Vehicle repair file processing method and related equipment

Abstract

The invention provides a vehicle repair file processing method and related equipment, and the method comprises the steps: receiving a vehicle repair file uploaded by a user, and recognizing the format of the vehicle repair file; converting the automobile repair file into a Markdown format by using a conversion method matched with the format of the automobile repair file to obtain a target automobile repair file; inputting the target automobile repair file into a preset language large model to obtain file content extracted by the language large model; and exporting the file content according to a preset business data format. In the scheme, the automobile repair file format is recognized and uniformly converted into the Markdown format, the problem of complex format analysis of multi-column typesetting, table nesting and image-text mixing is solved, the content is standardized, the standardized file is input into a large language model, the natural language understanding ability of the standardized file is utilized, professional terms are precisely recognized, and the structured content is extracted; the extracted content is exported according to the preset service format, so that unified management of the automobile repair file content is realized, and the data utilization rate is improved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a method and related equipment for processing automotive repair documents. Background Technology

[0002] With the continuous enrichment of automotive repair data, the demand for efficient knowledge management and intelligent diagnostics in automotive repair and after-sales service is becoming increasingly urgent. Automotive diagnostic equipment manufacturers and repair shops have accumulated massive amounts of unstructured data resources through long-term practice, and these data resources generally exist in the form of automotive repair documents.

[0003] Due to the specialized nature and data complexity in the automotive repair field, existing document processing methods are not suitable for handling automotive repair documents. For example, they have low accuracy in recognizing technical terms (such as "turbocharger seal failure"); they have difficulty parsing complex formats in PDF documents, such as multi-column layouts, nested tables, and mixed text and images; and they cannot perform semantic annotation and unified management of images in automotive repair documents.

[0004] In summary, how to provide a document processing method suitable for automotive repair documents, thereby improving the utilization rate of automotive repair data, is an urgent problem to be solved. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a method and related equipment for processing automotive repair documents, so as to provide a document processing method suitable for automotive repair documents, thereby improving the utilization rate of automotive repair data.

[0006] To achieve the above objectives, the embodiments of the present invention provide the following technical solutions:

[0007] The first aspect of this invention discloses a method for processing automotive repair documents, the method comprising:

[0008] Receive auto repair files uploaded by users and identify the format of the auto repair files;

[0009] Using a conversion method that matches the format of the auto repair file, the auto repair file is converted into Markdown format to obtain the target auto repair file;

[0010] The target auto repair file is input into a preset language model to obtain the file content extracted by the language model.

[0011] Export the file content according to the preset business data format.

[0012] Optionally, the method of converting the auto repair file to Markdown format using a format matching method to obtain the target auto repair file includes:

[0013] If the auto repair file is in audio format, the audio file is converted into audio content using a preset audio model; a target auto repair file in Markdown format is then generated based on the audio content.

[0014] If the auto repair document is a composite document format containing text and / or images, then the text content in the auto repair document is extracted, and the images in the auto repair document are semantically annotated to obtain image description information; the images in the auto repair document are stored in a cloud server to obtain the image storage location; based on the image description information and the image storage location, an image text link is generated; based on the text content and the image text link, a target auto repair document in Markdown format is generated.

[0015] Optionally, if the auto repair document is in PDF format within the composite document format, then the step of extracting the text content from the auto repair document and semantically annotating the images in the auto repair document to obtain image description information includes:

[0016] The auto repair file is processed into a JSON structured file using the preset MinerU file processing component;

[0017] Extract text content and images from the JSON structured file;

[0018] The image is semantically annotated using a pre-defined visual model to obtain image description information.

[0019] Optionally, before converting the auto repair file to Markdown format using a conversion method that matches the format of the auto repair file to obtain the target auto repair file, the method further includes:

[0020] If the auto repair file is in video format, then the preset video-to-audio component is used to convert the auto repair file into audio format.

[0021] Optionally, the step of inputting the target auto repair file into a preset language model to obtain the file content extracted by the language model includes:

[0022] Preset prompts are displayed on the front-end page;

[0023] If the user modifies the prompt word, the modified prompt word and the target auto repair file are input into the preset language model to obtain the file content extracted by the language model.

[0024] If the user does not modify the prompt word, the original prompt word and the target auto repair file are input into the preset language model to obtain the file content extracted by the language model.

[0025] Optionally, the step of inputting the target auto repair file into a preset language model to obtain the file content extracted by the language model includes:

[0026] The preset summary generation prompts and the target auto repair file are input into the preset language big model to obtain the file summary output by the language big model;

[0027] The preset question-and-answer pairs are generated into prompt words and the target auto repair file, which are then input into the language model to obtain multiple question-and-answer pairs output by the language model.

[0028] The preset key information extraction prompts and the target auto repair file are input into the language model to obtain the key information output by the language model.

[0029] Optionally, the method further includes:

[0030] The file content extracted from the large language model is displayed on the front-end page;

[0031] In response to the user's modification operation on the file content, the file content is modified to obtain the modified file content;

[0032] Export the modified file content according to the preset business data format.

[0033] A second aspect of this invention discloses an automotive repair document processing device, the device comprising:

[0034] The receiving unit is used to receive auto repair files uploaded by users and to identify the format of the auto repair files;

[0035] The processing unit is used to convert the auto repair file into Markdown format using a conversion method that matches the format of the auto repair file, thereby obtaining the target auto repair file;

[0036] The extraction unit is used to input the target auto repair file into a preset language model to obtain the file content extracted by the language model.

[0037] The export unit is used to export the file content according to a preset business data format.

[0038] A third aspect of this invention discloses an electronic device, comprising:

[0039] Memory, used to store computer programs;

[0040] A processor is configured to execute the computer program to implement any of the automotive repair document processing methods disclosed in the first aspect of the present invention.

[0041] A fourth aspect of the present invention discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements any of the automotive repair document processing methods disclosed in the first aspect of the present invention.

[0042] Based on the above embodiments of the present invention, a method and related equipment for processing automotive repair documents are provided. The method includes: receiving automotive repair documents uploaded by users and identifying the format of the automotive repair documents; converting the automotive repair documents into Markdown format using a conversion method that matches the format of the automotive repair documents to obtain a target automotive repair document; inputting the target automotive repair document into a preset language model to obtain the file content extracted by the language model; and exporting the file content according to a preset business data format. In this solution, by identifying the format of automotive repair documents and uniformly converting them into Markdown format, the problem of parsing complex formats such as multi-column layout, table nesting, and mixed text and graphics is solved, thus standardizing the content. The standardized document is then input into a language model, which utilizes its natural language understanding capabilities to accurately identify professional terms and extract structured content. The extracted content is then exported according to a preset business format, achieving unified management of automotive repair document content and improving data utilization. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0044] Figure 1 This is an architecture diagram of an automotive repair document processing system disclosed in an embodiment of the present invention;

[0045] Figure 2 This is an interface diagram of a file upload module disclosed in an embodiment of the present invention;

[0046] Figure 3 This is a flowchart of an automotive repair document processing method disclosed in an embodiment of the present invention;

[0047] Figure 4 This is a schematic diagram illustrating the conversion of PDF format auto repair documents as disclosed in an embodiment of the present invention;

[0048] Figure 5 This is an example diagram of a summary generation prompt and document summary disclosed in an embodiment of the present invention;

[0049] Figure 6This is an example diagram of a question-and-answer pair generation prompt and question-and-answer pair disclosed in an embodiment of the present invention;

[0050] Figure 7 This is a schematic diagram illustrating key information extraction according to an embodiment of the present invention;

[0051] Figure 8 This is a structural diagram of an automotive repair document processing device disclosed in an embodiment of the present invention;

[0052] Figure 9 This is a structural diagram of an electronic device disclosed in an embodiment of the present invention. Detailed Implementation

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] In this application, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0055] As the background technology indicates, due to the specialized nature and data complexity of the automotive repair field, existing document processing methods are unsuitable for handling automotive repair documents. For example, they have low accuracy in recognizing technical terms (such as "turbocharger seal failure"); difficulty in parsing complex formats in PDF documents, including multi-column layouts, nested tables, and mixed text and images; and an inability to semantically annotate and uniformly manage images in automotive repair documents. A detailed analysis follows:

[0056] 1. Inefficient parsing and extraction of heterogeneous documents from multiple sources: Automotive repair data comes from diverse sources, including large-scale fault code files, PDF documents, and other formats. Its content contains structured fields, unstructured text, and mixed-format information. Traditional methods can only achieve simple text extraction, lacking automatic identification, semantic analysis, and structured organization of key fields. This results in data that is difficult to use directly for retrieval, question answering, and reasoning, reducing document utilization. 2. Lack of intelligent annotation and unified management of visual content in automotive repair documents: Automotive repair documents typically contain visual information such as diagrams, flowcharts, and component images. Existing technologies have limited capabilities in image extraction, content understanding, and semantic annotation, lack a linkage mechanism with textual knowledge, and have not formed a unified cloud resource management system. This results in low utilization of visual data, failing to support the needs of intelligent document applications.

[0057] In summary, how to provide a document processing method suitable for automotive repair documents, thereby improving the utilization rate of automotive repair data, is an urgent problem to be solved.

[0058] Therefore, this invention discloses a method and related equipment for processing automotive repair documents. In this solution, by identifying the format of automotive repair documents and uniformly converting them into Markdown format, the problem of parsing complex formats such as multi-column layout, table nesting, and mixed text and graphics is solved, thus standardizing the content. The standardized documents are then input into a language model, which utilizes its natural language understanding capabilities to accurately identify professional terms and extract structured content. The extracted content is then exported according to a preset business format, achieving unified management of automotive repair document content and improving data utilization.

[0059] like Figure 1 The diagram shown is an architecture diagram of an automotive repair document processing system disclosed in an embodiment of the present invention. The system includes a front-end structure and a back-end structure.

[0060] The front-end structure includes: a user management interface, a file upload module, a text review module, and a file export module.

[0061] Specifically, the user management interface, used in the automotive repair document processing system, allows for user permission configuration, with each user having different operational functions. For example, the administrator role grants all operational permissions; domestic and international clients are assigned permissions as needed, including the ability to view, edit, and approve Q&A pairs; and annotation engineers are assigned permissions as needed, including the ability to view, edit, and approve Q&A pairs / document summaries / key information.

[0062] like Figure 2 The image shown is an interface diagram of a file upload module disclosed in an embodiment of the present invention.

[0063] In the file upload module, users can click the file upload button to upload files. Uploaded file formats include, but are not limited to: PDF, WORD, HTML, audio / video, TXT, JSON, etc.

[0064] The text review module is a front-end function, primarily responsible for manually reviewing content generated by the back-end. Users can manually modify content that does not meet their requirements. Reviewable content includes document summaries output by the language model, question-and-answer pairs, and key information.

[0065] The file export module is used to export file summaries, question-and-answer pairs, and key information from the backend language model. However, the main content exported is the question-and-answer pairs, and the exported format is CSV file, which is convenient for subsequent business data applications.

[0066] The backend structure includes: a file processing module, a summary generation module, a question-and-answer pair generation module, and a key information generation module.

[0067] Specifically, the file processing module starts running after the user uploads the file. Its main purpose is to process files of different formats into a Markdown target auto repair file.

[0068] It should be noted that the file processing module automatically extracts, classifies, and generates titles for the maintenance images embedded in the document (i.e., obtains image description information through visual or semantic annotation), and uploads the images to the cloud server to establish a unified image resource library, realizing the linkage management and retrieval of text knowledge and visual content.

[0069] Specifically, the summary generation module is used to input preset summary generation prompts and target auto repair files into a preset language model to obtain the file summary output by the language model.

[0070] Specifically, the question-answer pair generation module is used to input the preset question-answer pair generation prompts and the target auto repair file into the language big model, and obtain multiple question-answer pairs output by the language big model.

[0071] It should be noted that question-answer pairs can serve as semantic knowledge units for knowledge retrieval and large model training, supporting intelligent question answering and knowledge reasoning.

[0072] Specifically, the key information generation module is used to input preset key information extraction prompts and target auto repair files into the language big model, and obtain the key information output by the language big model.

[0073] Based on the above embodiments of the present invention, a vehicle repair document processing system is disclosed, such as... Figure 3The diagram shown is a flowchart of an automotive repair document processing method disclosed in an embodiment of the present invention, including the following steps:

[0074] Step S301: Receive the auto repair documents uploaded by the user and identify the format of the auto repair documents.

[0075] In step S301, the user can click the file upload button to upload a file. Uploaded file formats include, but are not limited to: PDF, WORD, HTML, audio / video, TXT, JSON, etc.

[0076] It should be noted that automotive repair documentation refers to all written or electronic records and documents generated and used in automotive repair, maintenance, diagnosis, and related business activities. It systematically records key data such as vehicle information, fault descriptions, repair items, replaced parts, labor costs, and quality inspection results.

[0077] Step S302: Using a conversion method that matches the format of auto repair files, convert the auto repair files to Markdown format to obtain the target auto repair file.

[0078] It should be noted that Markdown is a lightweight markup language that uses simple plain text symbols (such as #, *, -) to define the structure and format of a document.

[0079] In step S302, auto repair files of different formats are uniformly processed into Markdown text type. For example, PDF, WORD, HTML and other file types contain images. When processing files, the images are first stored on a cloud server (to facilitate users to view the corresponding image information and ensure the integrity of the content). Then, the images are saved as image text links. The image text links contain the image storage location and image description information (such as image title).

[0080] The image storage location is the address of the image on the cloud server.

[0081] For example, the format of an image text link is: .

[0082] like Figure 4 The diagram shown is a schematic diagram illustrating the conversion of PDF format auto repair documents disclosed in an embodiment of the present invention.

[0083] In the specific implementation of step S302, if the auto repair file is in audio format, the audio file is converted into audio content using a preset audio model; and a target auto repair file in Markdown format is generated based on the audio content.

[0084] If the auto repair document is a composite document format containing text and / or images, the text content in the auto repair document is extracted, and the images in the auto repair document are semantically annotated to obtain image description information; the images in the auto repair document are stored in a cloud server to obtain the image storage location; based on the image description information and image storage location, image text links are generated; based on the text content and image text links, a target auto repair document in Markdown format is generated.

[0085] The composite document formats include, but are not limited to, PDF, WORD, HTML, TXT, and JSON. Based on the text content and image text links, the target auto repair file is processed into Markdown format using Python dependency libraries.

[0086] The generation of image description information involves obtaining the first 100 characters and the last 100 characters of the original image location as text information, converting the image into base64 format information, and then sending it to the locally deployed visual model. The model then generates a short image title and description of the image.

[0087] In this embodiment of the invention, for images in auto repair files in composite document format, it is necessary to store them on a cloud server to obtain the corresponding storage address of the images on the cloud server, i.e., the image storage location. Then, based on the image storage location and image description information, an image text link is generated. This way, on the one hand, the images will be recorded in the target auto repair file in the format of image text links, and on the other hand, the image information of each file can still be used by other services.

[0088] In one embodiment, if the auto repair document is in PDF format, a composite document format, the auto repair document is processed into a JSON structured file using a preset MinerU file processing component; text content and images are extracted from the JSON structured file; and the images are semantically annotated using a preset visual big model to obtain image description information.

[0089] It should be noted that the MinerU file processing component processes PDF auto repair files into copyable and editable JSON structured files. The JSON structured file has a field 'type' that records different information from the PDF auto repair file. If 'type' is 'text', it records text information; if 'type' is 'image', it records images; and if 'type' is 'table', it records tables. Therefore, text content and images can be extracted through the corresponding fields.

[0090] Understandably, the text in the table can also be extracted as text content.

[0091] Optionally, for scanned PDF files, a visual large model, rule-based extraction technology (mainly using keywords to locate, for example, the position of images and tables), and an OCR text recognition model can be used to help extract some content that was not parsed from images and tables, in order to solve the problem that the MinerU file processing component cannot fully process some scanned PDF auto repair documents.

[0092] It should be noted that in scanned PDFs, text and images are scanned together, making it difficult to locate the images and thus hindering the use of large-scale visual models for semantic annotation. Therefore, a rule-based extraction technique is used to locate the images or tables within the scanned PDF using certain keywords. For example, the keyword " Figure 2 “Table 2”, etc.

[0093] In one embodiment, for auto repair documents in WORD format (a composite document format), WORD format is easier to process than PDF format. The entire text content can be directly read using Python libraries. Images are still semantically annotated using a pre-defined visual model to obtain image description information. For auto repair documents in HTML format (a composite document format), HTML information processing can be performed using Python libraries to extract the text content and finally process it into a Markdown format target auto repair document. For auto repair documents in TXT format (a composite document format), the text content can be directly read from the TXT file.

[0094] In one embodiment, if the auto repair file is in video format, a preset video-to-audio component is used to convert the auto repair file into an audio format, and then it is processed according to the above-mentioned processing method for auto repair files in audio format, finally obtaining the target auto repair file in Markdown format.

[0095] Video-to-audio conversion components, such as FFmpeg, are powerful open-source multimedia processing tools that support the conversion of various audio and video formats.

[0096] Step S303: Input the target auto repair file into the preset language model to obtain the file content extracted by the language model.

[0097] In step S303, preset prompts are displayed on the front-end page. If the user modifies the prompts, the modified prompts and the target auto repair file are input into the preset language model to obtain the file content extracted by the language model. If the user does not modify the prompts, the original prompts and the target auto repair file are input into the preset language model to obtain the file content extracted by the language model.

[0098] The extracted file content is mainly used to support applications such as intelligent customer service, fault diagnosis assistance, and knowledge retrieval. The corresponding file content includes: the file summary of the target auto repair file, the question-and-answer pairs generated based on the target auto repair file, and the key information in the target auto repair file.

[0099] In one embodiment, preset summary generation prompts and target auto repair documents are input into a preset language big model to obtain a document summary output by the language big model.

[0100] like Figure 5 The image shown is an example diagram of a summary generation prompt and document summary disclosed in an embodiment of the present invention.

[0101] In this embodiment of the invention, a large language model and summary generation prompts are mainly used to summarize the overall content of the target auto repair document and remove redundant information.

[0102] The summary generation prompts control the large language model to generate the desired results. Different users have different desired results, so this parameter is exposed on the front end. Users can customize it according to their needs, and the modified prompts are then sent to the language model to generate the corresponding results. If the user does not modify it, a default summary generation prompt will also be provided.

[0103] In one embodiment, preset question-and-answer pairs are generated by inputting prompt words and target auto repair files into a language big model, resulting in multiple question-and-answer pairs output by the language big model.

[0104] like Figure 6 The image shown is an example diagram of a question-and-answer pair generation prompt and question-and-answer pair disclosed in an embodiment of the present invention.

[0105] By using a large language model, question-and-answer pair prompts, and target auto repair documents, content-related question-and-answer pairs can be automatically generated. The number of question-and-answer pairs generated can be controlled to meet the data needs of subsequent business operations.

[0106] The question-and-answer pair generation prompts control the large model to generate the desired results. Different users have different desired results, so this parameter is exposed on the front end. Users can customize it according to their needs, and the modified prompts are then sent to the language model to generate the corresponding question-and-answer pair format and number of pairs based on the user's requirements. If the user does not modify it, there will also be a default question-and-answer pair prompt.

[0107] In one embodiment, preset key information extraction prompts and target auto repair documents are input into a language big data model to obtain key information output by the language big data model.

[0108] like Figure 7The diagram shown is a schematic diagram of key information extraction according to an embodiment of the present invention.

[0109] The method for extracting key information also utilizes a large language model and key information extraction prompts to automatically extract information from the target auto repair document. The extracted key information includes the vehicle's brand, model, year, fault codes, and component information. If no relevant key information is found in the target auto repair document, an empty value is returned.

[0110] In one embodiment, the file content extracted from the language big model is displayed through a front-end page; in response to the user's modification operation on the file content, the file content is modified to obtain the modified file content; and the modified file content is exported according to a preset business data format.

[0111] Users can manually modify the document summaries, question-and-answer pairs, and key information obtained above to identify any content that does not meet their needs.

[0112] Step S304: Export the file content according to the preset business data format.

[0113] It should be noted that the document summaries, question-and-answer pairs, and key information obtained above can all be exported, but the main content exported is the question-and-answer pairs, and the exported format is CSV file, which is convenient for subsequent business data application.

[0114] Based on the above-described embodiment of the present invention, the method for processing automotive repair documents overcomes the shortcomings of the prior art by performing structured parsing and semantic storage of multi-format data, thereby generating the following advantages:

[0115] First, existing document parsing methods often remain at the stage of simple OCR text recognition, easily losing information in complex tables and mixed text-image layouts, resulting in incomplete knowledge extraction. This invention introduces layout analysis, noise-resistant OCR, and semantic segmentation methods to transform the original document into high-quality structured knowledge units, thus significantly improving information integrity and extraction accuracy, providing a solid data foundation for subsequent retrieval and question answering.

[0116] Secondly, the extracted content is automatically converted into summaries, question-and-answer pairs, and key information, thus achieving efficient knowledge accumulation and reuse, reducing repetitive work and improving response speed.

[0117] Furthermore, existing systems often store embedded image information in documents only as attachments, making it impossible to participate in semantic retrieval. This invention designs an automatic image extraction, title generation, and semantic annotation mechanism, enabling visual content to be linked with the text knowledge base. This supports cross-modal retrieval and multi-dimensional knowledge utilization, greatly expanding the coverage of information retrieval.

[0118] Finally, combined with the front-end interface, functions can be completed efficiently and visually, and content can be freely modified, ensuring high utilization of file information. Therefore, the system not only surpasses traditional methods in accuracy, intelligence, and automation, but also ensures data security and compliance, and possesses good scalability and industry adaptability.

[0119] Based on the above embodiments of the present invention, a method for processing automotive repair documents is disclosed, such as... Figure 8 The diagram shown is a structural diagram of an automotive repair document processing device disclosed in an embodiment of the present invention, including: a receiving unit 801, a processing unit 802, an extraction unit 803, and an exporting unit 804.

[0120] The receiving unit 801 is used to receive the auto repair files uploaded by the user and identify the format of the auto repair files;

[0121] Processing unit 802 is used to convert the auto repair file into Markdown format using a conversion method that matches the format of the auto repair file, so as to obtain the target auto repair file;

[0122] Extraction unit 803 is used to input the target auto repair file into a preset language model to obtain the file content extracted by the language model.

[0123] Export unit 804 is used to export file content according to a preset business data format.

[0124] In one embodiment, the processing unit 802 is specifically used for:

[0125] If the auto repair file is in audio format, the audio file is converted into audio content using a preset audio model; a Markdown format target auto repair file is then generated based on the audio content.

[0126] If the auto repair document is a composite document format containing text and / or images, the text content in the auto repair document is extracted, and the images in the auto repair document are semantically annotated to obtain image description information; the images in the auto repair document are stored in a cloud server to obtain the image storage location; based on the image description information and image storage location, image text links are generated; based on the text content and image text links, a target auto repair document in Markdown format is generated.

[0127] In one embodiment, if the auto repair document is in PDF format (a composite document format), then the processing unit 802 is specifically used for:

[0128] Use the pre-set MinerU file processing component to process auto repair files into JSON structured files;

[0129] Extract text content and images from a JSON structured file;

[0130] The images are semantically annotated using a pre-defined visual model to obtain image description information.

[0131] In one embodiment, the device further includes:

[0132] The conversion unit is used to convert the auto repair file to Markdown format using a conversion method that matches the format of the auto repair file. Before obtaining the target auto repair file, if the auto repair file is in video format, it uses a preset video-to-audio component to convert the auto repair file to audio format.

[0133] In one embodiment, the extraction unit 803 is specifically used for:

[0134] Preset prompts are displayed on the front-end page;

[0135] If the user modifies the prompt words, the modified prompt words and the target auto repair file are input into the preset language model to obtain the file content extracted by the language model.

[0136] If the user does not modify the prompt words, the original prompt words and the target auto repair file will be input into the preset language model to obtain the file content extracted by the language model.

[0137] In one embodiment, the extraction unit 803 is specifically used for:

[0138] Input the preset summary generation prompts and the target auto repair file into the preset language model to obtain the file summary output by the language model;

[0139] Input the preset question-and-answer pairs, generated prompt words, and the target auto repair file into the language model to obtain multiple question-and-answer pairs output by the language model.

[0140] The preset key information extraction prompts and the target auto repair file are input into the language model to obtain the key information output by the language model.

[0141] In one embodiment, the device further includes:

[0142] The modification unit is used to display the file content extracted from the language model through the front-end page; respond to the user's modification operation on the file content, modify the file content, and obtain the modified file content; and export the modified file content according to the preset business data format.

[0143] Based on the above-disclosed automotive repair document processing device, this solution solves the problem of parsing complex formats such as multi-column layout, table nesting, and mixed text and graphics by recognizing the automotive repair document format and uniformly converting it into Markdown format. This standardizes the content, inputs the standardized document into a language model, and utilizes its natural language understanding capabilities to accurately identify professional terms and extract structured content. The extracted content is then exported according to a preset business format, achieving unified management of automotive repair document content and improving data utilization.

[0144] This invention also provides an electronic device, please refer to... Figure 9 The electronic device includes a memory 901 and a processor 902.

[0145] Among them, memory 901 is used to store computer programs;

[0146] The processor 902 is used to execute computer programs, specifically to implement an offline diagnostic method based on a large model provided in any embodiment of this application.

[0147] This application also provides a computer storage medium for storing a computer program, which, when executed, is specifically used to implement an offline diagnostic method based on a large model provided in any embodiment of this application.

[0148] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for system or system embodiments, since they are basically similar to method embodiments, the description is relatively simple, and relevant parts can be referred to the descriptions in the method embodiments. The systems and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0149] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0150] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for processing automotive repair documents, characterized in that, The method includes: Receive auto repair files uploaded by users and identify the format of the auto repair files; Using a conversion method that matches the format of the auto repair file, the auto repair file is converted into Markdown format to obtain the target auto repair file; The target auto repair file is input into a preset language model to obtain the file content extracted by the language model. Export the file content according to the preset business data format.

2. The method according to claim 1, characterized in that, The method of converting the auto repair file to Markdown format using a format matching the auto repair file to obtain the target auto repair file includes: If the auto repair file is in audio format, the audio file is converted into audio content using a preset audio model; a target auto repair file in Markdown format is then generated based on the audio content. If the auto repair document is a composite document format containing text and / or images, then the text content in the auto repair document is extracted, and the images in the auto repair document are semantically annotated to obtain image description information; the images in the auto repair document are stored in a cloud server to obtain the image storage location; based on the image description information and the image storage location, an image text link is generated; based on the text content and the image text link, a target auto repair document in Markdown format is generated.

3. The method according to claim 2, characterized in that, If the auto repair document is in PDF format within the composite document format, then the text content in the auto repair document is extracted, and the images in the auto repair document are semantically annotated to obtain image description information, including: The auto repair file is processed into a JSON structured file using the preset MinerU file processing component; Extract text content and images from the JSON structured file; The image is semantically annotated using a pre-defined visual model to obtain image description information.

4. The method according to claim 2, characterized in that, Before converting the auto repair file to Markdown format using a conversion method that matches the format of the auto repair file to obtain the target auto repair file, the method further includes: If the auto repair file is in video format, then the preset video-to-audio component is used to convert the auto repair file into audio format.

5. The method according to claim 1, characterized in that, The step of inputting the target auto repair file into a preset language model to obtain the file content extracted by the language model includes: Preset prompts are displayed on the front-end page; If the user modifies the prompt word, the modified prompt word and the target auto repair file are input into the preset language model to obtain the file content extracted by the language model. If the user does not modify the prompt word, the original prompt word and the target auto repair file are input into the preset language model to obtain the file content extracted by the language model.

6. The method according to claim 1, characterized in that, The step of inputting the target auto repair file into a preset language model to obtain the file content extracted by the language model includes: The preset summary generation prompts and the target auto repair file are input into the preset language big model to obtain the file summary output by the language big model; The preset question-and-answer pairs are generated into prompt words and the target auto repair file, which are then input into the language model to obtain multiple question-and-answer pairs output by the language model. The preset key information extraction prompts and the target auto repair file are input into the language model to obtain the key information output by the language model.

7. The method according to claim 1, characterized in that, The method further includes: The file content extracted from the large language model is displayed on the front-end page; In response to the user's modification operation on the file content, the file content is modified to obtain the modified file content; Export the modified file content according to the preset business data format.

8. A vehicle repair document processing device, characterized in that, The device includes: The receiving unit is used to receive auto repair files uploaded by users and to identify the format of the auto repair files; The processing unit is used to convert the auto repair file into Markdown format using a conversion method that matches the format of the auto repair file, thereby obtaining the target auto repair file; The extraction unit is used to input the target auto repair file into a preset language model to obtain the file content extracted by the language model. The export unit is used to export the file content according to a preset business data format.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the automotive repair document processing method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Used to store computer programs; wherein, when the computer programs are executed by a processor, they implement the automotive repair document processing method as described in any one of claims 1 to 7.

Images