A method and system for automatic analysis of audit data

Abstract

The present invention relates to the field of audit data analysis technology, specifically to an automated audit data analysis method and system, in which a correlation map analysis module in the system constructs an enterprise correlation map corresponding to each data element in the audit data based on the equity information of the enterprise to which each data element belongs, and constructs an enterprise correlation analysis set corresponding to the enterprise correlation map. The invention divides risk events according to the audit characteristics of the data elements in the audited data; combines the audit risk characteristics of the risk events and the enterprise correlation maps of the enterprises involved in the risk events and their corresponding enterprise correlation analysis sets to evaluate the risk tracking difficulty of the audit data of each risk event in the audited data; at the same time, combines the historical audit results of the enterprises involved in the risk events to achieve quantitative management of the audit priority of the risk events, construct and dynamically update the enterprise's risk event audit sequence, achieve effective management of audit data, and improve audit efficiency and accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of audit data analysis, and in particular to an automated audit data analysis method and system. Background Art

[0002] As enterprises expand in size and their business complexity increases, the volume and complexity of data audits face is also increasing dramatically. Traditional audit methods rely primarily on manual analysis, which is inefficient and prone to errors. Furthermore, the complex equity relationships between enterprises make it difficult for traditional methods to comprehensively and accurately identify and analyze the impact of these relationships on audits.

[0003] Existing audit systems can usually only process structured data, lack the ability to process unstructured data, and are unable to effectively combine corporate equity relationship maps for comprehensive analysis; this leads to auditors often relying on experience and intuition when determining audit priorities, lacking scientific and quantitative basis. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for automatic analysis of audit data to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solution: a method for automatic analysis of audit data, the method comprising the following steps:

[0006] S1, based on OCR Image recognition technology extracts the audit features of each data element in the audited data and generates the risk events to which the audit features of each data element belong. Data elements in the audit data that belong to the same risk event are aggregated into the same set, and the audit risk features of the aggregated set corresponding to each risk event are extracted.

[0007] S2. Based on the equity information of the enterprise to which each data element in the audit data belongs, construct an enterprise association map corresponding to the corresponding data element; and construct an enterprise association analysis set corresponding to the enterprise association map;

[0008] S3. Obtain each risk event belonging to the same enterprise in the audited data, combine the enterprise association analysis set corresponding to each enterprise involved in the risk event, and the audit risk characteristics of the corresponding summary set of the risk event, to obtain the risk tracking interference assessment value of the corresponding risk event; and combine the historical audit results of the enterprise involved in the corresponding risk event to generate the audit priority coefficient of the corresponding risk event, and construct the risk event audit sequence of the corresponding enterprise;

[0009] S4. The audit results for each risk event are fed back to the database in real time. The database retrieves the completed risk events and marks them. The risk event audit sequence of each enterprise stored in the current database is updated based on the marked risk events to obtain the unexecuted risk event audit sequence of each enterprise.

[0010] Furthermore, the data elements in the data to be audited in S1 include transaction invoices and transaction contracts; based on OCR When image recognition technology extracts the audit features of each data element in the audited data, OCR Image recognition technology automatically extracts the amount, execution time, transaction investor and transaction executor from the transaction invoice or transaction contract corresponding to the data element as the audit features of the corresponding data element; binds the data elements of the same enterprise with the same data element type, the same transaction investor, the same transaction executor and the corresponding execution time interval less than the preset value, and uses the binding result as a whole data element to replace each element in the binding result at the same time, and in the audit features corresponding to the obtained overall data element, the transaction investor and the transaction executor remain unchanged, the execution time is the average value of the execution time corresponding to each element in the corresponding binding result, and the amount is the cumulative value of the amount corresponding to each element in the corresponding binding result; data elements with the same amount, transaction investor and transaction executor in the corresponding audit features in the data to be audited are classified as the same risk event; the audit risk feature of the summary set corresponding to each risk event is the amount, transaction investor and transaction executor in the audit features corresponding to any element in the corresponding risk event.

[0011] Furthermore, the method steps for constructing the enterprise association graph corresponding to the corresponding data elements in S2 are specifically as follows:

[0012] S21. Obtain the summary set of enterprise names to which each data element in the audit data belongs, recorded as QS ; Get the equity information of the enterprise name to which each data element in the audit data belongs, and i The equity information in the enterprise name to which the data element belongs is recorded as G i , G i is a set, and each element in the set corresponds to a shareholder name, and the shareholder includes an individual or a company;

[0013] S22. i The name of the enterprise to which the data element belongs and G i The array after summarizing each element in is recorded as H i ; The audit datai Add the company name of each data element to a blank set to get the set Q i ;Will H i Each element in is regarded as a graph node. G i Each element in the audit data is respectively i The data elements are connected to get the first i The initial knowledge graph corresponding to the data elements is recorded as P i ;Will G i Each element in the j Level graph reference node, j The initial value of is 1;

[0014] S23, receiving P i 、 Q i Pass the exam j Level graph reference node, and jump to step S24;

[0015] S24, one by one for each j The reference nodes of the first level graph are analyzed. a No. j The reference node of the level graph is denoted as GR （i，j，a） ; Get QS Each element and its corresponding equity information contains GR （i，j，a） And the corresponding enterprise does not belong to Q i The aggregated set of enterprises is recorded as M （i，j，a） ;

[0016] S25, if M （i，j，a） If it is an empty set, stop a No. j Analysis of reference nodes in the level graph;

[0017] like M （i，j，a） If it is not an empty set, get M （i，j，a） The equity information corresponding to each element in the obtained equity information is compared with M （i，j，a） The knowledge graph obtained by connecting the corresponding elements in P i The splicing result is the new Pi ;Will M （i，j，a） The corresponding elements are added to Q i Get new Q i ;Will M （i，j，a） Each element in the corresponding equity information is used as a next-level graph reference node; j The corresponding value plus 1 is used as the new j value; jump to step S23 to iterate;

[0018] S25. Get the first i After all the graph reference nodes corresponding to the data elements stop analyzing P i , among the data to be audited i Enterprise association graph corresponding to each data element;

[0019] The set consisting of all enterprise names and shareholder names involved in the enterprise association map is used as the enterprise association analysis set corresponding to the corresponding enterprise association map.

[0020] Furthermore, the calculation formula for the risk tracking interference assessment value of the corresponding risk event obtained in S3 is as follows:

[0021] ;

[0022] in, RP （k，g） Indicates the first k The first g Risk tracking interference assessment value of each risk event; Num （k，g） Indicates the first k The first g The number of enterprises involved in the audit risk characteristics of each risk event; COM （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The complexity coefficient of the enterprise association graph of each enterprise;

[0023] ;

[0024] Level （k，g，d） Indicates that the data to be audited belongs to k The first gThe audit risk characteristics of a risk event involve d The maximum number of graph reference nodes in the enterprise association graph of an enterprise; B （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The maximum number of graph reference nodes at each level in the enterprise association graph of an enterprise; W （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The number of elements in the enterprise association analysis set corresponding to each enterprise; r represents the preset first weight coefficient; β Indicates the preset second weight coefficient.

[0025] Furthermore, the calculation formula for the audit priority coefficient of the corresponding risk event generated in S3 is as follows:

[0026] ;

[0027] in, F （k，g） Indicates the first k The corresponding enterprise g The audit priority coefficient of each risk event; E （k，g） Indicates the k The corresponding enterprise g The total number of risk events with unqualified audit results in the historical audit results of each enterprise included in the enterprise association analysis set of the enterprise association graph corresponding to the enterprises involved in each risk event; ES （k，g） Indicates the k The corresponding enterprise g The total number of risk events that have been audited in the historical audit results of each enterprise included in the enterprise association analysis set of the enterprise association graph corresponding to each enterprise involved in the risk event, and ES （k，g） >0; μ Indicates preset constants;

[0028] The risk event audit sequence of the corresponding enterprise is the result of arranging each risk event in the corresponding enterprise in descending order according to the audit priority coefficient.

[0029] Furthermore, the specific execution process in S4 includes:

[0030] The update result of the risk event audit sequence of each enterprise is the audit sequence composed of the remaining risk events after removing the risk events marked in the database from the risk event audit sequence of the corresponding enterprise.

[0031] An automated audit data analysis system comprising the following modules:

[0032] An audit risk feature extraction module, which extracts the audit features of each data element in the audited data based on OCR image recognition technology and generates the risk event to which the audit features of each data element belong; aggregates the data elements in the audit data that belong to the same risk event into the same set, and extracts the audit risk features of the aggregated set corresponding to each risk event;

[0033] A correlation map analysis module, which constructs an enterprise correlation map corresponding to each data element in the audit data based on the equity information of the enterprise to which the data element belongs; and constructs an enterprise correlation analysis set corresponding to the enterprise correlation map;

[0034] An audit sequence analysis module obtains each risk event belonging to the same enterprise in the audited data, combines the enterprise association analysis set corresponding to each enterprise involved in the risk event, and the audit risk characteristics of the corresponding summary set of the risk event, to obtain the risk tracking interference assessment value of the corresponding risk event; and combines the historical audit results of the enterprise involved in the corresponding risk event to generate the audit priority coefficient of the corresponding risk event, and construct the risk event audit sequence of the corresponding enterprise;

[0035] An audit sequence update management module feeds back the audit results of each risk event to the database in real time. The database retrieves completed risk events for marking and updates the risk event audit sequence of each enterprise stored in the current database based on the marked risk events to obtain the unexecuted risk event audit sequence of each enterprise.

[0036] Furthermore, the audit sequence analysis module includes a risk tracking interference assessment unit, an audit priority coefficient calculation unit and a risk event audit sequence construction unit.

[0037] The risk tracking interference assessment unit obtains each risk event belonging to the same enterprise in the audited data, combines the enterprise association analysis set corresponding to each enterprise involved in each risk event, and the audit risk characteristics of the summary set corresponding to the corresponding risk event, to obtain a risk tracking interference assessment value for the corresponding risk event;

[0038] The audit priority coefficient calculation unit generates an audit priority coefficient for the corresponding risk event based on historical audit results of the enterprise involved in the corresponding risk event;

[0039] The risk event audit sequence construction unit constructs a risk event audit sequence for the corresponding enterprise according to the result obtained by the audit priority coefficient calculation unit.

[0040] Compared with the existing technology, the beneficial effects achieved by the present invention are: the present invention divides risk events according to the audit characteristics of data elements in the data to be audited; and combines the audit risk characteristics of risk events and the enterprise association map of enterprises involved in risk events and their corresponding enterprise association analysis set to evaluate the risk tracking difficulty of audit data for each risk event in the audit data; at the same time, combined with the historical audit results of enterprises involved in risk events, it realizes quantitative management of audit priorities of risk events, constructs and dynamically updates the enterprise's risk event audit sequence, realizes effective management of audit data, and improves audit efficiency and audit accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0042] Figure 1 It is a structural diagram of an automatic audit data analysis system of the present invention;

[0043] Figure 2 It is a flow chart of an automatic analysis method of audit data of the present invention. DETAILED DESCRIPTION

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0045] See also Figure 1 The present invention provides a technical solution: an automated audit data analysis system, the system comprising the following modules:

[0046] An audit risk feature extraction module, which extracts the audit features of each data element in the audited data based on OCR image recognition technology and generates the risk event to which the audit features of each data element belong; aggregates the data elements in the audit data that belong to the same risk event into the same set, and extracts the audit risk features of the aggregated set corresponding to each risk event;

[0047] A correlation map analysis module, which constructs an enterprise correlation map corresponding to each data element in the audit data based on the equity information of the enterprise to which the data element belongs; and constructs an enterprise correlation analysis set corresponding to the enterprise correlation map;

[0048] An audit sequence analysis module, comprising a risk tracking interference assessment unit, an audit priority coefficient calculation unit, and a risk event audit sequence construction unit.

[0049] The risk tracking interference assessment unit obtains each risk event belonging to the same enterprise in the audited data, combines the enterprise association analysis set corresponding to each enterprise involved in each risk event, and the audit risk characteristics of the summary set corresponding to the corresponding risk event, to obtain a risk tracking interference assessment value for the corresponding risk event;

[0050] The audit priority coefficient calculation unit generates an audit priority coefficient for the corresponding risk event based on historical audit results of the enterprise involved in the corresponding risk event;

[0051] The risk event audit sequence construction unit constructs a risk event audit sequence for the corresponding enterprise based on the results obtained by the audit priority coefficient calculation unit;

[0052] An audit sequence update management module feeds back the audit results of each risk event to the database in real time. The database retrieves completed risk events for marking and updates the risk event audit sequence of each enterprise stored in the current database based on the marked risk events to obtain the unexecuted risk event audit sequence of each enterprise.

[0053] like Figure 2 As shown, a method for automatic analysis of audit data includes the following steps:

[0054] S1, based on OCR Image recognition technology extracts the audit features of each data element in the audited data and generates the risk events to which the audit features of each data element belong. Data elements in the audit data that belong to the same risk event are aggregated into the same set, and the audit risk features of the aggregated set corresponding to each risk event are extracted.

[0055] The data elements in the data to be audited in S1 include transaction invoices and transaction contracts; OCR When image recognition technology extracts the audit features of each data element in the audited data, OCRImage recognition technology automatically extracts the amount, execution time, transaction investor and transaction executor from the transaction invoice or transaction contract corresponding to the data element as the audit features of the corresponding data element; binds the data elements of the same enterprise with the same data element type, the same transaction investor, the same transaction executor and the corresponding execution time interval less than the preset value, and uses the binding result as a whole data element to replace each element in the binding result at the same time, and in the audit features corresponding to the obtained overall data element, the transaction investor and the transaction executor remain unchanged, the execution time is the average value of the execution time corresponding to each element in the corresponding binding result, and the amount is the cumulative value of the amount corresponding to each element in the corresponding binding result; data elements with the same amount, transaction investor and transaction executor in the corresponding audit features in the data to be audited are classified as the same risk event; the audit risk feature of the summary set corresponding to each risk event is the amount, transaction investor and transaction executor in the audit features corresponding to any element in the corresponding risk event.

[0056] S2. Based on the equity information of the enterprise to which each data element in the audit data belongs, construct an enterprise association map corresponding to the corresponding data element; and construct an enterprise association analysis set corresponding to the enterprise association map;

[0057] The method steps for constructing the enterprise association graph corresponding to the corresponding data elements in S2 are as follows:

[0058] S21. Obtain the summary set of enterprise names to which each data element in the audit data belongs, recorded as QS ; Get the equity information of the enterprise name to which each data element in the audit data belongs, and i The equity information in the enterprise name to which the data element belongs is recorded as G i , G i is a set, and each element in the set corresponds to a shareholder name, and the shareholder includes an individual or a company;

[0059] S22. i The name of the enterprise to which the data element belongs and G i The array after summarizing each element in is recorded as H i ; The audit data i Add the company name of each data element to a blank set to get the set Q i ;Will H i Each element in is regarded as a graph node. G iEach element in the audit data is respectively i The data elements are connected to get the first i The initial knowledge graph corresponding to the data elements is recorded as P i ;Will G i Each element in the j Level graph reference node, j The initial value of is 1;

[0060] S23, receiving P i 、 Q i Pass the exam j Level graph reference node, and jump to step S24;

[0061] S24, one by one for each j The reference nodes of the first level graph are analyzed. a No. j The reference node of the level graph is denoted as GR （i，j，a） ; Get QS Each element and its corresponding equity information contains GR （i，j，a） And the corresponding enterprise does not belong to Q i The aggregated set of enterprises is recorded as M （i，j，a） ;

[0062] S25, if M （i，j，a） If it is an empty set, stop a No. j Analysis of reference nodes in the level graph;

[0063] like M （i，j，a） If it is not an empty set, get M （i，j，a） The equity information corresponding to each element in the obtained equity information is compared with M （i，j，a） The knowledge graph obtained by connecting the corresponding elements in P i The splicing result is the new P i ;Will M （i，j，a） The corresponding elements are added to Q i Get new Q i ;Will M （i，j，a）Each element in the corresponding equity information is used as a next-level graph reference node; j The corresponding value plus 1 is used as the new j value; jump to step S23 to iterate;

[0064] S25. Get the first i After all the graph reference nodes corresponding to the data elements stop analyzing P i , among the data to be audited i Enterprise association graph corresponding to each data element;

[0065] The set consisting of all enterprise names and shareholder names involved in the enterprise association map is used as the enterprise association analysis set corresponding to the corresponding enterprise association map.

[0066] S3. Obtain each risk event belonging to the same enterprise in the audited data, combine the enterprise association analysis set corresponding to each enterprise involved in the risk event, and the audit risk characteristics of the corresponding summary set of the risk event, to obtain the risk tracking interference assessment value of the corresponding risk event; and combine the historical audit results of the enterprise involved in the corresponding risk event to generate the audit priority coefficient of the corresponding risk event, and construct the risk event audit sequence of the corresponding enterprise;

[0067] The calculation formula for the risk tracking interference assessment value of the corresponding risk event obtained in S3 is as follows:

[0068] ;

[0069] in, RP （k，g） Indicates the first k The first g Risk tracking interference assessment value of each risk event; Num （k，g） Indicates the first k The first g The number of enterprises involved in the audit risk characteristics of each risk event; COM （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The complexity coefficient of the enterprise association graph of each enterprise;

[0070] ;

[0071] Level （k，g，d） Indicates that the data to be audited belongs tok The first g The audit risk characteristics of a risk event involve d The maximum number of graph reference nodes in the enterprise association graph of an enterprise; B （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The maximum number of graph reference nodes at each level in the enterprise association graph of an enterprise; W （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The number of elements in the enterprise association analysis set corresponding to each enterprise; r represents the preset first weight coefficient; β Indicates the preset second weight coefficient.

[0072] The calculation formula for the audit priority coefficient of the corresponding risk event generated in S3 is as follows:

[0073] ;

[0074] in, F （k，g） Indicates the first k The corresponding enterprise g The audit priority coefficient of each risk event; E （k，g） Indicates the k The corresponding enterprise g The total number of risk events with unqualified audit results in the historical audit results of each enterprise included in the enterprise association analysis set of the enterprise association graph corresponding to the enterprises involved in each risk event; ES （k，g） Indicates the k The corresponding enterprise g The total number of risk events that have been audited in the historical audit results of each enterprise included in the enterprise association analysis set of the enterprise association graph corresponding to each enterprise involved in the risk event, and ES （k，g） >0; μ Indicates preset constants;

[0075] The risk event audit sequence of the corresponding enterprise is the result of arranging each risk event in the corresponding enterprise in descending order according to the audit priority coefficient.

[0076] S4. Feedback the audit results for each risk event to the database in real time. The database retrieves completed risk events and marks them. Based on the marked risk events, the audit sequence of risk events for each enterprise currently stored in the database is updated to obtain the audit sequence of unexecuted risk events for each enterprise.

[0077] The specific execution process in S4 includes:

[0078] The update result of the risk event audit sequence of each enterprise is the audit sequence composed of the remaining risk events after removing the risk events marked in the database from the risk event audit sequence of the corresponding enterprise.

[0079] In this embodiment, if there is a transaction transfer between two companies A and B, the transaction contract and transaction invoice corresponding to the transaction transfer between A and B constitute a risk event. The risk event exists in both companies A and B. Therefore, the risk event exists in the risk event audit sequences corresponding to companies A and B respectively. If the risk event in the risk event audit sequence of company A has been reviewed and completed, the database will mark the risk event, and at the same time, update the risk event audit sequences corresponding to companies A and B respectively containing the risk event, and delete the risk event in the risk event audit sequences corresponding to companies A and B respectively. Therefore, in this process, the risk event execution priority of the risk event in the risk event audit sequence corresponding to company A may remain unchanged, while the risk event is not executed in the risk event audit sequence corresponding to company B, and the execution priority of each risk event after the corresponding serial number of the risk event in the risk event audit sequence corresponding to company B is changed.

[0080] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0081] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for automated analysis of audit data, characterized in that: The method comprises the following steps: S1, based on OCR Image recognition technology extracts the audit features of each data element in the audited data and generates the risk events to which the audit features of each data element belong. Data elements in the audit data that belong to the same risk event are aggregated into the same set, and the audit risk features of the aggregated set corresponding to each risk event are extracted. S2. Based on the equity information of the enterprise to which each data element in the audit data belongs, construct an enterprise association map corresponding to the corresponding data element; and construct an enterprise association analysis set corresponding to the enterprise association map; S3. Obtain each risk event belonging to the same enterprise in the audited data, combine the enterprise association analysis set corresponding to each enterprise involved in the risk event, and the audit risk characteristics of the corresponding summary set of the risk event, to obtain the risk tracking interference assessment value of the corresponding risk event; and combine the historical audit results of the enterprise involved in the corresponding risk event to generate the audit priority coefficient of the corresponding risk event, and construct the risk event audit sequence of the corresponding enterprise; The calculation formula for the risk tracking interference assessment value is as follows: ； in, RP （k，g） Indicates the first k The first g Risk tracking interference assessment value of each risk event; Num （k，g） Indicates the first k The first g The number of enterprises involved in the audit risk characteristics of each risk event; COM （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The complexity coefficient of the enterprise association graph of each enterprise; ； Level （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The maximum number of graph reference nodes in the enterprise association graph of an enterprise; B （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The maximum number of graph reference nodes at each level in the enterprise association graph of an enterprise; W （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The number of elements in the enterprise association analysis set corresponding to each enterprise; r represents the preset first weight coefficient; β represents a preset second weight coefficient; The calculation formula of the audit priority coefficient is as follows: ； in, F （k，g） Indicates the first k The corresponding enterprise g The audit priority coefficient of each risk event; E （k，g） Indicates the k The corresponding enterprise g The total number of risk events with unqualified audit results in the historical audit results of each enterprise included in the enterprise association analysis set of the enterprise association graph corresponding to each enterprise involved in each risk event; ES （k，g） Indicates the k The corresponding enterprise g The total number of risk events that have been audited in the historical audit results of each enterprise included in the enterprise association analysis set of the enterprise association graph corresponding to each enterprise involved in the risk event, and ES （k，g） >0; μ Indicates preset constants; The risk event audit sequence of the corresponding enterprise is the result of arranging the risk events within the corresponding enterprise in descending order according to the audit priority coefficient; S4. The audit results for each risk event are fed back to the database in real time. The database retrieves the completed risk events and marks them. The risk event audit sequence of each enterprise stored in the current database is updated based on the marked risk events to obtain the unexecuted risk event audit sequence of each enterprise.

2. The method for automated analysis of audit data according to claim 1, characterized in that: The data elements in the data to be audited in S1 include transaction invoices and transaction contracts; OCR When image recognition technology extracts the audit features of each data element in the audited data, OCR Image recognition technology automatically extracts the amount, execution time, transaction investor and transaction executor from the transaction invoice or transaction contract corresponding to the data element as the audit features of the corresponding data element; binds the data elements of the same enterprise with the same data element type, the same transaction investor, the same transaction executor and the corresponding execution time interval less than the preset value, and uses the binding result as a whole data element to replace each element in the binding result at the same time, and in the audit features corresponding to the obtained overall data element, the transaction investor and the transaction executor remain unchanged, the execution time is the average value of the execution time corresponding to each element in the corresponding binding result, and the amount is the cumulative value of the amount corresponding to each element in the corresponding binding result; data elements with the same amount, transaction investor and transaction executor in the corresponding audit features in the data to be audited are classified as the same risk event; the audit risk feature of the summary set corresponding to each risk event is the amount, transaction investor and transaction executor in the audit features corresponding to any element in the corresponding risk event.

3. The method for automated analysis of audit data according to claim 2, characterized in that: The method steps for constructing the enterprise association graph corresponding to the corresponding data elements in S2 are as follows: S21. Obtain the summary set of enterprise names to which each data element in the audit data belongs, recorded as QS ; Get the equity information of the enterprise name to which each data element in the audit data belongs, and i The equity information in the enterprise name to which the data element belongs is recorded as G i , G i is a set, and each element in the set corresponds to a shareholder name, and the shareholder includes an individual or a company; S22. i The name of the enterprise to which the data element belongs and G i The array after summarizing each element in is recorded as H i ; The audit data i Add the company name of each data element to a blank set to get the set Q i ;Will H i Each element in is regarded as a graph node. G i Each element in the audit data is respectively i The data elements are connected to get the first i The initial knowledge graph corresponding to the data elements is recorded as P i ;Will G i Each element in the j Level graph reference node, j The initial value of is 1; S23, receiving P i 、 Q i Pass the exam j Level graph reference node, and jump to step S24; S24, one by one for each j The reference nodes of the first level graph are analyzed. a No. j The reference node of the level graph is denoted as GR （i，j，a） ; Get QS Each element and its corresponding equity information contains GR （i，j，a） And the corresponding enterprise does not belong to Q i The aggregated set of enterprises is recorded as M （i，j，a） ; S25, if M （i，j，a） If it is an empty set, stop a No. j Analysis of reference nodes in the level graph; like M （i，j，a） If it is not an empty set, get M （i，j，a） The equity information corresponding to each element in the obtained equity information is compared with M （i，j，a） The knowledge graph obtained by connecting the corresponding elements in P i The splicing result is the new P i ;Will M （i，j，a） The corresponding elements are added to Q i Get new Q i ;Will M （i，j，a） Each element in the corresponding equity information is used as a next-level graph reference node; j The corresponding value plus 1 is used as the new j value; Jump to step S23 to iterate; S25. Get the first i After all the graph reference nodes corresponding to the data elements stop analyzing P i , among the data to be audited i Enterprise association graph corresponding to each data element; The set consisting of all enterprise names and shareholder names involved in the enterprise association map is used as the enterprise association analysis set corresponding to the corresponding enterprise association map.

4. The method for automated analysis of audit data according to claim 1, wherein: The specific execution process in S4 includes: The update result of the risk event audit sequence of each enterprise is the audit sequence composed of the remaining risk events after removing the risk events marked in the database from the risk event audit sequence of the corresponding enterprise.

5. An automated audit data analysis system, characterized in that: The system includes the following modules: An audit risk feature extraction module, which extracts the audit feature of each data element in the audited data based on OCR image recognition technology and generates the risk event to which the audit feature of each data element belongs; Aggregate all data elements belonging to the same risk event in the audit data into the same set, and extract the audit risk characteristics of each risk event corresponding to the aggregated set; A correlation map analysis module, which constructs an enterprise correlation map corresponding to each data element in the audit data based on the equity information of the enterprise to which the data element belongs; and constructs an enterprise correlation analysis set corresponding to the enterprise correlation map; An audit sequence analysis module, which obtains risk events belonging to the same enterprise in the audited data, combines the enterprise association analysis set corresponding to each enterprise involved in each risk event with the audit risk characteristics of the summary set corresponding to the corresponding risk event, and obtains a risk tracking interference assessment value for the corresponding risk event; Combined with the historical audit results of the enterprises involved in the corresponding risk events, the audit priority coefficients of the corresponding risk events are generated, and the risk event audit sequence of the corresponding enterprises is constructed; An audit sequence update management module feeds back the audit results of each risk event to the database in real time. The database retrieves completed risk events, marks them, and updates the risk event audit sequence of each enterprise currently stored in the database based on the marked risk events, obtaining the unexecuted risk event audit sequence of each enterprise. The calculation formula for the risk tracking interference assessment value is as follows: ； in, RP （k，g） Indicates the first k The first g Risk tracking interference assessment value of each risk event; Num （k，g） Indicates the first k The first g The number of enterprises involved in the audit risk characteristics of each risk event; COM （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The complexity coefficient of the enterprise association graph of each enterprise; ； Level （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The maximum number of graph reference nodes in the enterprise association graph of an enterprise; B （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The maximum number of graph reference nodes at each level in the enterprise association graph of an enterprise; W （k，g，d） Indicates that the data to be audited belongs to k The first g The audit risk characteristics of a risk event involve d The number of elements in the enterprise association analysis set corresponding to each enterprise; r represents the preset first weight coefficient; β represents a preset second weight coefficient; The calculation formula of the audit priority coefficient is as follows: ； in, F （k，g） Indicates the first k The corresponding enterprise g The audit priority coefficient of each risk event; E （k，g） Indicates the k The corresponding enterprise g The total number of risk events with unqualified audit results in the historical audit results of each enterprise included in the enterprise association analysis set of the enterprise association graph corresponding to each enterprise involved in each risk event; ES （k，g） Indicates the k The corresponding enterprise g The total number of risk events that have been audited in the historical audit results of each enterprise included in the enterprise association analysis set of the enterprise association graph corresponding to each enterprise involved in the risk event, and ES （k，g） >0; μ Indicates preset constants; The risk event audit sequence of the corresponding enterprise is the result of arranging each risk event in the corresponding enterprise in descending order according to the audit priority coefficient.

6. The automated audit data analysis system according to claim 5, characterized in that: The audit sequence analysis module includes a risk tracking interference assessment unit, an audit priority coefficient calculation unit and a risk event audit sequence construction unit. The risk tracking interference assessment unit obtains each risk event belonging to the same enterprise in the audited data, combines the enterprise association analysis set corresponding to each enterprise involved in each risk event, and the audit risk characteristics of the summary set corresponding to the corresponding risk event, to obtain a risk tracking interference assessment value for the corresponding risk event; The audit priority coefficient calculation unit generates an audit priority coefficient for the corresponding risk event based on historical audit results of the enterprise involved in the corresponding risk event; The risk event audit sequence construction unit constructs a risk event audit sequence for the corresponding enterprise according to the result obtained by the audit priority coefficient calculation unit.

Images