A digital intelligent proofreading method and system based on text analysis

Abstract

This invention discloses a digital intelligent proofreading method based on text analysis, comprising: acquiring the identity information of the inspectors; identifying the identity information of the inspectors using entity recognition and keyword matching to obtain a data document of the corresponding relationships between the inspectors; correcting the data document and converting the data to obtain data text and backing it up, recording the archiving time of the backup data text; performing successive proofreading of the data text using a multi-text model, outputting abnormal text, and performing a second proofreading on the abnormal text; reviewing the abnormal text, recording entries that meet the review criteria in a dedicated appendix and sending them back to the multi-text model, exporting the abnormal text, and obtaining the final proofreading data, ensuring the accuracy of the text information proofreading results and supporting timely data updates.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a digital intelligent proofreading method and system based on text analysis. Background Technology

[0002] With the advent of the digital age, text processing and natural language processing technologies have developed rapidly. Manual document proofreading is prone to errors. With the widespread use of various office tools, people are using software for document proofreading, which has greatly improved the speed and accuracy of document proofreading.

[0003] By using document proofreading software, people can more quickly find and correct errors in documents, improving the efficiency and accuracy of proofreading. These software programs typically employ advanced text detection and correction algorithms, combined with large-scale lexicons and language models, to identify and correct various types of errors. Different software uses different text detection models and lexicons, leading to inconsistent results when proofreading documents. Since each model and lexicon has its specific advantages and limitations, relying solely on one text detection model for proofreading can easily result in false positives; some errors might be detected by one model but missed by another. The accuracy and timeliness of the lexicon are crucial in text proofreading; however, if the lexicon is not updated in a timely manner, it can lead to anomalies during proofreading. Due to the limitations of text detection models and lexicons, some anomalies may occur during proofreading, including incorrectly marking correct new words as errors or misjudging certain language structures. These anomalies may reduce proofreading accuracy and increase the need for manual intervention. Proofreading methods based on text detection models have certain dependencies and limitations; the system's performance and accuracy depend on the quality of the model and lexicon used. If the model or lexicon is of poor quality, the proofreading results may be affected. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is that: the present invention solves the problem that simply using a text detection model for proofreading is prone to false detections and lacks timely information updates, which leads to proofreading anomalies.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a digital intelligent proofreading method based on text analysis, comprising:

[0008] The identity information of the testing personnel is obtained, and entity recognition and keyword matching are used to identify the identity information of the testing personnel to obtain a data document of the corresponding relationship between the testing personnel;

[0009] The data document is corrected and transformed to obtain data text, which is then backed up. The archiving time of the backup data text is recorded.

[0010] The data text is checked successively using a multi-text model, and abnormal text is output. The abnormal text is then checked a second time.

[0011] The abnormal text is reviewed, and entries that meet the review criteria are recorded in a dedicated appendix and sent back to the multi-text model. The abnormal text is then exported to obtain the final proofreading data.

[0012] As a preferred embodiment of the text-based digital intelligent proofreading method described in this invention, the step of obtaining the identity information of the inspectors, identifying the identity information of the inspectors using entity recognition and keyword matching, and obtaining a data document of corresponding relationships among the inspectors includes:

[0013] By selecting a deep learning model to identify the identity information of the detection personnel, a list of keywords related to the identity information of the detection personnel is constructed.

[0014] The identity information of the person identified by the deep learning model is matched with the keyword list. If the identification result is consistent with the keyword matching result, the identity information of the detected person can be obtained.

[0015] As a preferred embodiment of the text-based digital intelligent proofreading method of the present invention, the method includes: correcting the data document, converting the data to obtain data text and backing it up, and recording the archiving time of the backed-up data text, including...

[0016] For paper documents, an edge detection algorithm is used to find the position of the edge by calculating the gradient magnitude and direction of each pixel in the image. After calculating the gradient, non-maximum suppression is used to refine the edge, dividing the edge pixels into strong edges and weak edges. The strong edges and weak edges are connected to form complete edge lines. Perspective transformation is applied to correct the document tilt, making the document horizontal or vertical and achieving the standard format.

[0017] For electronic documents, adjust the document style and format according to standard format requirements, and convert standard format text into PDF format data documents.

[0018] As a preferred embodiment of the text-based digital intelligent proofreading method described in this invention, the method includes: performing data transformation to obtain data text and backing it up, and recording the archiving time of the backed-up data text, including...

[0019] Export the data document as a data text file, back up the data text file, and record the archiving time of the backed-up data text file;

[0020] Set the save time to the first save time. When the backup data reaches the first save time, delete the corresponding backup data text.

[0021] As a preferred embodiment of the text analysis-based digital intelligent proofreading method of the present invention, the method includes: performing successive proofreading of the data text using a multi-text model, outputting abnormal text, and performing a second proofreading on the abnormal text, including...

[0022] By constructing a text processing pipeline, a data structure is established where both the input and output of the text processing pipeline are text files, and then the two are connected.

[0023] Several text detection models are connected in sequence to form a processing flow, and the output of each model will be used as the input of the next model.

[0024] The data text is transmitted to the first-ranked text detection model, and after the abnormal text feedback is output, the abnormal text is transmitted to the adjacent subsequent text detection models until abnormal text is output.

[0025] As a preferred embodiment of the text-based digital intelligent proofreading method described in this invention, the abnormal text is reviewed; entries that meet the review criteria are recorded in a dedicated appendix and sent back to the multi-text model; the abnormal text is exported to obtain the final proofreading data, including...

[0026] The abnormal texts are categorized and prioritized according to their type and level. After review, the entries in the abnormal texts that meet the review criteria are recorded in a dedicated supplementary section, and the entries are transmitted to the supplementary vocabulary of the text detection model.

[0027] As a preferred embodiment of the text-based intelligent proofreading method of the present invention, wherein: the abnormal text is exported to obtain the final proofreading data, including,

[0028] Set a second proofreading time limit. When the abnormal text reaches the second proofreading time limit, directly export the abnormal text as the final proofreading data.

[0029] Secondly, the present invention provides a system for a digital intelligent proofreading method based on text analysis, comprising:

[0030] The text input module is used to obtain the identity information of the testing personnel, and to identify the identity information of the testing personnel by using entity recognition and keyword matching to obtain a data document of the corresponding relationship between the testing personnel.

[0031] The record verification module is used to correct the data document, convert the data, obtain the data text and back it up, and record the archiving time of the backup data text.

[0032] The data verification module is used to perform multi-text model successive verification of the data text, output abnormal text, and perform secondary verification on the abnormal text.

[0033] The personalized analysis module is used to review the abnormal text, record the entries that meet the review criteria in a dedicated appendix and send them back to the multi-text model, and export the abnormal text to obtain the final proofreading data.

[0034] Thirdly, the present invention provides a computing device, comprising:

[0035] Memory and processor;

[0036] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the text analysis-based digital intelligent proofreading method.

[0037] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the text analysis-based digital intelligent proofreading method.

[0038] The beneficial effects of this invention are as follows: By identifying the information of the person responsible for text verification during text input, this invention determines the information of the person responsible for text verification. After sorting several text recognition models and exporting the abnormal identification results, a second verification is performed, reducing the probability of false detections and achieving the purpose of intelligent text verification. Furthermore, the invention allows for secondary manual verification by the person responsible for text verification, and the correct verification results are imported into a dictionary as the exclusive dictionary for the corresponding person responsible for text verification. This provides several text recognition models with data text verification results, ensuring the accuracy of the text information verification results. The preset archiving time avoids the situation where the initial state data text is stored for a long time and occupies data storage space, providing support for timely data updates. Attached Figure Description

[0039] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0040] Figure 1 The overall flowchart of a digital intelligent proofreading method based on text analysis provided by the present invention;

[0041] Figure 2 A system principle block diagram of a digital intelligent proofreading method based on text analysis provided by the present invention;

[0042] Figure 3 The system principle block diagram of the text input module in the digital intelligent proofreading method based on text analysis provided by the present invention;

[0043] Figure 4 The system principle block diagram of the recording verification module in the digital intelligent verification method based on text analysis provided by the present invention;

[0044] Figure 5 A system principle block diagram of the data verification module in a text-based intelligent digital verification method provided by the present invention;

[0045] Figure 6 This invention provides a system principle block diagram of the personalized analysis module in a text-based digital intelligent proofreading method. Detailed Implementation

[0046] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0047] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0048] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0049] This invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of this invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not adhering to the usual scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of this invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.

[0050] Furthermore, in the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used solely for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In addition, the terms "first," "second," or "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0051] Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0052] Example 1

[0053] Reference Figure 1 As an embodiment of the present invention, a digital intelligent proofreading method based on text analysis is provided, comprising:

[0054] S1: Obtain the identity information of the testing personnel, and use entity recognition and keyword matching to identify the identity information of the testing personnel to obtain a data document of the corresponding relationship between the testing personnel;

[0055] Furthermore, by selecting a deep learning model to identify the identity information of the detection personnel, a list of keywords related to the identity information of the detection personnel is constructed;

[0056] The identity information of the person identified by the deep learning model is matched with the keyword list. If the identification result is consistent with the keyword matching result, the identity information of the detected person is obtained.

[0057] It should be noted that a convolutional neural network (CNN) is chosen from deep learning models. The CNN is trained using labeled data, with preprocessed text data as input and entity recognition results as the target. Backpropagation and an optimizer are used to update the model's parameters to minimize entity recognition errors. The CNN model receives text or image data and outputs corresponding detection personnel identity information. A keyword list related to the detection personnel identity information is constructed, including common names, organization names, and job titles. For each identified entity, it is matched against the keyword list. If an entity matches any keyword in the list, the detection personnel identity information associated with that entity can be identified as a matching keyword. Entity recognition automatically extracts entity information from the text, and keyword matching is used to verify and supplement the entity recognition results. By combining these two methods, a balance can be achieved between the accuracy of entity recognition and the flexibility of keyword matching, improving the recognition effect of the detection personnel identity information in the text.

[0058] It should also be noted that, such as Figure 2 As shown, a digital intelligent proofreading system based on text analysis includes a text input module 100, a record verification module 200, a data proofreading module 300, and a personalized analysis module 400. The record verification module 200 is connected to the data proofreading module 300, and the data proofreading module 300 is connected to the personalized analysis module 400.

[0059] like Figure 3 As shown, the text input module 100 includes an information recognition module 101, a document correction module 102, and a PDF conversion module. The information recognition module 101 is connected to both the document correction module 102 and the PDF conversion module 103, and the document correction module 102 is connected to the PDF conversion module 103. The information recognition module 101 is used to identify the identity information of the person to whom the text belongs and to establish the corresponding relationship between the proofreader and the person to whom the text belongs.

[0060] S2: Correct the data document and convert the data to obtain data text and back it up, and record the archiving time of the backup data text;

[0061] Furthermore, for paper documents, edge detection algorithms are used to find the location of the edge by calculating the gradient magnitude and direction of each pixel in the image. After calculating the gradient, non-maximum suppression is used to refine the edge, dividing the edge pixels into strong edges and weak edges. The strong edges and weak edges are connected to form complete edge lines. Perspective transformation is applied to correct the document tilt, making the document horizontal or vertical, thus achieving the standard format.

[0062] For electronic documents, adjust the document style and format according to standard format requirements, and convert standard format text into PDF format data documents.

[0063] It should be noted that for paper documents, preprocessing the paper image improves the edge detection effect. This includes grayscale conversion, noise reduction, and smoothing. The color image is converted to a grayscale image, noise in the image is removed using a filter, and the image is smoothed. The image gradient of the Canny algorithm is used to detect edges. The purpose of using non-maximum suppression is to make the edge lines thinner, retaining only the pixels with the largest local gradient. For strong edge pixels, they are directly marked as edge pixels. For weak edge pixels, if there is at least one strong edge pixel around them, they are marked as edge pixels. In this way, strong edges and weak edges can be connected to form complete edge lines.

[0064] For electronic modules, the PDF conversion module 103 can be used directly to convert standard format text into PDF format data documents.

[0065] It should also be noted that the document correction module 102 is used to correct paper documents and obtain text in standard format; the PDF conversion module 103 is used to convert the text in standard format into a data document in PDF format and transmit the data document to the record verification module 200.

[0066] Furthermore, the data document is exported as a data text file, the data text is backed up, and the archiving time of the backed-up data text is recorded;

[0067] Set the save time to the first save time. When the backup data reaches the first save time, delete the corresponding backup data text.

[0068] It should be noted that the data document should be converted to an appropriate format (such as CSV, JSON, etc.) and saved as a text file, ensuring that the text file contains all necessary data. The exported data text should be backed up by copying it to a backup directory or storage device, ensuring that the backup data text is completely consistent with the original data document and has the same filename. A log file or database should be created to record the archive time of each backup data text, using a specific date and time format for later querying and management. The first archive time is the time of the first backup, which serves as the starting point for the archive time. Once the first archive time is reached, the corresponding backup data text should be deleted to save storage space.

[0069] It should also be noted that, such as Figure 4As shown, the record verification module 200 includes a data conversion module 201, a data backup module 202, and a backup processing module 203. The data conversion module 201 is connected to the data backup module 202, and the data backup module 202 is connected to the backup processing module 203.

[0070] The data conversion module 201 is used to receive data documents and convert them into data text, and then transmit the data text to the data verification module 300; the data backup module 202 is used to back up the data text; the backup processing module 203 is used to record the archiving time of the backed-up data text, and delete the corresponding backup data text after the first archiving time is reached.

[0071] S3: Perform multi-text model successive verification on the data text, output abnormal text, and perform a second verification on the abnormal text;

[0072] Furthermore, by constructing a text processing pipeline, the data structure that determines the input and output of the text processing pipeline to be text files is connected.

[0073] Several text detection models are connected in sequence to form a processing flow, and the output of each model will be used as the input of the next model.

[0074] The data text is transmitted to the first-ranked text detection model, and after the abnormal text feedback is output, the abnormal text is transmitted to the adjacent subsequent text detection models until abnormal text is output.

[0075] It should be noted that, as Figure 5 As shown, the data verification module 300 includes a verification link module 301 and a data verification module 302, with the data verification module 302 connected to the verification link module 301.

[0076] The proofreading link module 301 is used to connect with several text detection models and arrange the several text detection models in order; the data proofreading module 302 is used to receive data text and transmit the data text to the first text detection model in the order through the proofreading link module 301. After outputting abnormal text feedback, the segment to which the abnormal text belongs is transmitted to the adjacent subsequent text detection models until the proofreading abnormal text is output. Then the proofreading abnormal text is transmitted to the personalized analysis module 400.

[0077] It should also be noted that the text detection model includes various text classification, named entity recognition, sentiment analysis, keyword extraction, etc., ensuring that the text detection model can provide the required detection and analysis functions. The proofreading link module 301 connects with several text detection models, including the EAST model, CTPN model, TextBoxes model, SSD model, and YOLO model, and arranges these text detection models in a sequential order. The order can be random or customized. The data proofreading module 302 receives the data text and transmits it through the proofreading link module 301 to the text detection model at the top of the sorted list. After outputting abnormal text feedback, the segment to which the abnormal text belongs is transmitted to the adjacent subsequent text detection models until the proofreading abnormal text is output. Finally, the proofreading abnormal text is transmitted to the personalized analysis module 400. By sorting several text recognition models and exporting the abnormal recognition results, a second recognition is performed to reduce the probability of false detections and achieve the goal of intelligent text proofreading. Furthermore, the proofreader performs a second manual proofreading and imports the correct proofreading results into a dictionary, which serves as the exclusive dictionary for the corresponding proofreader. This dictionary is then used by several text recognition models to perform data text proofreading, ensuring the accuracy of the text information proofreading results.

[0078] S4: Review the abnormal text, record the entries that meet the review criteria in the dedicated appendix and send them back to the multi-text model, export the abnormal text, and obtain the final proofreading data;

[0079] Furthermore, the abnormal texts are categorized and prioritized according to their type and level. After review, entries that meet the review criteria in the abnormal texts are recorded in a dedicated supplementary section, and the entries are transmitted to the supplementary vocabulary of the text detection model.

[0080] Set a second proofreading time limit. When the abnormal text reaches the second proofreading time limit, directly export the abnormal text as the final proofreading data.

[0081] It should be noted that, as Figure 6 As shown, the personalized analysis module 400 includes a dedicated supplementary module 401, a secondary verification module 402, and a data export module 403. The dedicated supplementary module 401 is connected to the secondary verification module 402, and the secondary verification module 402 is connected to the data export module 403.

[0082] The dedicated supplementary module 401 is used to record entries that meet the proofreader's review standards and transmits these entries to the data proofreading module 300, serving as an supplementary vocabulary for several text detection models connected by the proofreading link module 301. The secondary proofreading module 402 receives proofreading error text, which is then manually reviewed by the proofreader, and records the entries in the error text that meet the proofreader's review standards. The data export module 403 is used to preset the secondary proofreading time, and after the time is reached, directly exports the proofreading error text as the final proofreading data.

[0083] It should also be noted that the secondary proofreading module 402 receives proofreading error text, which is then manually reviewed by the proofreader. The entries in the proofreading error text that meet the proofreader's review standards are recorded in the dedicated supplementary module 401, and the entries are transmitted to the data proofreading module 300. The supplementary vocabulary of several text detection models provides an accuracy guarantee for subsequent text proofreading. In order to reduce the waiting time for the output of proofreading error text, the data export module 403 presets a secondary proofreading time of 24 hours. After the secondary proofreading time is reached, the proofreading error text is directly exported as the final proofreading data.

[0084] The above is an illustrative scheme of a text-based intelligent digital proofreading method according to this embodiment. It should be noted that the technical solution of the device for this text-based intelligent digital proofreading method belongs to the same concept as the technical solution of the text-based intelligent digital proofreading method described above. Details not described in detail in the technical solution of the device for this text-based intelligent digital proofreading method in this embodiment can be found in the description of the technical solution of the text-based intelligent digital proofreading method described above.

[0085] The apparatus for the digital intelligent proofreading method based on text analysis in this embodiment includes:

[0086] The text input module is used to obtain the identity information of the testing personnel, and to identify the identity information of the testing personnel by using entity recognition and keyword matching to obtain a data document of the corresponding relationship between the testing personnel.

[0087] The record verification module is used to correct the data document, convert the data, obtain the data text and back it up, and record the archiving time of the backup data text.

[0088] The data verification module is used to perform multi-text model successive verification of the data text, output abnormal text, and perform secondary verification on the abnormal text.

[0089] The personalized analysis module is used to review the abnormal text, record the entries that meet the review criteria in a dedicated appendix and send them back to the multi-text model, and export the abnormal text to obtain the final proofreading data.

[0090] This embodiment also provides a computing device suitable for text-based digital intelligent proofreading methods, including:

[0091] The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement the text analysis-based digital intelligent proofreading method as described in the above embodiments.

[0092] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the text analysis-based digital intelligent proofreading method as proposed in the above embodiments.

[0093] The storage medium proposed in this embodiment and the digital intelligent proofreading method based on text analysis proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0094] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0095] Example 2

[0096] This embodiment differs from the first embodiment in that it provides a verification test of a digital intelligent proofreading method based on text analysis, to verify and illustrate the technical effectiveness of the method.

[0097] A publishing house is conducting data text proofreading. Assume that the information of the text files to be compared by the publishing house corresponds to the information of the personnel being checked, as shown in Table 1.

[0098] Table 1 shows the correspondence between the text files and the testing personnel.

[0099] 1 Here are some incorrect sentences. Zhang 2 Ensure the file is correct. Li 3 The first paragraph needs to be rephrased. Wang 4 digital intelligent proofreading methods Liu

[0100] A deep learning model is selected to identify the corresponding information between text files and detection personnel. A keyword list is constructed, which includes the text fields to be compared and the corresponding detection personnel. The personnel identity information identified by the deep learning model is matched with the keyword list. If the identification result is consistent with the keyword matching result, the information of the detection personnel corresponding to the text file is obtained.

[0101] Correct the data document to ensure correct formatting, remove unnecessary spaces, punctuation marks and other special characters, back up the corrected data text, and record the archiving time of the backup data text to help track the data history and version control;

[0102] The data text is input into a multi-text model for successive proofreading, and anomalous text is identified from the model's output. For anomalous text, a second proofreading is performed to ensure accuracy and completeness. The anomalous text that has undergone the second proofreading is then reviewed; entries that meet the review criteria are recorded in a dedicated appendix and returned to the multi-text model for further processing.

[0103] The proofread and reviewed text is merged with the original data text to generate the final proofread data. The final proofread data is then exported for subsequent publishing and printing processes.

[0104] Table 2 shows the correspondence between the optimized comparison text files and the testing personnel.

[0105]

[0106] The "Abnormal Text" column displays the corrected text for text that has passed proofreading, and the corrected text for text that has failed proofreading.

[0107] The review criteria column indicates whether the problematic text meets the review criteria. For text that passes proofreading, the review criteria column displays "Proofreading Passed"; for text that fails proofreading, the review criteria column displays the corresponding review criteria.

[0108] By using multi-text models and natural language processing techniques, text analysis-based proofreading methods can capture and correct spelling errors, grammatical errors, punctuation errors, and other problems in text, thereby improving proofreading accuracy. Digital intelligent proofreading methods can automate proofreading tasks, greatly reducing the workload and time of manual proofreading. Multi-text models can process multiple data points simultaneously, enabling batch proofreading and thus improving proofreading efficiency.

[0109] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A digital intelligent proofreading method based on text analysis, characterized in that, include: The identity information of the testing personnel is obtained, and entity recognition and keyword matching are used to identify the identity information of the testing personnel to obtain a data document of the corresponding relationship between the testing personnel; The data document is corrected and transformed to obtain data text, which is then backed up. The archiving time of the backup data text is recorded. The data text is checked successively using a multi-text model, and abnormal text is output. The abnormal text is then checked a second time. The abnormal text is reviewed, and entries that meet the review criteria are recorded in a dedicated appendix and sent back to the multi-text model. The abnormal text is then exported to obtain the final proofreading data. The process involves acquiring the identity information of the testing personnel, using entity recognition and keyword matching to identify the personnel's identity information, and obtaining a data document detailing the corresponding relationships between the testing personnel. By selecting a deep learning model to identify the identity information of the detection personnel, a list of keywords related to the identity information of the detection personnel is constructed. The identity information of the person identified by the deep learning model is matched with the keyword list. If the identification result is consistent with the keyword matching result, the identity information of the detected person is obtained. The data document is corrected and transformed to obtain data text, which is then backed up. The archiving time of the backed-up data text is recorded, including... For paper documents, an edge detection algorithm is used to find the position of the edge by calculating the gradient magnitude and direction of each pixel in the image. After calculating the gradient, non-maximum suppression is used to refine the edge, dividing the edge pixels into strong edges and weak edges. The strong edges and weak edges are connected to form complete edge lines. Perspective transformation is applied to correct the document tilt, making the document horizontal or vertical and achieving the standard format. For electronic documents, adjust the document style and format according to standard format requirements, and convert standard format text into PDF format data documents; The data text is successively proofread using a multi-text model, and abnormal text is output. The abnormal text is then subjected to a second proofreading, including... By constructing a text processing pipeline, a data structure is established where both the input and output of the text processing pipeline are text files, and then the two are connected. Several text detection models are connected in sequence to form a processing flow, and the output of each model will be used as the input of the next model. The data text is transmitted to the first text detection model in the ranking. After the abnormal text feedback is output, the abnormal text is transmitted to the adjacent subsequent text detection models until the abnormal text is output. The abnormal text is reviewed, and entries that meet the review criteria are recorded in a dedicated appendix and sent back to the multi-text model. The abnormal text is then exported to obtain the final proofreading data, including... The abnormal texts are classified and prioritized according to their type and level. After review, the words in the abnormal texts that meet the review criteria are recorded in a dedicated supplementary section and transmitted to the supplementary vocabulary of the text detection model.

2. The digital intelligent proofreading method based on text analysis as described in claim 1, characterized in that, Perform data transformation to obtain data text and back it up, recording the archiving time of the backup data text, including... Export the data document as a data text file, back up the data text file, and record the archiving time of the backed-up data text file; Set the save time to the first save time. When the backup data reaches the first save time, delete the corresponding backup data text.

3. The digital intelligent proofreading method based on text analysis as described in claim 1, characterized in that, Export the abnormal text to obtain the final proofreading data, including: Set a second proofreading time limit. When the abnormal text reaches the second proofreading time limit, directly export the abnormal text as the final proofreading data.

4. A system for a text-based intelligent digital proofreading method, employing the text-based intelligent digital proofreading method as described in any one of claims 1-3, characterized in that, include: The text input module is used to obtain the identity information of the testing personnel, and to identify the identity information of the testing personnel by using entity recognition and keyword matching to obtain a data document of the corresponding relationship between the testing personnel. The verification module is used to correct the data document, convert the data into text, back it up, and record the archiving time of the backup data text. The data verification module is used to perform multi-text model successive verification of the data text, output abnormal text, and perform secondary verification on the abnormal text. The personalized analysis module is used to review the abnormal text, record the entries that meet the review criteria in a dedicated appendix and send them back to the multi-text model, and export the abnormal text to obtain the final proofreading data.

5. An electronic device, characterized in that, The device includes: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method described in any one of claims 1 to 3.

6. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 3.

Images