Document auditing method and device, computer readable storage medium and electronic equipment
By determining the semantic layout template and extracting the text to be reviewed during document review, the problem of low review efficiency in existing technologies is solved, and efficient and accurate cross-document information association review is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHONGQING SOKON IND GRP CO LTD
- Filing Date
- 2025-12-30
- Publication Date
- 2026-05-01
AI Technical Summary
In existing technologies, document review methods rely on human experience, which leads to low review efficiency and makes it difficult to systematically identify information inconsistencies between different documents to be reviewed, thus affecting the accuracy of the review results.
By determining the semantic layout templates corresponding to different documents to be reviewed in the document review task, extracting the corresponding text to be reviewed, and determining the review result based on the text to be reviewed, cross-document recognition and content association review are achieved.
It improves the efficiency and accuracy of cross-document review, reduces reliance on manual verification, and enhances the systematicness and reliability of the review process.
Smart Images
Figure CN121960448A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of document processing technology, and more specifically, to a document review method, apparatus, computer-readable storage medium, and electronic device. Background Technology
[0002] With the popularization of digital management in enterprises, electronic verification of various materials submitted by new employees (such as ID cards, academic certificates, and proof of termination of employment) has become a routine part of human resource management.
[0003] In related technologies, common document review methods typically employ a sequential processing approach. This involves first independently identifying and reviewing each document to be reviewed. After all documents have been processed, reviewers manually compare and verify the consistency of information between different documents. This fragments the entire document review process into two disconnected stages: "independent identification" and "post-review manual comparison." This not only heavily relies on human experience and effort, leading to low review efficiency, but also makes it difficult to systematically identify information inconsistencies between different documents, thus compromising the accuracy and reliability of the review results. Summary of the Invention
[0004] To address the aforementioned technical problems, embodiments of this application provide a document review method, apparatus, computer-readable storage medium, and electronic device that can eliminate the need for manual verification of information inconsistencies between different documents to be reviewed.
[0005] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part from practice of this application.
[0006] According to one aspect of the embodiments of this application, a document review method is provided, comprising: determining a semantic layout template corresponding to each of the different documents to be reviewed in a document review task; extracting the corresponding text to be reviewed based on the semantic layout template of each document to be reviewed; and determining the review result of the document review task based on the text to be reviewed corresponding to the different documents to be reviewed.
[0007] According to one aspect of the embodiments of this application, a document review apparatus is provided, comprising: an acquisition module configured to determine a semantic layout template corresponding to each of the different documents to be reviewed in a document review task; an extraction module configured to extract the corresponding text to be reviewed for each of the documents to be reviewed based on the semantic layout template of each document to be reviewed; and an review module configured to determine the review result of the document review task based on the text to be reviewed corresponding to the different documents to be reviewed.
[0008] In some embodiments of this application, based on the foregoing scheme, the acquisition module is further configured to: determine the target document type corresponding to the document to be reviewed; and obtain the semantic layout template corresponding to the document to be reviewed from a preset template library according to the target document type.
[0009] In some embodiments of this application, based on the foregoing scheme, the extraction module is further configured to: determine a target extraction region in the document to be reviewed based on the semantic layout template corresponding to the document to be reviewed; wherein, the target extraction region is the position of the text to be reviewed in the document to be reviewed indicated by the semantic layout template; extract the target text content of the document to be reviewed in the target extraction region; and use the target text content as the text to be reviewed corresponding to the document to be reviewed.
[0010] In some embodiments of this application, based on the foregoing scheme, the review module is further configured to: determine a target review rule according to the review text type corresponding to the text to be reviewed; review the text to be reviewed in different documents to be reviewed based on the target review rule; if the review is passed, determine the review result of the document review task as a first review result; wherein, the first review result indicates that the document review task has passed the review; if the review is not passed, determine the review result of the document review task as a second review result; wherein, the second review result indicates that the document review task has not passed the review.
[0011] In some embodiments of this application, based on the foregoing scheme, the review module is further configured to: before determining the review result of the document review task based on the text to be reviewed corresponding to different documents to be reviewed, evaluate the recognition quality level of the text to be reviewed corresponding to the document to be reviewed; if it is determined that the recognition quality level does not reach the preset quality level, then locate the current location of the text to be reviewed in the document to be reviewed; adjust the text to be reviewed corresponding to the document to be reviewed based on the current location.
[0012] In some embodiments of this application, based on the foregoing scheme, the review module is further configured to: after locating the current location of the text to be reviewed in the document to be reviewed, increment the historical correction count of the text to be reviewed in the document to be reviewed by one; if it is determined that the historical correction count is greater than the preset correction count, adjust the semantic layout template corresponding to the document to be reviewed according to the current location.
[0013] In some embodiments of this application, based on the foregoing scheme, the acquisition module is further configured to: acquire the current document format corresponding to the document to be reviewed; determine a target preprocessing method according to the current document format; wherein, the target preprocessing method is used to convert the current document format into the review document format; and preprocess the document to be reviewed according to the target preprocessing method so that the document to be reviewed is converted from the current document format into the review document format.
[0014] According to one aspect of the embodiments of this application, a computer-readable storage medium is provided, on which computer-readable instructions are stored, which, when executed by a computer's processor, cause the computer to perform the document review method as described in the above embodiments.
[0015] According to one aspect of the embodiments of this application, an electronic device is provided, including: one or more processors; and a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the electronic device enables the document review method as described in the above embodiments.
[0016] In the technical solution of this application embodiment, the semantic layout template corresponding to each of the different documents to be reviewed in the document review task can be determined first. Then, the corresponding text to be reviewed can be extracted according to the semantic layout template corresponding to each document to be reviewed. Then, the review result of the document review task can be determined based on the text to be reviewed corresponding to the different documents to be reviewed. This achieves the purpose of identifying and reviewing the content association of multiple documents to be reviewed, so that there is no need for manual verification of information contradictions between different documents to be reviewed, thereby improving the efficiency of cross-document review and the accuracy of review results. Attached Figure Description
[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort. In the drawings: Figure 1 This is a flowchart illustrating a document review method as shown in an exemplary embodiment of this application; Figure 2 yes Figure 1 The flowchart in an example embodiment preceding step S130 in the illustrated embodiment; Figure 3 Is Figure 2 A flowchart of another document review method provided based on the above; Figure 4 yes Figure 1 The flowchart in an example embodiment preceding step S110 in the illustrated embodiment; Figure 5 This is a block diagram illustrating a document review apparatus according to an exemplary embodiment of this application; Figure 6 This is a schematic diagram of the structure of an electronic device shown in an exemplary embodiment of this application. Detailed Implementation
[0018] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this application more comprehensive and complete, and to fully convey the concept of the exemplary embodiments to those skilled in the art.
[0019] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a thorough understanding of embodiments of this application. However, those skilled in the art will recognize that the technical solutions of this application can be practiced without one or more of the specific details, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of this application.
[0020] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.
[0021] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.
[0022] It should be noted that "multiple" in this article refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.
[0023] The technical solution of this application embodiment proposes a text review method, which is detailed below. Figure 1As shown. The method includes at least steps S110 to S130, which are described in detail below: In step S110, the semantic layout templates corresponding to different documents to be reviewed in the document review task are determined.
[0024] It should be noted that when reviewers conduct cross-document reviews, they can create document review tasks and include all documents involved in the task as pending documents for review. For example, when reviewing onboarding materials, reviewers can create a document review task and include documents such as ID cards, educational certificates, and resignation certificates as pending documents for review.
[0025] The semantic layout template corresponding to the document to be reviewed includes the text type contained in the document (e.g., "name", "ID number", "address" and other text types), as well as the position of that text type in the document to be reviewed.
[0026] In some embodiments of this application, the method for determining the semantic layout template corresponding to different documents to be reviewed in a document review task can be to first determine the target document type corresponding to the document to be reviewed, and then obtain the semantic layout template corresponding to the document to be reviewed from the preset template library according to the target document type. That is, different document types and different semantic layout templates can be associated and stored in the preset template library in advance, so as to determine the corresponding semantic layout template based on the document type corresponding to the document to be reviewed.
[0027] For example, when the target document type corresponding to the document to be reviewed is an ID card, the semantic layout template corresponding to the ID card can be obtained from the preset template library as the semantic layout template corresponding to the document to be reviewed.
[0028] In some embodiments of this application, when determining the target document type corresponding to the document to be reviewed, the document to be reviewed can be input into a lightweight document classification model (such as MobileNetV3 or CLIP image encoder) so that the target document type corresponding to the document to be reviewed can be quickly output by the model.
[0029] In addition, in some embodiments of this application, after determining the target document type corresponding to the document to be reviewed, the classification confidence of the target document type can also be obtained. If it is determined that the classification confidence does not reach the preset classification confidence threshold, the preset default type is used as the document type of the document to be reviewed. In the subsequent process of obtaining the semantic layout template corresponding to the document to be reviewed from the preset template library according to the target document type, the semantic layout template corresponding to the preset default type in the preset template library is used as the semantic layout template corresponding to the document to be reviewed, so as to improve the matching degree between the document to be reviewed and its corresponding semantic layout template.
[0030] In step S120, the corresponding text to be reviewed is extracted according to the semantic layout template of each document to be reviewed.
[0031] In the embodiments of this application, after determining the semantic layout templates corresponding to different documents to be reviewed, the corresponding text to be reviewed can be extracted according to the semantic layout templates of each document to be reviewed.
[0032] Among them, the text to be reviewed represents the text content that needs to be reviewed in the document review task for different documents.
[0033] In some embodiments of this application, the target extraction region in the document to be reviewed can be determined first based on the semantic layout template corresponding to the document to be reviewed. The target extraction region is the position of the text to be reviewed in the document to be reviewed as indicated by the semantic layout template. Then, the target text content of the document to be reviewed in the target extraction region is extracted, and then the target text content is used as the text to be reviewed corresponding to the document to be reviewed.
[0034] In some embodiments of this application, the method of determining the target extraction region based on the semantic layout template corresponding to the document to be reviewed can first determine the review text type corresponding to the text to be reviewed, then obtain the target text region corresponding to the review text type from the semantic layout template, and then use the target text region as the target extraction region to achieve the purpose of determining the target extraction region.
[0035] In some embodiments of this application, the method for extracting the target text content of the document to be reviewed in the target extraction area can be as follows: first, perform OCR (Optical Character Recognition) processing on the document to be reviewed to obtain the OCR text corresponding to the document to be reviewed; then, based on the OCR text of the document to be reviewed and the target extraction area, output the OCR text of the document to be reviewed in the target extraction area; and finally, use the OCR text of the document to be reviewed in the target extraction area as the target text content, thereby achieving the purpose of extracting the target text content of the document to be reviewed in the target extraction area.
[0036] In some embodiments of this application, the method of extracting the target text content of the document to be reviewed in the target extraction area can also be to generate an extraction instruction based on the target extraction area and a preset extraction instruction structure, and then extract the target text content based on the extraction instruction and the document to be reviewed through a preset multimodal large model.
[0037] For example, if the document to be reviewed is an ID card, and the text type corresponding to the text to be reviewed is name and ID number, then the target extraction area is the position of the name and ID number on the ID card, i.e., the name is located in the first line to the right of the national emblem, and the ID number is located in the lower right area. Correspondingly, if the preset extraction instruction structure is "extract the text to be reviewed from the document to be reviewed, and the text to be reviewed is located in the target extraction area," then the extraction instruction generated based on the target extraction area and the preset extraction instruction structure is "extract the name and ID number from the ID card, and the name is located in the first line to the right of the national emblem, and the ID number is located in the lower right area." Then, using a preset multimodal large model based on the extraction instruction and the document to be reviewed, the target text content as the text to be reviewed can be extracted.
[0038] In step S130, the review result of the document review task is determined based on the text to be reviewed corresponding to different documents to be reviewed.
[0039] In the embodiments of this application, after extracting the text to be reviewed for each document to be reviewed, the review result of the document review task can be determined based on the text to be reviewed for each document to be reviewed.
[0040] In some embodiments of this application, the method of determining the review result of a document review task based on the text to be reviewed corresponding to different documents to be reviewed can first determine the target review rule according to the type of the text to be reviewed, and then review the text to be reviewed in different documents to be reviewed based on the target review rule. If the review is passed, the review result of the document review task is determined as the first review result indicating that the document review task has passed the review. If the review is not passed, the review result of the document review task is determined as the second review result indicating that the document review task has not passed the review.
[0041] For example, if the text to be reviewed corresponds to a name, the review rule corresponding to the name can be used as the target review rule. That is, determine whether the text content corresponding to the name in different documents to be reviewed is consistent. If it is found to be consistent, the review is approved. Conversely, if it is found to be inconsistent, the review is rejected.
[0042] When the type of text to be reviewed is date, the review rules corresponding to the date can be used as the target review rules. This involves determining whether the text content corresponding to the date in different documents to be reviewed conforms to preset logical rules (e.g., whether the award date in a document for a degree certificate is later than the graduation date in a document for an academic certificate; whether the birth date in a document for an ID card is the same as the birth date in a document for a medical examination report; whether the award date in a document for a degree certificate and the graduation date in a document for an academic certificate are later than the birth date in a document for an ID card, etc.). If they match, the review is approved. Conversely, if they do not match, the review is rejected.
[0043] In some embodiments of this application, after determining the review result of the document review task as the first review result, the text to be reviewed in different documents to be reviewed can be integrated to merge the text to be reviewed with the same text type in different documents to be reviewed, thereby obtaining the review result text, which is convenient for reviewers to view.
[0044] In some embodiments of this application, after determining the review result of the document review task as the second review result, abnormal review text can be determined based on the text to be reviewed in different documents to be reviewed. Then, an abnormal report is generated based on the abnormal review text and the document to be reviewed corresponding to the abnormal review text, so that the reviewers can view the document to be reviewed that has abnormalities and the abnormal text content in the document through the abnormal report.
[0045] Through the above implementation method, the semantic layout templates corresponding to different documents to be reviewed in the document review task can be determined first. Then, the corresponding text to be reviewed can be extracted according to the semantic layout templates of each document to be reviewed. Finally, the review result of the document review task can be determined based on the text to be reviewed of different documents to be reviewed. This achieves the purpose of identifying and reviewing the content association of multiple documents to be reviewed, so that there is no need for manual verification of information contradictions between different documents to be reviewed, thereby improving the efficiency of cross-document review and the accuracy of review results.
[0046] See Figure 2 , Figure 2 This is a flowchart illustrating a document review method according to another exemplary embodiment. For example... Figure 2 As shown, in Figure 1 Before step S130 in the illustrated embodiment, the method may further include steps S210 to S230, which are described in detail below: In step S210, the recognition quality level of the text corresponding to the document to be reviewed is evaluated.
[0047] In the embodiments of this application, before determining the review result of the document review task based on the text to be reviewed corresponding to different documents to be reviewed, the recognition quality level of the text to be reviewed corresponding to the documents to be reviewed can also be evaluated.
[0048] Among them, the recognition quality level represents the recognition status of the text content in the text to be reviewed.
[0049] In some embodiments of this application, the method for assessing the recognition quality level of the text corresponding to the document to be reviewed can be to obtain the recognition confidence level of the text content in the text to be reviewed, and determine the recognition quality level of the text to be reviewed based on the recognition confidence level. In this case, the recognition confidence level and the recognition quality level are positively correlated, that is, the higher the recognition confidence level, the higher the recognition quality level.
[0050] In some embodiments of this application, the correlation between the text content and text type in the text to be reviewed can be obtained, and the recognition quality level of the text to be reviewed can be determined based on the correlation. The correlation and the recognition quality level are positively correlated, that is, the higher the correlation, the higher the recognition quality level.
[0051] In some embodiments of this application, the distance between the position of the text to be reviewed in the document to be reviewed and the position of the text to be reviewed in the semantic layout template can be obtained, and the recognition quality level of the text to be reviewed can be determined based on the distance. The distance and the recognition quality level are negatively correlated, that is, the greater the distance, the lower the recognition quality level.
[0052] In some embodiments of this application, the recognition quality level of the text to be reviewed can also be determined by the above-mentioned recognition confidence, relevance and distance, so as to improve the accuracy when evaluating the recognition quality level.
[0053] In step S220, if it is determined that the recognition quality level has not reached the preset quality level, the current location of the text to be reviewed in the document to be reviewed is located.
[0054] In the embodiments of this application, after evaluating the recognition quality level, if it is determined that the recognition quality level has not reached the preset quality level, it indicates that the effectiveness of the text to be reviewed extracted from the document to be reviewed through the semantic layout template is low. In order to ensure that effective text to be reviewed is obtained from the document to be reviewed, the current location of the text to be reviewed can be located in the document to be reviewed.
[0055] Considering the different priorities of various documents awaiting review in document review tasks—for example, in a task involving reviewing employment application materials, an ID card would obviously have a higher priority than a medical examination report—some embodiments of this application first determine the priority of each document in the review task, then use the corresponding quality level threshold as a preset quality level to dynamically configure the preset quality level for different document types.
[0056] In some embodiments of this application, the method for locating the current location of the text to be reviewed in the document to be reviewed can be as follows: first, obtain the OCR text corresponding to the document to be reviewed; then, determine the relevant fields corresponding to the text to be reviewed from the OCR text; and finally, determine the current location of the text to be reviewed in the document to be reviewed based on the relevant fields using a visual detection model, so as to achieve the purpose of locating the current location of the text to be reviewed.
[0057] Among them, visual detection models include, but are not limited to, Grad-CAM (Gradient-weighted ClassActivation Mapping), Score-CAM (Score-weighted ClassActivation Mapping), or Layer-CAM (Layer-wise ClassActivation Mapping).
[0058] In step S230, the text to be reviewed for the document to be reviewed is adjusted based on the current region.
[0059] In the embodiments of this application, after locating the current location of the text to be reviewed, the text to be reviewed corresponding to the document to be reviewed can be adjusted based on the current location to improve the accuracy of the document to be reviewed.
[0060] The method of adjusting the text of the document to be reviewed based on the current region can be flexibly set as needed. In one example, the text content of the relevant fields of the document to be reviewed in the current region can be used as the text of the document to be reviewed.
[0061] In another example, an extraction instruction can be generated based on the current region and a preset extraction instruction structure. Then, a preset multimodal large model can be used to extract the text content corresponding to the text to be reviewed based on the extraction instruction and the document to be reviewed, so as to further improve the accuracy of the document to be reviewed corresponding to the document to be reviewed.
[0062] See Figure 3 , Figure 3 Is Figure 2 A flowchart of another document review method provided based on this. For example... Figure 3 As shown, the method may further include steps S310 to S320, which are described in detail below: Step S310: After locating the current location of the text to be reviewed in the document to be reviewed, the historical correction count of the text to be reviewed in the document to be reviewed is incremented by one. Step S320: If it is determined that the number of historical corrections is greater than the preset number of corrections, then adjust the semantic layout template corresponding to the document to be reviewed according to the current region.
[0063] In the above process, after locating the current location of the text to be reviewed in the document to be reviewed, the historical correction count of the text to be reviewed in the document to be reviewed is incremented by one. If it is determined that the historical correction count is greater than the preset correction count, it indicates that after extracting the text to be reviewed from the semantic layout template of the document to be reviewed, it is often necessary to relocate the position of the text to be reviewed in the document to be reviewed. The semantic layout template of the document to be reviewed can be adjusted according to the current location, that is, the current location is taken as the location area of the text to be reviewed in the semantic layout template, so as to correct the location area of the text to be reviewed in the semantic layout template of the document to be reviewed, thereby facilitating the correct extraction of the text to be reviewed from the document to be reviewed through the semantic layout template of the document to be reviewed.
[0064] See Figure 4 , Figure 4 This is a flowchart illustrating a document review method according to another exemplary embodiment. For example... Figure 4 As shown, in Figure 1 Before step S110 in the illustrated embodiment, the method may further include steps S410 to S430, which are described in detail below: Step S410: Obtain the current document format corresponding to the document to be reviewed; Step S420: Determine the target preprocessing method based on the current document format; Step S430: Preprocess the document to be reviewed according to the target preprocessing method.
[0065] In the above process, before determining the semantic layout template corresponding to each of the different documents to be reviewed in the document review task, the current document format corresponding to the document to be reviewed can be obtained first. Then, the target preprocessing method can be determined according to the current document format. Then, the document to be reviewed can be preprocessed according to the target preprocessing method so that the document to be reviewed can be converted from the current document format to the review document format, which makes it easier to obtain the text to be reviewed from the document to be reviewed later.
[0066] The target preprocessing method is used to convert the current document format corresponding to the document to be reviewed into the review document format.
[0067] For example, when reviewing documents in image format, if the document to be reviewed is in Word format, the Pillow (image processing library) method can be chosen as the target preprocessing method to convert the document from Word to image format. If the document to be reviewed is in PDF format, the Pypdfium2 method can be chosen as the target preprocessing method to convert the document from PDF to image format.
[0068] In addition, in some embodiments of this application, after the document to be reviewed is preprocessed according to the target preprocessing method, a preset image processing method can be used to perform secondary processing on the document to be reviewed, such as orientation correction processing, image clarity processing, image resolution processing, image white edge cropping processing, etc., to further facilitate the subsequent extraction of the text to be reviewed from the document to be reviewed.
[0069] The following describes an apparatus embodiment of this application, which can be used to execute the document review method in the above embodiments of this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the document review method described above.
[0070] Figure 5 A block diagram of a document review apparatus 100 according to an embodiment of this application is shown.
[0071] Reference Figure 5 As shown, a document review apparatus according to an embodiment of this application includes: an acquisition module 110 configured to determine the semantic layout template corresponding to each of the different documents to be reviewed in a document review task; an extraction module 120 configured to extract the corresponding text to be reviewed for each document based on its semantic layout template; and an review module 130 configured to determine the review result of the document review task based on the text to be reviewed corresponding to the different documents to be reviewed.
[0072] In some embodiments of this application, based on the foregoing scheme, the acquisition module 110 is further configured to: determine the target document type corresponding to the document to be reviewed; and obtain the semantic layout template corresponding to the document to be reviewed from the preset template library according to the target document type.
[0073] In some embodiments of this application, based on the foregoing scheme, the extraction module 120 is further configured to: determine the target extraction region in the document to be reviewed based on the semantic layout template corresponding to the document to be reviewed; wherein, the target extraction region is the position of the text to be reviewed indicated by the semantic layout template in the document to be reviewed; extract the target text content of the document to be reviewed in the target extraction region; and use the target text content as the text to be reviewed corresponding to the document to be reviewed.
[0074] In some embodiments of this application, based on the foregoing scheme, the review module 130 is further configured to: determine a target review rule according to the review text type corresponding to the text to be reviewed; review the text to be reviewed in different documents to be reviewed based on the target review rule; if the review is passed, the review result of the document review task is determined as the first review result; wherein, the first review result indicates that the document review task has passed the review; if the review is not passed, the review result of the document review task is determined as the second review result; wherein, the second review result indicates that the document review task has not passed the review.
[0075] In some embodiments of this application, based on the foregoing scheme, the review module 130 is further configured to: before determining the review result of the document review task based on the text to be reviewed corresponding to different documents to be reviewed, evaluate the recognition quality level of the text to be reviewed corresponding to the document to be reviewed; if it is determined that the recognition quality level does not reach the preset quality level, locate the current area of the text to be reviewed in the document to be reviewed; and adjust the text to be reviewed corresponding to the document to be reviewed based on the current area.
[0076] In some embodiments of this application, based on the foregoing scheme, the review module 130 is further configured to: after locating the current location of the text to be reviewed in the document to be reviewed, increment the historical correction count of the text to be reviewed in the document to be reviewed by one; if it is determined that the historical correction count is greater than the preset correction count, adjust the semantic layout template corresponding to the document to be reviewed according to the current location.
[0077] In some embodiments of this application, based on the foregoing scheme, the acquisition module 110 is further configured to: acquire the current document format corresponding to the document to be reviewed; determine a target preprocessing method according to the current document format; wherein the target preprocessing method is used to convert the current document format into the review document format; and preprocess the document to be reviewed according to the target preprocessing method so that the document to be reviewed is converted from the current document format into the review document format.
[0078] It should be noted that the document review device 100 provided in the above embodiments and the document review method provided in the above embodiments belong to the same concept. The specific way in which each module and unit performs operations has been described in detail in the method embodiments, and will not be repeated here.
[0079] Embodiments of this application also provide an electronic device, including a processor and a memory, wherein the memory stores computer-readable instructions that, when executed by the processor, implement the document review method as described above.
[0080] Figure 6 A schematic diagram of the structure of a computer system suitable for implementing the electronic device of the present application is shown.
[0081] It should be noted that, Figure 6 The computer system 200 of the electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.
[0082] like Figure 6 As shown, the computer system 200 includes a Central Processing Unit (CPU) 201, which can perform various appropriate actions and processes based on programs stored in Read-Only Memory (ROM) 202 or programs loaded from storage portion 208 into Random Access Memory (RAM) 203, such as performing the methods described in the above embodiments. The RAM 203 also stores various programs and data required for system operation. The CPU 201, ROM 202, and RAM 203 are interconnected via a bus 204. An Input / Output (I / O) interface 205 is also connected to the bus 204.
[0083] The following components are connected to I / O interface 205: an input section 206 including a keyboard, mouse, etc.; an output section 207 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 208 including a hard disk, etc.; and a communication section 209 including a network interface card such as a LAN (Local Area Network) card, modem, etc. The communication section 209 performs communication processing via a network such as the Internet. A drive 210 is also connected to I / O interface 205 as needed. Removable media 211, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 210 as needed so that computer programs read from them can be installed into storage section 208 as needed.
[0084] Specifically, according to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program including a computer program for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 209, and / or installed from removable medium 211. When the computer program is executed by central processing unit (CPU) 201, it performs various functions defined in the system of this application.
[0085] It should be noted that the computer-readable medium shown in the embodiments of this application can be a computer-readable signal medium, a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying a computer-readable computer program. The transmitted data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The computer program contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to wireless, wired, etc., or any suitable combination thereof.
[0086] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. Each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.
[0087] The units described in the embodiments of this application can be implemented in software or hardware, and the described units can also be located in a processor. The names of these units do not necessarily limit the specific unit itself.
[0088] In another aspect, this application also provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The computer-readable storage medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to perform the methods described in the above embodiments.
[0089] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to the embodiments of this application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.
[0090] Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, touch terminal, or network device, etc.) to execute the methods according to the embodiments of this application.
[0091] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the embodiments disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.
[0092] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.
Claims
1. A document review method, characterized in that, The method includes: Determine the semantic layout templates corresponding to different documents to be reviewed in the document review task; Extract the corresponding text to be reviewed for each document based on its respective semantic layout template. The review result of the document review task is determined based on the text to be reviewed corresponding to the different documents to be reviewed.
2. The method according to claim 1, characterized in that, The process of determining the semantic layout templates corresponding to different documents to be reviewed in the document review task includes: Determine the target document type corresponding to the document to be reviewed; The semantic layout template corresponding to the document to be reviewed is obtained from the preset template library according to the target document type.
3. The method according to claim 1, characterized in that, The step of extracting the corresponding text to be reviewed for each document based on its respective semantic layout template includes: The target extraction region in the document to be reviewed is determined based on the semantic layout template corresponding to the document to be reviewed; wherein, the target extraction region is the position of the text to be reviewed in the document to be reviewed as indicated by the semantic layout template; Extract the target text content of the document to be reviewed from the target extraction area; The target text content is used as the text to be reviewed corresponding to the document to be reviewed.
4. The method according to claim 1, characterized in that, The step of determining the review result of the document review task based on the text to be reviewed corresponding to different documents to be reviewed includes: The target review rule is determined based on the type of review text corresponding to the text to be reviewed; The text to be reviewed in different documents to be reviewed is reviewed based on the target review rules. If the review is successful, the review result of the document review task is determined as the first review result; wherein, the first review result indicates that the document review task has passed the review. If the review fails, the review result of the document review task is determined as the second review result; wherein, the second review result indicates that the document review task has failed the review.
5. The method according to claim 1, characterized in that, Before determining the review result of the document review task based on the text to be reviewed corresponding to different documents to be reviewed, the following steps are included: Assess the recognition quality level of the text corresponding to the document to be reviewed; If it is determined that the recognition quality level has not reached the set quality level, then the current location of the text to be reviewed in the document to be reviewed is located. Adjust the text of the document to be reviewed based on the current location.
6. The method according to claim 5, characterized in that, The method further includes: After locating the current location of the text to be reviewed in the document to be reviewed, the historical revision count of the text to be reviewed in the document to be reviewed is incremented by one; If it is determined that the number of historical corrections is greater than the preset number of corrections, then the semantic layout template corresponding to the document to be reviewed is adjusted according to the current region.
7. The method according to any one of claims 1, characterized in that, Before determining the semantic layout templates corresponding to different documents to be reviewed in the document review task, the following steps are included: Obtain the current document format corresponding to the document to be reviewed; A target preprocessing method is determined based on the current document format; wherein, the target preprocessing method is used to convert the current document format into an audit document format; The document to be reviewed is preprocessed according to the target preprocessing method to convert the document to be reviewed from the current document format to the review document format.
8. A document review device, characterized in that, include: The acquisition module is configured to determine the semantic layout template corresponding to each different document to be reviewed in the document review task. The extraction module is configured to extract the corresponding text to be reviewed for each document based on its respective semantic layout template. The review module is configured to determine the review result of the document review task based on the text to be reviewed corresponding to different documents to be reviewed.
9. A computer-readable storage medium, characterized in that, It stores computer-readable instructions that, when executed by a computer's processor, cause the computer to perform the document review method according to any one of claims 1 to 7.
10. An electronic device, characterized in that, include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the electronic device to implement the document review method as described in any one of claims 1 to 7.