A document review method and related equipment

Abstract

The present application discloses a document review method and related equipment, the method comprising: after obtaining a document to be tested (for example, financial statements used to generate an audit report, and documents such as report notes), first determining the review rules to be used based on the attribute description information of the document to be tested (for example, an audit template or a company industry, etc.); then reviewing the document to be tested according to the review rules to be used to obtain a document review result, so that automatic document review processing can be achieved for the document data, thereby effectively overcoming the defects of manual review, and further facilitating improving the document review effect.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a document review method and related equipment. Background Art

[0002] For some industries (e.g., auditing industry, etc.), practitioners in these industries (e.g., accountants, lawyers, etc.) can provide clients with some document writing services (e.g., audit reports, etc.). In addition, in order to improve the effect of document writing services, after these practitioners have completed writing a document, relevant personnel can review the document, so as to avoid errors in the documents received by clients as much as possible, thereby improving the effect of document writing services.

[0003] However, due to the defects in the above-mentioned manual review process, the document review effect is poor, which leads to poor document writing service effect. Summary of the invention

[0004] The main purpose of the embodiments of the present application is to provide a document review method and related equipment, which can improve the document review effect.

[0005] The present application provides a document review method, the method comprising:

[0006] Get the document to be tested;

[0007] Determining the audit rules to be used according to the attribute description information of the document to be detected;

[0008] According to the audit rules to be used, the document to be detected is audited to obtain a document audit result.

[0009] In a possible implementation manner, determining the audit rules to be used according to the attribute description information of the document to be detected includes:

[0010] The candidate audit rule corresponding to the attribute description information is searched from the preset mapping relationship and determined as the audit rule to be used; wherein the preset mapping relationship includes the correspondence between the attribute description information and the candidate audit rule corresponding to the attribute description information.

[0011] In a possible implementation manner, the attribute description information includes document template description information and / or company industry affiliation information.

[0012] In a possible implementation, the method further includes:

[0013] Acquire the attribute information to be used provided by the target user for the document to be detected; wherein the attribute information to be used includes attribute description information of the document to be detected.

[0014] In a possible implementation manner, the process of determining the document review result includes:

[0015] Performing element extraction processing on the document to be detected to obtain element information to be used;

[0016] The document review result is determined according to the element information to be used and the review rule to be used.

[0017] In a possible implementation manner, the extracting elements from the document to be detected to obtain element information to be used includes:

[0018] According to the pre-constructed document content index template, element extraction processing is performed on the document to be detected to obtain element information to be used.

[0019] In a possible implementation manner, the process of determining the element information to be used includes:

[0020] If it is determined that there is a fifth table satisfying the third condition in the document to be detected, extracting at least one table element information from the fifth table according to a preset table element extraction rule corresponding to the third condition;

[0021] The element information to be used is determined according to the at least one table element information.

[0022] In a possible implementation, the table element information includes a first table value and identification information of the first table value; wherein the identification information of the first table value is determined based on the table identification, the row name to be used, and the column name to be used of the fifth table; the row name to be used refers to the row name of the table row where the first table value is located in the fifth table; the column name to be used refers to the column name of the table column where the first table value is located in the fifth table.

[0023] In a possible implementation manner, the audit rules to be used include M rules to be confirmed; wherein M is a positive integer;

[0024] The step of determining the document review result according to the element information to be used and the review rule to be used includes:

[0025] Querying the element description information corresponding to the mth rule to be confirmed from the element information to be used; wherein m is a positive integer, m≤M;

[0026] Determine, according to the mth rule to be confirmed and the element description information corresponding to the mth rule to be confirmed, a rule confirmation result corresponding to the mth rule to be confirmed; wherein m is a positive integer, m≤M;

[0027] The document review result is determined according to the rule confirmation result corresponding to the 1st rule to be confirmed to the rule confirmation result corresponding to the Mth rule to be confirmed.

[0028] In a possible implementation, the method further includes:

[0029] Initializing the content of the document to be processed according to the document to be detected;

[0030] Determining whether there is a first table satisfying a first condition in the content of the document to be processed;

[0031] When it is determined that there is a first table satisfying the first condition in the content of the document to be processed, determining a second table satisfying the second condition from the content of the document to be processed according to the document position of the first table;

[0032] Merge the first table with the second table to obtain a third table;

[0033] According to the third table, updating the content of the document to be processed, and continuing to perform the step of determining whether there is a first table satisfying the first condition in the content of the document to be processed;

[0034] The step of auditing the document to be detected according to the audit rule to be used to obtain the document audit result includes:

[0035] When it is determined that the first table satisfying the first condition does not exist in the content of the document to be processed, the content of the document to be processed is audited according to the audit rule to be used to obtain a document audit result.

[0036] In a possible implementation manner, the process of determining the third table includes:

[0037] Compare the header of the first table with the header of the second table to obtain a common header;

[0038] Delete the common table headers in the first table to obtain a fourth table;

[0039] According to the preset splicing position corresponding to the common table header, the fourth table and the second table are spliced ​​to obtain a third table.

[0040] In a possible implementation, the method further includes:

[0041] According to the document review result, error marking processing is performed on the document to be detected to obtain a marked document;

[0042] and / or,

[0043] Based on the document review results, an audit report to be used is generated.

[0044] The present application also provides a document review device, including:

[0045] A document acquisition unit, used to acquire the document to be detected;

[0046] A rule determination unit, used to determine the audit rules to be used according to the attribute description information of the document to be detected;

[0047] The document review unit is used to review the document to be detected according to the review rule to be used to obtain a document review result.

[0048] The embodiment of the present application further provides a device, the device comprising: a processor, a memory, and a system bus;

[0049] The processor and the memory are connected via the system bus;

[0050] The memory is used to store one or more programs, and the one or more programs include instructions. When the instructions are executed by the processor, the processor executes any implementation of the document review method provided in the embodiments of the present application.

[0051] An embodiment of the present application also provides a computer-readable storage medium, in which instructions are stored. When the instructions are executed on a terminal device, the terminal device executes any implementation of the document review method provided in the embodiment of the present application.

[0052] The embodiment of the present application also provides a computer program product. When the computer program product is run on a terminal device, the terminal device executes any implementation of the document review method provided in the embodiment of the present application.

[0053] Based on the above technical solution, this application has the following beneficial effects:

[0054] In the technical solution provided by the present application, after obtaining the document to be tested (for example, financial statements used to generate audit reports, and documents such as notes to the statements), the audit rules to be used are first determined based on the attribute description information of the document to be tested (for example, the audit template or the company industry, etc.); then, the document to be tested is audited and processed according to the audit rules to be used to obtain the document audit results. In this way, automatic document audit processing can be performed on the document data, thereby effectively overcoming the defects of manual audit, and further helping to improve the document audit effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0056] Figure 1 A flowchart of a document review method provided in an embodiment of the present application;

[0057] Figure 2 A schematic diagram of a document review process provided in an embodiment of the present application;

[0058] Figure 3 A schematic diagram of a document content index template provided in an embodiment of the present application;

[0059] Figure 4 A schematic diagram of a table provided in an embodiment of the present application;

[0060] Figure 5 A schematic diagram of a marked document provided in an embodiment of the present application;

[0061] Figure 6 A schematic diagram of an extra-long table provided in an embodiment of the present application;

[0062] Figure 7 A schematic diagram of a third table provided in an embodiment of the present application;

[0063] Figure 8 A schematic diagram of a preset table element extraction rule corresponding to a third condition provided in an embodiment of the present application;

[0064] Fig. 9 A schematic diagram of another table provided in an embodiment of the present application;

[0065] Fig.10 A schematic diagram of an operator provided in an embodiment of the present application;

[0066] Fig.11 A schematic diagram of the structure of a document review device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0067] The inventor discovered in his research on document review that for an audit report (or, financial statements used to generate an audit report, and notes to the statements, etc.), because the audit report has a lot of content and complex information, the audit report needs to be reviewed collaboratively by multiple accountants, resulting in a relatively low review efficiency for the audit report, which in turn leads to a relatively poor manual review effect on the audit report.

[0068] It should be noted that the business background of the above-mentioned "audit report" is as follows: accounting firms in the audit industry must issue audit reports corresponding to annual reports, semi-annual reports, and quarterly reports for various corporate clients every year to prove to the market that the information in annual reports, semi-annual reports and quarterly reports is authentic and reliable, and to disclose any problems discovered during the audit process in a timely manner.

[0069] Based on the above findings, in order to solve the technical problems shown in the background technology part, an embodiment of the present application provides a document review method, which may include: after obtaining the document to be tested (for example, financial statements used to generate audit reports, and documents such as report notes), first determine the review rules to be used based on the attribute description information of the document to be tested (for example, audit templates or company industries, etc.); then, according to the review rules to be used, review the document to be tested to obtain the document review results, so that automatic document review processing can be achieved for the document data, thereby effectively overcoming the defects of manual review, and further helping to improve the document review effect.

[0070] In addition, the embodiments of the present application do not limit the execution subject of the document review method. For example, the execution subject of the document review method provided in the embodiments of the present application may be a data processing device such as a terminal device or a server. The terminal device may be a smart phone, a computer, a personal digital assistant (PDA) or a tablet computer. The server may be an independent server, a cluster server or a cloud server.

[0071] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0072] Method Example 1

[0073] See also Figure 1 , which is a flow chart of a document review method provided in an embodiment of the present application.

[0074] The document review method provided in the embodiment of the present application includes S1-S3:

[0075] S1: Get the document to be tested.

[0076] The above-mentioned "document to be tested" refers to document data that needs to be reviewed and processed; and the embodiment of the present application does not limit the "document to be tested". For example, when the document review method provided by the embodiment of the present application is applied to the audit field, the "document to be tested" may include audit documents.

[0077] The above-mentioned "audit document" refers to document data in the audit field; and the embodiment of the present application does not limit the "audit document", for example, the "audit document" may include an audit report. For another example, the above-mentioned "audit document" may also include financial statements and notes to the statements used to generate audit reports. Among them, the above-mentioned "financial statements and notes to the statements" refer to document data generated by a large number of audit working papers; and the "financial statements and notes to the statements" can be used to generate audit reports.

[0078] It should be noted that since the above-mentioned “financial statements and notes to the financial statements” not only record the information recorded in the audit report, but also record a large amount of information that is not recorded in the audit report, the “financial statements and notes to the financial statements” can more accurately represent the audit information carried by the audit report. Therefore, in order to improve the audit effect of the audit report, the financial statements and notes to the financial statements used to generate the audit report can be identified as the above-mentioned “documents to be tested” so that the audit process of the audit report can be realized with the help of the audit process of the “financial statements and notes to the financial statements”.

[0079] It should also be noted that the above-mentioned "report" can be an Excel document; and the above-mentioned "report notes" can be a Word document.

[0080] In addition, the embodiment of the present application does not limit the acquisition process of the above-mentioned "document to be detected", for example, it may specifically include: reading the above-mentioned "document to be detected" from a preset storage space. For another example, the above-mentioned "document to be detected" may also be uploaded to the terminal device by the user through a human-computer interaction method, so that the terminal device (or a server capable of data communication with the terminal device) can obtain the "document to be detected".

[0081] S2: Determine the audit rules to be used based on the attribute description information of the document to be detected.

[0082] The above-mentioned "attribute description information of the document to be detected" is used to describe the attributes of the document to be detected; and the embodiment of the present application does not limit the "attribute description information of the document to be detected", for example, it may specifically include the document template description information of the document to be detected, and at least one of the company industry affiliation information of the document to be detected.

[0083] The above-mentioned "document template description information of the document to be detected" is used to describe the document template used by the document to be detected; and the embodiment of the present application does not limit the "document template description information of the document to be detected", for example, it may include at least one of the characteristics of the document template used by the document to be detected, the name of the document template, and the identifier of the document template.

[0084] It should be noted that the embodiments of the present application do not limit the above-mentioned “document template used for the document to be tested”. For example, when the document review method provided in the embodiments of the present application is applied to the audit field, the “document template used for the document to be tested” can be any audit template (for example, SASAC / listed company template, IPO (initial public offering) template, or small and medium-sized enterprise template and other audit templates).

[0085] The above-mentioned "company industry affiliation information of the document to be tested" is used to describe the industry of the enterprise to which the document to be tested belongs; and the embodiment of the present application does not limit the "industry of the enterprise to which the document to be tested belongs", for example, it can be real estate, finance, agriculture, forestry, animal husbandry and fishery, white appliances, or investment operations and other industries.

[0086] It should be noted that, for the above-mentioned “enterprise to which the document to be tested belongs”, if the document to be tested includes an audit report, then the audit report is issued for the “enterprise to which the document to be tested belongs”; if the document to be tested includes financial statements and notes to the financial statements, then the audit report issued for the “enterprise to which the document to be tested belongs” can be generated using the “document to be tested”.

[0087] In addition, the embodiments of the present application do not limit the method for obtaining the above-mentioned "attribute description information of the document to be detected". For example, it may specifically include: obtaining the attribute information to be used provided by the target user for the document to be detected; wherein the attribute information to be used includes the attribute description information of the document to be detected.

[0088] The above-mentioned “target user” is used to indicate the provider of the attribute information to be used; and the embodiment of the present application does not limit the “target user”, for example, it may be the writer, uploader or reviewer of the above-mentioned “document to be detected”.

[0089] It should be noted that "writer" refers to the person who writes the document to be detected; "uploader" refers to the person who uploads the document to be detected to the terminal device; "reviewer" refers to the person responsible for reviewing the document to be detected. In addition, the embodiment of the present application does not limit the association relationship between the "writer", "uploader" and "reviewer", for example, these three can be the same person; or, these three are different. For another example, the above-mentioned "writer" and "uploader" are the same person (for example, Zhang San), and the above-mentioned "reviewer" is another person (for example, Li Si). For another example, the "uploader" and "reviewer" are the same person (for example, Li Si), and the above-mentioned "writer" is another person (for example, Zhang San).

[0090] The above-mentioned "attribute information to be used" is used to describe the relevant attributes of the document to be detected (for example, document template description information, company industry affiliation information, etc.), so that the "attribute information to be used" includes the above-mentioned "attribute description information of the document to be detected".

[0091] In addition, the present application does not limit the provision method of the above-mentioned "attribute information to be used". For example, the target user can select some attributes of the document to be detected on the terminal device (for example, Figure 2 It can be implemented by clicking "Select Audit Template" and "Select Company Industry" as shown in the figure.

[0092] Based on the above-mentioned “attribute description information of the document to be tested”, it can be known that when it is necessary to audit the document to be tested, the target user can use a preset input device (for example, a keyboard, a mouse, a touch screen, etc.) on the terminal device to provide the attribute information to be used of the document to be tested to the terminal device, so that the terminal device (or a server capable of data communication with the terminal device) can obtain the attribute description information of the document to be tested from the attribute information to be used, so that the audit rules to be used when auditing the document to be tested (that is, the above-mentioned “audit rules to be used”) can be determined based on the attribute description information of the document to be tested.

[0093] The above-mentioned "audit rules to be used" refer to the audit rules required to be used when auditing the documents to be inspected; and the embodiments of the present application do not limit the above-mentioned "audit rules to be used", for example, they may include: document format check rules, entity name consistency check rules, report / note consistency check rules, digital cross-reference relationship check rules, and at least one of element integrity check rules.

[0094] The above “document format check rules” are used to check the document format for the document to be tested (such as Figure 2The "Document Format Check" shown in the figure); and the "Document Format Check" may include: checking whether the use of the first to fourth level title formats is correct, checking the correctness of the title serial number encoding, checking the correctness of the page number encoding, checking whether the cross-page display of extra-long tables is correct, etc.

[0095] The above “Entity Name Consistency Check Rules” are used to check the entity names of documents to be tested (such as Figure 2 The “entity name check” shown); and the “entity name check” may include: checking whether the entity names in the document to be tested are consistent (that is, determining whether the name and abbreviation of the same company in the document to be tested are always consistent in the document to be tested).

[0096] It should be noted that when the document audit method provided in the embodiment of the present application is applied to the audit field, the above-mentioned "entity name" can be the name (or abbreviation) of various enterprises and institutions involved in the document to be detected. In addition, the audit document requires that the name and abbreviation of the same enterprise should always be consistent in the document. In addition, the entity name can be extracted with the help of natural language processing (Natural Language Processing, NLP) technology.

[0097] The above “Report / Note Consistency Check Rules” are used to perform cross-document consistency checks on the documents to be tested (such as Figure 2 and the “cross-document consistency check” may include: checking whether the contents recorded in the financial statements in the document to be tested are consistent with the contents recorded in the notes to the statements.

[0098] It should be noted that when the document review method provided in the embodiment of the present application is applied to the auditing field, it is usually required that the contents recorded in the financial statements (for example, cash and cash equivalents, accounts receivable) be consistent with the contents recorded in the notes to the financial statements.

[0099] The above “digital cross-reference check rules” are used to check the cross-reference relationship of the document to be tested (such as Figure 2 and the “cross-reference check” may include: checking the calculation logic between the numbers in various tables and across tables in the document to be tested to ensure that the numbers before and after are aligned and the accounts are balanced.

[0100] The above-mentioned “element integrity check rules” are used to perform integrity checks on the documents to be tested (such as Figure 2The “completeness check” shown in the figure); and the “completeness check” may include: checking whether the document to be tested covers the audit document mandatory elements corresponding to the enterprise to which the document to be tested belongs. It should be noted that the above “mandatory elements of the audit report” are determined in advance based on the enterprise classification and industry of the above “enterprise to which the document to be tested belongs”.

[0101] In addition, the embodiments of the present application do not limit the determination process of the above-mentioned "audit rules to be used" (that is, the implementation method of S2). For example, it may specifically include: searching for the candidate audit rules corresponding to the above-mentioned "attribute description information of the document to be detected" from the pre-constructed preset mapping relationship, and determining them as the audit rules to be used.

[0102] The above-mentioned "preset mapping relationship" is used to record the correspondence between each attribute description information and each candidate audit rule. For example, the above-mentioned "preset mapping relationship" may include: the correspondence between the first attribute description information and the first candidate audit rule, the correspondence between the second attribute description information and the second candidate audit rule, ..., and the correspondence between the Ith attribute description information and the Ith candidate audit rule. Wherein, I is a positive integer. Wherein, the above-mentioned "attribute description information" includes at least one of the document template description information and the company industry affiliation information.

[0103] It can be seen that after obtaining the above-mentioned "attribute description information of the document to be detected", the "attribute description information of the document to be detected" can be matched with the I attribute description information in the preset mapping relationship respectively, so that when it is determined that the "attribute description information of the document to be detected" matches the i-th attribute description information in the preset mapping relationship successfully, the i-th candidate audit rule corresponding to the i-th attribute description information is determined as the audit rule to be used. Among them, i∈{1, 2, 3, ..., I}. Based on this, it can be seen that the above-mentioned "preset mapping relationship" includes the correspondence between the above-mentioned "attribute description information of the document to be detected" and the audit rule to be used (that is, the candidate audit rule corresponding to the above-mentioned "attribute description information of the document to be detected").

[0104] It should be noted that the embodiments of the present application do not limit the generation process of the above-mentioned "candidate audit rules", for example, it can be implemented by manual aggregation. For example, when the document audit method provided by the embodiments of the present application is applied to the audit field, relevant personnel can classify and aggregate thousands of checkpoints required for quality inspection of audit documents to obtain the above-mentioned "candidate audit rules".

[0105] It should also be noted that the embodiments of the present application do not limit the construction process of the above-mentioned "preset mapping relationship". For example, it can be implemented by manual construction.

[0106] Based on the relevant content of S2 above, it can be known that for the document to be detected, after obtaining the attribute description information of the document to be detected, the candidate review rules that have a corresponding relationship with the "attribute description information of the document to be detected" can be searched from the pre-constructed preset mapping relationship; and then the found candidate review rules are determined as the review rules to be used, so that the document to be detected can be reviewed and processed in accordance with the review rules to be used.

[0107] S3: According to the audit rules to be used, the document to be tested is audited and processed to obtain the document audit result.

[0108] The above-mentioned "document review results" are used to describe the errors existing in the document to be tested (for example, document format errors, inconsistent entity names, inconsistent content between financial statements and notes to the financial statements, errors in digital cross-references, missing required elements in the audit document, etc.).

[0109] In addition, the embodiments of the present application do not limit the determination process of the above-mentioned "document review results". For example, it can be implemented with the help of a pre-built machine learning model with document review function.

[0110] In addition, in order to improve the document review effect, the embodiment of the present application also provides another possible implementation method for determining the above-mentioned "document review result", which may specifically include step 11-step 12:

[0111] Step 11: Extract elements from the document to be detected to obtain the element information to be used.

[0112] The above-mentioned “element extraction processing” is used to structurally index the document to be detected so that the document to be detected can be represented by a large amount of structured data, thereby achieving the structuring of the document content of the document to be detected.

[0113] The above-mentioned “element information to be used” is used to represent the element content recorded in the document to be detected; and the “element information to be used” can represent the document to be detected in a structured form.

[0114] In addition, the embodiments of the present application do not limit the above-mentioned "element information to be used", for example, it may include at least one element description information. The above-mentioned "element description information" may include at least one of the element to be extracted, the value of the element to be extracted, and the position description data of the element to be extracted.

[0115] The above-mentioned “elements to be extracted” are used to describe the parameter information that needs to be extracted from the document to be detected; and the embodiment of the present application does not limit the “elements to be extracted”, for example, it can be parameter information such as company name, amount, etc.

[0116] The above-mentioned “value of the element to be extracted” refers to the value of the above-mentioned “element to be extracted” in the document to be detected (for example, XXX Co., Ltd., ten thousand yuan, etc.).

[0117] It should be noted that the above-mentioned “elements to be extracted” and the above-mentioned “values ​​of the elements to be extracted” can constitute a key-value pair, so that the key-value pair can be used in the subsequent document review process.

[0118] The above-mentioned "position description data of the element to be extracted" is used to describe the position of the above-mentioned "value of the element to be extracted" in the document to be detected; and the embodiment of the present application does not limit the "position description data of the element to be extracted", for example, it can be the 10th to 15th characters of the 2nd line of the 3rd paragraph on page 10, etc.

[0119] The embodiment of the present application does not limit the implementation method of step 11. For example, it can be implemented with the help of a pre-built machine learning model with feature extraction function. For another example, in order to improve the feature extraction effect, step 11 can specifically include: according to the pre-built document content index template, perform feature extraction processing on the document to be detected to obtain the feature information to be used.

[0120] The above-mentioned "document content index template" is used to record the parameter information to be extracted under each title in the document to be detected. That is, the above-mentioned "document content index template" may include: the elements to be extracted under each title in the document to be detected.

[0121] In addition, the embodiment of the present application does not limit the representation method of the above-mentioned "document content index template". For example, the above-mentioned "document content index template" can be in the form of a tree index table (such as Figure 3 It should be noted that for Figure 3 For example, the document content index template shown in the figure may include at least one first-level title (e.g., a first-level title with T06 as the title identifier). Each first-level title may be divided into at least one second-level title (e.g., a second-level title with T0601 as the title identifier), and these second-level titles are respectively attached to their corresponding first-level titles. Each second-level title may be divided into at least one third-level title (e.g., a third-level title with T060101 as the title identifier), and these third-level titles are respectively attached to their corresponding second-level titles.

[0122] In addition, the embodiments of the present application do not limit the construction process of the above-mentioned "document content index template". For example, it can specifically be: relevant personnel summarize, standardize, and organize the document content indexes of a large number of audit documents to obtain the "document content index template", so that the "document content index template" can describe the information skeleton of any audit document, so that the "document content index template" can represent the document content index of any audit document, so that the "document content index template" can be used to perform element extraction processing on any audit document in the future.

[0123] The embodiments of the present application also do not limit the use process of the above-mentioned "document content index template". For example, it may specifically include: after obtaining the document to be detected, according to the parameter information that needs to be extracted under each title recorded in the document content index template, the element extraction processing is performed on the document to be detected to obtain the element information to be used, so that the element information to be used includes a large amount of structured data, so that the element information to be used can represent the document to be detected in a structured manner.

[0124] Based on the relevant content of the above step 11, it can be known that after obtaining the document to be detected, the element extraction processing of the document to be detected can be performed according to the pre-constructed document content index template to obtain the element information to be used, so that the element information to be used can include a large amount of structured data, so that the element information to be used can represent the document to be detected in a structured manner, thereby realizing the structuring of the document content for the document to be detected.

[0125] Step 12: Determine the document review result based on the element information to be used and the review rules to be used.

[0126] The embodiment of the present application does not limit the implementation method of step 12. For example, when the above-mentioned "audit rules to be used" include M rules to be confirmed, step 12 may specifically include steps 121 to 123:

[0127] Step 121: Query the element description information corresponding to the mth rule to be confirmed from the element information to be used, where m is a positive integer, m≤M.

[0128] The above-mentioned “mth rule to be confirmed” is used to perform the mth check on the document to be detected, so as to confirm whether the document to be detected can pass the mth check.

[0129] The above-mentioned “element description information corresponding to the mth rule to be confirmed” refers to the document content related to the mth rule to be confirmed recorded in the document to be detected. For ease of understanding, the following is an explanation with reference to an example.

[0130] Assume that the above “Article m to be confirmed rule” is: the amount of bad debt provision for each item plus the amount of bad debt provision for each combination under the period-end balance of the accounts receivable statement should be equal to the total; the above “documents to be tested” include Figure 4 The table shown; the above-mentioned "element information to be used" includes the first element description information, the second element description information, and the third element description information; the first element description information includes the first key-value pair (the amount of bad debt provision for each item under the beginning balance of the accounts receivable table in the notes to the financial statements, 8,729,708,190.67+587,164,022.77); the second element description information includes the second key-value pair (the amount of bad debt provision for each combination under the beginning balance of the accounts receivable table in the notes to the financial statements, 28,147,520,514.69); the third element description information includes the third key-value pair (the total under the beginning balance of the accounts receivable table in the notes to the financial statements, 37,464,392,728.13).

[0131] As an example, based on the above assumptions, it can be known that the above "element description information corresponding to the mth rule to be confirmed" may include first element description information, second element description information and third element description information, so that it can be determined whether the document to be detected satisfies the mth rule to be confirmed based on these three element description information.

[0132] Step 122: Determine the rule confirmation result corresponding to the mth rule to be confirmed according to the mth rule to be confirmed and the element description information corresponding to the mth rule to be confirmed, where m is a positive integer, m≤M.

[0133] The above-mentioned “rule confirmation result corresponding to the mth rule to be confirmed” is used to indicate whether the document to be detected satisfies the mth rule to be confirmed, thereby indicating whether the document to be detected passes the mth check.

[0134] It can be seen that after obtaining the element description information corresponding to the mth rule to be confirmed, the element description information corresponding to the mth rule to be confirmed can be substituted into the mth rule to be confirmed to obtain the rule confirmation result corresponding to the mth rule to be confirmed, so that the rule confirmation result is used to indicate whether the above-mentioned "document to be detected" satisfies the mth rule to be confirmed. For ease of understanding, the following is an explanation with examples.

[0135] As an example, when the above-mentioned "mth rule to be confirmed" is that the amount of bad debt provision for each item plus the amount of bad debt provision for combination under the beginning balance of the accounts receivable statement should be equal to the total, and the element description information corresponding to the above-mentioned "mth rule to be confirmed" can include the first element description information, the second element description information and the third element description information, because (8,729,708,190.67+587,164,022.77)+28,147,520,514.69=37,464,392,728.13, the above-mentioned "document to be detected" satisfies the mth rule to be confirmed, so the rule confirmation result corresponding to the mth rule to be confirmed can indicate that the "document to be detected" passes the inspection item represented by the mth rule to be confirmed (that is, the mth inspection).

[0136] Step 123: Determine the document review result according to the rule confirmation result corresponding to the first rule to be confirmed to the rule confirmation result corresponding to the Mth rule to be confirmed.

[0137] In an embodiment of the present application, after obtaining the rule confirmation results corresponding to M rules to be confirmed, the rule confirmation results corresponding to the M rules to be confirmed can be summarized (especially, those to be confirmed rules that are not satisfied) to obtain a document review result, so that the document review result can accurately indicate which document contents in the document to be detected have problems.

[0138] In addition, in order to further improve the document review effect, the embodiment of the present application also adopts another possible implementation of step 123, which may specifically include steps 1231-1233:

[0139] Step 1231: According to the rule confirmation results corresponding to the M rules to be confirmed, F target rules satisfying the fourth condition are determined from the M rules to be confirmed, where F is a positive integer.

[0140] The above-mentioned "fourth condition" can be preset. For example, the "fourth condition" can be: the rule confirmation result indicates that the inspection has not passed. It can be seen that for the mth rule to be confirmed, if the rule confirmation result corresponding to the mth rule to be confirmed indicates that the document to be detected has not passed the inspection item represented by the mth rule to be confirmed, then the mth rule to be confirmed is determined as the target rule; if the rule confirmation result corresponding to the mth rule to be confirmed indicates that the document to be detected has passed the inspection item represented by the mth rule to be confirmed, then the mth rule to be confirmed can be discarded. Wherein, m is a positive integer, m≤M.

[0141] The above-mentioned “target rule” refers to the rule to be confirmed that meets the fourth condition.

[0142] Step 1232: Find the importance level corresponding to the f-th target rule from the mapping relationship to be used, where f is a positive integer, f≤F.

[0143] The above-mentioned "mapping relationship to be used" is used to record the importance level corresponding to each candidate audit rule. It can be seen that the above-mentioned "mapping relationship to be used" includes the corresponding relationship between the e-th candidate audit rule and the importance level corresponding to the e-th candidate audit rule. Among them, e is a positive integer, e≤E, E is a positive integer, and E represents the number of the above-mentioned "candidate audit rules".

[0144] The embodiments of the present application do not limit the implementation method of step 1232. For example, it may specifically include: matching the f-th target rule with at least one candidate audit rule in the mapping relationship to be used, so that when it is determined that there is a candidate audit rule in the mapping relationship to be used that successfully matches the f-th target rule, the importance level corresponding to the successfully matched candidate audit rule is determined as the importance level corresponding to the f-th target rule.

[0145] Step 1233: Determine the error alarm level of the f-th target rule according to the importance level corresponding to the f-th target rule, where f is a positive integer and f≤F.

[0146] The above-mentioned “error warning level” is used to describe the severity of a document error; and the embodiment of the present application does not limit the “error warning level”, for example, it can be doubtful, normal, serious, or high-risk, etc.

[0147] In addition, the above “error warning level of the f-th target rule” is positively correlated with the above “importance level corresponding to the f-th target rule”. That is, the higher the “importance level corresponding to the f-th target rule”, the higher the “error warning level of the f-th target rule”, which indicates that the severity of the error presented by the f-th target rule is higher.

[0148] Step 1233: Determine the document review result based on the F target rules and the error warning levels of the F target rules.

[0149] In an embodiment of the present application, after obtaining F target rules that are not satisfied by the document to be detected and their error warning levels, the document review result can be determined based on the F target rules and the error warning levels of the F target rules, so that the document review result can not only indicate which document contents in the document to be detected have problems, but also indicate the severity of these problems in the document to be detected.

[0150] Based on the relevant contents of the above steps 11 to 12, it can be known that after obtaining the audit rules to be used corresponding to the document to be detected, the audit process for the document to be detected can be implemented with the help of factor extraction processing, which is conducive to improving the document audit effect.

[0151] Based on the relevant contents of S1 to S3 above, it can be known that for the document review method provided in the embodiment of the present application, after obtaining the document to be tested (for example, financial statements used to generate audit reports, and documents such as report notes), first determine the review rules to be used based on the attribute description information of the document to be tested (for example, audit template or company industry, etc.); then, according to the review rules to be used, review the document to be tested and obtain the document review results. In this way, it is possible to automatically perform document review processing on the document data, thereby effectively overcoming the defects of manual review, and thus helping to improve the document review effect.

[0152] Method Example 2

[0153] In fact, in order to facilitate the writer (or reviewer) of the above-mentioned "document to be tested" to accurately know where the "document to be tested" has errors, the embodiment of the present application also provides another possible implementation of the document review method. In this implementation, in addition to the above-mentioned S1-S3, the document review method may also include S4:

[0154] S4: According to the document review result, the document to be detected is error marked to obtain a marked document.

[0155] The above-mentioned “marked document” refers to a document to be detected that carries an error mark; and the embodiment of the present application does not limit the “marked document”, for example, it may include Figure 5 The document contents are shown.

[0156] In addition, the embodiment of the present application does not limit the implementation of S4. For example, when there are errors to be marked in the document review result, S4 may include S41-S42:

[0157] S41: Determine the document marking position where the error is to be marked.

[0158] The above-mentioned “Errors to be Marked” is used to represent any document error in the document review results.

[0159] The above-mentioned “document marking position of the error to be marked” is used to indicate the position where the error to be marked is marked in the document to be detected.

[0160] In addition, the embodiments of the present application do not limit the determination process of the above-mentioned "document marking position of the error to be marked". For example, when the above-mentioned "error to be marked" indicates that the document to be detected does not satisfy the f-th target rule, the element description information corresponding to the f-th target rule includes the fourth element description information, and the fourth element description information includes the fourth key-value pair and the document position corresponding to the fourth key-value pair, the above-mentioned "document position corresponding to the fourth key-value pair" is determined as the document marking position of the error to be marked.

[0161] It should be noted that the above “document position corresponding to the fourth key-value pair” is used to represent the value in the “fourth key-value pair” (for example, Figure 5 The position of "15,801,164.63") in the document to be detected.

[0162] S42: According to the document marking position of the error to be marked, the error to be marked is added to the document to be detected to obtain a marked document.

[0163] Based on the relevant content of S4 above, it can be known that after obtaining the document review results, the errors recorded in the document review results can be marked in the document to be tested respectively to obtain a marked document, so that the subsequent writer (or reviewer) of the document to be tested can accurately know from the marked document which document contents in the document to be tested have errors, which is conducive to improving user experience.

[0164] Method Example 3

[0165] In some cases, it is sometimes necessary to summarize and report the audit results of multiple document data. Therefore, in order to meet the user's needs, the embodiment of the present application also provides another possible implementation of the document audit method. In this implementation, in addition to all or part of the above steps, the document audit method may also include S5:

[0166] S5: Generate an audit report to be used based on the document audit results.

[0167] The above-mentioned "audit report to be used" is used to describe the audit results of at least one document to be tested; and the embodiment of the present application does not limit the representation method of the "audit report to be used", for example, it can be represented by a visual chart + detailed summary.

[0168] In addition, the embodiment of the present application does not limit the implementation method of S5. For example, it may specifically include: after obtaining the document review results of D documents to be detected, statistical analysis and processing may be performed on the document review results of the D documents to be detected to obtain a to-be-used review report, so that the to-be-used review report can vividly represent the errors existing in the D documents to be detected, so that the to-be-used review report can be used to perform other evaluation processes (for example, the writing level change trend of a writer, the comprehensive writing level of multiple writers, etc.) in the future. Wherein, D is a positive integer.

[0169] It should be noted that the embodiments of the present application are not limited to the above-mentioned "D documents to be detected". For example, they may include: document data written by one writer in D different time periods. For another example, "D documents to be detected" may include: document data written by multiple writers in the same time period.

[0170] Based on the relevant content of S5 above, it can be known that for some situations, after obtaining the document review results of some documents to be tested, statistical analysis and processing can be performed on the document review results of these documents to be tested to obtain an audit report to be used, so that the audit report to be used can accurately indicate the errors existing in these documents to be tested, thereby enabling the audit report to be used to accurately indicate the writing level of these documents to be tested.

[0171] Method Example 4

[0172] In fact, some document data may contain an overlong table. However, due to the limited page size of these document data, the overlong table cannot be fully displayed in one page of the document. Therefore, in order to better display the overlong table, the overlong table is usually divided into multiple sub-tables, and these sub-tables are displayed in different pages of the document (such as Figure 6 As shown), this may easily lead to a very difficult review process for the overlong form. Based on this, in order to overcome this problem, the present application embodiment also provides another possible implementation of the document review method. In this implementation, in addition to the above steps, the document review method may also include S7-S12:

[0173] S7: Initialize the content of the document to be processed according to the document to be detected.

[0174] The above-mentioned “document content to be processed” is used to indicate the document content (for example, character content and table content, etc.) required to be used when reviewing the document to be detected.

[0175] In addition, the embodiment of the present application does not limit the implementation method of S7. For example, it may specifically be: directly determining the document to be detected as the document content to be processed.

[0176] It should be noted that since the document to be detected is always subject to page size restrictions, it is impossible to achieve the purpose of recovering an overlong table in the document to be detected. Therefore, in order to overcome this difficulty, the document to be detected can be first used to initialize a document content to be processed that is not restricted by the page size, so that the document content to be processed includes the document content in the document to be detected, so that the purpose of recovering the overlong table can be achieved in the document content to be processed later.

[0177] S8: Determine whether there is a first table satisfying the first condition in the content of the document to be processed; if so, execute S9-S11; if not, execute S12.

[0178] The above-mentioned "first condition" refers to the condition met when a document is displayed in a sub-table other than the first one in an extra-long table; and the embodiment of the present application does not limit the "first condition", for example, it may specifically include: there are preset detection characters at the preset detection position corresponding to the first table (for example, continued from the previous table, continued from the previous table, continued, continued, etc.).

[0179] It should be noted that the embodiments of the present application do not limit the above-mentioned "preset detection position". Figure 6 As shown, when the above-mentioned "preset detection position" refers to Figure 6 The position of "Continued from the previous table" in the table, and the first table refers to Figure 6 When the second table is in the document to be processed, the "preset detection position" is adjacent to the table position of the first table, and the "preset detection position" is ahead of the table position of the first table. The "table position of the first table" is used to indicate the position of the first table in the document content to be processed.

[0180] The above-mentioned “first table” refers to a table that exists in the content of the document to be processed and satisfies the first condition, so that the “first table” is used to represent a sub-table that is not the first in an extra-long table.

[0181] S9: Determine a second table that meets a second condition from the content of the document to be processed according to the table position of the first table.

[0182] The above-mentioned "second condition" can be preset. For example, the "second condition" can be specifically: the distance difference between the table position of the second table and the above-mentioned "table position of the first table" is the smallest, and the table position of the second table is ahead of the above-mentioned "table position of the first table".

[0183] The above-mentioned “second table” refers to a table existing in the content of the document to be processed and satisfying the second condition, so that the “second table” is used to record the table content located before the first table in an extra-long table.

[0184] In addition, the embodiment of the present application does not limit the above-mentioned "second table". For example, when the above-mentioned "first table" is the second sub-table in an extra-long table (such as Figure 6 , etc.), then the "second table" refers to the first subtable in the super-long table (such as Figure 6 For another example, when the above-mentioned "first table" is the y-th sub-table in an extra-long table, the "second table" is obtained by concatenating the 1st to y-1th sub-tables in the extra-long table; wherein y is a positive integer, 3≤y≤Y, and Y represents the number of sub-tables into which the above-mentioned "one extra-long table" is divided.

[0185] In addition, the embodiments of the present application do not limit the determination process of the above-mentioned "second table". For example, it may specifically include: in the content of the document to be processed, the table position of the first table is used as the starting position of the table search, and the table search is performed from the beginning to the front, so that the first table searched is determined as the above-mentioned "second table".

[0186] S10: merging the first table with the second table to obtain a third table.

[0187] The above-mentioned "third table" is obtained by splicing the first table and the second table, so that the "third table" is used to represent the table content carried by the first table and the second table (such as Figure 7 For example, when the above "first table" is the xth subtable in an extra-long table, the "third table" is obtained by concatenating the first to the xth subtables in the extra-long table. Where x is a positive integer, 2≤x≤Y, and Y represents the number of subtables into which the above "extra-long table" is divided.

[0188] In addition, the embodiment of the present application does not limit the determination process of the above-mentioned "third table", for example, it may specifically include steps 21 to 23:

[0189] Step 21: Compare the header of the first table with the header of the second table to obtain a common header.

[0190] The above-mentioned “common header” refers to the header columns that are common to the first table and the second table (for example, Figure 6 The headers shown in column 1 of the two tables shown) or header rows.

[0191] Step 22: Delete the common table headers in the first table to obtain the fourth table.

[0192] In an embodiment of the present application, after determining the common header, the common header can be deleted from the first table to obtain a fourth table, so that the fourth table does not include the common header and only includes other table contents in the first table except the non-common header.

[0193] Step 23: According to the preset splicing position corresponding to the common table header, the fourth table is spliced ​​with the second table to obtain a third table.

[0194] The above-mentioned "preset splicing position corresponding to the common header" can be preset. For example, if the "common header" is a column of data in the second table, the right border position of the last column in the second table can be determined as the "preset splicing position corresponding to the common header". For another example, if the "common header" is a row of data in the second table, the bottom border position of the last row in the second table can be determined as the "preset splicing position corresponding to the common header".

[0195] It can be seen that Figure 6 and Figure 7 As shown, if the "public header" is a column of data in the second table, after obtaining the fourth table, the fourth table can be spliced ​​to the right of the last column in the second table to obtain the third table (such as Figure 7 ), so that the third table is used to accurately represent the table contents of the above-mentioned "first table" and "second table".

[0196] Based on the relevant content of S10 above, it can be known that after the second table is obtained, the second table can be merged with the first table to obtain a third table, so that the third table includes the table contents of the above "first table" and "second table".

[0197] S11: Update the content of the document to be processed according to the third table, and return to execute S8.

[0198] The embodiments of the present application do not limit the updating process of the content of the document to be processed. For example, it may specifically include: after obtaining the third table, the first table, the second table, and the preset detection characters between the first table and the second table may be first deleted from the content of the document to be processed; then, according to the table position of the above-mentioned "first table" (or, the table position of the above-mentioned "second table"), the third table is added to the content of the document to be processed to obtain the updated content of the document to be processed, so that the updated content of the document to be processed includes the third table, but no longer includes the first table, the second table, and the preset detection characters between the first table and the second table, so that S8 and its subsequent steps can be continued to be executed based on the updated content of the document to be processed, thereby achieving the purpose of recovering the overlong table.

[0199] S12: According to the audit rules to be used, the content of the document to be processed is audited and processed to obtain the document audit result.

[0200] In an embodiment of the present application, after determining that there is no first table satisfying the first condition in the content of the document to be processed, it can be determined that there is no overlong table that is split in the content of the document to be processed, and thus it can be determined that each table in the content of the document to be processed is a complete table. Therefore, the content of the document to be processed can be directly audited and processed according to the audit rules to be used to obtain the document audit result, so that the document audit result can better represent the errors existing in the above "document to be detected".

[0201] It should be noted that S12 can be implemented using any implementation of S3, and it is only necessary to replace the "document to be detected" in any implementation of S3 with the "document content to be processed".

[0202] Based on the relevant contents of S7 to S12 above, it can be known that after obtaining the document to be detected, the recovery process of the extra-long table in the document to be detected can be implemented by merging multiple sub-tables in the content of the document to be processed that are not restricted by the page size. This can effectively avoid the audit difficulties caused by the extra-long table being divided into multiple sub-tables, thereby helping to improve the audit effect.

[0203] Method Example 5

[0204] In fact, some document data may contain tables with very complex headers (such as Figure 6 Based on this, in order to overcome this problem, the embodiment of the present application also provides a possible implementation method for determining the above-mentioned "elements information to be used", which may specifically include steps 31-32:

[0205] Step 31: If it is determined that a fifth table that meets the third condition exists in the document to be detected, at least one table element information is extracted from the fifth table according to a preset table element extraction rule corresponding to the third condition.

[0206] The above-mentioned "third condition" can be preset. For example, the "third condition" can be specifically: including multiple layers of nested headers (for example, similar to Figure 6 The three-level nested header of "Amount under Book Balance under Beginning Balance" is shown).

[0207] The above-mentioned “fifth table” refers to a table that satisfies the third condition.

[0208] The above-mentioned "preset table element extraction rule corresponding to the third condition" refers to the element extraction rule adopted for the table that meets the third condition; and the embodiment of the present application does not limit the "preset table element extraction rule corresponding to the third condition", for example, it can be specifically: according to the row name and column name in the fifth table, determine the identification information of each table value in the fifth table. For ease of understanding, the following is combined with Figure 6 as well as Figure 8 Provide explanation.

[0209] As an example, for Figure 6 For example, the table value "27,577,113,388.43" can be used to Figure 6 The row name and column name of the row in the table shown are identified to obtain the identification information of the table value (that is, Figure 8 The "Note_T060402_Bad Debt Provision by Combination_Beginning Balance##Book Balance##Amount" shown above) is used to make the identification information of the table value uniquely represent the table value, so that the table value and the identification information of the table value can subsequently form a key-value pair (Note_T060402_Bad Debt Provision by Combination_Beginning Balance##Book Balance##Amount, 27,577,113,388.43), and based on the key-value pair and the document position corresponding to the key-value pair, generate an element description information (that is, the "table element information" shown above).

[0210] The above-mentioned "table element information" refers to the element description information extracted from the fifth table, so that the "table element information" is used to describe the element content recorded in the fifth table; and the embodiment of the present application does not limit the "table element information", for example, it may include a first table value, identification information of the first table value, and at least one of the location description information of the first table value.

[0211] The above-mentioned “first table value” refers to the numerical value recorded in the fifth table (for example, “27,577,113,388.43”).

[0212] The above-mentioned “identification information of the first table value” is used to uniquely identify the first table value; and the “identification information of the first table value” can be based on the table identifier of the fifth table (for example, Figure 8 Note_T060402 as shown), the row name to be used (for example, Figure 8 "Bad debt provision by group" as shown), and the column name to be used (for example, Figure 8 The row name to be used refers to the row name of the first table value in the fifth table. The column name to be used refers to the column name of the first table value in the fifth table.

[0213] It should be noted that the above “identification information of the first table value” is equivalent to the above “element to be extracted”, and the above “first table value” is equivalent to the above “value of the element to be extracted”.

[0214] The above-mentioned “position description information of the first table value” is used to describe the position of the first table value in the document to be detected.

[0215] Step 32: Determine the element information to be used based on at least one table element information.

[0216] In the implementation of the present application, after extracting at least one table element information from the fifth table, the at least one table element information and the element description information extracted from the document content other than the fifth table in the document to be detected can be summarized to obtain the element information to be used, so that the element information to be used can more accurately represent the document to be detected.

[0217] Based on the relevant contents of steps 31 to 32 above, it can be known that in some cases, a special element extraction method can be used for special document content in the document to be detected (for example, a table with a complex header) to obtain the element content carried by the special document content. This is conducive to improving the element extraction effect, thereby helping to improve the document review effect.

[0218] Method Example 6

[0219] In some cases, it may be necessary to specify some tables (such as Fig. 9 For example, for Fig. 9 For example, the fixed asset status table shown in the figure includes three levels of headings; (1), (2), (3), etc. are all third-level headings, lowercase Arabic numerals (e.g., 1, 2, 3, ..., etc.) are second-level headings, and uppercase Chinese numerals (e.g., one, two, three, ..., etc.) are first-level headings; and the audit rules used for the "fixed asset status" table can generally include the following three rules:

[0220] Rule 1: The sum of the values ​​under a level 3 heading should equal the sum of the values ​​under the level 2 heading above it. Fig. 9 As shown, in the Houses and Buildings column, the sum of the values ​​under "(1) Purchase" (e.g., 135,922.33), "(2) Transfer from construction in progress" (e.g., 15,665,242.31), and "(3) Increases from business combination" (e.g., 0) should equal "2. Increases in this period" (e.g., 15,801,164.63).

[0221] Rule 2: The horizontal sum of the rows equals the total.

[0222] Rule 3: The beginning balance under each first-level heading plus or minus the increase or decrease in this period equals the ending balance.

[0223] Based on the above analysis, we can see that when reviewing a document (for example, a document to be tested), we may encounter Fig. 9 The table shown in Figure 1 shows a special operator (for example, Fig.10 levelSum as shown in the figure) so that the operator can implement the calculation logic of "the sum of the values ​​of each subheading is equal to the sum of the subheadings of the previous level (that is, the sum of the three levels is equal to its corresponding second level, and the sum of the second level is equal to its corresponding first level)", so that it can be used for subsequent operations similar to Fig. 9 When the table shown is reviewed and processed, the operator can be directly called, which is helpful to improve the efficiency of document review.

[0224] Method Example 7

[0225] In fact, some document data may have version compatibility issues. Based on this, in order to improve the document review effect, the embodiment of the present application also provides another possible implementation method for obtaining the above-mentioned "document to be detected", which may specifically include steps 41 to 44:

[0226] Step 41: after acquiring the first document data, determine whether the version information of the first document data meets the preset version condition, if yes, execute step 42; if no, execute step 43-step 44.

[0227] The above-mentioned “version information of the first document data” is used to describe the document version of the first document data (for example, 2003 version Word, 2007 version Word, 2010 version Word, 2003 version Excel, 2007 version Excel, 2010 version Excel, etc.).

[0228] The above-mentioned "preset version condition" can be preset. For example, the above-mentioned "preset version condition" can include 2010 version of Word and 2010 version of Excel. It can be seen that when the above-mentioned "first document data" includes financial statements and statement notes, it can be determined whether the version information of the financial statements is 2010 version of Excel, and whether the version information of the statement notes is 2010 version of Word; if it is determined that the version information of the financial statements is 2010 version of Excel and the version information of the statement notes is 2010 version of Word, it can be determined that the version information of the first document data meets the preset version condition; if it is determined that the version information of the financial statements is not 2010 version of Excel, or it is determined that the version information of the statement notes is not 2010 version of Word, it can be determined that the version information of the first document data does not meet the preset version condition.

[0229] Step 42: Determine the document to be detected according to the first document data.

[0230] The embodiment of the present application does not limit the implementation of step 42. For example, it may specifically include: directly determining the first document data as the document to be detected. For another example, it may be implemented in any implementation of steps 51 to 53 shown below, and it only needs to replace the "second document data" with "first document data" in any implementation of steps 51 to 53 shown below.

[0231] Step 43: Perform compatibility processing on the first document data to obtain compatible document data.

[0232] The above-mentioned "compatibility processing" is used to convert a document version of document data; and the embodiment of the present application does not limit the "compatibility processing", for example, it may include: converting a financial statement to a target Excel version (for example, Excel 2010 version); and / or, converting a report note to a target Word version (for example, Word 2010 version).

[0233] It can be seen that for the first document data that does not meet the preset version conditions, if the financial statements in the first document data are not the target Excel version, the financial statements in the first document data can be converted to the target Excel version; if the report notes in the first document data are not the target Word version, the report notes in the first document data can be converted to the target Word version.

[0234] The above-mentioned “compatible document data” refers to document data obtained by performing compatibility processing on the first document data, so that the “compatible document data” meets the preset version condition.

[0235] Step 44: Determine the document to be detected based on the compatible document data.

[0236] The embodiment of the present application does not limit the implementation of step 44. For example, it may specifically include: directly determining the compatible document data as the document to be detected. For another example, it may be implemented in any implementation of steps 51 to 53 shown below, and it only needs to replace the "second document data" with "compatible document data" in any implementation of steps 51 to 53 shown below.

[0237] Based on the relevant contents of the above steps 41 to 44, it can be known that after obtaining a document data, compatibility processing can be performed on the document data to obtain the document to be tested. This can solve the adverse effects on document review caused by compatibility issues of different versions of office software (office), which is conducive to improving the document review effect.

[0238] In fact, some document data may not be audited due to some reasons (for example, it does not belong to the audit report, financial statements and financial notes; or the document content is incomplete, etc.). Based on this, in order to avoid wasting audit resources, the embodiment of the present application also provides another possible implementation method for obtaining the above-mentioned "document to be tested", which may specifically include steps 51 to 53:

[0239] Step 51: Determine whether the second document data meets the preset review condition, if yes, execute step 52; if no, execute step 53.

[0240] The above-mentioned "preset audit conditions" can be preset. For example, the above-mentioned "preset audit conditions" can include: belonging to the target document type (for example, audit report, financial statement and statement notes), and the document content is complete.

[0241] Step 52: Determine the second document data as the document to be detected.

[0242] Step 53: Generate warning information, and send the warning information to relevant personnel (eg, writer or reviewer) of the second document data.

[0243] The above-mentioned “warning information” is used to indicate that the second document data does not meet the preset review conditions.

[0244] Based on the relevant contents of the above steps 51 to 53, it can be known that after obtaining a document data, the document type and document content integrity of the document data can be checked so as to intercept and alarm the document data that does not meet the preset review conditions. This can avoid reviewing invalid documents, thereby effectively avoiding the waste of review resources, and thus helping to improve the document review experience.

[0245] It should be noted that Figure 2The “document preprocessing” shown may include the above steps 41 to 44, and / or the above steps 51 to 53.

[0246] Based on the document review method provided in the above method embodiment, the embodiment of the present application also provides a document review device, which is explained and illustrated below in conjunction with the accompanying drawings.

[0247] Device Embodiment

[0248] The device embodiment introduces the document review device, and for related content, please refer to the above method embodiment.

[0249] See also Fig.11 , which is a structural diagram of a document review device provided in an embodiment of the present application.

[0250] The document review device 1100 provided in the embodiment of the present application includes:

[0251] The document acquisition unit 1101 is used to acquire the document to be detected;

[0252] A rule determination unit 1102, configured to determine the audit rules to be used according to the attribute description information of the document to be detected;

[0253] The document review unit 1103 is used to review the document to be detected according to the review rule to be used, and obtain a document review result.

[0254] In a possible implementation, the rule determination unit 1102 is specifically used to: search for candidate audit rules corresponding to the attribute description information from a preset mapping relationship, and determine them as audit rules to be used; wherein the preset mapping relationship includes a correspondence between the attribute description information and the candidate audit rules corresponding to the attribute description information.

[0255] In a possible implementation manner, the attribute description information includes document template description information and / or company industry affiliation information.

[0256] In a possible implementation manner, the document review device 1100 further includes:

[0257] The attribute acquisition unit is used to acquire the to-be-used attribute information provided by the target user for the to-be-detected document; wherein the to-be-used attribute information includes attribute description information of the to-be-detected document.

[0258] In a possible implementation, the document review unit 1103 includes:

[0259] The element extraction subunit is used to perform element extraction processing on the document to be detected to obtain element information to be used;

[0260] The result determination subunit is used to determine the document review result according to the element information to be used and the review rule to be used.

[0261] In a possible implementation manner, the element extraction subunit is specifically used to: perform element extraction processing on the document to be detected according to a pre-constructed document content index template to obtain element information to be used.

[0262] In a possible implementation, the element extraction subunit is specifically used to: if it is determined that there is a fifth table that meets the third condition in the document to be detected, extract at least one table element information from the fifth table according to a preset table element extraction rule corresponding to the third condition; and determine the element information to be used based on the at least one table element information.

[0263] In a possible implementation, the table element information includes a first table value and identification information of the first table value; wherein the identification information of the first table value is determined based on the table identification, the row name to be used, and the column name to be used of the fifth table; the row name to be used refers to the row name of the table row where the first table value is located in the fifth table; the column name to be used refers to the column name of the table column where the first table value is located in the fifth table.

[0264] In a possible implementation manner, the audit rules to be used include M rules to be confirmed; wherein M is a positive integer;

[0265] The result determination subunit is specifically used to: query the element description information corresponding to the mth rule to be confirmed from the element information to be used; wherein m is a positive integer, m≤M; determine the rule confirmation result corresponding to the mth rule to be confirmed based on the mth rule to be confirmed and the element description information corresponding to the mth rule to be confirmed; wherein m is a positive integer, m≤M; determine the document review result based on the rule confirmation results corresponding to the 1st rule to be confirmed to the rule confirmation results corresponding to the Mth rule to be confirmed.

[0266] In a possible implementation manner, the document review device 1100 further includes:

[0267] An overlong table recovery unit is used to initialize the content of the document to be processed according to the document to be detected; determine whether there is a first table that meets a first condition in the content of the document to be processed; when it is determined that there is a first table that meets the first condition in the content of the document to be processed, determine a second table that meets a second condition from the content of the document to be processed according to the document position of the first table; perform table merging processing on the first table and the second table to obtain a third table; update the content of the document to be processed according to the third table, and continue to perform the step of determining whether there is a first table that meets the first condition in the content of the document to be processed;

[0268] The document review unit 1103 is specifically used to: when it is determined that there is no first table satisfying the first condition in the content of the document to be processed, review the content of the document to be processed according to the review rule to be used to obtain a document review result.

[0269] In a possible implementation, the process of determining the third table includes: comparing the header of the first table and the header of the second table to obtain a common header; deleting the common header in the first table to obtain a fourth table; and splicing the fourth table with the second table according to a preset splicing position corresponding to the common header to obtain a third table.

[0270] In a possible implementation manner, the document review device 1100 further includes:

[0271] The error marking unit is used to perform error marking processing on the document to be detected according to the document review result to obtain a marked document.

[0272] In a possible implementation manner, the document review device 1100 further includes:

[0273] The report generating unit is used to generate a to-be-used audit report according to the document audit result.

[0274] Furthermore, an embodiment of the present application also provides a device, including: a processor, a memory, and a system bus;

[0275] The processor and the memory are connected via the system bus;

[0276] The memory is used to store one or more programs, and the one or more programs include instructions. When the instructions are executed by the processor, the processor executes any one of the implementation methods of the above-mentioned document review method.

[0277] Furthermore, an embodiment of the present application also provides a computer-readable storage medium, in which instructions are stored. When the instructions are executed on a terminal device, the terminal device executes any one of the implementation methods of the above-mentioned document review method.

[0278] Furthermore, an embodiment of the present application also provides a computer program product, which, when executed on a terminal device, enables the terminal device to execute any one of the implementation methods of the above-mentioned document review method.

[0279] It can be known from the description of the above implementation mode that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product can be stored in a storage medium such as ROM / RAM, a disk, an optical disk, etc., including several instructions for enabling a computer device (which can be a personal computer, a server, or a network communication device such as a media gateway, etc.) to execute the methods described in the various embodiments of the present application or certain parts of the embodiments.

[0280] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0281] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0282] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A document review method, characterized in that: The method comprises: Get the document to be tested; Determining the audit rules to be used according to the attribute description information of the document to be detected; According to the audit rules to be used, the document to be tested is audited to obtain a document audit result; The step of auditing the document to be detected according to the audit rule to be used to obtain the document audit result includes: Perform element extraction processing on the document to be detected to obtain element information to be used, and the process of determining the element information to be used includes: if it is determined that there is a table that meets the preset conditions in the document to be detected, then according to the preset table element extraction rule corresponding to the preset conditions, at least one table element information is extracted from the table that meets the preset conditions, and the element information to be used is determined according to the at least one table element information; the preset conditions include multiple layers of nested table headers; the table element information includes a first table value and identification information of the first table value; wherein the identification information of the first table value is determined according to the table identification of the table that meets the preset conditions, the row name to be used, and the column name to be used; the row name to be used refers to the row name of the table row where the first table value is located in the table that meets the preset conditions; the column name to be used refers to the column name of the table column where the first table value is located in the table that meets the preset conditions; The document review result is determined according to the element information to be used and the review rule to be used.

2. The method according to claim 1, characterized in that The step of determining the audit rules to be used according to the attribute description information of the document to be detected includes: The candidate audit rule corresponding to the attribute description information is searched from the preset mapping relationship and determined as the audit rule to be used; wherein the preset mapping relationship includes the correspondence between the attribute description information and the candidate audit rule corresponding to the attribute description information.

3. The method according to claim 1, characterized in that The attribute description information includes document template description information and / or company industry affiliation information.

4. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Acquire the attribute information to be used provided by the target user for the document to be detected; wherein the attribute information to be used includes attribute description information of the document to be detected.

5. The method according to claim 1, characterized in that The extracting elements of the document to be detected to obtain element information to be used includes: According to the pre-constructed document content index template, element extraction processing is performed on the document to be detected to obtain element information to be used.

6. The method according to any one of claims 1 or 5, characterized in that: The audit rules to be used include M rules to be confirmed; wherein M is a positive integer; The step of determining the document review result according to the element information to be used and the review rule to be used includes: Querying the element description information corresponding to the mth rule to be confirmed from the element information to be used; wherein m is a positive integer, m≤M; Determine, according to the mth rule to be confirmed and the element description information corresponding to the mth rule to be confirmed, a rule confirmation result corresponding to the mth rule to be confirmed; The document review result is determined according to the rule confirmation result corresponding to the 1st rule to be confirmed to the rule confirmation result corresponding to the Mth rule to be confirmed.

7. The method according to claim 1, characterized in that The method further comprises: Initializing the content of the document to be processed according to the document to be detected; Determining whether there is a first table satisfying a first condition in the content of the document to be processed, wherein the first condition is that a preset detection character exists at a preset detection position corresponding to the first table; When it is determined that there is a first table satisfying a first condition in the content of the document to be processed, determining a second table satisfying a second condition from the content of the document to be processed according to the document position of the first table, wherein the second condition is that a difference in distance between a table position of the second table and a table position of the first table is the smallest, and the table position of the second table is ahead of the table position of the first table; Merge the first table with the second table to obtain a third table; According to the third table, updating the content of the document to be processed, and continuing to perform the step of determining whether there is a first table satisfying the first condition in the content of the document to be processed; The step of auditing the document to be detected according to the audit rule to be used to obtain the document audit result includes: When it is determined that the first table satisfying the first condition does not exist in the content of the document to be processed, the content of the document to be processed is audited according to the audit rule to be used to obtain a document audit result.

8. The method according to claim 7, characterized in that The process of determining the third table includes: Compare the header of the first table with the header of the second table to obtain a common header; Delete the common table headers in the first table to obtain a fourth table; According to the preset splicing position corresponding to the common table header, the fourth table and the second table are spliced ​​to obtain a third table.

9. The method according to claim 1, characterized in that: The method further comprises: According to the document review result, error marking processing is performed on the document to be detected to obtain a marked document; and / or, Based on the document review results, an audit report to be used is generated.

10. A document review device, characterized in that: include: A document acquisition unit, used to acquire the document to be detected; A rule determination unit, used to determine the audit rules to be used according to the attribute description information of the document to be detected; A document review unit, used to review the document to be tested according to the review rule to be used, and obtain a document review result; The document review unit is specifically used for: performing element extraction processing on the document to be detected to obtain element information to be used, and the process of determining the element information to be used includes: if it is determined that there is a table that meets the preset conditions in the document to be detected, then according to the preset table element extraction rule corresponding to the preset conditions, at least one table element information is extracted from the table that meets the preset conditions, and the element information to be used is determined according to the at least one table element information; the preset conditions include multiple layers of nested table headers; the table element information includes a first table value and identification information of the first table value; wherein the identification information of the first table value is determined according to the table identification, the row name to be used, and the column name to be used of the table that meets the preset conditions; the row name to be used refers to the row name of the table row where the first table value is located in the table that meets the preset conditions; the column name to be used refers to the column name of the table column where the first table value is located in the table that meets the preset conditions; the document review result is determined according to the element information to be used and the review rule to be used.

11. A document review device, characterized in that: The device includes: a processor, a memory, and a system bus; The processor and the memory are connected via the system bus; The memory is used to store one or more programs, wherein the one or more programs include instructions, and when the instructions are executed by the processor, the processor executes the method according to any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed on a terminal device, the terminal device executes the method according to any one of claims 1 to 9.

13. A computer program product, characterized in that When the computer program product is executed on a terminal device, the terminal device executes the method according to any one of claims 1 to 9.

Images