A document generation method, device, and medium based on multi-agent collaboration

Abstract

This application discloses a document generation method, device, and medium based on multi-agent collaboration, relating to the field of artificial intelligence technology. The method includes: receiving a natural language research topic input by a user; inputting the research topic into a pre-defined planning agent to generate a keyword set for performing web searches based on the research topic; inputting the keyword set into a pre-defined search agent to perform several web searches using web search tools to obtain several key-point summaries; inputting the research topic and several key-point summaries into a writing agent to generate a structured research report, and outputting the structured research report to a user interface.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a document generation method, device and medium based on multi-agent collaboration. Background Technology

[0002] In the field of computer software technology, especially in the area of ​​artificial intelligence-driven information processing, existing technologies generally adopt single-dimensional solutions to address document generation needs. Traditional manual methods rely on researchers manually performing the entire process of searching, filtering, integrating, and writing, which inherently suffers from inefficiency and difficulty in ensuring information integrity. Although automated search tools and basic document generation models exist, their core limitation lies in the fragmentation of functional modules—search engines only return raw webpage links or simple summaries, lacking the ability to strategically decompose complex research topics; while independently operating writing models, due to the lack of precise structured input, often result in logical gaps or missing key dimensions in the generated reports.

[0003] Existing technologies attempt to integrate search and writing functions through a single intelligent agent architecture, but due to the lack of a specialized division of labor mechanism, they still exhibit significant shortcomings when handling complex research tasks. On the one hand, a single intelligent agent struggles to simultaneously address multiple objectives such as search strategy optimization, information density extraction, and report structure planning, resulting in output content often containing redundant information or deviating from the core topic. On the other hand, the lack of collaborative protocols between modules creates data silos between the search and writing stages, with search results being directly input into the writing module without deep compression, leading to information overload and reducing the professionalism of the final report.

[0004] Therefore, how to establish a multi-agent collaborative mechanism to achieve seamless integration of the entire process from research topic analysis and precise search to structured report generation has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] This application provides a document generation method, device, and medium based on multi-agent collaboration to solve the following technical problem: how to establish a multi-agent collaboration mechanism to achieve seamless integration of the entire process from research topic analysis and precise search to structured report generation.

[0006] In a first aspect, embodiments of this application provide a document generation method based on multi-agent collaboration. The method includes: receiving a natural language research topic input by a user; inputting the research topic into a preset planning agent to generate a keyword set for performing web searches based on the research topic; inputting the keyword set into a preset search agent to perform several web searches using web search tools to obtain several key-point summaries; inputting the research topic and several key-point summaries into a writing agent to generate a structured research report, and outputting the structured research report to a user interface.

[0007] In one embodiment of this application, generating a keyword set for performing web searches based on a research topic specifically includes: using natural language processing technology to parse the grammatical structure and semantic relationships of the research topic, and extracting core concept entities based on the parsing results; combining the core concept entities to generate several keywords to obtain a keyword set.

[0008] In one embodiment of this application, a web search tool is used to perform several web searches to obtain several key summary items. Specifically, this includes: generating several keyword combinations based on keywords in a keyword set; calling a web search tool to perform a search for each keyword combination to obtain several search results; and compressing and refining the content of the several search results to obtain several key summary items. The key summary item is a text summary containing multiple paragraphs, containing only core information and no subjective comments.

[0009] In one embodiment of this application, several search results are compressed and refined to obtain several key summary items. Specifically, this includes: analyzing the text content in the search results to identify core information related to keyword combinations; deleting duplicate statements and non-substantive descriptions in the search results; and reorganizing the retained core information into text units containing multiple paragraphs according to logical relationships. The text units retain only objective factual descriptions and do not contain subjective evaluation statements.

[0010] In one embodiment of this application, generating a structured research report specifically includes: generating a report outline to be applied based on the research topic and several key point summaries; and filling the report outline with main text content based on the several key point summaries to generate a structured research report.

[0011] In one embodiment of this application, an outline of the report to be applied is generated based on the research topic and several key point summaries. Specifically, this includes: analyzing the research topic and several key point summaries, and based on the analysis results, planning the overall structural framework and writing logic of the structured research report; and constructing the outline of the report to be applied based on the overall structural framework and writing logic of the structured research report.

[0012] In one embodiment of this application, after constructing the outline of the report to be applied based on the overall structural framework and writing logic of the structured research report, the method further includes: comparing the matching degree of several key point summaries with the node content of the outline of the report to be applied; when it is identified that there are missing nodes in the summary that do not cover the core dimensions of the research topic, a supplementary search instruction is sent to the search agent; based on the supplementary search instruction, a web search tool is used to search to obtain the key point summaries to be supplemented.

[0013] In one embodiment of this application, outputting the structured research report to the user interface specifically includes: during the process of the writing agent generating the structured research report, monitoring the generation status of each chapter of the report in real time; when a new complete chapter is detected, immediately converting the content of that chapter into a streaming data transmission format; and pushing the converted chapter data to the user interface in segments for dynamic rendering through a server push event protocol.

[0014] Secondly, embodiments of this application also provide a document generation device based on multi-agent collaboration, the device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute a document generation method based on multi-agent collaboration as described above.

[0015] Thirdly, embodiments of this application also provide a non-volatile computer storage medium storing computer-executable instructions, which, when executed, implement a document generation method based on multi-agent collaboration as described above.

[0016] This application provides a document generation method, device, and medium based on multi-agent collaboration, which has the following beneficial effects: By establishing a specialized division of labor and collaborative mechanism among planning agents, search agents, and writing agents, it fundamentally solves the long-standing technical contradiction in document generation between efficiency and quality. The planning agent's semantic parsing of the research topic enables precise formulation of search strategies, significantly improving the targeting of information acquisition; the search agent's point-based summary generation ensures high information density and objectivity of the input content through forced information compression and redundancy filtering; the writing agent's two-stage generation mode combined with a dynamic outline adjustment mechanism maintains the rigor of the report structure while achieving in-depth content coverage, ultimately achieving a synergistic gain of improved efficiency and strong output quality throughout the entire process. This multi-agent collaborative architecture not only minimizes the need for manual intervention but also achieves end-to-end automated closed-loop from research topic input to structured report output in industrial applications, providing a revolutionary knowledge production tool for academic research, business analysis, and other fields. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart illustrating a document generation method based on multi-agent collaboration provided in this application embodiment; Figure 2This is a schematic diagram of the internal structure of a document generation device based on multi-agent collaboration, provided as an embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] This application provides a document generation method, device, and medium based on multi-agent collaboration to solve the following technical problem: how to establish a multi-agent collaboration mechanism to achieve seamless integration of the entire process from research topic analysis and precise search to structured report generation.

[0020] The technical solutions proposed in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0021] Figure 1 This document is a flowchart illustrating a document generation method based on multi-agent collaboration, as provided in an embodiment of this application. Figure 1 As shown in the figure, the document generation method based on multi-agent collaboration provided in this application embodiment specifically includes the following steps: Step 101: Receive the natural language research topic input by the user.

[0022] In this embodiment, the user submits an unstructured description of the research topic, such as "the latest advances of artificial intelligence in medical image diagnosis," through a graphical user interface (such as a web chat box or a mobile application input area).

[0023] Understandably, "natural language" here specifically refers to unformatted, everyday human expression. Its core principle is to allow users to express their needs in free text form, without adhering to specific grammatical templates. It's important to note that the system immediately initiates a preprocessing mechanism upon receiving input: first, it filters out irrelevant characters (such as advertising code and special symbols); then, it detects and blocks inappropriate content (such as sensitive words); and finally, it encapsulates the purified text into a structured data object (such as a JSON-formatted {"query":"The latest advancements in artificial intelligence in medical image diagnosis"}).

[0024] Step 102: Input the research topic into the preset planning agent to generate a set of keywords for performing web searches based on the research topic.

[0025] In one embodiment of this application, generating a keyword set for performing web searches based on a research topic specifically includes: using natural language processing technology to parse the grammatical structure and semantic relationships of the research topic, and extracting core concept entities based on the parsing results; combining the core concept entities to generate several keywords to obtain a keyword set.

[0026] In this embodiment, the planning agent acts as a task decomposition engine, receiving the research topic text after preprocessing in step 101. It is understood that this agent employs hierarchical semantic parsing technology: first, it constructs a syntax tree through dependency parsing (e.g., identifying "artificial intelligence" as the subject and "medical image diagnosis" as the object); then, it locates core concept entities based on a domain knowledge graph (e.g., mapping "artificial intelligence" to a technology entity set ["machine learning", "deep learning"], and associating "medical image diagnosis" to an application entity set ["CT image analysis", "MRI lesion identification"]). This process directly achieves the feature of "parsing syntactic structures and semantic relationships to extract core concept entities".

[0027] It should be noted that the entity combination strategy employs a dynamic weighting mechanism: frequently co-occurring entities (such as the strong correlation between "deep learning" and "medical imaging" in literature) are automatically merged into composite keywords (generating "deep learning medical imaging applications" instead of independent keywords), while weakly related entities retain their independent search value (such as retaining "medical data privacy" separately). Specifically, this combination mechanism significantly differs from the mechanical enumeration mode of traditional keyword generation tools. Each keyword generated carries logical relationship information between entities (such as "artificial intelligence + medical imaging diagnosis + ethical challenges" forming a triple structure), providing semantic guidance for the subsequent search agent to perform keyword combination searches. For example, when the input topic is "bottleneck of fast charging technology for new energy batteries," the planning agent not only extracts basic entities such as "lithium-ion battery" and "fast charging protocol," but also generates combinations of professional terms such as "solid electrolyte interface growth" and "lithium plating effect" through semantic association, ensuring that the keyword set covers multiple perspectives from the technical principle level and the problem level.

[0028] Step 103: Input the keyword set into the preset search agent to perform several web searches using web search tools to obtain several key summary points.

[0029] In one embodiment of this application, a web search tool is used to perform several web searches to obtain several key summary items. Specifically, this includes: generating several keyword combinations based on keywords in a keyword set; calling a web search tool to perform a search for each keyword combination to obtain several search results; and compressing and refining the content of the several search results to obtain several key summary items. The key summary item is a text summary containing multiple paragraphs, containing only core information and no subjective comments.

[0030] In one embodiment of this application, several search results are compressed and refined to obtain several key summary items. Specifically, this includes: analyzing the text content in the search results to identify core information related to keyword combinations; deleting duplicate statements and non-substantive descriptions in the search results; and reorganizing the retained core information into text units containing multiple paragraphs according to logical relationships. The text units retain only objective factual descriptions and do not contain subjective evaluation statements.

[0031] In this embodiment, the search agent initiates an adaptive grouping strategy based on the semantic keyword set output by the planning agent (such as ["solid electrolyte interface growth", "fast charging protocol lithium plating risk"] in the new energy battery case). Understandably, this strategy first calculates the semantic coupling degree between entities, automatically grouping highly coupled keywords (such as "interface growth" and "dendritic formation" having a strong causal relationship) into the same search group (generating the compound search instruction "solid electrolyte interface growth AND dendritic formation"), while retaining independent search groups for isolated entities (such as "charging pile compatibility").

[0032] It should be noted that each keyword search uses the deep crawling mode of the web search tool: it prioritizes authoritative sources (such as academic journals and patent databases) and uses DOM tree parsing to remove non-essential content such as advertisements and navigation bars.

[0033] Specifically, the compression and refinement of the original content adopts a three-stage process: the first stage identifies the core technical parameters (such as extracting key data such as "SEI film thickening rate" and "lithium deposition critical current density" from long papers); the second stage removes duplicate discussions (such as retaining only the most complete version of experimental methods that are similar in multiple papers); and the final stage reorganizes the information according to the "problem-mechanism-impact" logical chain (such as integrating fragmented data into: "1. Problem: Fast charging induces interface growth; 2. Mechanism: Lithium ion concentration gradient leads to phase transition; 3. Impact: Battery cycle life decreases"). The final generated summary strictly meets the requirement of "multi-paragraph text units containing only objective facts" as defined in claim 4.

[0034] For example, when processing the "fast charging safety standards" search group, the summary forcibly excludes subjective evaluations (such as deleting "a certain manufacturer's solution is highly innovative"), and only retains regulatory provisions ("International Electrotechnical Commission IEC 62660-3 stipulates: additional overcharge protection is required for fast charging rates > 3C") and technical facts ("graphite anode lithium intercalation potential plateau < 0.1V is prone to lithium plating").

[0035] Step 104: Input the research topic and several key points into the writing agent to generate a structured research report, and output the structured research report to the user interface.

[0036] In one embodiment of this application, generating a structured research report specifically includes: generating a report outline to be applied based on the research topic and several key point summaries; and filling the report outline with main text content based on the several key point summaries to generate a structured research report.

[0037] In one embodiment of this application, an outline of the report to be applied is generated based on the research topic and several key point summaries. Specifically, this includes: analyzing the research topic and several key point summaries, and based on the analysis results, planning the overall structural framework and writing logic of the structured research report; and constructing the outline of the report to be applied based on the overall structural framework and writing logic of the structured research report.

[0038] In one embodiment of this application, after constructing the outline of the report to be applied based on the overall structural framework and writing logic of the structured research report, the method further includes: comparing the matching degree of several key point summaries with the node content of the outline of the report to be applied; when it is identified that there are missing nodes in the summary that do not cover the core dimensions of the research topic, a supplementary search instruction is sent to the search agent; based on the supplementary search instruction, a web search tool is used to search to obtain the key point summaries to be supplemented.

[0039] In one embodiment of this application, outputting the structured research report to the user interface specifically includes: during the process of the writing agent generating the structured research report, monitoring the generation status of each chapter of the report in real time; when a new complete chapter is detected, immediately converting the content of that chapter into a streaming data transmission format; and pushing the converted chapter data to the user interface in segments for dynamic rendering through a server push event protocol.

[0040] In this embodiment, the writing agent initiates a two-stage generation protocol: in the first stage, when constructing a dynamic outline, the report framework template ("current situation analysis → technical conflicts → regulatory comparison → governance suggestions") is automatically matched based on the domain attributes of the research topic (such as "difficulties in regulating blockchain technology" being identified as a policy research category).

[0041] Understandably, the process simultaneously performs summary-outline matching detection: key semantic nodes of each summary (such as "EU MiCA regulatory requirements") are mapped to the corresponding chapters of the outline ("regulatory comparison" node). When a key dimension is found to be missing (such as the "governance recommendations" node not being supported by a summary), a supplementary instruction is immediately sent to the search agent. The instruction content integrates the original research topic and the missing dimension identifier (generating a new search term: "blockchain regulatory sandbox implementation case").

[0042] It should be noted that during the text generation stage, the agent does not simply copy the summary content: for the first occurrence of technical terms (such as "zero-knowledge proof"), it automatically inserts parenthetical comments to explain the basic principles (such as "cryptographic protocol where the verifier does not know the specific data"); and adds logical connectors at the transition between chapters (such as "in view of the above regulatory differences" connecting the chapter on regulatory comparison and governance recommendations).

[0043] In this embodiment, the output stage adopts a real-time segmented streaming mechanism: when the writing agent completes any chapter (such as "Technical Conflict Analysis"), the content generation monitoring module immediately converts it into an event stream data packet (including chapter ID, MD format text, and completion timestamp), and transmits it to the front end in batches through the server push event protocol. The user interface dynamically renders chapter cards and supports independent editing. For example, when generating the "New Energy Battery Recycling Technology" report, the user can annotate and revise the previously completed "Hydrometallurgical Process" chapter while the writing agent is still writing the "Economic Analysis" chapter, thus realizing a human-machine collaborative document production closed loop.

[0044] The above are embodiments of the method proposed in this application. Based on the same inventive concept, embodiments of this application also provide a document generation device based on multi-agent collaboration, the structure of which is as follows: Figure 2 As shown.

[0045] Figure 2 This is a schematic diagram of the internal structure of a document generation device based on multi-agent collaboration, provided as an embodiment of this application. Figure 2 As shown, the device includes: At least one processor 201; And a memory 202 that is communicatively connected to at least one processor; The memory 202 stores instructions executable by at least one processor, which are executed by at least one processor 201 to enable at least one processor 201 to: Natural language research topics are received from user input; Input the research topic into a pre-defined planning agent to generate a set of keywords for performing web searches based on the research topic; Input the keyword set into the preset search agent to perform several web searches using web search tools to obtain several key summary points; Input the research topic and several key points into the writing agent to generate a structured research report, and then output the structured research report to the user interface.

[0046] Some embodiments of this application provide corresponding to Figure 1 A non-volatile computer storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured as follows: Natural language research topics are received from user input; Input the research topic into a pre-defined planning agent to generate a set of keywords for performing web searches based on the research topic; Input the keyword set into the preset search agent to perform several web searches using web search tools to obtain several key summary points; Input the research topic and several key points into the writing agent to generate a structured research report, and then output the structured research report to the user interface.

[0047] The various embodiments in this application 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, the embodiments for IoT devices and media are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0048] The systems, media, and methods provided in this application are one-to-one correspondences. Therefore, the systems and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the systems and media will not be repeated here.

[0049] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0050] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0051] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0052] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0053] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0054] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0055] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0056] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further 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.

[0057] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A document generation method based on multi-agent collaboration, characterized in that, The method includes: Natural language research topics are received from user input; The research topic is input into a preset planning agent to generate a set of keywords for performing web searches based on the research topic; The keyword set is input into a preset search agent to perform several web searches using web search tools to obtain several key summary points. The research topic and the several key points of the abstract are input into the writing agent to generate a structured research report, and the structured research report is output to the user interface. Perform several web searches using web search tools to obtain several key-point summaries, specifically including: Based on the keywords in the keyword set, generate several keyword combinations; For each keyword combination, a web search tool is invoked to perform a search, resulting in several search results; The search results are compressed and refined to obtain the key summary; wherein the key summary is a text summary containing multiple paragraphs, containing only core information and no subjective comments; The content of the search results is compressed and refined to obtain the key summary, specifically including: Analyze the text content in the search results to identify key information related to the keyword combination; Remove duplicate statements and non-substantive descriptions from the search results. The retained core information is logically reorganized into text units containing multiple paragraphs; wherein, the text units retain only objective factual descriptions and do not contain subjective evaluation statements; Generate structured research reports, specifically including: Based on the research topic and the aforementioned key points summary, generate an outline for the report to be applied; Based on the aforementioned key-point summaries, the report outline is filled with main text content to generate a structured research report; Based on the research topic and the aforementioned key-point summaries, an outline for the application report is generated, specifically including: Analyze the research topic and the key points of the abstract, and based on the analysis results, plan the overall structural framework and writing logic of the structured research report. Based on the overall structural framework and writing logic of the structured research report, construct the outline of the report to be applied; After constructing the outline of the report to be applied based on the overall structural framework and writing logic of the structured research report, the method further includes: The matching degree of the aforementioned key point summaries is compared with the node content of the report outline to be applied; When a missing node in the summary that does not cover the core dimensions of the research topic is identified, a supplementary search instruction is sent to the search agent; Based on the supplementary search instructions, a search is performed using the web search tool to obtain a summary of the key points to be supplemented.

2. The document generation method based on multi-agent collaboration according to claim 1, characterized in that, Generate a set of keywords for performing web searches based on the research topic, specifically including: Using natural language processing technology, the grammatical structure and semantic relationships of the research topic are analyzed, and core conceptual entities are extracted based on the analysis results; The core concept entities are combined to generate several keywords, thereby obtaining the keyword set.

3. The document generation method based on multi-agent collaboration according to claim 1, characterized in that, Outputting the structured research report to the user interface specifically includes: During the process of generating a structured research report by the writing agent, the generation status of each chapter of the report is monitored in real time; When a new complete chapter is detected to be generated, the chapter content is immediately converted into a streaming data transmission format; The converted chapter data is pushed to the user interface in segments for dynamic rendering using a server push event protocol.

4. A document generation device based on multi-agent collaboration, characterized in that, The device includes: At least one processor; And, a memory communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform a document generation method based on multi-agent collaboration as described in any one of claims 1-3.

5. A non-volatile computer storage medium storing computer-executable instructions, characterized in that, When the computer-executable instructions are executed, they implement a document generation method based on multi-agent collaboration as described in any one of claims 1-3.

Images