Document processing method and device, equipment, storage medium and program product

Abstract

The invention discloses an official document processing method and device, equipment, a storage medium and a program product, and relates to the technical field of data processing.The official document processing method comprises the steps that a to-be-processed official document file is received; based on the official document file to be processed, extracting official document text content; according to the official document text content, performing to-do processing through a preset official document processing model to generate official document to-do opinions, the official document processing model comprising a plurality of to-do process sub-models, and the plurality of to-do process sub-models at least comprising an abstract generation model, an official document transfer model and a reviewing and auditing model. According to the method and the device, official document planning process processing based on the multi-agent model is realized, the problem of low processing efficiency of an existing official document processing method is solved, and the overall efficiency of official document processing is improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a document processing method, apparatus, device, storage medium and program product. Background Art

[0002] Currently, common document processing usually relies on manual segmentation and experience-based judgment. Official documents are summarized and summarized manually or with simple tools. Then, secretarial staff find the transfer department according to the responsibility division manual, and report it for approval step by step, and finally form the proposed handling opinions.

[0003] This entire process of document processing is highly dependent on manual experience and repetitive operations, which is time-consuming, labor-intensive, and prone to errors. It is difficult to cope with large-scale document processing needs and has become a significant bottleneck restricting the improvement of administrative efficiency.

[0004] The above content is only used to assist in understanding the technical solution of this application and does not constitute an admission that the above content is prior art. Summary of the Invention

[0005] The main purpose of this application is to provide a document processing method, device, equipment, storage medium and program product, aiming to solve the technical problem of low processing efficiency of existing document processing methods.

[0006] To achieve the above objectives, the present application proposes a document processing method, which includes:

[0007] Receive official documents to be processed;

[0008] Extracting the text content of the official document based on the official document to be processed;

[0009] According to the content of the official document text, the document is processed through a preset official document processing model to generate an official document processing opinion. The official document processing model includes several processing flow sub-models, and the several processing flow sub-models include at least a summary generation model, an official document transfer model and a review and audit model.

[0010] In one embodiment, the step of extracting the text content of the official document based on the official document to be processed includes:

[0011] Performing layout analysis and text recognition on the official document to be processed to obtain structured text;

[0012] Segmenting the structured text into semantically complete text blocks of fixed length, wherein the segmentation boundaries of the semantically complete text blocks are determined by contextual coherence and semantic integrity;

[0013] The semantically complete text block is converted into official document text content in the form of semantic vectors.

[0014] In one embodiment, before the step of preparing a handling opinion based on the content of the document using a preset document processing model and generating a handling opinion, the step further includes:

[0015] Obtain historical proposed opinion data;

[0016] Dividing the historical proposed opinion data into a number of proposed process training sets according to the proposed process;

[0017] Based on the several proposed process training sets, several proposed process sub-models in the document processing model are fine-tuned.

[0018] In one embodiment, the step of performing processing based on the content of the document using a preset document processing model to generate a document processing opinion includes:

[0019] Inputting the official document text content into the summary generation model to generate an official document summary;

[0020] According to the document summary, a transfer prediction is made through the document transfer model to obtain the transfer department and handling opinions;

[0021] According to the transfer department, personnel are screened using the review and approval model to generate review and approval personnel;

[0022] Based on the document summary, transfer department, handling opinions and review and audit personnel, the proposed handling opinions of the document are generated.

[0023] In one embodiment, the step of performing transfer prediction based on the document summary using the document transfer model to obtain the transfer department and handling opinions includes:

[0024] Through the document transfer model, the document summary is matched with a preset responsibility knowledge base to obtain a knowledge matching result. The responsibility knowledge base includes department function mapping relationships and historical handling rules.

[0025] Based on the knowledge matching results, the referral department and handling opinions are generated.

[0026] In one embodiment, the step of screening personnel based on the transfer department using the review and approval model to generate review and approval personnel includes:

[0027] According to the transfer department, the review and approval model is input in combination with the preset management hierarchy relationship map, and personnel are screened through the review and approval model to obtain review and approval personnel. The management hierarchy relationship map includes an organizational structure tree and authority association rules.

[0028] In addition, to achieve the above-mentioned purpose, the present application also proposes a document processing device, which includes:

[0029] A receiving module, used for receiving official documents to be processed;

[0030] An extraction module, configured to extract the text content of an official document based on the official document to be processed;

[0031] The processing module is used to prepare for processing according to the content of the official document through a preset official document processing model to generate official document processing opinions. The official document processing model includes several processing flow sub-models, and the several processing flow sub-models include at least a summary generation model, an official document transfer model and a review and audit model.

[0032] In addition, to achieve the above-mentioned purpose, the present application also proposes a document processing device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the document processing method described above.

[0033] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium. A computer program is stored on the storage medium, and when the computer program is executed by the processor, the steps of the document processing method described above are implemented.

[0034] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the document processing method described above.

[0035] This application provides a method for processing official documents. First, it receives official documents to be processed and automatically extracts text content, replacing the tedious operations of manual text entry and understanding, and avoiding the risk of semantic fragmentation due to manual intervention; then, it uses an official document processing model that integrates multiple official document processing subtasks such as summary generation, transfer department recommendation, and review and approval to prepare for processing the text content. The traditional official document processing link that relies on manual experience and judgment is replaced by model automation. By integrating scattered manual operations into a coherent automated processing link, the efficiency loss caused by repetitive operations and experience dependence in the traditional process is significantly reduced. At the same time, the risk of deviation in manual segmentation and rule matching is avoided, and the full process optimization of official document processing from input to decision-making is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0037] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0038] Figure 1 A flowchart of the first embodiment of the document processing method of this application is provided;

[0039] Figure 2 A flowchart of the second embodiment of the document processing method of this application is provided;

[0040] Figure 3 This is a schematic diagram of the module structure of the document processing device according to an embodiment of the present application;

[0041] Figure 4 This is a schematic diagram of the device structure of the hardware operating environment involved in the document processing method in the embodiment of the present application.

[0042] The purpose, features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0043] It should be understood that the specific embodiments described herein are merely used to explain the technical solutions of the present application and are not intended to limit the present application.

[0044] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0045] The main solution of the embodiment of the present application is: receiving official documents to be processed; extracting the text content of the official documents based on the official documents to be processed; and generating opinions on the official document processing by using a preset official document processing model according to the text content of the official document. The official document processing model includes several processing flow sub-models, and the several processing flow sub-models include at least a summary generation model, an official document transfer model, and a review and audit model.

[0046] Because existing document processing relies heavily on manual experience and repetitive operations, secretarial staff must manually handle each step, including summary writing, department matching, and personnel selection. For example, it's difficult to maintain semantic coherence when writing summaries, leading to subsequent processing errors due to semantic fragmentation. Department recommendations require repeated reference to static responsibilities manuals, which is time-consuming and susceptible to lags in rule updates. Personnel selection relies on escalating reports and level verification, resulting in a lengthy process with a low tolerance for errors. This efficiency bottleneck is particularly prominent in large-scale document processing scenarios, which is not only time-consuming and costly, but also increases the risk of error due to the uncertainty of manual operations.

[0047] In this application, first, the official document files to be processed are received and the text content is automatically extracted, replacing the tedious operations of manual text entry and understanding, and avoiding the risk of semantic fragmentation due to manual intervention; then, the official document processing model that integrates multiple official document processing subtasks such as summary generation, transfer department recommendation, and review and approval is used to prepare for processing of the text content. The traditional official document processing link that relies on manual experience and judgment is replaced by model automation. By integrating the scattered manual operations into a coherent automated processing link, the efficiency loss caused by repetitive operations and experience dependence in the traditional process is significantly reduced. At the same time, the risk of deviation in manual segmentation and rule matching is avoided, and the full process optimization of official document processing from input to decision-making is achieved.

[0048] It should be noted that the execution subject of this embodiment may be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, mobile phone, etc., or an electronic device or document processing device capable of performing the above functions. This embodiment and the following embodiments will be described below using a document processing device as an example.

[0049] Based on this, the embodiment of the present application provides a document processing method, referring to Figure 1 , Figure 1 This is a flowchart of the first embodiment of the document processing method of this application.

[0050] In this embodiment, the document processing method includes steps S10 to S30:

[0051] Step S10, receiving the official document to be processed;

[0052] It is understood that the official documents to be processed refer to various official documents that need to be processed, such as letters from superiors, requests from subordinates, etc. In order to obtain the original input for official document processing, by receiving the official documents to be processed, basic data can be provided for subsequent text extraction and preparation for processing.

[0053] Specifically, receiving pending official documents can be accomplished in a variety of ways. For example, documents can be received via email, a convenient and fast method suitable for most government agencies. Alternatively, documents can be received via a dedicated document transmission system, a highly secure method that ensures the confidentiality and integrity of documents. Different receiving methods can be selected based on the specific circumstances of the organization to meet diverse needs.

[0054] For example, a unit receives a request document in PDF format from a subordinate unit through an internal document transmission system. The system automatically detects the file and marks it as a pending document for subsequent processing.

[0055] Step S20, extracting the official document text content based on the official document to be processed;

[0056] It should be noted that extracting the text content of an official document refers to obtaining the text information therein from the official document file and converting the text information into a text form suitable for processing so as to carry out subsequent processing.

[0057] It is understandable that in order to obtain valid text information from official documents so that subsequent processing can be based on accurate text content, the accuracy and effectiveness of subsequent processing can be ensured by extracting the official document text content.

[0058] Specifically, document text extraction from different file formats can be achieved through different technical means. For text-based documents, the text content can be directly read; for non-text-based documents such as scanned copies, OCR (Optical Character Recognition) technology is required to recognize the text in the image.

[0059] Step S30, based on the content of the official document, a preset official document processing model is used to perform processing to generate an official document processing opinion. The official document processing model includes several processing flow sub-models, and the several processing flow sub-models include at least a summary generation model, an official document transfer model and a review and audit model.

[0060] It should be noted that the official document processing model refers to a pre-trained intelligent agent model that can prepare and process official documents. It can include multiple intelligent agent models related to the preparation process, namely the summary generation model, the official document transfer model and the review and audit model. These sub-models are responsible for the generation of official document summaries, the prediction of transfer departments and processing opinions, and the screening of review and audit personnel, and together complete the generation of proposed opinions on official documents.

[0061] As you can see, once the document content is fed into the agent-based model, it will sequentially process it through the summary generation model, the document transfer model, and the review and approval model to generate draft document processing opinions, thereby improving the efficiency and quality of document processing. By using an agent-based model, document processing can be automated, reducing manual intervention and improving processing speed and accuracy.

[0062] In a feasible implementation manner, the step of extracting the text content of the official document based on the official document to be processed includes:

[0063] Step S201: performing layout analysis and text recognition on the official document to be processed to obtain structured text;

[0064] It's important to note that layout analysis involves analyzing the page layout of official documents, identifying the position and arrangement of elements like text, images, and tables. This helps provide accurate regional positioning for subsequent text recognition. Text recognition converts the text within official documents into an editable format. Structured text refers to processed text with a clear structure and format, such as paragraphs, headings, and lists.

[0065] It can be understood that by converting unstructured official documents into structured text through layout analysis and text recognition, the text information in the official documents can be accurately extracted, and the original structure and semantic relationships in the text can be retained, providing a high-quality data foundation for subsequent text segmentation and semantic analysis processing.

[0066] Specifically, layout analysis tools can be used to perform in-depth layout analysis of PDF files, identifying the location and attributes of elements such as text blocks, tables, and images. Text recognition allows direct text extraction for official documents in text format; for documents in image formats such as scanned copies, OCR technology is required for text recognition.

[0067] For example, for a PDF document containing text and tables, a layout analysis tool is first used to analyze the document's layout and identify the locations of text blocks and tables. Then, for text blocks, the text content is directly extracted; for tables, a specialized table recognition tool is used to process them and convert them into a structured data format. The result is a structured document containing the document's main text, title, tables, and other content, while preserving the original semantic structure.

[0068] Step S202: Segment the structured text into semantically complete text blocks of fixed length, where the segmentation boundaries of the semantically complete text blocks are determined by contextual coherence and semantic integrity;

[0069] It should be noted that segmentation refers to dividing a long text into several shorter blocks according to certain rules. Fixed-length means that each block has roughly the same number of words, set here at approximately 500. However, the semantic integrity of the blocks must be taken into account, ensuring that each block is semantically coherent and complete. Semantically complete blocks mean that the logical relationships and semantic information between sentences in the segmented blocks are intact, and the segmentation does not destroy the original meaning.

[0070] It can be understood that by dividing the structured text into blocks, long texts can be divided into semantically complete text blocks of appropriate length, so that the length of each text block is moderate, which will neither affect the processing efficiency due to being too long nor destroy the semantic integrity due to being too short, thereby improving the model's understanding and processing effect of the text.

[0071] Specifically, text segmentation can be performed based on the following methods. One method is to segment according to the number of words. When the text length reaches the set number of words (such as 500 words), it is segmented. However, when segmenting, it is necessary to check whether the current sentence is complete. If it is incomplete, the segmentation point is moved to the next punctuation mark to ensure semantic integrity. Another method is to segment based on semantic units. For example, each paragraph or each topic-related paragraph is combined into a text block. It is also possible to combine machine learning algorithms to train a text segmentation model to automatically determine the segmentation points based on the semantic features of the text.

[0072] For example, for a structured official document text, the length is 2000 words. According to the word count segmentation method, it is segmented into 4 text blocks, each of which is about 500 words. During the segmentation process, when the last text block is less than 500 words, it is merged with the previous text block to ensure the semantic integrity of each text block. For example, when segmenting to the fourth text block, it is found that there are only 300 words, and these 300 words are closely related to the third text block in semantics, then the fourth text block is merged into the third text block, so that the third text block becomes about 800 words, but its semantic integrity is ensured.

[0073] Step S203: converting the semantically complete text block into official document text content in the form of semantic vectors.

[0074] It's important to note that converting to semantic vectors involves mapping each word in a text block to a fixed-dimensional vector space, thereby representing the entire text block as a vector or a set of vectors. The purpose of this step is to convert the original text into a numerical form that the model can process for subsequent processing.

[0075] It can be understood that the semantic information of the text can be encoded into a vector, and the text block can be converted into a numerical form that the model can understand and process, so that the model can capture the semantic similarity and correlation between the texts, so as to use the intelligent agent model for semantic analysis and processing, thereby improving the accuracy and effectiveness of the processing.

[0076] Specifically, the process of converting text blocks into semantic vectors can be implemented using a word embedding model. For each semantically complete text block, each word in it can be converted into a corresponding word vector, and then all word vectors can be aggregated into a text block vector by taking the average, summing, or using an attention mechanism.

[0077] For example, for a semantically complete text block such as "Notice on Strengthening Community Environmental Governance: All communities should strengthen routine inspections of environmental sanitation...", a pre-trained word embedding model is used to convert each word into a 300-dimensional word vector. The average of all word vectors is then calculated to obtain the semantic vector for the text block, which is then used as input for the subsequent document processing model.

[0078] In this implementation, structured text is obtained by performing layout analysis and text recognition on the official document to be processed, and then the structured text is divided into semantically complete text blocks of fixed length and converted into official document text content in the form of semantic vectors, providing high-quality data input for subsequent preparation and processing, further improving the automation and intelligence level of official document processing, enhancing the model's understanding and processing capabilities of official document texts, and helping to generate more accurate and reasonable official document preparation opinions, effectively solving the inefficiency and error-prone problems caused by manual segmentation and experience judgment in existing official document processing methods.

[0079] In a feasible implementation manner, before the step of preparing a handling opinion based on the content of the document using a preset document processing model, the step further includes:

[0080] Step S3011, obtaining historical proposed action data;

[0081] It should be noted that historical proposed opinion data refers to the records of official documents processed in the past and their corresponding proposed opinions, which contain rich experience in official document processing and can reflect the proposed handling methods and results under different official document types, contents and situations.

[0082] It is understandable that in order to provide rich sample data for model training, so that the model can learn effective processing patterns and rules, thereby improving the accuracy and generalization ability of processing, by obtaining historical data, we can make full use of existing document processing experience, so that the model can learn processing methods in different situations, thereby improving the model's processing effect on new documents.

[0083] Specifically, historical proposed action data can be extracted from databases storing document processing records, such as the unit's document management system and archives system. For example, SQL queries can be used to extract documents, their corresponding proposed actions, and processing results from the document management system's historical database. This data also needs to be cleaned and preprocessed to remove invalid, duplicate, or erroneous records, and formatted and standardized for subsequent model training.

[0084] Step S3012, dividing the historical proposed action opinion data into a number of proposed action process training sets according to the proposed action process;

[0085] It should be noted that the preparation process refers to all aspects of the preparation and processing of official documents, including summary generation, document transfer, and review and approval.

[0086] It can be understood that the purpose of this step is to classify and organize historical data according to different links of the proposed process to form multiple training sets. Each training data set corresponds to the training of a proposed process sub-model, thereby providing targeted training data for each sub-model, so that the model can better learn the proposed rules and characteristics of each link, so as to effectively train the corresponding sub-model and improve the training effect of the model and the accuracy of the proposed processing.

[0087] Specifically, it can be divided according to the following method: for the training set of the summary generation model, the main content of the official document and the corresponding summary are extracted from the historical proposed opinion data; for the training set of the official document transfer model, the main content of the official document, the summary, the transfer department and the handling opinions are extracted; for the training set of the review and audit model, information such as the transfer department and the review and audit personnel are extracted.

[0088] Step S3013: fine-tune several proposed process sub-models in the document processing model based on the several proposed process training sets.

[0089] It should be noted that fine-tuning refers to further training and optimizing the model using training data from a specific field based on the pre-trained model to adapt to specific tasks and data distribution, and improve its performance and effectiveness in document processing.

[0090] It is understandable that the historical draft opinion data can be used to further fine-tune the various sub-models of the official document processing model. Through fine-tuning, the model can learn the specific laws and characteristics of the official document drafting field on the basis of existing pre-training, so that the model can better adapt to the actual needs of official document drafting and processing, and improve the accuracy and applicability of the model.

[0091] Specifically, the fine-tuning process can adopt the transfer learning method. First, use large-scale general corpus to pre-train each sub-model to obtain the initial weights and parameters of the model. Then, use the corresponding proposed process training set to fine-tune the model. For example, use the training set of the summary generation model to fine-tune the summary generation model. By adjusting the weights and parameters of the model, the model can accurately generate official document summaries. During the fine-tuning process, you can use optimization algorithms such as gradient descent, and set appropriate hyperparameters such as learning rate and batch size. For example, for the summary generation model, you can use a sequence-to-sequence model structure, use general corpus for training in the pre-training stage, and then use the official document summary generation training set for training in the fine-tuning stage. Adjust the model's attention mechanism and decoder parameters to adapt to the characteristics of the official document summary generation task.

[0092] For example, the summary generation model was pre-trained using one million official documents and their summaries, resulting in a preliminary summary generation model. Next, a unit extracted 1,000 official documents processed over the past year and their proposed handling opinions from its official document management system's historical database. This data included information such as the document's main content, proposed handling summaries, referral departments, handling opinions, and reviewers. After data cleaning and pre-processing, a dataset containing 980 valid document handling opinions was obtained. These 980 document handling opinion data were divided into three training sets based on the handling process. For the training set used to fine-tune the summary generation model, the document's main text and abstracts were extracted, yielding 980 training examples. For the training set used to fine-tune the document transfer model, the document's main text, abstracts, referral departments, and handling opinions were extracted, yielding a similar number of training examples. For the training set used to train the review model, the document's main text, abstracts, referral departments, and handling opinions were extracted, yielding a similar number of training examples. For the training set used to train the review model, the referral departments and reviewers were extracted, yielding a similar number of training examples. Each training set was designed with input and output data for training the corresponding sub-model. For the initial summary generation model, the divided official document summary model training set was used to fine-tune the model. During fine-tuning, the learning rate was set to 0.001, the batch size was set to 32, and training was performed for 10 epochs. After fine-tuning, the model generated more accurate official document summaries that adhered to the style and requirements of official documents, better meeting the needs of document preparation and processing.

[0093] In this implementation, by dividing the acquired historical proposed opinion data into training sets according to the proposed processing process, and fine-tuning the sub-model in the official document processing model, the model can learn the domain knowledge and rules of proposed document processing, thereby improving the model's proposed processing capabilities and accuracy, further enhancing the intelligent level of official document processing, making the proposed opinions more in line with actual needs, and effectively solving the problem of existing official document processing methods relying on manual experience and being difficult to cope with large-scale official document processing, thereby improving the overall efficiency and quality of official document processing.

[0094] This embodiment provides a method for processing official documents. By receiving official documents to be processed, extracting the text content of the official documents, and using a preset official document processing model to perform preliminary processing, the method realizes automation and intelligence of official document processing. It effectively solves the problems of existing official document processing methods that are highly dependent on manual experience and repetitive operations and have low processing efficiency, improves the speed and accuracy of official document processing, and can better cope with large-scale official document processing needs.

[0095] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the first embodiment can be referred to the above introduction and will not be described in detail later. Figure 2 , Figure 2This is a flow chart of the second embodiment of the document processing method of this application.

[0096] In this embodiment, the step of performing processing based on the content of the document using a preset document processing model to generate a document processing opinion includes:

[0097] Step S3021: Input the official document text content into the summary generation model to generate an official document summary;

[0098] It's important to note that the summary generation model is a trained and fine-tuned machine learning model that generates concise summaries based on text input. Generating a summary summarizes the main content of a document, extracting key information and providing a concise overview for subsequent transfer prediction and review.

[0099] It is understandable that the summary generation model condenses the content of the official document into a concise official document summary to highlight the core points of the official document, making it easier for subsequent links to quickly understand the main content of the official document and improve processing efficiency.

[0100] Specifically, the summary generation model can be implemented using a variety of technologies. A common approach is to use a sequence-to-sequence model based on an attention mechanism. The model consists of an encoder and a decoder. The encoder encodes the input text vector sequence into a context vector, and the decoder generates a summary text based on the context vector. For example, an LSTM (long short-term memory network) can be used as the structure of the encoder and decoder, and through training, it can learn how to extract key information from official document text and generate a summary. When inputting the content of the official document text, the semantic vector sequence is input into the encoder, and the context vector is obtained after processing by the encoder. The decoder then gradually generates each word in the summary based on the context vector until a complete summary is generated.

[0101] For example, a sequence of semantic vectors from an official document is input into a summary generation model. The encoder portion of the model processes the input vector sequence, capturing the semantic information and contextual relationships within the text and generating a context vector. Based on this context vector, the decoder portion, combined with an attention mechanism, focuses on the parts of the input text most relevant to summary generation, gradually generating a summary. For example, the generated summary might read: "Regarding the funding application for the XX project, the application amount is XX million yuan, to be used for project equipment procurement and personnel training. The reasons for the application are detailed in the official document."

[0102] Step S3022: Based on the document summary, the document transfer model is used to predict the transfer process and obtain the transfer department and handling opinions.

[0103] It's important to note that the document transfer model is a machine learning model specifically designed to predict the department to which a document should be transferred and provide corresponding handling opinions. The "referring department" refers to the specific department responsible for handling the document, and the "handling opinions" are suggestions and requirements for how the document should be handled.

[0104] It is understandable that according to the key information in the official document summary, the appropriate processing department and specific handling measures can be determined through transfer prediction, so that the official document can be accurately assigned to the corresponding department, and specific handling opinions can be given to ensure that the official document is handled in a timely and effective manner, which helps to improve the efficiency and pertinence of official document processing, provide a clear direction for the effective handling of official documents, and avoid delays in official documents between departments.

[0105] Specifically, the document transfer model can be trained based on a supervised learning algorithm. Input features include keywords, topics, and the type of business involved in the document abstract, and the output is the transfer department and handling opinions. For example, a classification algorithm can be used to classify document abstracts and determine the transfer department. Furthermore, to generate handling opinions, a text generation model, such as a variant of GPT (Generative Pre-Training Model), can be used to generate specific handling opinion text based on the document abstract.

[0106] For example, based on the generated document summary, "Regarding the funding application for Project XX, the application amount is XX million yuan, to be used for project equipment procurement and personnel training. The reasons for the application are detailed in the document," the document transfer model analyzes and determines that the transfer department is the Finance Department and generates a handling opinion: "Please ask the Finance Department to review the rationale of the project budget, approve it according to the financial approval process, and provide feedback on the approval result within 5 working days."

[0107] Step S3023: Based on the transfer department, personnel are screened using the review and approval model to generate review and approval personnel;

[0108] It should be noted that the review and approval model is a model used to determine the review and approval personnel of official documents. Review and approval personnel refer to leaders or relevant personnel who need to review, approve and approve official documents.

[0109] Understandably, in order to identify the person responsible for reviewing and approving official documents, clarify responsibilities and authorities, and ensure the standardization and authority of document processing, appropriate reviewers are selected based on the information of the referring department, combined with the unit's organizational structure and management hierarchy, to ensure effective document approval and decision-making. This ensures that documents are approved promptly and correctly, improving the quality and efficiency of document processing.

[0110] Specifically, the review and audit model can perform personnel screening based on the organizational structure and management hierarchy relationships stored in the graph database. For example, the organizational structure of the unit is represented as a graph structure, with nodes representing departments and personnel, and edges representing management relationships and authority associations. According to the transferring department, its superiors and relevant review and audit personnel are searched in the graph. Graph traversal algorithms, such as depth-first search or breadth-first search, can be used to start from the node of the transferring department and search upward for leaders with review and audit authority. At the same time, it can also be combined with a rule engine to further screen out suitable personnel based on preset review and audit rules, such as the leader's scope of responsibilities, approval authority, etc.

[0111] For example, given that the referral department is the Finance Department, the review and approval model searches the organizational chart for the Finance Department's supervisor. Traversing the chart, it finds the Finance Department's head and the deputy director in charge of finance. Based on the review and approval rules, the deputy director and the head of the department are identified as the reviewers, ensuring they review and approve the funding request.

[0112] Step S3024, generating proposed handling opinions for the document based on the document summary, transfer department, handling opinions and review and approval personnel.

[0113] It should be noted that the proposed handling opinions for official documents are complete proposed handling suggestions formed by integrating information such as the official document summary, the transferring department, handling opinions, and the review and audit personnel. It contains comprehensive arrangements and suggestions for official document processing, including the summary content, responsible department, handling measures, and approval personnel.

[0114] It is understandable that the information obtained from the previous summary generation model, official document transfer model and review and audit model is integrated and summarized to form a complete official document preparation opinion, which provides a comprehensive and systematic reference for leaders and assists them in approval and decision-making.

[0115] Specifically, generating proposed opinions on official documents can be achieved through text template filling and natural language generation technology. For example, a text template for proposed opinions on official documents can be designed, including fields such as summary, transfer department, handling opinions, and reviewer. The corresponding information obtained from the previous model processing process is filled into the template to generate the text of the proposed opinions. Alternatively, a natural language generation model can be used to take various information as input to generate a smooth and natural text of proposed opinions. During the generation process, the model will organize the language based on the input information to ensure the coherence and professionalism of the text.

[0116] For example, the following official document summary is input into the natural language generation model: "Regarding the funding application for the XX project, the application amount is XX million yuan, which will be used for project equipment procurement and personnel training. The reasons for the application are detailed in the official document," the transferring department is "Finance Department," the handling opinion is "Please ask the Finance Department to review the rationality of the project budget, approve it according to the financial approval process, and feedback the approval result within 5 working days," and the reviewers are "Deputy Director and Section Chief." The final generated official document handling opinion is: "Regarding the funding application for the XX project, after review, it is proposed to be transferred to the Finance Department. Please ask the Finance Department to review the project budget, approve it according to the financial approval process, and feedback the approval result to the Deputy Director and Section Chief within 5 working days. The Deputy Director and Section Chief will review the document and provide approval opinions. The detailed handling opinions are attached."

[0117] In a feasible implementation, the step of performing transfer prediction based on the document summary using the document transfer model to obtain the transfer department and handling opinions includes:

[0118] Step S30221: Using the document transfer model, the document summary is matched with a preset responsibility knowledge base to obtain a knowledge matching result. The responsibility knowledge base includes departmental function mapping relationships and historical handling rules.

[0119] It should be noted that the responsibility knowledge base contains mappings between departmental functions and historical processing rules. Knowledge matching involves using specific algorithms and rules to identify the corresponding relationships between the content in the document summary and the departmental functions and processing rules in the responsibility knowledge base.

[0120] It can be understood that it is through knowledge matching that the business content in the official document summary is associated with the functions of the department, so as to determine which department is most suitable for handling the official document, and put forward reasonable handling opinions based on historical handling rules, thereby improving the accuracy and efficiency of the transfer and avoiding incorrect handling or delay of official documents.

[0121] Specifically, knowledge matching can be achieved through the following methods. First, keywords are extracted from the official document abstract, and natural language processing techniques such as TF-IDF (term frequency-inverse document frequency) or text rank algorithm are used to extract the key entities and concepts in the abstract. Then, the similarity between these keywords and the department function keywords in the responsibility knowledge base is calculated. The similarity between keywords can be calculated using methods such as cosine similarity to find the department function that best matches the official document abstract. At the same time, refer to historical handling rules, analyze the transferred departments and handling opinions of similar official documents in the past, and further verify and determine the knowledge matching results. For example, for the keywords "project funding application" and "equipment procurement", a search in the responsibility knowledge base finds that the functions of the Finance Department include "funding management" and "budget review", then it can be determined that the transferred department of the official document abstract is the Finance Department.

[0122] For example, for the official document abstract "Regarding the funding application for Project XX, the application amount is XX million yuan, to be used for project equipment procurement and personnel training. The reasons for the application are detailed in the official document," the document transfer model extracts the keywords "funding application" and "equipment procurement." In the responsibilities knowledge base, the Finance Department's responsibilities include "responsible for budget review and management of project funds" and "handling financial matters related to equipment procurement." The cosine similarity between these keywords and the Finance Department's responsibilities is 0.85, higher than that of other departments, so the knowledge match result is the Finance Department.

[0123] Step S30222: Generate the transfer department and handling opinions based on the knowledge matching results.

[0124] It should be noted that the transferring department is the department responsible for handling the document determined based on the knowledge matching results, and the handling opinion is a suggestion for the document based on the department's functions and historical handling rules.

[0125] It is understandable that based on the results of knowledge matching, the department to which the official document should be transferred is clarified, and specific handling measures and requirements are given, so that the official document processing has a clear direction and operational guidelines, which helps to improve the standardization and effectiveness of official document processing and ensure that the official document is handled in a timely and correct manner.

[0126] Specifically, the transfer department can be directly determined based on the department function with the highest similarity in the knowledge matching results. For the generation of handling opinions, the department function description and historical handling rules in the knowledge matching results can be combined to use rule templates or text generation models to create handling opinions.

[0127] For example, based on the knowledge matching results, the Finance Department is identified as the transfer department. Combining the Finance Department's responsibilities and historical processing rules, a handling opinion is generated: "Please have the Finance Department conduct a detailed review of the budget for Project XX, focusing on the rationality of equipment procurement and personnel training expenses. Organize relevant financial personnel to review and approve the application in accordance with the financial approval process, and provide feedback to the office within five working days. If additional materials are required, please communicate with the requesting department in a timely manner." The handling opinion clarifies the specific work content and time requirements of the Finance Department, providing clear guidance for document processing.

[0128] In this implementation, by matching the document summary with the preset responsibility knowledge base and generating the transfer department and handling opinions based on the matching results, the intelligent generation of transfer departments and handling opinions is achieved, the accuracy and rationality of the transfer are improved, the subjectivity and error rate of manual judgment are reduced, and it is ensured that the document can be assigned to the most appropriate department and effectively handled. It effectively solves the problems of inaccurate determination of transfer departments and unreasonable handling opinions in existing document processing methods, and improves the overall quality and efficiency of document processing.

[0129] In a feasible implementation, the step of screening personnel according to the transfer department using the review and approval model to generate review and approval personnel includes:

[0130] Step S30231: input the review and approval model according to the transfer department and the preset management hierarchy relationship map, and screen personnel through the review and approval model to obtain review and approval personnel. The management hierarchy relationship map includes an organizational structure tree and authority association rules.

[0131] It's important to note that the management hierarchy diagram is a knowledge graph that includes an organizational structure tree and permission association rules. The organizational structure tree describes the hierarchical relationships between departments and positions within the organization, while the permission association rules define the scope of authority and responsibilities of different positions and departments in document processing. Personnel screening involves identifying individuals with the appropriate review and approval permissions based on the organizational structure and permission rules.

[0132] It is understandable that according to the transferring department, determining which leaders or personnel have the right to review and examine the document in the management hierarchy relationship map, clarifying the approval process and division of responsibilities, and ensuring that the document is effectively approved and decided will help improve the efficiency of document processing and avoid confusion and delays in the approval process.

[0133] Specifically, the review and approval model can be implemented based on a graph database and a rules engine. First, the organizational structure tree and permission association rules are stored in the graph database. Nodes represent departments, positions, and personnel, and edges represent hierarchical relationships and permission associations. Based on the transferring department, the graph database searches for its superior leadership node, and based on the permission association rules, selects individuals with review and approval permissions. A graph traversal algorithm, such as breadth-first search, can be used to search upward from the node of the transferring department to identify leaders with approval authority. Simultaneously, the rules engine is combined with pre-defined review and approval rules, such as the leadership's scope of responsibility and approval limits, to further determine the final review and approval personnel. For example, for a document transferred from the Finance Department, according to the organizational structure, the direct superior is the Director of the Finance Department, whose superior is the Deputy Director in charge of Finance. According to the permission association rules, funding applications exceeding a certain limit require approval from the Deputy Director, so the review and approval personnel are the Deputy Director and the Director.

[0134] For example, based on the fact that the referral department is the Finance Department, the review and approval model is input based on the preset management hierarchy relationship map. The model first locates the Finance Department in the organizational structure tree and then searches upwards for its direct supervisor, the Director of the Finance Department. According to the authority association rules, documents requesting funding exceeding 500,000 yuan require approval by the Deputy Director. Assuming the application amount is 800,000 yuan, the model selects the Deputy Director in charge of Finance and the Director of the Finance Department as the reviewers. The Deputy Director is responsible for final approval, while the Director is responsible for preliminary review and summary of opinions.

[0135] In this implementation, personnel screening is performed by combining the relationship between the transfer department and the management hierarchy, achieving precise identification of reviewers. This ensures the standardization and correctness of the document approval process, improving the efficiency and quality of document processing. This method effectively addresses the issues of inaccurate identification of reviewers and chaotic approval processes in traditional document processing methods, strengthens the management and supervision of document processing, and improves the standardization of administrative work.

[0136] In this embodiment, the document summary is generated by inputting the document text content into the summary generation model, and the transfer department and handling opinions are predicted based on the summary through the document transfer model. Then, the review and approval personnel are screened according to the transfer department through the review and approval model, and finally the proposed handling opinions of the document are integrated and generated, thereby realizing the automation and intelligence of the entire process of the proposed handling of the document. It not only improves the efficiency of the document processing, but also ensures the accuracy and rationality of the proposed handling opinions, effectively solving the problems of cumbersome and error-prone manual operations in traditional document processing methods, providing strong support for leadership decision-making, and improving administrative efficiency.

[0137] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the document processing method of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0138] This application also provides a document processing device, please refer to Figure 3 , the document processing device includes:

[0139] Receiving module 10, for receiving official documents to be processed;

[0140] An extraction module 20 is used to extract the text content of the official document based on the official document to be processed;

[0141] The processing module 30 is used to prepare the processing according to the content of the official document through a preset official document processing model to generate an official document processing opinion. The official document processing model includes several processing flow sub-models, and the several processing flow sub-models include at least a summary generation model, an official document transfer model and a review and audit model.

[0142] Optionally, the extraction module 20 is further configured to:

[0143] Obtain historical proposed opinion data;

[0144] Dividing the historical proposed opinion data into a number of proposed process training sets according to the proposed process;

[0145] Based on the several proposed process training sets, several proposed process sub-models in the document processing model are fine-tuned.

[0146] Optionally, the processing module 30 is further configured to:

[0147] Inputting the official document text content into the summary generation model to generate an official document summary;

[0148] According to the document summary, a transfer prediction is made through the document transfer model to obtain the transfer department and handling opinions;

[0149] According to the transfer department, personnel are screened using the review and approval model to generate review and approval personnel;

[0150] Based on the document summary, transfer department, handling opinions and review and audit personnel, the proposed handling opinions of the document are generated.

[0151] Optionally, the processing module 30 is further configured to:

[0152] Through the document transfer model, the document summary is matched with a preset responsibility knowledge base to obtain a knowledge matching result. The responsibility knowledge base includes department function mapping relationships and historical handling rules.

[0153] Based on the knowledge matching results, the referral department and handling opinions are generated.

[0154] Optionally, the processing module 30 is further configured to:

[0155] According to the transfer department, the review and approval model is input in combination with the preset management hierarchy relationship map, and personnel are screened through the review and approval model to obtain review and approval personnel. The management hierarchy relationship map includes an organizational structure tree and authority association rules.

[0156] The document processing device provided in this application utilizes the document processing method of the aforementioned embodiment, and can address the technical issue of low processing efficiency associated with existing document processing methods. Compared to the prior art, the document processing device provided in this application achieves the same beneficial effects as the document processing method of the aforementioned embodiment, and other technical features of the document processing device are the same as those disclosed in the aforementioned embodiment, and are not further detailed here.

[0157] The present application provides a document processing device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the document processing method in the above-mentioned first embodiment.

[0158] Reference below Figure 4 , which shows a schematic diagram of the structure of a document processing device suitable for implementing the embodiments of the present application. The document processing device in the embodiments of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), in-vehicle terminals (such as in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 4 The document processing device shown is merely an example and should not limit the functions and scope of use of the embodiments of the present application.

[0159] like Figure 4As shown, the document processing device may include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to programs stored in a read-only memory 1002 or programs loaded from a storage device 1003 into a random access memory 1004. Random access memory 1004 also stores various programs and data required for the operation of the document processing device. The processing device 1001, the read-only memory 1002, and the random access memory 1004 are interconnected via a bus 1005. An input / output interface 1006 is also connected to the bus. Typically, the following systems can be connected to the input / output interface 1006: an input device 1007 including, for example, a touch screen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the document processing device to communicate with other devices wirelessly or wired to exchange data. Although the figure shows a document processing device with various systems, it should be understood that it is not required to implement or have all of the systems shown. More or fewer systems can be implemented or provided instead.

[0160] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program comprising program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device 1003, or installed from a read-only memory 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are performed.

[0161] The document processing device provided in this application utilizes the document processing method of the aforementioned embodiment, thereby resolving the technical issue of low processing efficiency associated with existing document processing methods. Compared to the prior art, the document processing device provided in this application achieves the same beneficial effects as the document processing method of the aforementioned embodiment. Other technical features of the document processing device are the same as those disclosed in the aforementioned embodiment and are not further detailed here.

[0162] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0163] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0164] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer program) stored thereon, and the computer-readable program instructions are used to execute the document processing method in the above-mentioned embodiment.

[0165] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared or semiconductor systems, systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0166] The computer-readable storage medium may be included in the document processing device; or it may exist independently without being assembled into the document processing device.

[0167] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the official document processing device, the official document processing device is enabled to: receive the official document file to be processed; extract the official document text content based on the official document file to be processed; and perform processing according to the official document text content through a preset official document processing model to generate official document processing opinions. The official document processing model includes several processing process sub-models, and the several processing process sub-models include at least a summary generation model, an official document transfer model and a review and audit model.

[0168] Computer program code for performing the operations of the present application may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on the remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0169] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, program segment or part of code, and the module, program segment or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, as well as the combination of boxes in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or can be implemented using a combination of dedicated hardware and computer instructions.

[0170] The modules described in the embodiments of the present application may be implemented in software or hardware, wherein the name of a module does not necessarily limit the unit itself.

[0171] The computer-readable storage medium provided in this application is a computer-readable storage medium storing computer-readable program instructions (i.e., a computer program) for executing the aforementioned document processing method. This computer-readable storage medium can address the technical issue of low processing efficiency associated with existing document processing methods. Compared to the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the document processing method provided in the aforementioned embodiments, and are not further elaborated upon here.

[0172] The present application also provides a computer program product, comprising a computer program, which implements the steps of the above-mentioned document processing method when executed by a processor.

[0173] The computer program product provided by this application can solve the technical problem of low efficiency of existing document processing methods. Compared with the existing technology, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the document processing method provided by the above embodiment, and will not be repeated here.

[0174] The above description is only part of the embodiments of the present application and does not limit the patent scope of the present application. All equivalent structural transformations made by using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect application in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A document processing method, characterized in that: The document processing method includes: Receive official documents to be processed; Extracting the text content of the official document based on the official document to be processed; According to the content of the official document text, the document is processed through a preset official document processing model to generate an official document processing opinion. The official document processing model includes several processing flow sub-models, and the several processing flow sub-models include at least a summary generation model, an official document transfer model and a review and audit model.

2. The document processing method according to claim 1, wherein: The step of extracting the text content of the official document based on the official document to be processed includes: Performing layout analysis and text recognition on the official document to be processed to obtain structured text; Segmenting the structured text into semantically complete text blocks of fixed length, wherein the segmentation boundaries of the semantically complete text blocks are determined by contextual coherence and semantic integrity; The semantically complete text block is converted into official document text content in the form of semantic vectors.

3. The document processing method according to claim 1, wherein: Before the step of preparing a handling opinion based on the content of the document by using a preset document processing model and generating a handling opinion for the document, the method further includes: Obtain historical proposed opinion data; Dividing the historical proposed opinion data into a number of proposed process training sets according to the proposed process; Based on the several proposed process training sets, several proposed process sub-models in the document processing model are fine-tuned.

4. The document processing method according to claim 1, wherein: The step of performing processing based on the content of the document using a preset document processing model to generate a document processing opinion includes: Inputting the official document text content into the summary generation model to generate an official document summary; According to the document summary, a transfer prediction is made through the document transfer model to obtain the transfer department and handling opinions; According to the transfer department, personnel are screened using the review and approval model to generate review and approval personnel; Based on the document summary, transfer department, handling opinions and review and audit personnel, the proposed handling opinions of the document are generated.

5. The document processing method according to claim 4, characterized in that: The step of performing transfer prediction based on the document summary using the document transfer model to obtain the transfer department and handling opinions includes: Through the document transfer model, the document summary is matched with a preset responsibility knowledge base to obtain a knowledge matching result. The responsibility knowledge base includes department function mapping relationships and historical handling rules. Based on the knowledge matching results, the referral department and handling opinions are generated.

6. The document processing method according to claim 4, characterized in that: The step of screening personnel according to the transfer department through the review and approval model to generate review and approval personnel includes: According to the transfer department, the review and approval model is input in combination with the preset management hierarchy relationship map, and personnel are screened through the review and approval model to obtain review and approval personnel. The management hierarchy relationship map includes an organizational structure tree and authority association rules.

7. A document processing device, characterized in that: The document processing device comprises: A receiving module, used for receiving official documents to be processed; An extraction module, configured to extract the text content of an official document based on the official document to be processed; The processing module is used to prepare for processing according to the content of the official document through a preset official document processing model to generate official document processing opinions. The official document processing model includes several processing flow sub-models, and the several processing flow sub-models include at least a summary generation model, an official document transfer model and a review and audit model.

8. A document processing device, characterized in that: The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the document processing method according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the document processing method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, the steps of the document processing method according to any one of claims 1 to 6 are implemented.

Images