A document auditing method, device, equipment and storage medium

Abstract

The application provides a document auditing method, device and equipment and a storage medium, wherein the document auditing method comprises the following steps: obtaining a to-be-audited document as a target document; searching for an auditing reference document of the target document from a document library; if the auditing reference document of the target document is found, auditing the target document by referring to the auditing reference document of the target document; and if the auditing reference document of the target document is not found, extracting an element from the target document and auditing the extracted element. The document auditing method provided by the application is a document auditing method with high intelligence, has low dependence on manual work, can automatically determine a strategy for auditing the to-be-audited document, and has high auditing efficiency and good auditing effect.

Description

Technical Field

[0001] This invention relates to the field of document review technology, and in particular to a document review method, apparatus, device, and storage medium. Background Technology

[0002] Document review is a common scenario in corporate work. Currently, most document review solutions are based on purely manual review. Understandably, purely manual review is not only inefficient, but also prone to overlooking information due to the tedious and demanding nature of the work, leading to potential risks. Therefore, current document review solutions rely heavily on human intervention, resulting in low review efficiency and unsatisfactory review outcomes. Summary of the Invention

[0003] In view of this, the present invention provides a document review method, apparatus, device, and storage medium to solve the problem that existing document review schemes rely heavily on manual labor, resulting in low review efficiency and poor review results. The technical solution is as follows:

[0004] A document review method includes:

[0005] Obtain the document to be reviewed and use it as the target document;

[0006] Locate the review reference document for the target document in the document library;

[0007] If a review reference document for the target document is found, the target document is reviewed with reference to the review reference document.

[0008] If no review reference document is found for the target document, elements are extracted from the target document and reviewed.

[0009] Optionally, the step of searching for the review reference document of the target document in the document library includes:

[0010] If the review status of the target document is known, then the review reference document for the target document is retrieved from the document library based on the review status of the target document.

[0011] If the review status of the target document is unknown, the review reference document for the target document is directly searched from the document library.

[0012] Optionally, when the review status of the target document is known, searching for a review reference document for the target document from the document library based on the review status of the target document includes:

[0013] When the review status of the target document is known:

[0014] If the target document is in the initial review stage, then the template document corresponding to the target document is searched from the template document sub-library in the document library and used as the review reference document for the target document.

[0015] If the target document is under review, then an older version of the target document is retrieved from the document draft sub-library in the document library and used as a reference document for the review of the target document.

[0016] Optionally, when the review status of the target document is unknown, directly searching for a review reference document for the target document from the document library includes:

[0017] When the review status of the target document is unknown:

[0018] Search for older versions of the target document in the document draft sub-library of the document library, and use them as reference documents for reviewing the target document;

[0019] If no older version of the target document is found, the template document corresponding to the target document is searched from the template document sub-library in the document library and used as the review reference document for the target document.

[0020] Optionally, the step of searching for the template document corresponding to the target document from the template document sub-library in the document library includes:

[0021] The sentence repetition ratio between each template document in the template document sub-library and the target document is calculated to obtain the sentence repetition ratio corresponding to each template document.

[0022] Based on the sentence repetition ratio of each template document, the template document corresponding to the target document is determined from the template documents.

[0023] Optionally, the step of searching for an older version of the target document from the document draft sub-library in the document library includes:

[0024] Obtain target information from the target document as the target information corresponding to the target document. The target information is information that can distinguish the corresponding document from other documents.

[0025] The document draft sub-library is searched for documents whose target information matches the target information of the target document, and these documents are used as older versions of the target document.

[0026] Optionally, the review of the target document by referring to the review reference document includes:

[0027] When the reference document for reviewing the target document is the template document corresponding to the target document:

[0028] By comparing the target document with the template document corresponding to the target document textually, the differences between the target document and the template document corresponding to the target document are determined.

[0029] Determine whether the difference points are filled in at the corresponding blank positions marked in the template document corresponding to the target document;

[0030] If there is no filling rule for the blank position, and the difference point is filled in the blank position, then the difference point is determined to be a risk-free difference point; otherwise, the difference point is determined to be a risky difference point.

[0031] If there are filling rules for the blank position, and the difference point is filled in the blank position and the filling of the difference point meets the filling rules, then the difference point is determined to be a risk-free difference point; otherwise, the difference point is determined to be a risk-bearing difference point.

[0032] If the reference document for reviewing the target document is an older version of the target document:

[0033] By comparing the target document with its older version, the differences between the target document and its older version are identified and used as risk difference points.

[0034] Optionally, the review of the extracted elements includes:

[0035] Based on preset element rules, the extracted elements are reviewed, and elements that do not meet the element rules are identified as candidate risk elements to obtain a set of candidate risk elements.

[0036] For each candidate risk element in the candidate risk element set, determine whether the text item in the target document containing that candidate risk element is a risk-free text item;

[0037] If the text item containing the candidate risk element in the target document is a risk-free text item, then the candidate risk element is determined to be a risk-free element.

[0038] Optionally, determining whether a text item containing the candidate risk element in the target document is a risk-free text item includes:

[0039] Construct structured information corresponding to the text item containing the candidate risk element;

[0040] The structured information that matches the structured information corresponding to the text item containing the candidate risk element in the target document is searched in the maintained structured information database and is used as the target structured information; wherein, the structured information database includes several different structured information, each structured information is constructed based on the text item in the historical document, and each structured information corresponds to a number of occurrences;

[0041] If the target structured information is found, and the number of occurrences of the target structured information is greater than or equal to a preset number, then the text item in the target document containing the candidate risk element is determined to be a risk-free text item.

[0042] A document review device includes: a document acquisition module, a review reference document search module, a first document review module, and a second document review module;

[0043] The document acquisition module is used to acquire the document to be reviewed as the target document;

[0044] The review reference document search module is used to search for the review reference document of the target document from the document library;

[0045] The first document review module is used to review the target document by referring to the review reference document when the review reference document search module finds the review reference document of the target document;

[0046] The second document review module is used to extract elements from the target document and review the extracted elements when the review reference document search module does not find a review reference document for the target document.

[0047] A document review device includes: a memory and a processor;

[0048] The memory is used to store programs;

[0049] The processor is used to execute the program to implement each step of the document review method described above.

[0050] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the document review method described in any of the preceding claims.

[0051] This invention provides a document review method, apparatus, device, and storage medium. After obtaining the document to be reviewed (target document), the method searches for a review reference document for the target document in a document library. If a review reference document is found, the target document is reviewed with reference to the review reference document. If no review reference document is found, elements are extracted from the target document, and the extracted elements are reviewed. The document review method provided by this invention can automatically search for review reference documents for the target document in the document library and can automatically decide which strategy to use for reviewing the document. The document review method provided by this invention has a high degree of intelligence, low dependence on manual labor, high review efficiency, and good review results. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0053] Figure 1 This is a schematic diagram of the hardware architecture involved in this invention;

[0054] Figure 2 A flowchart illustrating the document review method provided in an embodiment of the present invention;

[0055] Figure 3 This is a schematic diagram illustrating the process of finding the template document corresponding to the target document from the template document sub-library in the document library, as provided in an embodiment of the present invention.

[0056] Figure 4 A schematic diagram illustrating the process of searching for an older version of a target document from a document draft sub-library in a document library, as provided in an embodiment of the present invention;

[0057] Figure 5 This is a flowchart illustrating the process of reviewing a target document by referring to a template document corresponding to the target document when the review status of the target document is in the initial review stage, as provided in an embodiment of the present invention.

[0058] Figure 6 This is a flowchart illustrating an implementation method for reviewing extracted elements according to an embodiment of the present invention.

[0059] Figure 7 A schematic diagram of the document review device provided in an embodiment of the present invention;

[0060] Figure 8 A schematic diagram of the structure of a document review device provided in an embodiment of the present invention. Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] Considering that the review process based on pure manual review is inefficient and the results are greatly affected by subjective factors, the inventors of this case attempted to propose a review scheme that can reduce the reliance on manual review and improve review efficiency. To this end, research was conducted, and it was found that the review of the document to be reviewed can adopt one of the following two strategies: first, obtain a comparison document of the document to be reviewed and compare the document to be reviewed with the comparison document; second, extract elements from the document to be reviewed and review the elements.

[0063] Based on the above findings, the inventors of this case conceived of designing an automatic review scheme based on comparison and an automatic review scheme based on elements. For the document to be reviewed, a human decides which scheme to use for review, and then uses the human-selected automatic review scheme to automatically review the document to be reviewed. When using the automatic review scheme based on comparison, the user selects and uploads the comparison document.

[0064] The inventors of this case studied the above-mentioned solution and found that although the solution reduced the reliance on human intervention to a certain extent, the decision on which review strategy to use for reviewing the documents still required human intervention and could not be made intelligently. Furthermore, it still required human intervention to select and upload documents for comparison. In other words, the above-mentioned solution was not highly intelligent and still relied heavily on human intervention.

[0065] In view of this, the inventors of this case continued their research based on the above-mentioned solutions. Through continuous research, they finally proposed a document review method with a high degree of intelligence. This document review method can automatically decide which strategy to use to review the document, and does not require manual selection and uploading of comparison documents.

[0066] Before introducing the document review method provided by this invention, the hardware architecture involved in this invention will be described first.

[0067] In one possible implementation, such as Figure 1 As shown, the hardware architecture involved in this invention may include: electronic device 101 and server 102.

[0068] For example, electronic device 101 can be any electronic product that can interact with a user through one or more means such as a keyboard, touchpad, touch screen, remote control, voice interaction or handwriting device, such as mobile phone, laptop, tablet computer, PDA, personal computer, wearable device, smart TV, PAD, etc.

[0069] It should be noted that, Figure 1 This is just one example; there can be many types of electronic devices, not limited to... Figure 1 The laptop in the middle.

[0070] For example, server 102 can be a single server, a server cluster consisting of multiple servers, or a cloud computing server center. Server 102 may include processors, memory, and network interfaces, etc.

[0071] For example, electronic device 101 can establish a connection and communicate with server 102 through a wireless communication network; for example, electronic device 101 can establish a connection and communicate with server 102 through a wired network.

[0072] Electronic device 101 obtains the document to be reviewed and sends it to server 102. The server reviews the document to be reviewed according to the document review method provided by the present invention and sends the review result to electronic device 101.

[0073] In another possible implementation, the hardware architecture involved in this invention may include an electronic device. The electronic device is one with strong data processing capabilities.

[0074] For example, an electronic device can be any electronic product that can interact with a user through one or more means such as a keyboard, touchpad, touch screen, remote control, voice interaction or handwriting device, such as a PC, mobile phone, laptop, tablet computer, PDA, personal computer, etc.

[0075] Electronic devices can review documents to be reviewed using the document review method provided by this invention and output the review results.

[0076] Those skilled in the art should understand that the above-described electronic devices and servers are merely examples, and other existing or future electronic devices or servers that are applicable to this invention should also be included within the scope of protection of this invention, and are hereby incorporated by reference.

[0077] The document review method provided by the present invention will be described below through the following embodiments.

[0078] First Embodiment

[0079] Please seeFigure 2 The diagram illustrates a flowchart of a document review method provided in an embodiment of the present invention, which may include:

[0080] Step S201: Obtain the document to be reviewed as the target document.

[0081] There are three types of documents pending review: first, documents that are to be reviewed initially and are rewritten from template documents; second, documents that are to be reviewed initially but are not rewritten from template documents; and third, documents that are to be reviewed again.

[0082] Step S202: Locate the review reference document for the target document in the document library.

[0083] When reviewing a target document, there are two situations: first, the review status of the target document (either preliminary review or secondary review) is known; second, the review status of the target document is unknown. This invention adopts different processing strategies for different situations.

[0084] Specifically, the process of finding the review reference document for the target document from the document library may include: if the review status of the target document is known, finding the review reference document for the target document from the document library based on the review status of the target document; if the review status of the target document is unknown, directly finding the review reference document for the target document from the document library.

[0085] When the review status of the target document is known, the process of finding the review reference document for the target document from the document library based on the review status may include: if the review status of the target document is preliminary review, then finding the template document corresponding to the target document from the template document sub-library in the document library as the review reference document for the target document; if the review status of the target document is secondary review, then finding the old version of the target document from the document draft sub-library in the document library as the review reference document for the target document.

[0086] In this embodiment, the document library includes a template document sub-library and a document draft sub-library. The template document sub-library contains several template documents, and the document draft sub-library contains several drafted but not yet finalized documents. If the target document's review status is "initial review," it means that there is no older version of the target document; it is the first version. In this case, the template document sub-library is searched for to find the corresponding template document. If the target document's review status is "second review," it means that the target document is not the first version; an older version exists. Therefore, the document draft sub-library is searched for to find the older version.

[0087] In cases where the review status of the target document is unknown, the process of directly searching for the review reference document of the target document in the document library may include: searching for an older version of the target document in the document draft sub-library of the document library as the review reference document of the target document; if no older version of the target document is found, then searching for the corresponding template document in the template document sub-library of the document library as the review reference document of the target document.

[0088] It should be noted that when the review status of the target document is unknown, this invention determines the review status of the target document by searching for an older version of the target document in the document draft sub-library of the document library. If an older version of the target document is found in the document draft sub-library, it indicates that the review status of the target document is under review, and the found older version document is used as the review reference document for the target document. If no older version of the target document is found in the document draft sub-library, it indicates that the review status of the target document is not under review. In this case, the review status of the target document can be further determined by searching for a template document corresponding to the target document in the template document sub-library of the document library. If a template document corresponding to the target document is found in the template document sub-library, it indicates that the review status of the target document is under review, and the template document corresponding to the target document is used as the review reference document for the target document.

[0089] Step S203a: If a review reference document for the target document is found, the target document is reviewed with reference to the review reference document.

[0090] When a review reference document for the target document is found, the review strategy adopted by this invention is to review the target document with reference to the review reference document. If the review reference document for the target document is a template document corresponding to the target document (i.e., the target document is a document to be initially reviewed and rewritten from the template document), then the target document is reviewed with reference to the template document. If the review reference document for the target document is an older version of the target document (i.e., the target document is a document to be re-reviewed), then the older version of the target document is reviewed with reference to the older version of the target document.

[0091] Step S203b: If no review reference document is found for the target document, extract elements from the target document and review the extracted elements.

[0092] If no review reference document for the target document is found, it means that the target document may be a document intended for initial review but not rewritten from a template.

[0093] In the absence of a review reference document for the target document, the review strategy adopted by this invention is to extract elements from the target document and review the extracted elements.

[0094] After obtaining the audit results, the audit results can be output, and the output audit results can be manually verified.

[0095] The document review method provided in this embodiment of the invention, after obtaining the document to be reviewed (i.e., the target document), searches for a review reference document for the target document in the document library. If a review reference document is found, the target document is reviewed with reference to the review reference document. If no review reference document is found, elements are extracted from the target document, and the extracted elements are reviewed. The document review method provided in this embodiment of the invention can automatically decide which strategy to use for reviewing the document to be reviewed, and can also automatically search for a review reference document for the target document in the document library. The document review method provided by this invention has a high degree of intelligence, a low degree of dependence on manual labor, high review efficiency, and good review effect.

[0096] Second Embodiment

[0097] This embodiment describes the specific implementation process of "finding the template document corresponding to the target document from the template document sub-library in the document library" and "finding the old version document of the target document from the document draft sub-library in the document library" in the above embodiment.

[0098] First, we will introduce the specific implementation process of finding the template document corresponding to the target document from the template document sub-library of the document library.

[0099] Please see Figure 3 This illustrates a flowchart of the process for finding the template document corresponding to the target document from the template document sub-library in the document library, which may include:

[0100] Step S301: Calculate the sentence repetition ratio between each template document in the template document sub-library and the target document to obtain the sentence repetition ratio for each template document in the template document sub-library.

[0101] This invention pre-indexes the sentences in each template document within the template document sub-library. Optionally, the method for indexing the sentences in each template document is as follows: first, the text in the template document is split into sentences; then, punctuation marks and whitespace are removed; short sentences with fewer than a preset word count threshold and sentences with fill-in-the-blank fields are removed; finally, an index is created for the remaining sentences.

[0102] When calculating the sentence repetition rate between each template document and the target document, for each sentence in the target document, the index built for the sentences in the template document is used to retrieve sentences in the template document that are identical to that sentence, thus obtaining the corresponding retrieval result. After obtaining the retrieval results for each sentence in the target document, the number of repetitive sentences is determined based on the retrieval results for each sentence in the target document. This allows us to obtain the proportion of repetitive sentences in the template document, i.e., the sentence repetition rate between the template document and the target document. Through this method, the sentence repetition rate between each template document in the template document sub-library and the target document can be obtained.

[0103] Step S302: Based on the sentence repetition ratio of each template document in the template document sub-library, determine the template document corresponding to the target document from each template document in the template document sub-library.

[0104] Specifically, from the sentence repetition ratios of each template document in the template document sub-library, the largest sentence repetition ratio is determined. It is then determined whether the largest sentence repetition ratio is greater than a preset sentence repetition ratio threshold. If so, the template document corresponding to the largest sentence repetition ratio is determined as the template document corresponding to the target document.

[0105] The following section describes the specific implementation process of finding older versions of the target document from the document draft sub-library of the document library.

[0106] Please see Figure 4 This illustrates a flowchart of a process for finding an older version of a target document from a document draft sub-library within a document library, which may include:

[0107] Step S401: Obtain target information from the target document and use it as the target information corresponding to the target document.

[0108] The target information is the information that can distinguish the corresponding document from other documents.

[0109] For example, if the target document is a contract, information such as the contract number, the name of Party A, the name of Party B, the subject matter, and the project name can be extracted from the target document as the target information corresponding to the target document.

[0110] Step S402: Find the document whose target information matches the target information of the target document in the document draft sub-library, and use it as the old version document of the target document.

[0111] This invention pre-acquires the target information corresponding to each document in the document draft sub-library. When searching for an old version of the target document in the document draft sub-library, the target information corresponding to the target document is obtained. Then, the target information that matches the target information corresponding to the target document is determined from the target information corresponding to each document in the document draft sub-library. The document corresponding to the target information that matches the target information corresponding to the target document in the document draft sub-library is determined as the old version of the target document.

[0112] For example, the target document is a contract, and each document in the document draft sub-library is also a contract. The contract number, Party A's name, Party B's name, and subject matter are obtained from each document in the document draft sub-library beforehand as target information. This yields the target information corresponding to each document in the document draft sub-library. When searching for an older version of the target document in the document draft sub-library, the contract number, Party A's name, Party B's name, and subject matter are obtained from the target document to obtain the target information corresponding to the target document. After obtaining the target information corresponding to the target document, the target information matching the target information corresponding to the target document is searched from the target information corresponding to each document in the document draft sub-library. For example, if the contract number, Party A's name, Party B's name, and subject matter corresponding to a document in the document draft sub-library match the contract number, Party A's name, Party B's name, and subject matter corresponding to the target document, then that document in the document draft sub-library is determined to be an older version of the target document.

[0113] Third Embodiment

[0114] In the document review method provided in the above embodiments, when a review reference document for the target document is found, the target document is reviewed with reference to the review reference document. This embodiment describes the specific implementation process of reviewing the target document with reference to the review reference document.

[0115] The above embodiments mention that when the target document's review status is preliminary review, the reference document for reviewing the target document is the template document corresponding to the target document; when the target document's review status is secondary review, the reference document for reviewing the target document is the old version of the target document. This embodiment first introduces the specific implementation process of reviewing the target document by referring to the template document corresponding to the target document when the target document's review status is preliminary review.

[0116] When the target document is in the initial review stage, there are several ways to review the target document by referring to the template document corresponding to the target document. In one possible approach, the differences between the target document and the template document can be determined by comparing the text of the target document with the template document. These differences can then be directly identified as risk discrepancies.

[0117] Considering that further manual verification of the audit results (the aforementioned risk difference points) is generally required, and that there are usually many differences between the target document and the template document corresponding to the target document, directly identifying the differences between the target document and the template document corresponding to the target document as risk difference points would lead to a heavy burden on subsequent manual review and affect the review efficiency. Therefore, this embodiment provides another preferred implementation method.

[0118] Please see Figure 5 This illustrates a flowchart illustrating another implementation method for reviewing a target document by referring to the corresponding template document when the target document's review status is in the initial review stage. This method may include:

[0119] Step S501: By comparing the text of the target document with the template document corresponding to the target document, the differences between the target document and the template document corresponding to the target document are determined.

[0120] Step S502: Determine whether the differences have been filled in the blanks marked in the template document corresponding to the target document.

[0121] This invention pre-marks the blank positions in each template document in the template document sub-library.

[0122] Step S503a: If there is no filling rule for the blank position, and the difference point is filled in the blank position marked in the template document corresponding to the target document, then the difference point is determined to be a risk-free difference point; otherwise, the difference point is determined to be a risky difference point.

[0123] Step S503b: If there are filling rules for the blank position, and the difference point is filled in the blank position marked in the template document corresponding to the target document, and the filling of the difference point meets the filling rules of the blank position, then the difference point is determined to be a risk-free difference point; otherwise, the difference point is determined to be a risky difference point.

[0124] The table below provides examples of the clauses to be filled in, the rules for filling in the blanks, and prompts in the template contract:

[0125]

[0126] After identifying the risk difference points, the risk difference points can be output, and prompts can be generated for each risk difference point, indicating why the difference point poses a risk.

[0127] The second method for reviewing the target document by referring to the template document corresponding to the target document involves identifying the differences between the target document and the template document corresponding to the target document, then assessing the risks of these differences, and filtering out the differences that are deemed risk-free. This filtering of the differences reduces the burden of subsequent manual review and improves review efficiency.

[0128] Fourth embodiment

[0129] In the document review method provided in the above embodiments, when no review reference document for the target document is found, "elements are extracted from the target document and the extracted elements are reviewed". This embodiment introduces the specific implementation process of "reviewing the extracted elements".

[0130] There are several ways to review the extracted elements. One possible approach is to review the extracted elements based on preset element rules and identify elements that do not meet the rules as risk elements.

[0131] In some cases, considering that certain elements that do not meet the element rules need to be ignored as non-risk elements, for example, in the case of contracts, a major source of contracts not based on the company's template is Party A. That is, the document to be reviewed uses Party A's template, in which some clauses are slightly biased towards Party A and cannot be changed, but usually will not cause disputes. Party B can only accept such clauses. In other words, there is no need to review such elements later, and they can be directly ignored as non-risk elements. In view of this, this embodiment provides another way to review the extracted elements.

[0132] Please see Figure 6 The diagram illustrates an alternative implementation for reviewing extracted elements, which may include:

[0133] Step S601: Based on the preset element rules, the extracted elements are reviewed, and elements that do not meet the element rules are identified as candidate risk elements to obtain a set of candidate risk elements.

[0134] For example, if the target document is a contract, the preset element rules may include, but are not limited to, the following: capitalization, consistency of contract amount, reasonableness of liquidated damages, completeness of contact information, etc.

[0135] Step S602: For each candidate risk element in the candidate risk element set, determine whether the text item in the target document containing the candidate risk element is a risk-free text item. If yes, proceed to step S603a; otherwise, proceed to step S603b.

[0136] The process of determining whether a text item in the target document containing the candidate risk element is a risk-free text item may include:

[0137] Step S6021: Construct structured information corresponding to the text item containing the candidate risk element.

[0138] For example, if the target document is a contract, then the text items can be the clauses in the contract, and the structured information corresponding to the clauses can be a triple (Party A, Party B, clause content).

[0139] Step S6022: Search the maintained structured information database for structured information that matches the structured information corresponding to the text item containing the candidate risk element, and use it as the target structured information.

[0140] The structured information database includes several different structured information entries. Each structured information entry is constructed based on text items in historical documents, and each structured information entry corresponds to a number of occurrences.

[0141] For example, the target document is a contract, and the maintained structured information database includes several triples. These triples are continuously added to the structured information database as follows: Whenever a contract passes the review process and is signed and archived, a triple (Party A, Party B, clause content) is constructed for each clause in the contract. After constructing each triple, it is determined whether the triple exists in the structured information database. If the triple does not exist in the structured information database, it is stored in the structured database, and the occurrence count of the triple is recorded as 1. If the triple exists in the structured information database, the occurrence count of the triple is incremented by 1. If the occurrence count of a certain triple in the structured information database is greater than or equal to a preset count, the clause corresponding to that triple is considered a risk-free clause.

[0142] Step S6023a: If target structured information is found, and the number of occurrences of the target structured information is greater than or equal to the preset number, then the text item containing the candidate risk element is determined to be a risk-free text item.

[0143] Step S6023b: If no target structured information is found, the structured information corresponding to the text item containing the candidate risk element is stored in the structured information database, and the occurrence count of the currently stored structured information is set to 1.

[0144] Step S603a: Determine that the candidate risk element is a non-risk element.

[0145] Step S603b: Determine the candidate risk factor as a risk factor.

[0146] After identifying the risk factors, they can be output for further manual verification. During the manual verification stage, if an auditor adds a risk annotation or modifies a risk-free text item (such as a risk-free clause), the system will automatically prompt "This is a risk-free text item; please confirm whether you want to modify it." If the auditor confirms the modification, the structured information (such as triples) corresponding to that text item in the structured information database will be marked as risky, and the occurrence count of the structured information (such as triples) corresponding to that text item will be reset to zero.

[0147] Fifth embodiment

[0148] This invention provides a document review device. The document review device provided in this invention is described below. The document review device described below can be referred to in correspondence with the document review method described above.

[0149] Please see Figure 7 The diagram shows a structural schematic of a document review device provided in an embodiment of the present invention. The document review device may include: a document acquisition module 701, a review reference document search module 702, a first document review module 703a, and a second document review module 703b.

[0150] The document acquisition module 701 is used to acquire documents to be reviewed as target documents.

[0151] The review reference document search module 702 is used to search for the review reference document of the target document from the document library;

[0152] The first document review module 703a is used to review the target document by referring to the review reference document when the review reference document search module finds the review reference document of the target document.

[0153] The second document review module 703b is used to extract elements from the target document and review the extracted elements when the review reference document search module does not find a review reference document for the target document.

[0154] Optionally, the audit reference document search module 702 includes: a first search submodule and a second search submodule.

[0155] The first search submodule is used to search for the review reference document of the target document from the document library based on the review status of the target document, when the review status of the target document is known.

[0156] The second search submodule is used to directly search for the review reference document of the target document from the document library when the review status of the target document is unknown.

[0157] Optionally, when the review status of the target document is known, the first search submodule, when searching for the review reference document of the target document from the document library based on the review status of the target document, is specifically used for:

[0158] When the review status of the target document is known:

[0159] If the target document is in the initial review stage, then the template document corresponding to the target document is searched from the template document sub-library in the document library and used as the review reference document for the target document.

[0160] If the target document is under review, then an older version of the target document is retrieved from the document draft sub-library in the document library and used as a reference document for the review of the target document.

[0161] Optionally, when the review status of the target document is unknown, the second search submodule, when directly searching for the review reference document of the target document from the document library, is specifically used for:

[0162] When the review status of the target document is unknown:

[0163] Search for older versions of the target document in the document draft sub-library of the document library, and use them as reference documents for reviewing the target document;

[0164] If no older version of the target document is found, the template document corresponding to the target document is searched from the template document sub-library in the document library and used as the review reference document for the target document.

[0165] Optionally, when the first search submodule or the second submodule searches for the template document corresponding to the target document in the template document sub-library of the document library, it is specifically used for:

[0166] The sentence repetition ratio between each template document in the template document sub-library and the target document is calculated to obtain the sentence repetition ratio corresponding to each template document.

[0167] Based on the sentence repetition ratio of each template document, the template document corresponding to the target document is determined from the template documents.

[0168] Optionally, when the first search submodule or the second submodule searches for an older version of the target document from the document draft sub-library in the document library, it is specifically used for:

[0169] Obtain target information from the target document as the target information corresponding to the target document. The target information is information that can distinguish the corresponding document from other documents.

[0170] The document draft sub-library is searched for documents whose target information matches the target information of the target document, and these documents are used as older versions of the target document.

[0171] Optionally, the first document review module 703a may include: a first review submodule and a second review submodule.

[0172] The first review submodule is used when the review reference document for the target document is the template document corresponding to the target document:

[0173] By comparing the target document with the template document corresponding to the target document textually, the differences between the target document and the template document corresponding to the target document are determined.

[0174] Determine whether the difference points are filled in at the corresponding blank positions marked in the template document corresponding to the target document;

[0175] If there is no filling rule for the blank position, and the difference point is filled in the blank position, then the difference point is determined to be a risk-free difference point; otherwise, the difference point is determined to be a risky difference point.

[0176] If there are filling rules for the blank position, and the difference point is filled in the blank position and the filling of the difference point meets the filling rules, then the difference point is determined to be a risk-free difference point; otherwise, the difference point is determined to be a risk-bearing difference point.

[0177] The second review submodule is used when the review reference document for the target document is an older version of the target document:

[0178] By comparing the target document with its older version, the differences between the target document and its older version are identified and used as risk difference points.

[0179] Optionally, the second document review module 703b may include: an element review submodule, a text item review submodule, and an element discrimination submodule.

[0180] The element review submodule is used to review the extracted elements based on preset element rules, and identify elements that do not meet the element rules as candidate risk elements to obtain a set of candidate risk elements.

[0181] The text item review submodule is used to determine whether a text item in the target document containing a candidate risk element is a risk-free text item for each candidate risk element in the candidate risk element set.

[0182] The element discrimination submodule is used to determine that the candidate risk element is a risk-free element when the text item review submodule determines that the text item containing the candidate risk element in the target document is a risk-free text item.

[0183] Optionally, when determining whether a text item containing the candidate risk element in the target document is a risk-free text item, the text item review submodule is specifically used for:

[0184] Construct structured information corresponding to the text item containing the candidate risk element;

[0185] The structured information that matches the structured information corresponding to the text item containing the candidate risk element in the target document is searched in the maintained structured information database and is used as the target structured information; wherein, the structured information database includes several different structured information, each structured information is constructed based on the text item in the historical document, and each structured information corresponds to a number of occurrences;

[0186] If the target structured information is found, and the number of occurrences of the target structured information is greater than or equal to a preset number, then the text item in the target document containing the candidate risk element is determined to be a risk-free text item.

[0187] This invention provides a document review device. After obtaining the document to be reviewed (target document), it searches for a review reference document for the target document in a document library. If a review reference document is found, the target document is reviewed with reference to the review reference document. If no review reference document is found, elements are extracted from the target document, and the extracted elements are reviewed. The document review device provided by this invention can automatically search for review reference documents for the target document in the document library and can automatically decide which strategy to use to review the document. The document review method provided by this invention has a high degree of intelligence, low dependence on manual labor, high review efficiency, and good review effect.

[0188] Sixth Embodiment

[0189] This invention provides a document review device; please refer to [link / reference]. Figure 8The diagram shows the structure of the document review device, which may include: a processor 801, a communication interface 802, a memory 803, and a communication bus 804.

[0190] In this embodiment of the invention, the number of processor 801, communication interface 802, memory 803, and communication bus 808 is at least one, and processor 801, communication interface 802, and memory 803 communicate with each other through communication bus 804.

[0191] The processor 801 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention.

[0192] The memory 803 may include high-speed RAM, or it may also include non-volatile memory, such as at least one disk storage device;

[0193] The memory stores a program, which the processor can call. The program is used for:

[0194] Obtain the document to be reviewed and use it as the target document;

[0195] Locate the review reference document for the target document in the document library;

[0196] If a review reference document for the target document is found, the target document is reviewed with reference to the review reference document.

[0197] If no review reference document is found for the target document, elements are extracted from the target document and reviewed.

[0198] Optionally, the refined and extended functions of the program can be found in the description above.

[0199] Seventh Embodiment

[0200] This invention also provides a readable storage medium that stores a program suitable for execution by a processor, the program being used for:

[0201] Obtain the document to be reviewed and use it as the target document;

[0202] Locate the review reference document for the target document in the document library;

[0203] If a review reference document for the target document is found, the target document is reviewed with reference to the review reference document.

[0204] If no review reference document is found for the target document, elements are extracted from the target document and reviewed.

[0205] Optionally, the refined and extended functions of the program can be found in the description above.

[0206] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0207] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0208] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily 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 invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A document review method, characterized in that, include: Obtain the document to be reviewed and use it as the target document; Locate the review reference document for the target document in the document library; If a review reference document for the target document is found, the target document is reviewed with reference to the review reference document. If no review reference document is found for the target document, elements are extracted from the target document and reviewed. The review of the extracted elements includes: Based on preset element rules, the extracted elements are reviewed, and elements that do not meet the element rules are identified as candidate risk elements to obtain a set of candidate risk elements. For each candidate risk element in the candidate risk element set, determine whether the text item in the target document containing that candidate risk element is a risk-free text item; If the text item containing the candidate risk element in the target document is a risk-free text item, then the candidate risk element is determined to be a risk-free element.

2. The document review method according to claim 1, characterized in that, The step of searching the document library for the review reference document of the target document includes: If the review status of the target document is known, the review reference document for the target document is retrieved from the document library based on the review status of the target document; If the review status of the target document is unknown, the review reference document for the target document is directly searched from the document library.

3. The document review method according to claim 2, characterized in that, When the review status of the target document is known, the step of searching for a review reference document for the target document from the document library based on the review status of the target document includes: When the review status of the target document is known: If the target document is in the initial review stage, then the template document corresponding to the target document is searched from the template document sub-library in the document library and used as the review reference document for the target document. If the target document is under review, then an older version of the target document is retrieved from the document draft sub-library in the document library and used as a reference document for the review of the target document.

4. The document review method according to claim 2, characterized in that, When the review status of the target document is unknown, directly searching for the review reference document of the target document in the document library includes: When the review status of the target document is unknown: Search for older versions of the target document in the document draft sub-library of the document library, and use them as reference documents for reviewing the target document; If no older version of the target document is found, the template document corresponding to the target document is searched from the template document sub-library in the document library and used as the review reference document for the target document.

5. The document review method according to claim 3 or 4, characterized in that, The step of searching for the template document corresponding to the target document from the template document sub-library in the document library includes: The sentence repetition ratio between each template document in the template document sub-library and the target document is calculated to obtain the sentence repetition ratio corresponding to each template document. Based on the sentence repetition ratio of each template document, the template document corresponding to the target document is determined from the template documents.

6. The document review method according to claim 3 or 4, characterized in that, The step of searching for an older version of the target document from the document draft sub-library in the document library includes: Obtain target information from the target document as the target information corresponding to the target document. The target information is information that can distinguish the corresponding document from other documents. The document draft sub-library is searched for documents whose target information matches the target information of the target document, and these documents are used as older versions of the target document.

7. The document review method according to claim 3 or 4, characterized in that, The review of the target document by referring to the review reference document includes: When the reference document for reviewing the target document is the template document corresponding to the target document: By comparing the target document with the template document corresponding to the target document textually, the differences between the target document and the template document corresponding to the target document are determined. Determine whether the difference points are filled in at the corresponding blank positions marked in the template document corresponding to the target document; If there is no filling rule for the blank position, and the difference point is filled in the blank position, then the difference point is determined to be a risk-free difference point; otherwise, the difference point is determined to be a risky difference point. If there are filling rules for the blank position, and the difference point is filled in the blank position and the filling of the difference point meets the filling rules, then the difference point is determined to be a risk-free difference point; otherwise, the difference point is determined to be a risk-bearing difference point. If the reference document for reviewing the target document is an older version of the target document: By comparing the target document with its older version, the differences between the target document and its older version are identified and used as risk difference points.

8. The document review method according to claim 1, characterized in that, Determining whether a text item in the target document containing the candidate risk element is a risk-free text item includes: Construct structured information corresponding to the text item containing the candidate risk element; The structured information that matches the structured information corresponding to the text item containing the candidate risk element in the target document is searched in the maintained structured information database and is used as the target structured information; wherein, the structured information database includes several different structured information, each structured information is constructed based on the text item in the historical document, and each structured information corresponds to a number of occurrences; If the target structured information is found, and the number of occurrences of the target structured information is greater than or equal to a preset number, then the text item in the target document containing the candidate risk element is determined to be a risk-free text item.

9. A document review device, characterized in that, include: Document acquisition module, reference document search module, first document review module, and second document review module; The document acquisition module is used to acquire the document to be reviewed as the target document; The review reference document search module is used to search for the review reference document of the target document from the document library; The first document review module is used to review the target document by referring to the review reference document when the review reference document search module finds the review reference document of the target document; The second document review module is used to extract elements from the target document and review the extracted elements when the review reference document search module does not find a review reference document for the target document. Specifically, the second document review module, when reviewing the extracted elements, is used for: Based on preset element rules, the extracted elements are reviewed, and elements that do not meet the element rules are identified as candidate risk elements to obtain a set of candidate risk elements. For each candidate risk element in the candidate risk element set, determine whether the text item in the target document containing that candidate risk element is a risk-free text item; If the text item containing the candidate risk element in the target document is a risk-free text item, then the candidate risk element is determined to be a risk-free element.

10. A document review device, characterized in that, include: Memory and processor; The memory is used to store programs; The processor is used to execute the program to implement each step of the document review method as described in any one of claims 1 to 8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements each step of the document review method as described in any one of claims 1 to 8.

Images