterprise to be analyzed to obtain the weak points of the informatization of the enterprise to be analyzed includes the following steps: Judge the business data in the business data set to be analyzed. When the business data is not electronic, the business data is used as the weak point of the informatization of the enterprise to be analyzed.

8. A method for analyzing potential risks of enterprise taxation based on cloud computing and data mining according to claim 7, characterized in that, The steps for forming manual accounting work for the weak points of informatization include the following steps: The data items consistent with the weak points of informatization are used as characteristic data items; The accounting indicators for using the characteristic data items during accounting are used as characteristic accounting indicators; When the characteristic accounting indicators are accounted, the usage steps involving the characteristic data items are used as characteristic usage steps; Summarize all the characteristic usage steps to obtain the manual accounting work steps.

9. A method for analyzing potential risks of enterprise tax based on cloud computing and data mining according to claim 8, characterized in that The steps for calculating the risk coefficient of the manual accounting work steps include the following steps: Set a benchmark accounting step, and the benchmark accounting step is composed of the usage steps of any data item; Obtain the first difficulty characteristic of the benchmark accounting step, and the first difficulty characteristic is equal to the result of multiplying the occupied space of the data processed by the benchmark accounting step by the number of operations of the processed data; Obtain the second difficulty characteristic of the manual accounting work steps, and the second difficulty characteristic is equal to the result of multiplying the occupied space of the data processed by the manual accounting work steps by the number of operations of the processed data; Based on big data, obtain the average error rate of the benchmark accounting step under the condition of manual accounting; Use the risk calculation formula to calculate the risk coefficient of the manual accounting work steps; The risk calculation formula is as follows: Where B is the risk coefficient, C is the second difficulty characteristic, D is the first difficulty characteristic, and E is the average error rate.

10. A method for analyzing potential risks of enterprise taxation based on cloud computing and data mining according to claim 9, characterized in that, The steps for forming a critical value for risk coefficient judgment include the following steps: Obtain the upper limit of the enterprise compensation ability of the enterprise to be analyzed; Based on historical data, obtain the value range of the accounting errors of the characteristic accounting indicators; Based on the fine grading standard, divide the value range of the accounting errors to obtain at least one identification interval, such that when the accounting result of the characteristic accounting indicator belongs to the identification interval, the fine amount is the same; Based on the fine grading standard, obtain the fine amount of the identification interval; Use the fine comprehensive formula to calculate the comprehensive fine value of the characteristic accounting indicators; Accumulate the comprehensive fine values of all the characteristic accounting indicators to obtain the total fine value; Divide the upper limit of the enterprise compensation ability by the total fine value to obtain the critical value for risk coefficient judgment; The fine comprehensive formula is as follows: Among them, F is the comprehensive fine value of the feature accounting index, i is the subscript, n is the number of identification intervals, G i is the proportion of the i-th identification interval in the value range of accounting errors, H i is the fine amount of the i-th identification interval.

Images