A method for sending audit analysis model results to a suspicious point database

Abstract

The present invention provides a method for implementing sending the audit analysis model result to a doubt point library, including: conducting a first doubt point audit on the target audit analysis model result based on a preset audit rule, and determining the doubt point to be verified based on the first doubt point audit result; sending the doubt point to be verified to the corresponding subordinate verification terminal, and receiving the supplementary data submitted by the subordinate verification terminal based on the doubt point to be verified in real time, and conducting a second doubt point audit on the doubt point to be verified based on the supplementary data; determining the target doubt point based on the second doubt point audit result, parsing the target doubt point, determining the target field characteristics, and at the same time, determining the target doubt point library corresponding to the target doubt point and the project doubt point of the target doubt point in the target doubt point library based on the target field characteristics, and inserting the display field corresponding to the target doubt point into the project doubt point. The efficiency and accuracy of the review of the audit analysis model results are improved, and the efficiency and accuracy of sending the target doubt point obtained after the audit to the corresponding doubt point library are also guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of digital data processing, and in particular to a method for implementing sending audit analysis model results to a doubt database. Background Art

[0002] As digital auditing is increasingly valued and applied by audit institutions, how to improve the quality and efficiency of auditing has become a focus of more and more audit organizations;

[0003] At present, when verifying the audit results, the auditors usually conduct a secondary verification of the audit results manually, and after the verification, the audit results need to be manually stored in the corresponding suspicious point library. Due to the need for human participation, the efficiency and accuracy of the secondary verification cannot be guaranteed. At the same time, when storing the audit results in the corresponding suspicious point library, the auditors need to consult a large number of directory indexes to query the corresponding suspicious point library, which leads to low storage efficiency of the audit results and storage reliability cannot be guaranteed.

[0004] Therefore, in order to overcome the above-mentioned defects, the present invention provides an implementation method for sending the audit analysis model results to a doubt library. By processing and analyzing the audit analysis model results, the results with doubts in the audit analysis model results are locked and accurately sent to the corresponding doubt library, thereby overcoming the above-mentioned defects of low storage efficiency, low verification efficiency and low verification accuracy. Summary of the invention

[0005] The present invention provides a method for implementing sending the audit analysis model results to a doubt point library, which is used to quickly and accurately lock the doubt points to be verified in the target audit analysis model results by performing a doubt point review on the target audit analysis model results. Secondly, by sending the determined doubt points to be verified to the corresponding lower-level verification terminal, and processing and analyzing the supplementary data submitted by the lower-level verification terminal, a secondary review of the doubt points to be verified is achieved, which improves the efficiency and accuracy of resolving the doubt points to be verified. Finally, the target audit analysis model results that still have problems after the second doubt point review are determined as target doubt points, and the target field characteristics of the target doubt points are analyzed to quickly and effectively determine the target doubt point library corresponding to the target doubt point, and insert the display field corresponding to the target doubt point into the corresponding target doubt point library, which improves the efficiency and accuracy of reviewing the audit analysis model results, and also ensures the efficiency and accuracy of sending the target doubt points obtained after the review to the corresponding doubt point library.

[0006] The present invention provides a method for sending the audit analysis model result to a doubt database, comprising:

[0007] Step 1: Based on the result check request, the target audit analysis model result is locked, and the first suspicious point audit is performed on the target audit analysis model result based on the preset audit rules, and the suspicious points to be verified are determined based on the first suspicious point audit results;

[0008] Step 2: Send the doubtful points to be verified to the corresponding lower-level verification terminal, and receive the supplementary data submitted by the lower-level verification terminal based on the doubtful points to be verified in real time, and conduct a second doubtful point review on the doubtful points to be verified based on the supplementary data;

[0009] Step 3: Determine the target doubt based on the second doubt review result, and analyze the target doubt to determine the target field characteristics. At the same time, determine the target doubt library corresponding to the target doubt and the project doubt in the target doubt library based on the target field characteristics, and insert the display field corresponding to the target doubt into the project doubt.

[0010] Preferably, a method for implementing sending the audit analysis model result to the doubt database, in step 1, locking the target audit analysis model result based on the result viewing request, includes:

[0011] Obtain a result viewing request, parse the result viewing request, and obtain the target query field corresponding to the result viewing request;

[0012] Input the target query field into the preset convolutional neural network for analysis, obtain the target semantics corresponding to the target query field, generate a query statement based on the target semantics, and query the audit analysis model results in the preset result cache based on the search engine according to the query statement;

[0013] Based on the query results, the audit analysis model results to be viewed corresponding to the result viewing request are obtained, and the audit analysis model results to be viewed are encapsulated to obtain the final target audit analysis model, wherein there is at least one audit analysis model result to be viewed.

[0014] Preferably, a method for implementing sending the audit analysis model result to the doubt database, obtaining the audit analysis model result to be viewed corresponding to the result viewing request based on the query result, includes:

[0015] Obtain the audit analysis model results to be viewed, and extract the identity identifier of the audit analysis model results to be viewed. At the same time, determine the subject identifier to be viewed based on the result viewing request, match the identity identifier with the subject identifier to be viewed, and obtain the subordinate relationship between the identity identifier and the subject identifier to be viewed based on the matching result;

[0016] Determine the audit business scope corresponding to the subject to be checked based on the subject identifier to be checked, and determine whether the audit analysis model result to be checked is within the audit business scope corresponding to the subject to be checked based on the subordinate relationship;

[0017] When the audit analysis model results to be reviewed are within the audit business scope corresponding to the subject to be reviewed, the audit analysis model results to be reviewed are locked; otherwise, the audit analysis model results to be reviewed are eliminated.

[0018] Preferably, a method for implementing sending the audit analysis model result to the doubt point library, in step 1, a first doubt point audit is performed on the target audit analysis model result based on a preset audit rule, and a doubt point to be verified is determined based on the first doubt point audit result, including:

[0019] Obtain the target audit analysis model results, extract the business tags of the target audit analysis model results, and retrieve the corresponding preset audit rules from the preset rule library based on the business tags;

[0020] Parse the preset audit rules, determine the audit dimensions included in the preset audit rules, and determine the audit strategy and business parameter comparison table corresponding to each audit dimension, where there is at least one audit dimension;

[0021] The target audit analysis model results are layered based on the audit dimensions to obtain the sub-audit analysis model results corresponding to each audit dimension, and the sub-audit analysis model results are audited for the first doubt points based on the audit strategy of the corresponding audit dimension to obtain the corresponding audit parameters under each audit dimension;

[0022] Compare the audit parameters corresponding to each audit dimension with the corresponding business parameter comparison table, and determine the sub-audit analysis model results whose audit parameters are not within the preset limit range of the business parameter comparison table as sub-doubts to be verified;

[0023] The sub-doubts to be verified corresponding to different audit dimensions are summarized to obtain the doubts to be verified corresponding to the audit result of the first doubt.

[0024] Preferably, a method for implementing sending the audit analysis model results to the doubt point library, determining the sub-audit analysis model results whose audit parameters are not within the preset limited range of the business parameter comparison table as sub-doubts to be verified, includes:

[0025] Obtain the comparison results of the corresponding audit parameters and the corresponding business parameter comparison table under each audit dimension, and determine the sub-audit analysis model results whose audit parameters are within the preset limited range of the business parameter comparison table as unquestionable results;

[0026] The results without doubt shall be removed from the list corresponding to the target audit analysis model results, and the results without doubt shall be transferred to the processed results list at the same time.

[0027] Preferably, a method for implementing sending the audit analysis model results to a doubt database, in step 2, performing a second doubt review on the doubts to be verified based on the supplementary data, comprises:

[0028] Acquire the obtained supplementary data, parse the supplementary data, extract data features of the supplementary data, determine the first structural features of the supplementary data based on the data features, and extract the second structural features corresponding to the doubtful points to be verified;

[0029] The supplementary data is split based on the first structural feature and the second structural feature, and sub-supplementary data corresponding to each audit item in the doubtful point to be verified is obtained based on the splitting result;

[0030] The sub-supplementary data is split into terms to obtain N business terms, and a key business term set is selected from the N business terms based on the business characteristics of the audit project, and the element information corresponding to each key business term in the key business term set is extracted;

[0031] Based on the element information, the proof parameters of the audit items in the doubtful points to be verified are determined using the supplementary data, and a second doubtful point audit is performed based on the proof parameters and the doubtful points to be verified based on the preset audit rules.

[0032] Preferably, a method for implementing sending the audit analysis model results to a doubt point library, in step 3, determining the target doubt point library corresponding to the target doubt point and the project doubt point of the target doubt point in the target doubt point library based on the target field characteristics, and inserting the display field corresponding to the target doubt point into the project doubt point, including:

[0033] Acquire the target field features, and determine the class identifier of the target suspicious point based on the target field features;

[0034] Extract the library identifier corresponding to the preset doubtful point library, and match the class identifier with the library identifier to obtain the target doubtful point library corresponding to the target field feature. At the same time, determine the target feature value corresponding to the target doubtful point based on the target field feature, and traverse the stored project doubtful points in the target doubtful point library based on the target feature value, and determine the project doubtful point based on the traversal result;

[0035] Extracting the project characteristic value of the stored doubtful point in the project doubtful point, and determining the theoretical insertion position of the target doubtful point in the project doubtful point based on the project characteristic value and the target characteristic value corresponding to the target doubtful point;

[0036] The display fields corresponding to the target doubtful points are inserted into the theoretical insertion positions, and the display fields in the project doubtful points are smoothed based on the insertion results to complete the insertion of the display fields of the target doubtful points into the project doubtful points.

[0037] Preferably, a method for implementing sending the audit analysis model results to the doubt point library, inserting the display field corresponding to the target doubt point into the project doubt point, includes:

[0038] Obtain the insertion result of the display field of the target suspicious point, and after completing the insertion of the display field of the target suspicious point, extract the flow information corresponding to the target suspicious point in each processing link, wherein the processing link includes the first suspicious point review, the second suspicious point review and inserting the display field corresponding to the target suspicious point into the project suspicious point;

[0039] The flow information corresponding to each processing link is recorded to obtain a record report of the target suspicious point, and the record report is stored to complete the record of the complete flow process of the target suspicious point.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. Through the review of doubts in the target audit analysis model results, the doubts to be verified in the target audit analysis model results can be quickly and accurately locked. Secondly, by sending the determined doubts to be verified to the corresponding lower-level verification terminal, and processing and analyzing the supplementary data submitted by the lower-level verification terminal, a second review of the doubts to be verified is achieved, which improves the efficiency and accuracy of resolving the doubts to be verified. Finally, the target audit analysis model results that still have problems after the second doubt review are determined as target doubts, and the target field characteristics of the target doubts are analyzed to quickly and effectively determine the target doubt library corresponding to the target doubt, and insert the display fields corresponding to the target doubt into the corresponding target doubt library, which improves the efficiency and accuracy of the review of the audit analysis model results, and also ensures the efficiency and accuracy of sending the target doubts obtained after the review to the corresponding doubt library.

[0042] 2. By parsing the obtained target audit analysis model results, the business tags corresponding to the target audit analysis model results can be accurately and effectively determined. Secondly, the corresponding preset audit rules are retrieved from the preset rule path through the business tags to lock the audit dimensions according to the preset audit rules, and the target audit analysis model results are stratified according to the audit dimensions. The target audit analysis model results are analyzed one-to-one through the preset audit rules to obtain the corresponding audit parameters under each audit dimension. Finally, the corresponding audit parameters under each audit dimension are compared with the corresponding business parameter comparison table to accurately and effectively determine the doubts to be verified in the target audit analysis model results, thereby improving the accuracy of the audit of the target audit analysis model results and the efficiency and reliability of sending the doubts to be verified to the corresponding doubt library.

[0043] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0044] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] 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:

[0046] Figure 1 It is a flow chart of a method for implementing sending the audit analysis model result to the doubt database in an embodiment of the present invention;

[0047] Figure 2 This is a flowchart of step 1 in a method for implementing sending the audit analysis model results to a doubt database in an embodiment of the present invention;

[0048] Figure 3 This is a flowchart of step 2 in a method for implementing sending audit analysis model results to a doubt database in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0050] Embodiment 1:

[0051] This embodiment provides a method for sending the audit analysis model results to the doubt database, such as Figure 1 As shown, including:

[0052] Step 1: Based on the result check request, the target audit analysis model result is locked, and the first suspicious point audit is performed on the target audit analysis model result based on the preset audit rules, and the suspicious points to be verified are determined based on the first suspicious point audit results;

[0053] Step 2: Send the doubtful points to be verified to the corresponding lower-level verification terminal, and receive the supplementary data submitted by the lower-level verification terminal based on the doubtful points to be verified in real time, and conduct a second doubtful point review on the doubtful points to be verified based on the supplementary data;

[0054] Step 3: Determine the target doubt based on the second doubt review result, and analyze the target doubt to determine the target field characteristics. At the same time, determine the target doubt library corresponding to the target doubt and the project doubt in the target doubt library based on the target field characteristics, and insert the display field corresponding to the target doubt into the project doubt.

[0055] In this embodiment, the result viewing request is generated by the management terminal and is used to query the target audit analysis model result that needs to be analyzed from the result repository, wherein the target audit analysis model result is the audit analysis model result that needs to be viewed.

[0056] In this embodiment, the preset audit rules are set in advance and are used to verify the target audit analysis model results in order to verify whether there are any audit doubts in the obtained audit analysis model results.

[0057] In this embodiment, the first doubtful point review refers to reviewing the doubtful points existing in the target audit analysis model results through preset review rules, with the purpose of accurately and reliably confirming the doubtful points existing in the target audit analysis model results.

[0058] In this embodiment, the doubtful points to be verified refer to the audit items with problems determined from the target audit analysis model results after the first doubtful points are reviewed.

[0059] In this embodiment, the subordinate verification terminal is known in advance, and represents the source corresponding to different audit items.

[0060] In this embodiment, the supplementary data refers to the certification data submitted by the lower-level verification terminal for problems existing in the doubtful points to be verified, which can prove that there are no abnormalities in the doubtful points to be verified.

[0061] In this embodiment, the second doubtful point review refers to reviewing the doubtful points to be verified based on the supplementary data, with the purpose of checking whether the problems existing in the doubtful points to be verified can be solved based on the supplementary data.

[0062] In this embodiment, the target doubtful point refers to an audit item or audit result that still has problems after a second doubtful point review is performed on the doubtful point to be verified based on the supplementary data.

[0063] In this embodiment, the target field feature refers to the specific data type and the corresponding specific value situation corresponding to the target suspicious point.

[0064] In this embodiment, the target doubt database refers to a database in which target doubts can be stored, which is consistent with the data type of the target doubts.

[0065] In this embodiment, the project doubtful point refers to a data group in the target doubtful point library that belongs to the same audit project as the target doubtful point, so as to facilitate storing the target doubtful point in the corresponding data group in the target doubtful point library.

[0066] In this embodiment, the display field refers to the specific data content corresponding to the target suspicious point.

[0067] The beneficial effects of the above technical solution are: by reviewing the doubts in the target audit analysis model results, the doubts to be verified in the target audit analysis model results can be quickly and accurately locked; secondly, by sending the determined doubts to be verified to the corresponding lower-level verification terminal, and processing and analyzing the supplementary data submitted by the lower-level verification terminal, a second review of the doubts to be verified is achieved, which improves the efficiency and accuracy of resolving the doubts to be verified; finally, the target audit analysis model results that still have problems after the second doubt review are determined as target doubts, and the target field characteristics of the target doubts are analyzed to quickly and effectively determine the target doubt library corresponding to the target doubt, and insert the display fields corresponding to the target doubts into the corresponding target doubt library, which improves the efficiency and accuracy of reviewing the audit analysis model results, and also ensures the efficiency and accuracy of sending the target doubts obtained after the review to the corresponding doubt library.

[0068] Embodiment 2:

[0069] Based on Example 1, this example provides a method for sending the audit analysis model results to the doubt database, such as Figure 2 As shown, in step 1, based on the results, the audit analysis model results of the request lock target are viewed, including:

[0070] Step 101: Obtain a result viewing request, and parse the result viewing request to obtain a target query field corresponding to the result viewing request;

[0071] Step 102: Input the target query field into a preset convolutional neural network for analysis, obtain the target semantics corresponding to the target query field, generate a query statement based on the target semantics, and query the audit analysis model results in the preset result cache based on the search engine according to the query statement;

[0072] Step 103: Based on the query result, the audit analysis model result to be viewed corresponding to the result viewing request is obtained, and the audit analysis model result to be viewed is encapsulated to obtain the final target audit analysis model, wherein there is at least one audit analysis model result to be viewed.

[0073] In this embodiment, the target query field refers to the specific query content corresponding to the result viewing request, that is, the specific data content that can be queried.

[0074] In this embodiment, the preset convolutional neural network is set in advance and is used to parse and obtain the semantics of the target query field.

[0075] In this embodiment, the target semantics refers to the specific content corresponding to the target query field, that is, the main content corresponding to the target query field.

[0076] In this embodiment, the query statement refers to a statement generated according to the target semantics, which can be directly input into a search engine for query.

[0077] In this embodiment, the preset result cache library is set in advance and is used to store different audit analysis model results.

[0078] In this embodiment, the audit analysis model results to be viewed refer to the audit analysis model results in the preset result cache library that match the viewing request.

[0079] The beneficial effects of the above technical solution are: by parsing the viewing request, the target query field can be accurately and effectively determined according to the viewing request; secondly, the audit analysis model results in the preset demerit cache library are queried through the target query field to accurately and effectively lock the required audit analysis model results; finally, the obtained audit analysis model results are encapsulated to accurately and effectively obtain the target audit analysis model, which provides convenience and guarantee for the corresponding suspicious point review.

[0080] Embodiment 3:

[0081] Based on Example 2, this embodiment provides a method for implementing sending the audit analysis model results to a doubt database, and obtaining the audit analysis model results to be viewed corresponding to the result viewing request based on the query results, including:

[0082] Obtain the audit analysis model results to be viewed, and extract the identity identifier of the audit analysis model results to be viewed. At the same time, determine the subject identifier to be viewed based on the result viewing request, match the identity identifier with the subject identifier to be viewed, and obtain the subordinate relationship between the identity identifier and the subject identifier to be viewed based on the matching result;

[0083] Determine the audit business scope corresponding to the subject to be checked based on the subject identifier to be checked, and determine whether the audit analysis model result to be checked is within the audit business scope corresponding to the subject to be checked based on the subordinate relationship;

[0084] When the audit analysis model results to be reviewed are within the audit business scope corresponding to the subject to be reviewed, the audit analysis model results to be reviewed are locked; otherwise, the audit analysis model results to be reviewed are eliminated.

[0085] In this embodiment, the identity identifier refers to a marking symbol that can represent the identity information of the audit analysis model result, and different audit analysis model results correspond to different identity identifiers.

[0086] In this embodiment, the subject identifier to be viewed refers to a marking symbol corresponding to the audit object or audit business to be viewed as determined by the result viewing request.

[0087] In this embodiment, the subordinate relationship refers to the inclusive relationship between the identity identifier and the subject identifier to be checked, and specifically may represent whether the identity identifier is within the range defined by the subject identifier to be checked.

[0088] In this embodiment, the audit business scope refers to all audit items included in the subject to be reviewed.

[0089] The beneficial effect of the above technical solution is: by parsing the obtained audit analysis model results to be reviewed, the obtained audit analysis model results to be reviewed are matched with the audit business scope corresponding to the result review request, and accurate and effective verification of whether the audit analysis model results to be reviewed meet the review request is achieved, thereby ensuring the accuracy and reliability of the final target audit analysis model results, and facilitating the corresponding doubt review, so that when doubts exist, the doubts can be accurately and effectively sent to the corresponding doubt library.

[0090] Embodiment 4:

[0091] Based on Example 1, this embodiment provides a method for implementing sending the audit analysis model results to the doubt point library. In step 1, a first doubt point audit is performed on the target audit analysis model results based on a preset audit rule, and a doubt point to be verified is determined based on the first doubt point audit result, including:

[0092] Obtain the target audit analysis model results, extract the business tags of the target audit analysis model results, and retrieve the corresponding preset audit rules from the preset rule library based on the business tags;

[0093] Parse the preset audit rules, determine the audit dimensions included in the preset audit rules, and determine the audit strategy and business parameter comparison table corresponding to each audit dimension, where there is at least one audit dimension;

[0094] The target audit analysis model results are layered based on the audit dimensions to obtain the sub-audit analysis model results corresponding to each audit dimension, and the sub-audit analysis model results are audited for the first doubt points based on the audit strategy of the corresponding audit dimension to obtain the corresponding audit parameters under each audit dimension;

[0095] Compare the audit parameters corresponding to each audit dimension with the corresponding business parameter comparison table, and determine the sub-audit analysis model results whose audit parameters are not within the preset limit range of the business parameter comparison table as sub-doubts to be verified;

[0096] The sub-doubts to be verified corresponding to different audit dimensions are summarized to obtain the doubts to be verified corresponding to the audit result of the first doubt.

[0097] In this embodiment, the business label refers to the marking symbol of the business type corresponding to the target audit analysis model result, for example, the target audit analysis model result corresponds to the marking symbol of the business type of financial audit or management audit.

[0098] In this embodiment, the preset rule library is set in advance and is used to store known audit rules, and the preset audit rules stored internally are not unique.

[0099] In this embodiment, the audit dimension refers to the audit angle corresponding to the preset audit rules, so as to facilitate a comprehensive audit of the target audit analysis model results through the preset audit rules, in order to determine the doubts therein.

[0100] In this embodiment, the audit strategy refers to the method and means of auditing the results of the target audit analysis model, which may specifically be the audit steps corresponding to each dimension, and each dimension corresponds to an audit strategy.

[0101] In this embodiment, the business parameter comparison table is known in advance and is used to record the standard audit result parameters corresponding to the audit analysis model results under each audit dimension. It is a reference for determining whether there are any doubts in the target audit analysis model.

[0102] In this embodiment, stratifying the target audit analysis model results based on the audit dimension means splitting the target audit analysis model results into sub-results corresponding to the audit dimension one by one according to the determined audit dimension, in order to facilitate accurate and effective audit of the corresponding sub-results according to the audit strategy corresponding to each audit dimension and the business parameter comparison table.

[0103] In this embodiment, the sub-audit analysis model result refers to the result corresponding to the audit dimension obtained by splitting the target audit analysis model result according to the audit dimension, and is a part of the target audit analysis model result.

[0104] In this embodiment, the first doubtful point review refers to reviewing the sub-audit analysis model results according to the audit strategy under the corresponding dimension, with the purpose of accurately determining the doubtful points existing in the target audit analysis model results.

[0105] In this embodiment, the audit parameter refers to the result obtained after auditing the sub-audit analysis model result of the corresponding dimension through the audit strategy corresponding to each dimension, so as to facilitate determining whether there are doubts in the target audit analysis model result based on the audit parameter.

[0106] In this embodiment, the preset limited range is the standard value range corresponding to the target audit analysis model result under the current audit dimension specified in the business parameter comparison table under normal circumstances, and can be adjusted according to actual conditions.

[0107] In this embodiment, the doubtful points to be verified refer to the audit analysis model results in which the audit parameters determined under each audit dimension are not within the preset limit range after comparing the audit parameters obtained under each audit dimension with the preset limit range of the corresponding business parameter comparison table.

[0108] The beneficial effects of the above technical solution are: by parsing the obtained target audit analysis model results, the business tags corresponding to the target audit analysis model results can be accurately and effectively determined; secondly, the corresponding preset audit rules are retrieved from the preset rule path through the business tags, and the audit dimensions are locked according to the preset audit rules, and the target audit analysis model results are stratified according to the audit dimensions, so that the target audit analysis model results are analyzed one-to-one through the preset audit rules to obtain the corresponding audit parameters under each audit dimension; finally, the corresponding audit parameters under each audit dimension are compared with the corresponding business parameter comparison table, so as to accurately and effectively determine the doubts to be verified in the target audit analysis model results, thereby improving the accuracy of the audit of the target audit analysis model results, and also improving the efficiency and reliability of sending the doubts to be verified to the corresponding doubt library.

[0109] Embodiment 5:

[0110] Based on Example 4, this embodiment provides a method for implementing sending the audit analysis model results to the doubt point library, and determining the sub-audit analysis model results whose audit parameters are not within the preset limited range of the business parameter comparison table as sub-doubts to be verified, including:

[0111] Obtain the comparison results of the corresponding audit parameters and the corresponding business parameter comparison table under each audit dimension, and determine the sub-audit analysis model results whose audit parameters are within the preset limited range of the business parameter comparison table as unquestionable results;

[0112] The results without doubt shall be removed from the list corresponding to the target audit analysis model results, and the results without doubt shall be transferred to the processed results list at the same time.

[0113] In this embodiment, the unquestionable result refers to the sub-audit analysis model result in which the audit parameters obtained under each audit dimension are within the preset limit range of the business parameter comparison table, that is, the audit analysis model result without any doubt.

[0114] The beneficial effect of the above technical solution is: by removing the sub-audit analysis model results whose audit parameters are within the preset limited range of the business parameter comparison table from the list and transferring them to the processed results list, it is convenient to effectively distinguish between the audit analysis model results of doubtful points to be verified and those without doubts, thereby improving the efficiency of the review of the target audit analysis model results and avoiding repeated reviews.

[0115] Embodiment 6:

[0116] Based on Example 1, this embodiment provides a method for implementing sending the audit analysis model results to the doubt point library. In step 2, the doubt points to be verified are sent to the corresponding subordinate verification terminal, and the supplementary data submitted by the subordinate verification terminal based on the doubt points to be verified is received in real time, including:

[0117] Obtain the doubts to be verified, analyze the doubts to be verified, and extract the project features and business entities corresponding to the doubts to be verified;

[0118] Determine the audit items included in the doubts to be verified based on the item characteristics, determine the information source corresponding to each audit item based on the item attributes of the audit item, and determine the subordinate verification terminal corresponding to the audit item in the business entity based on the information source;

[0119] Constructing a distributed wireless transmission link between the audit terminal and the lower-level verification terminal, pre-transmitting preset test data in the distributed wireless transmission link, and determining a conduction result of the distributed wireless transmission link based on the pre-transmission result;

[0120] When all distributed wireless transmission links are connected, the audit items contained in the doubtful points to be verified are sent to the corresponding lower-level verification terminals based on the distributed wireless transmission links. At the same time, a monitoring thread is allocated to each distributed wireless transmission link, and the corresponding lower-level verification terminal is monitored in real time based on the monitoring thread for the reception process of the audit items. When it is monitored that the lower-level verification terminal has completed the reception, the communication node in the distributed wireless transmission link is authorized for uplink communication.

[0121] Based on the uplink communication authorization result, the supplementary data submitted by each lower-level verification terminal for the audit project is received in real time according to the communication node, and the supplementary data is forwarded based on the communication node. The received supplementary data is summarized based on the audit terminal to obtain the supplementary data submitted by the lower-level verification terminal based on the doubts to be verified.

[0122] In this embodiment, the project characteristics refer to the audit project type corresponding to the doubtful point to be verified and the specific audit result parameters corresponding to the current audit project.

[0123] In this embodiment, the business subject refers to the specific audit object corresponding to the doubtful point to be verified.

[0124] In this embodiment, the project attributes refer to the project type corresponding to each audit project and the operating environment of the project, etc., wherein the audit project is the specific business or audit data that needs to be audited, such as the total amount of expenditure in a certain period of time.

[0125] In this embodiment, there is at least one lower-level verification terminal, which corresponds one-to-one to each audit item.

[0126] In this embodiment, the preset test data is known in advance and is used to test whether the constructed distributed wireless transmission link is conductive, wherein pre-transmission is a process of performing a performance test on the distributed wireless transmission link through the preset test data.

[0127] In this embodiment, the monitoring thread is constructed in advance and is used to monitor the transmission progress of each transmission link in the distributed transmission link to the audit item.

[0128] In this embodiment, the communication node is an intermediary for receiving and forwarding data in a distributed wireless transmission link, and its purpose is to transmit the data to be transmitted to the corresponding terminal.

[0129] In this embodiment, the uplink communication authorization refers to the condition for granting data uploading to the subordinate verification terminal, so as to facilitate the audit terminal to accurately and effectively receive the supplementary data submitted by the subordinate verification terminal.

[0130] The beneficial effects of the above technical solution are: by parsing the obtained doubtful points to be verified, the project characteristics and business subjects corresponding to the doubtful points to be verified can be accurately and effectively determined; secondly, the audit items contained in the doubtful points to be verified can be accurately and effectively determined according to the project characteristics, and the subordinate verification terminals of each audit project can be accurately and effectively determined according to the information source of each audit project; finally, a distributed wireless transmission link is constructed between the audit terminal and the subordinate verification terminal, so that the audit items existing in the doubtful points to be verified can be sent to the corresponding subordinate verification terminals at the same time; at the same time, after the sending is completed, the distributed wireless transmission link is authorized for uplink communication, so as to effectively receive the supplementary data submitted by the subordinate verification terminal, so as to facilitate the accurate and effective review of the doubtful points to be verified based on the received supplementary data, thereby ensuring the accuracy of the analysis of the target audit analysis model results, and also facilitating the accuracy of sending the target audit analysis model results with doubtful points to the corresponding doubtful point library.

[0131] Embodiment 7:

[0132] Based on Example 1, this example provides a method for sending the audit analysis model results to the doubt database, such as Figure 3 As shown, in step 2, a second doubt point review is performed on the doubt points to be verified based on the supplementary data, including:

[0133] Step 201: Obtain the obtained supplementary data, parse the supplementary data, extract data features of the supplementary data, and determine the first structural features of the supplementary data based on the data features, and at the same time, extract the second structural features corresponding to the doubtful points to be verified;

[0134] Step 202: split the supplementary data based on the first structural feature and the second structural feature, and obtain sub-supplementary data corresponding to each audit item in the doubtful point to be verified based on the splitting result;

[0135] Step 203: split the sub-supplementary data into N business terms, and select a key business term set from the N business terms based on the business characteristics of the audit project, and extract the element information corresponding to each key business term in the key business term set;

[0136] Step 204: Determine the certification parameters of the audit items in the doubtful points to be verified by the supplementary data based on the element information, and conduct a second doubtful point audit on the doubtful points to be verified according to the certification parameters based on the preset audit rules.

[0137] In this embodiment, the data feature refers to the specific size value and data composition corresponding to the supplementary data.

[0138] In this embodiment, the first structural feature refers to the association relationship between the supplementary data corresponding to the doubtful points existing in different audit items contained in the supplementary data and the proof parameters corresponding to the different doubtful points in the audit items.

[0139] In this embodiment, the second structural feature refers to all audit items included in the doubtful point to be verified, so as to facilitate the determination of the corresponding relationship between each doubtful point in the audit item and the supplementary data.

[0140] In this embodiment, the sub-supplementary data refers to the supplementary data corresponding to each audit item submitted by the subordinate verification terminal.

[0141] In this embodiment, the business vocabulary refers to data segments obtained by discretizing the sub-supplementary data and splitting the sub-supplementary data into multiple vocabulary.

[0142] In this embodiment, the business characteristics refer to the audit requirements of the audit project and the specific audit steps involved.

[0143] In this embodiment, the key business vocabulary set refers to the key data segments that can prove that there is no doubt about the doubts to be verified, which are screened out from the N business vocabulary obtained according to the business characteristics of the audit project.

[0144] In this embodiment, the element information refers to the specific content corresponding to each key business word.

[0145] In this embodiment, the proof parameter refers to a data parameter that can be used to prove the doubt point to be verified.

[0146] The beneficial effects of the above technical solution are: by parsing the obtained supplementary data, the first structural feature of the supplementary data and the second structural feature of the doubtful point to be verified are determined, and the supplementary data is split according to the first structural feature and the second structural feature, so as to facilitate the one-to-one correspondence between the supplementary data and the audit items contained in the doubtful point to be verified. Secondly, the sub-supplementary data corresponding to each audit item is parsed to achieve accurate and effective acquisition of the proof parameters corresponding to each audit item. Finally, the second doubtful point audit is performed on the doubtful point to be verified according to the proof parameters through preset audit rules, thereby ensuring the accuracy and reliability of the review of the doubtful point to be verified, and facilitating the accurate and effective sending of the target audit analysis model results with doubtful points to the corresponding doubtful point library according to the review results, thereby ensuring the reliability of data transmission.

[0147] Embodiment 8:

[0148] Based on Example 1, this embodiment provides a method for implementing sending the audit analysis model results to a doubt point library. In step 3, a target doubt point is determined based on the second doubt point audit result, and the target doubt point is analyzed to determine the target field characteristics, including:

[0149] Obtain the second suspicious point audit result, and determine the unverified suspicious points that do not meet the preset audit requirement parameters in the second suspicious point audit result as target suspicious points. At the same time, transfer the unverified suspicious points that meet the preset audit requirement parameters to the processed result list;

[0150] Convert the target suspicious point into a format to obtain the text content corresponding to the target suspicious point, and split the text content to obtain M data segments corresponding to the text content;

[0151] Determine the field type and the corresponding field value of each data segment respectively, and determine the field feature vector corresponding to each data segment based on the field type and the corresponding field value;

[0152] Based on the position information of each data fragment in the text content, the influence weight corresponding to each field feature vector is determined, and based on the influence weight, the field feature vectors corresponding to each data fragment are weighted averaged to obtain the feature vector of the target suspicious point, and the target field feature corresponding to the target suspicious point is obtained based on the feature vector.

[0153] In this embodiment, the preset audit requirement parameters are set in advance, used to characterize the standard audit requirements, are parameters used to measure whether the results meet the audit requirements, and can be adjusted.

[0154] In this embodiment, converting the format of the target suspicious point to obtain the text content corresponding to the target suspicious point refers to converting the target suspicious point into text content capable of field feature analysis, wherein the text content is the specific data parameter corresponding to the target suspicious point.

[0155] In this embodiment, the data segments refer to data parameters obtained after splitting the text content.

[0156] In this embodiment, the field type refers to the data parameter type contained in the data segment, for example, it may be a process parameter type and a material parameter type.

[0157] In this embodiment, the field value refers to the specific value size corresponding to the data segment.

[0158] In this embodiment, the field feature vector refers to a vector that can characterize the field features of a data segment, including values ​​and associations between data.

[0159] In this embodiment, the influence weight refers to the importance of different data segments in the entire text content, and a larger value indicates a greater influence.

[0160] In this embodiment, the feature vector is obtained by weighted averaging the field feature vectors of each data segment, and is used to characterize the parameter values ​​of the target suspicious point and the vector parameters of the association relationship between the data.

[0161] The beneficial effects of the above technical solution are: by analyzing the results of the second suspicious point review, the target suspicious points that are still in doubt after the second review can be accurately and effectively determined; secondly, the target suspicious points are split and analyzed to accurately and effectively lock the feature vectors of the target suspicious points, and accurately and effectively analyze and determine the target field characteristics of the target suspicious points based on the feature vectors, so that the target suspicious points can be sent to the corresponding suspicious point library according to the determined target field characteristics, thereby ensuring the reliability and accuracy of the sending of the target suspicious points.

[0162] Embodiment 9:

[0163] Based on Example 1, this embodiment provides a method for implementing sending the audit analysis model results to a doubt point library. In step 3, the target doubt point library corresponding to the target doubt point and the project doubt point of the target doubt point in the target doubt point library are determined based on the target field characteristics, and the display field corresponding to the target doubt point is inserted into the project doubt point, including:

[0164] Acquire the target field features, and determine the class identifier of the target suspicious point based on the target field features;

[0165] Extract the library identifier corresponding to the preset doubtful point library, and match the class identifier with the library identifier to obtain the target doubtful point library corresponding to the target field feature. At the same time, determine the target feature value corresponding to the target doubtful point based on the target field feature, and traverse the stored project doubtful points in the target doubtful point library based on the target feature value, and determine the project doubtful point based on the traversal result;

[0166] Extracting the project characteristic value of the stored doubtful point in the project doubtful point, and determining the theoretical insertion position of the target doubtful point in the project doubtful point based on the project characteristic value and the target characteristic value corresponding to the target doubtful point;

[0167] The display fields corresponding to the target doubtful points are inserted into the theoretical insertion positions, and the display fields in the project doubtful points are smoothed based on the insertion results to complete the insertion of the display fields of the target doubtful points into the project doubtful points.

[0168] In this embodiment, the class identifier refers to a marking symbol that can represent the type of the target.

[0169] In this embodiment, the preset doubtful point library is known in advance and is used to store different types of doubtful points, wherein the library identifier is a marking symbol used to characterize the identity of the preset doubtful point library.

[0170] In this embodiment, the target feature value refers to the mean or variance value corresponding to the target suspicious point.

[0171] In this embodiment, the stored project doubtful points are doubtful points that have been stored in the target doubtful point library and are not unique. The purpose is to determine the storage location in the target doubtful point library according to the target feature value corresponding to the target doubtful point.

[0172] In this embodiment, the item characteristic value refers to the mean or variance of the stored doubtful points.

[0173] In this embodiment, the theoretical insertion position refers to the specific insertion position of the target doubt point in the corresponding project doubt point in the target doubt point library.

[0174] The beneficial effects of the above technical solution are: by parsing the obtained target field characteristics, the target doubt library of the target doubt and the project doubts in the target doubt library can be accurately and effectively locked according to the target field characteristics; secondly, according to the project characteristic value of the target doubt, the specific insertion position of the target doubt in the project doubt can be accurately and effectively analyzed; finally, according to the analysis results, the display field of the target doubt is accurately and effectively stored in the project doubt, thereby ensuring the reliability and accuracy of sending the target doubt to the doubt library.

[0175] Embodiment 10:

[0176] Based on Example 1, this example provides a method for sending the audit analysis model results to a doubt point library, inserting the display field corresponding to the target doubt point into the project doubt point, including:

[0177] Obtain the insertion result of the display field of the target suspicious point, and after completing the insertion of the display field of the target suspicious point, extract the flow information corresponding to the target suspicious point in each processing link, wherein the processing link includes the first suspicious point review, the second suspicious point review and inserting the display field corresponding to the target suspicious point into the project suspicious point;

[0178] The flow information corresponding to each processing link is recorded to obtain a record report of the target suspicious point, and the record report is stored to complete the record of the complete flow process of the target suspicious point.

[0179] In this embodiment, the flow information refers to each step information of the target doubt point from being determined as the final doubt point to being sent to the doubt point library.

[0180] In this embodiment, the record report refers to a report obtained by recording the flow information through the record template.

[0181] The beneficial effect of the above technical solution is: by determining the flow information corresponding to the target doubts in each processing link and recording and storing the flow information, the process of sending the audit analysis model results to the doubt library can be accurately and effectively recorded, which is convenient for management personnel to retrieve when verifying the data.

[0182] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for sending the audit analysis model results to a doubt database, characterized in that: include: Step 1: Based on the result check request, the target audit analysis model result is locked, and the first suspicious point audit is performed on the target audit analysis model result based on the preset audit rules, and the suspicious points to be verified are determined based on the first suspicious point audit results; Step 2: Send the doubtful points to be verified to the corresponding lower-level verification terminal, and receive the supplementary data submitted by the lower-level verification terminal based on the doubtful points to be verified in real time, and conduct a second doubtful point review on the doubtful points to be verified based on the supplementary data; Step 3: Determine the target doubt based on the second doubt review result, analyze the target doubt, determine the target field characteristics, and at the same time, determine the target doubt library corresponding to the target doubt and the project doubt of the target doubt in the target doubt library based on the target field characteristics, and insert the display field corresponding to the target doubt into the project doubt; Among them, in step 3, the target doubt library corresponding to the target doubt and the project doubt of the target doubt in the target doubt library are determined based on the target field characteristics, and the display field corresponding to the target doubt is inserted into the project doubt, including: Acquire the target field features, and determine the class identifier of the target suspicious point based on the target field features; Extract the library identifier corresponding to the preset doubtful point library, and match the class identifier with the library identifier to obtain the target doubtful point library corresponding to the target field feature. At the same time, determine the target feature value corresponding to the target doubtful point based on the target field feature, and traverse the stored project doubtful points in the target doubtful point library based on the target feature value, and determine the project doubtful point based on the traversal result; Extracting the project characteristic value of the stored doubtful point in the project doubtful point, and determining the theoretical insertion position of the target doubtful point in the project doubtful point based on the project characteristic value and the target characteristic value corresponding to the target doubtful point; The display fields corresponding to the target doubtful points are inserted into the theoretical insertion positions, and the display fields in the project doubtful points are smoothed based on the insertion results to complete the insertion of the display fields of the target doubtful points into the project doubtful points.

2. A method for implementing sending the audit analysis model results to a doubt database according to claim 1, characterized in that: In step 1, based on the results, view the audit analysis model results of the request lock target, including: Obtain a result viewing request, parse the result viewing request, and obtain the target query field corresponding to the result viewing request; Input the target query field into the preset convolutional neural network for analysis, obtain the target semantics corresponding to the target query field, generate a query statement based on the target semantics, and query the audit analysis model results in the preset result cache based on the search engine according to the query statement; Based on the query results, the audit analysis model results to be viewed corresponding to the result viewing request are obtained, and the audit analysis model results to be viewed are encapsulated to obtain the final target audit analysis model, wherein there is at least one audit analysis model result to be viewed.

3. A method for sending the audit analysis model results to the doubt database according to claim 2, characterized in that: Based on the query results, the audit analysis model results to be viewed corresponding to the result viewing request are obtained, including: Obtain the audit analysis model results to be viewed, and extract the identity identifier of the audit analysis model results to be viewed. At the same time, determine the subject identifier to be viewed based on the result viewing request, match the identity identifier with the subject identifier to be viewed, and obtain the subordinate relationship between the identity identifier and the subject identifier to be viewed based on the matching result; Determine the audit business scope corresponding to the subject to be checked based on the subject identifier to be checked, and determine whether the audit analysis model result to be checked is within the audit business scope corresponding to the subject to be checked based on the subordinate relationship; When the audit analysis model results to be reviewed are within the audit business scope corresponding to the subject to be reviewed, the audit analysis model results to be reviewed are locked; otherwise, the audit analysis model results to be reviewed are eliminated.

4. The method for sending the audit analysis model result to the doubt database according to claim 1, characterized in that: In step 1, the first suspicious point audit is performed on the target audit analysis model result based on the preset audit rules, and the suspicious points to be verified are determined based on the first suspicious point audit results, including: Obtain the target audit analysis model results, extract the business tags of the target audit analysis model results, and retrieve the corresponding preset audit rules from the preset rule library based on the business tags; Parse the preset audit rules, determine the audit dimensions included in the preset audit rules, and determine the audit strategy and business parameter comparison table corresponding to each audit dimension, where there is at least one audit dimension; The target audit analysis model results are layered based on the audit dimensions to obtain the sub-audit analysis model results corresponding to each audit dimension, and the sub-audit analysis model results are audited for the first doubt points based on the audit strategy of the corresponding audit dimension to obtain the corresponding audit parameters under each audit dimension; Compare the audit parameters corresponding to each audit dimension with the corresponding business parameter comparison table, and determine the sub-audit analysis model results whose audit parameters are not within the preset limit range of the business parameter comparison table as sub-doubts to be verified; The sub-doubts to be verified corresponding to different audit dimensions are summarized to obtain the doubts to be verified corresponding to the audit result of the first doubt.

5. A method for implementing sending the audit analysis model result to the doubt database according to claim 4, characterized in that: The sub-audit analysis model results whose audit parameters are not within the preset limit range of the business parameter comparison table are determined as sub-doubts to be verified, including: Obtain the comparison results of the corresponding audit parameters and the corresponding business parameter comparison table under each audit dimension, and determine the sub-audit analysis model results whose audit parameters are within the preset limited range of the business parameter comparison table as unquestionable results; The results without doubt shall be removed from the list corresponding to the target audit analysis model results, and the results without doubt shall be transferred to the processed results list at the same time.

6. A method for implementing sending the audit analysis model result to the doubt database according to claim 1, characterized in that: In step 2, a second review of the doubts to be verified is conducted based on the supplementary data, including: Acquire the obtained supplementary data, parse the supplementary data, extract data features of the supplementary data, determine the first structural features of the supplementary data based on the data features, and extract the second structural features corresponding to the doubtful points to be verified; The supplementary data is split based on the first structural feature and the second structural feature, and sub-supplementary data corresponding to each audit item in the doubtful point to be verified is obtained based on the splitting result; The sub-supplementary data is split into terms to obtain N business terms, and a key business term set is selected from the N business terms based on the business characteristics of the audit project, and the element information corresponding to each key business term in the key business term set is extracted; Based on the element information, the proof parameters of the audit items in the doubtful points to be verified are determined using the supplementary data, and a second doubtful point audit is performed based on the proof parameters and the doubtful points to be verified based on the preset audit rules.

7. The method for sending the audit analysis model result to the doubt database according to claim 1, characterized in that: Insert the display fields corresponding to the target doubt into the project doubt, including: Obtain the insertion result of the display field of the target suspicious point, and after completing the insertion of the display field of the target suspicious point, extract the flow information corresponding to the target suspicious point in each processing link, wherein the processing link includes the first suspicious point review, the second suspicious point review and inserting the display field corresponding to the target suspicious point into the project suspicious point; The flow information corresponding to each processing link is recorded to obtain a record report of the target suspicious point, and the record report is stored to complete the record of the complete flow process of the target suspicious point.

Images