Document generation system

Abstract

A document generation system for generating an output document describing a physical apparatus includes a data structure configured to store information about physical components of the physical apparatus in a standardized format using a set of terms, each term having a term name and a term number. The system also includes a text processing module configured to generate the output document using a large language model based on the data structure and set of terms. The output document is used to describe the physical apparatus in natural language.

Description

[0001] Document Generation System

[0002] Field

[0003] The technology relates to the field of automated document generation, specifically focusing on generating descriptive documents for physical apparatuses using standardized terms and data structures. This field involves natural language processing, machine learning, and data management techniques to create accurate and comprehensive documents, such as patent documents, technical specifications, user manuals, and other related materials.

[0004] In various industries, it is essential to create documents that accurately describe physical apparatuses, such as machines, devices, or systems. These documents can include patent applications, technical specifications, user manuals, design documents, manufacturing instructions, maintenance guides, assembly instructions, product descriptions, regulatory compliance documentation, training materials, and marketing materials. The quality and accuracy of these documents are crucial for effective communication, collaboration, and compliance with industry standards and regulations.

[0005] However, creating such documents can be a complex and time-consuming process, often involving multiple stakeholders, such as engineers, designers, technical writers, and legal professionals. One of the challenges in creating these documents is ensuring consistency in the description of the physical components of the apparatus, their attributes, and their relationships with other components. Inconsistencies in component labeling or naming can lead to confusion, misinterpretation of information, and potential errors in the generated output.

[0006] In addition, the increasing use of artificial intelligence (Al) models, such as large language models (LLMs) or natural language processing (NLP) engines, for document analysis and generation has highlighted the need for consistent and standardized component descriptions. Inaccurate or inconsistent component descriptions can result in incorrect analysis by Al models, leading to potential errors in the generated output and decreased output quality.

[0007] The lack of standardized naming conventions for components within a document can lead to different labels or names being used for the same component. This inconsistency can cause confusion and misinterpretation of the information, ultimately affecting the Al model's analysis and output quality. Furthermore, manual efforts to maintain consistency in component descriptions can be labor-intensive and prone to human error.

[0008] Therefore, there is a need for an improved method and system for generating documents describing physical apparatuses that can address the issues of inconsistent component labeling, inaccurate document analysis, and decreased output quality.

[0009] Summary

[0010] According to a first aspect of the disclosure, a document generation system is provided for generating an output document describing a physical apparatus. The system includes a data structure configured to store information about physical components of the physical apparatus in a standardized format using a set of terms. Each term has a term name and a term number. The system also includes a text processing module configured to generate the output document using a large language model based on the data structure and set of terms. This aspect of the disclosure provides the advantage of creating a standardized and structured way of describing a physical apparatus, which can be easily understood and interpreted by a large language model to generate a comprehensive output document.

[0011] Optionally in some examples, the document generation system further includes a register of terms configured to provide the set of terms for use in the data structure. This provides the advantage of having a centralized and standardized set of terms that can be used across different data structures, ensuring consistency and uniformity in the description of various physical apparatuses. Optionally in some examples, each term in the register of terms further includes a term description providing a detailed description of the term. This provides the advantage of giving more context and clarity to each term, making it easier for users to understand and use the terms correctly.

[0012] Optionally in some examples, the data structure includes a tree structure representing a hierarchical organization of the terms, attributes of the terms, and relationships between different terms. This provides the advantage of visualizing the relationships and dependencies between different terms and attributes, making it easier to understand and navigate the data structure.

[0013] Optionally in some examples, the tree structure includes tree nodes configured to contain the terms, attributes of the terms, and relationships between different terms. This provides the advantage of encapsulating related information in a single node, making the data structure more organized and easier to navigate.

[0014] Optionally in some examples, the document generation system further includes a user interface configured to allow a user to edit the data structure and the set of terms. This provides the advantage of giving users the flexibility to customize the data structure and terms according to their specific needs, making the system more adaptable and user-friendly.

[0015] Optionally in some examples, the user interface provides a visual interface for the user to add, modify, or delete tree nodes within the data structure, specify relationships between different tree nodes within the data structure, add descriptions or additional attributes to the tree nodes within the data structure, or organize and structure the tree nodes within the data structure through drag-and-drop or hierarchical tools. This provides the advantage of making the user interface more intuitive and easy to use, improving user experience and productivity.

[0016] Optionally in some examples, the user interface is configured to facilitate the user in searching and selecting existing terms from the register of terms to assign to the components within the data structure. This provides the advantage of making it easier for users to find and use the correct terms, ensuring accuracy and consistency in the description of the physical apparatus.

[0017] Optionally in some examples, the document generation system further includes a document augmentor configured to automatically generate new tree nodes within the tree structure describing the physical apparatus. This provides the advantage of automating the process of creating new nodes, saving time and effort for the user and ensuring a comprehensive description of the physical apparatus.

[0018] Optionally in some examples, the document augmentor is configured to automatically generate new tree nodes comprising terms describing the physical components of the physical apparatus, including their names, attributes, and relationships with other components. This provides the advantage of creating a detailed and structured description of each physical component, making it easier for the large language model to generate a comprehensive output document.

[0019] Optionally in some examples, the output document is selected from the group consisting of a disclosure document, a technical specification document, a user manual, a design document, a manufacturing document, a maintenance guide, assembly instructions, a product description, documentation for regulatory compliance, training materials, and marketing materials. This provides the advantage of versatility, as the system can generate a wide range of documents to suit different needs, from technical specifications for engineers to user manuals for end users.

[0020] Optionally in some examples, the large language model is connected to a remote server or operated within the text processing module. This provides the advantage of flexibility in the deployment of the large language model, allowing it to be used in different environments and scenarios, from cloud-based applications to standalone systems.

[0021] Optionally in some examples, the document augmentor utilizes a rule-based system or machine learning model to automatically generate new tree nodes. Optionally in some examples, the LLM generates text by processing the terms and their hierarchical relationships within the tree structure using template-based prompting or context-aware prompting.

[0022] Optionally in some examples, the user interface provides a drag-and-drop interface for manipulating tree nodes within the tree structure.

[0023] Optionally in some examples, the system integrates with on or more external systems via an API using a standardized data format.

[0024] Brief Description of the Drawings

[0025] Examples are described in more detail below with reference to the appended drawings. Figure 1 is a block diagram illustrating the overall architecture of the document generation system 100.

[0026] Figure 2 is a schematic representation of the data structure 10 and its components, including the tree structure 12 and register of terms 20.

[0027] Figure 3 is a flowchart depicting the process of generating an output document 54 using the text processing module 50 and large language model (LLM) 52.

[0028] Figure 4 is a user interface 30 for editing the data structure 10 and interacting with the document generation system 100.

[0029] Figure 5 is an example of a physical apparatus 200 with its physical components 210 and methods of operation 220.

[0030] Figure 6 is a detailed view of the terms 22 within the register of terms 20, including term name 24, term number 26, and term description 28.

[0031] Detailed Description

[0032] The detailed description set forth below provides information and examples of the disclosed technology with sufficient detail to enable those skilled in the art to practice the disclosure.

[0033] Figure 1 shows a block diagram illustrating the overall architecture of the document generation system 100. The system includes a data structure 10, a register of terms 20, a user interface 30, a document augmentor 40, and a text processing module 50. The data structure 10 is configured to store information about physical components 210 of a physical apparatus 200 in a standardized format using a set of terms 22. The text processing module 50 generates an output document 54 using a large language model LLM 52 based on the data structure 10 and the set of terms 22.

[0034] Figure 2 is a schematic representation of the data structure 10 and its components, including the tree structure 12 and register of terms 20. The tree structure 12 represents a hierarchical organization of the terms 22, attributes of the terms 22, and relationships between different terms 22. The register of terms 20 provides a standardized set of terms 22 for use in the data structure 10.

[0035] Figure 3 is a flowchart depicting the process of generating an output document 54 using the text processing module 50 and large language model (LLM) 52. The process begins with the data structure 10 and the set of terms 22, which are used as input for the LLM 52. The LLM 52 generates the output document 54 based on the information provided by the data structure 10 and the set of terms 22.

[0036] Figure 4 is a user interface 30 for editing the data structure 10 and interacting with the document generation system 100. The user interface 30 allows a user to add, modify, or delete tree nodes 14 within the data structure 10, specify relationships between different tree nodes 14 within the data structure 10, add descriptions or additional attributes to the tree nodes 14 within the data structure 10, and organize and structure the tree nodes 14 within the data structure 10 through drag-and-drop or hierarchical tools.

[0037] Figure 5 is an example of a physical apparatus 200 with its physical components 210 and methods of operation 220. The physical components 210 operate as individual elements or parts within the physical apparatus 200, while the methods of operation 220 describe the various procedures or processes by which the physical apparatus 200 operates.

[0038] Figure 6 is a detailed view of the terms 22 within the register of terms 20, including term name 24, term number 26, and term description 28. The term name 24 is a standardized name or label assigned to a specific physical component 210 or method of operation 220. The term number 26 is used to uniquely identify and reference a specific term within the register of terms 20 and data structure 10. The term description 28 provides a more detailed description of the terms 22.

[0039] 1 . Document Generation System Details

[0040] The document generation system 100 is a comprehensive solution designed to facilitate the creation of detailed, standardized documents describing a physical apparatus 200. The system 100 leverages a unique combination of data structures, standardized terms, and advanced text processing capabilities to generate output documents 54 that accurately and comprehensively describe the physical apparatus 200. The system 100 is designed to be user-friendly, allowing users to easily input, edit, and manage information about the physical apparatus 200. The system 100 also includes features that automate certain aspects of the document generation process, reducing the time and effort required to produce high-quality output documents 54.

[0041] 1.1. Data Structure

[0042] The data structure 10 is a key component of the document generation system 100. In some examples, the data structure 10 is configured to store information about the physical components 210 of the physical apparatus 200 in a standardized format using a set of terms 22. The data structure 10 provides a structured and organized way to store and manage information about the physical apparatus 200, facilitating efficient and accurate document generation. The data structure 10 may comprise various components and features that support the storage and management of information about the physical apparatus 200.

[0043] 1 .1 .1 . Tree Structure

[0044] The tree structure 12 is an integral part of the data structure 10. In some examples, the tree structure 12 represents a hierarchical organization of the terms 22, attributes of the terms 22, and relationships between different terms 22. The tree structure 12 provides a visual and logical representation of the information stored in the data structure 10, making it easier for users to understand and navigate the information. The tree structure 12 may comprise various components and features that support the hierarchical organization of the terms 22 and other information about the physical apparatus 200.

[0045] 1 .1 .1 .1 . Tree Nodes

[0046] The tree nodes 14 are fundamental elements of the tree structure 12. In some examples, the tree nodes 14 are configured to contain the terms 22, attributes of the terms 22, and relationships between different terms 22. Each tree node 14 represents a specific piece of information about the physical apparatus 200, such as a physical component 210 or a method of operation 220. The tree nodes 14 provide a flexible and scalable way to store and manage information about the physical apparatus 200, allowing for easy addition, modification, or deletion of information as needed.

[0047] 1 .1 .1 .1 .1 . Containing Terms

[0048] In some examples, the tree nodes 14 contain the terms 22. The terms 22 provide standardized names or labels for the physical components 210, methods of operation 220, or other attributes of the physical apparatus 200. By storing the terms 22 in the tree nodes 14, the system 100 ensures that the information about the physical apparatus 200 is represented in a consistent and standardized manner across all output documents 54. This enhances the readability and comprehensibility of the output documents 54, making them more useful and effective for their intended audiences.

[0049] 1 .1 .1 .1 .2. Containing Attributes Of terms

[0050] In some examples, the tree nodes 14 contain attributes of the terms 22. The attributes of the terms 22 provide additional details or specifications about the physical components 210, methods of operation 220, or other aspects of the physical apparatus 200. By storing the attributes of the terms 22 in the tree nodes 14, the system 100 allows for a more detailed and nuanced representation of the physical apparatus 200 in the output documents 54. This enhances the accuracy and completeness of the output documents 54, ensuring that they provide a comprehensive and detailed description of the physical apparatus 200.

[0051] 1.1.1.1.3. Containing Relationships Between Different Terms

[0052] In some examples, the tree nodes 14 contain relationships between different terms 22. The relationships between different terms 22 provide information about how the physical components 210, methods of operation 220, or other aspects of the physical apparatus 200 interact or relate to each other. By storing the relationships between different terms 22 in the tree nodes 14, the system 100 allows for a more interconnected and holistic representation of the physical apparatus 200 in the output documents 54. This enhances the coherence and comprehensibility of the output documents 54, ensuring that they provide a clear and understandable description of the physical apparatus 200.

[0053] 1 .2. Register Of Terms

[0054] The register of terms 20 is another key component of the document generation system 100. In some examples, the register of terms 20 is configured to provide a standardized set of terms 22 for use in the data structure 10. The register of terms 20 serves as a central repository for the terms 22, ensuring that they are consistently used across all output documents 54. The register of terms 20 may comprise various components and features that support the storage, management, and retrieval of the terms 22.

[0055] 1 .2.1 . Terms

[0056] The terms 22 are fundamental elements of the register of terms 20. In some examples, the terms 22 provide standardized names or labels for the physical components 210, methods of operation 220, or other attributes of the physical apparatus 200. Each term 22 represents a specific piece of information about the physical apparatus 200, such as a physical component 210 or a method of operation 220. The terms 22 provide a consistent and standardized way to represent information about the physical apparatus 200, enhancing the readability and comprehensibility of the output documents 54. 1 .2.1 .1 . Term Name

[0057] In some examples, each term 22 includes a term name 24. The term name 24 is a standardized name or label assigned to a specific physical component 210 or method of operation 220. The term name 24 provides a clear and concise way to refer to the physical components 210, methods of operation 220, or other attributes of the physical apparatus 200. By using standardized term names 24, the system 100 ensures that the information about the physical apparatus 200 is represented in a consistent and understandable manner across all output documents 54.

[0058] 1 .2.1 .2. Term Number

[0059] In some examples, each term 22 includes a term number 26. The term number 26 is used to uniquely identify and reference a specific term within the register of terms 20 and data structure 10. The term number 26 provides a reliable and efficient way to reference the terms 22, facilitating quick and accurate retrieval of information about the physical apparatus 200. By using unique term numbers 26, the system 100 ensures that the terms 22 can be accurately referenced and retrieved, enhancing the efficiency and accuracy of the document generation process.

[0060] 1.2.1.3. Term Description

[0061] In some examples, each term 22 includes a term description 28. The term description 28 provides a more detailed description of the term 22, offering additional information about the physical component 210, method of operation 220, or other attribute of the physical apparatus 200 that the term 22 represents. The term description 28 enhances the comprehensibility of the terms 22, providing users with a deeper understanding of the physical apparatus 200. By including detailed term descriptions 28, the system 100 ensures that the output documents 54 provide a comprehensive and detailed description of the physical apparatus 200.

[0062] 1 .3. User Interface The user interface 30 is a component of the document generation system 100. In some examples, the user interface 30 is configured to allow a user to edit the data structure 10 and the set of terms 22. The user interface 30 provides a visual and interactive platform for users to input, edit, and manage information about the physical apparatus 200. The user interface 30 is designed to be user-friendly and intuitive, making it easy for users to interact with the system 100 and generate high-quality output documents 54. The user interface 30 may comprise various features and tools that support the editing and management of the data structure 10 and the set of terms 22.

[0063] 1.4. Text Processing Module

[0064] The text processing module 50 is a key component of the document generation system 100. In some examples, the text processing module 50 is configured to generate the output document 54 using a large language model (LLM) 52 based on the data structure 10 and the set of terms 22. The text processing module 50 leverages advanced text processing capabilities to generate output documents 54 that accurately and comprehensively describe the physical apparatus 200. The text processing module 50 may comprise various components and features that support the generation of output documents 54.

[0065] 1.4.1. Large Language Model (LLM)

[0066] The large language model (LLM) 52 is a component of the text processing module 50. In some examples, the LLM 52 is used to generate the output document 54 based on the information provided by the data structure 10 and the set of terms 22. The LLM 52 leverages advanced machine learning techniques to generate natural language text that accurately and comprehensively describes the physical apparatus 200. By using the LLM 52, the system 100 can generate high-quality output documents 54 that are readable, comprehensible, and accurate, enhancing the effectiveness and utility of the output documents 54. Alternatives to a large language model (LLM) 52 may altenatively used, including more modern language processing methods.

[0067] 1.4.2. Output Document The output document 54 is the final product of the document generation system 100. In some examples, the output document 54 is used to describe the physical apparatus 200 in natural language. The output document 54 is generated by the text processing module 50 using the LLM 52 based on the information stored in the data structure 10 and the set of terms 22. The output document 54 provides a detailed, comprehensive, and standardized description of the physical apparatus 200, making it a valuable resource for various purposes, such as patent documentation, technical specifications, user manuals, design documents, manufacturing documents, maintenance guides, assembly instructions, product descriptions, documentation for regulatory compliance, training materials, and marketing materials.

[0068] 1.5. Document Augmentor

[0069] The document augmentor 40 is an optional feature of the document generation system 100. In some examples, the document augmentor 40 is configured to automatically generate new tree nodes 14 within the tree structure 12 describing the physical apparatus 200. The document augmentor 40 leverages advanced algorithms and techniques to automatically generate detailed and accurate descriptions of the physical apparatus 200, reducing the time and effort required to create the data structure 10. By using the document augmentor 40, the system 100 can automatically generate new tree nodes 14 comprising terms 22 describing the physical components 210 of the physical apparatus 200, including their names, attributes, and relationships with other components. This enhances the comprehensiveness and accuracy of the data structure 10, ensuring that the output documents 54 provide a detailed and comprehensive description of the physical apparatus 200. The document augmentor 40 can be controlled by the user interface 30, allowing users to easily generate new tree nodes 14 as needed.

[0070] 2. Physical Apparatus Details

[0071] The physical apparatus 200 is the subject of the output document 54 generated by the document generation system 100. In some examples, the physical apparatus 200 comprises various physical components 210 and methods of operation 220. The physical apparatus 200 can be any type of physical device or system, such as a machine, a device, a vehicle, a building, a structure, a system, or any other type of physical entity. The physical apparatus 200 is described in detail in the output document 54, with the information about the physical apparatus 200 being stored and managed in the data structure 10 using the set of terms 22.

[0072] 2.1. Physical Components

[0073] The physical components 210 are integral parts of the physical apparatus 200. In some examples, the physical components 210 operate as individual elements or parts within the physical apparatus 200. Each physical component 210 is represented by a specific term 22 in the data structure 10, with the term 22 providing a standardized name or label for the physical component 210. The physical components 210 can include various types of elements or parts, such as mechanical parts, electrical parts, electronic parts, structural parts, functional parts, or any other type of physical elements or parts. The physical components 210 are described in detail in the output document 54, with the information about the physical components 210 being stored and managed in the data structure 10 using the set of terms 22.

[0074] 2.2. Methods Of Operation

[0075] The methods of operation 220 describe the various procedures or processes by which the physical apparatus 200 operates. In some examples, each method of operation 220 is represented by a specific term 22 in the data structure 10, with the term 22 providing a standardized name or label for the method of operation 220. The methods of operation 220 can include various types of procedures or processes, such as operational procedures, functional processes, control processes, maintenance procedures, or any other type of procedures or processes related to the operation of the physical apparatus 200. The methods of operation 220 are described in detail in the output document 54, with the information about the methods of operation 220 being stored and managed in the data structure 10 using the set of terms 22.

[0076] 3. Operational Process The operational process describes the various steps and procedures involved in the operation of the document generation system 100. In some examples, the operational process includes various stages, such as the creation and management of the data structure 10, the input of information about the physical apparatus 200, the standardization of information using the set of terms 22, the generation of the output document 54 using the text processing module 50 and the LLM 52, and the automatic generation of new tree nodes 14 by the document augmentor 40. The operational process is designed to be efficient and user-friendly, allowing users to easily operate the document generation system 100 and generate high-quality output documents 54.

[0077] 3.1. Data Structure Creation and Management

[0078] The creation and management of the data structure 10 is a stage in the operational process. In some examples, the data structure 10 is created and managed using the user interface 30, with the user inputting information about the physical apparatus 200 into the data structure 10 using the set of terms 22. The data structure 10 is designed to be flexible and scalable, allowing for easy addition, modification, or deletion of information as needed. The data structure 10 is also designed to support the hierarchical organization of the terms 22 and other information about the physical apparatus 200, making it easier for users to understand and navigate the information.

[0079] 3.1 .1 . Input of Physical Apparatus Information

[0080] The input of information about the physical apparatus 200 into the data structure 10 is a key step in the operational process. In some examples, the user inputs information about the physical components 210 and methods of operation 220 of the physical apparatus 200 into the data structure 10 using the set of terms 22. The user interface 30 provides a visual and interactive platform for the user to input the information, making it easy for the user to input accurate and detailed information about the physical apparatus 200.

[0081] 3.1.2. Standardization of Information Using Terms The standardization of information using the set of terms 22 is another important step in the operational process. In some examples, the information about the physical apparatus 200 is standardized using the set of terms 22, with each term 22 providing a standardized name or label for a specific physical component 210 or method of operation 220. The standardization of information using the set of terms 22 ensures that the information about the physical apparatus 200 is represented in a consistent and standardized manner across all output documents 54, enhancing the readability and comprehensibility of the output documents 54.

[0082] 3.1.3. Hierarchical Organization of Terms in Tree Structure

[0083] The hierarchical organization of terms 22 in the tree structure 12 is a step in the operational process. In some examples, the terms 22, attributes of the terms 22, and relationships between different terms 22 are organized hierarchically in the tree structure 12. This hierarchical organization provides a visual and logical representation of the information stored in the data structure 10, making it easier for users to understand and navigate the information. The tree structure 12 supports parent-child relationships between the terms 22, allowing for a more interconnected and holistic representation of the physical apparatus 200.

[0084] 3.2. User Interaction with the System

[0085] User interaction with the document generation system 100 is a key aspect of the operational process. In some examples, the user interacts with the system 100 through the user interface 30, which allows the user to edit the data structure 10 and the set of terms 22. The user interface 30 provides a visual and interactive platform for the user to input, edit, and manage information about the physical apparatus 200, making it easy for the user to operate the system 100 and generate high-quality output documents 54.

[0086] 3.2.1 . Editing of Data Structure via User Interface

[0087] The editing of the data structure 10 via the user interface 30 is a step in the operational process. In some examples, the user interface 30 allows the user to add, modify, or delete tree nodes 14 within the data structure 10, specify relationships between different tree nodes 14 within the data structure 10, add descriptions or additional attributes to the tree nodes 14 within the data structure 10, and organize and structure the tree nodes 14 within the data structure 10 through drag-and-drop or hierarchical tools. This provides the user with a high degree of control over the information stored in the data structure 10, allowing the user to tailor the data structure 10 to the specific needs and requirements of the physical apparatus 200.

[0088] 3.2.2. Modification of Terms in Register of Terms

[0089] The modification of terms 22 in the register of terms 20 is another important step in the operational process. In some examples, the user interface 30 allows the user to add new terms 22 to the register of terms 20, modify the term name 24, term number 26, or term description 28 of existing terms 22, and organize and structure the terms 22 within the register of terms 20 through drag-and-drop or hierarchical tools. This provides the user with a high degree of control over the set of terms 22, allowing the user to tailor the set of terms 22 to the specific needs and requirements of the physical apparatus 200.

[0090] 3.3. Document Generation Process

[0091] The document generation process is the final stage in the operational process. In some examples, the text processing module 50 generates the output document 54 using the LLM 52 based on the information stored in the data structure 10 and the set of terms 22. The document generation process leverages advanced text processing capabilities to generate output documents 54 that accurately and comprehensively describe the physical apparatus 200.

[0092] 3.3.1. Utilization of Large Language Model

[0093] The utilization of the LLM 52 is a key step in the document generation process. In some examples, the LLM 52 generates the output document 54 based on the information provided by the data structure 10 and the set of terms 22. The LLM 52 leverages advanced machine learning techniques to generate natural language text that accurately and comprehensively describes the physical apparatus 200. By using the LLM 52, the system 100 can generate high-quality output documents 54 that are readable, comprehensible, and accurate.

[0094] 3.3.2. Generation of Output Document Based on Data Structure and Terms

[0095] The generation of the output document 54 based on the data structure 10 and the set of terms 22 is the final step in the document generation process. In some examples, the text processing module 50 generates the output document 54 using the LLM 52 based on the information stored in the data structure 10 and the set of terms 22. The output document 54 provides a detailed, comprehensive, and standardized description of the physical apparatus 200, making it a valuable resource for various purposes.

[0096] 3.4. Automatic Augmentation of Data Structure

[0097] The automatic augmentation of the data structure 10 is an optional feature of the operational process. In some examples, the document augmentor 40 automatically generates new tree nodes 14 within the tree structure 12 describing the physical apparatus 200. This feature reduces the time and effort required to create the data structure 10, enhancing the efficiency of the document generation process. The document augmentor 40 can be controlled by the user interface 30, allowing users to easily generate new tree nodes 14 as needed.

[0098] 3.4.1 . Generation of New Tree Nodes by Document Augmentor

[0099] The generation of new tree nodes 14 by the document augmentor 40 is a key feature of the automatic augmentation of the data structure 10. In some examples, the document augmentor 40 automatically generates new tree nodes 14 comprising terms 22 describing the physical components 210 of the physical apparatus 200, including their names, attributes, and relationships with other components. This feature enhances the comprehensiveness and accuracy of the data structure 10, ensuring that the output documents 54 provide a detailed and comprehensive description of the physical apparatus 200. 4. Description of Examples of the Disclosure

[0100] The following sections provide examples of the disclosure, illustrating how the document generation system 100 can be used to generate output documents 54 describing a physical apparatus 200. These examples are provided for illustrative purposes and are not intended to limit the scope of the disclosure.

[0101] 4.1 . Example of Physical Apparatus Description

[0102] In one example, the document generation system 100 is used to generate an output document 54 describing a physical apparatus 200. The physical apparatus 200 comprises various physical components 210 and methods of operation 220, which are represented by specific terms 22 in the data structure 10.

[0103] 4.1.1. Detailed Description of Physical Components

[0104] In this example, the output document 54 provides a detailed description of the physical components 210 of the physical apparatus 200. Each physical component 210 is represented by a specific term 22 in the data structure 10, with the term 22 providing a standardized name or label for the physical component 210. The output document 54 includes detailed specifications for each physical component 210, including dimensions, materials, and functionalities.

[0105] 4.1.2. Detailed Description of Methods of Operation

[0106] In this example, the output document 54 also provides a detailed description of the methods of operation 220 of the physical apparatus 200. Each method of operation 220 is represented by a specific term 22 in the data structure 10, with the term 22 providing a standardized name or label for the method of operation 220. The output document 54 includes detailed descriptions of the various procedures or processes by which the physical apparatus 200 operates.

[0107] 4.2. Example of User Interaction with the System In another example, a user interacts with the document generation system 100 through the user interface 30. The user interface 30 allows the user to input, edit, and manage information about the physical apparatus 200, facilitating the creation and management of the data structure 10 and the set of terms 22.

[0108] 4.2.1 . Use of User Interface for Editing Data Structure

[0109] In this example, the user uses the user interface 30 to edit the data structure 10. The user interface 30 provides a visual and interactive platform for the user to add, modify, or delete tree nodes 14 within the data structure 10, specify relationships between different tree nodes 14 within the data structure 10, add descriptions or additional attributes to the tree nodes 14 within the data structure 10, and organize and structure the tree nodes 14 within the data structure 10 through drag-and-drop or hierarchical tools. This provides the user with a high degree of control over the information stored in the data structure 10, allowing the user to tailor the data structure 10 to the specific needs and requirements of the physical apparatus 200.

[0110] 4.2.2. Use of User Interface for Editing Terms

[0111] In this example, the user uses the user interface 30 to edit the set of terms 22 in the register of terms 20. The user interface 30 provides a visual and interactive platform for the user to add new terms 22 to the register of terms 20, modify the term name 24, term number 26, or term description 28 of existing terms 22, and organize and structure the terms 22 within the register of terms 20 through drag-and-drop or hierarchical tools. This provides the user with a high degree of control over the set of terms 22, allowing the user to tailor the set of terms 22 to the specific needs and requirements of the physical apparatus 200.

[0112] 4.3. Example of Document Generation

[0113] In another example, the document generation system 100 is used to generate an output document 54 describing a physical apparatus 200. The output document 54 is generated by the text processing module 50 using the LLM 52 based on the information stored in the data structure 10 and the set of terms 22. 4.3.1 . Use of Large Language Model for Document Generation

[0114] In this example, the LLM 52 is used to generate the output document 54 based on the information provided by the data structure 10 and the set of terms 22. The LLM 52 leverages advanced machine learning techniques to generate natural language text that accurately and comprehensively describes the physical apparatus 200. By using the LLM 52, the system 100 can generate high-quality output documents 54 that are readable, comprehensible, and accurate.

[0115] 4.3.2. Example of Generated Output Document

[0116] In this example, the output document 54 generated by the document generation system 100 provides a detailed, comprehensive, and standardized description of the physical apparatus 200. The output document 54 includes detailed descriptions of the physical components 210 and methods of operation 220 of the physical apparatus 200, represented by specific terms 22 in the data structure 10. The output document 54 also includes detailed specifications for each physical component 210, including dimensions, materials, and functionalities, as well as detailed descriptions of the various procedures or processes by which the physical apparatus 200 operates.

[0117] 5. Potential Applications

[0118] The document generation system 100 has potential applications in various fields and contexts, including patent documentation, technical specifications, user manuals, design documents, manufacturing documents, maintenance guides, assembly instructions, product descriptions, documentation for regulatory compliance, training materials, and marketing materials.

[0119] 5.1. Application in Patent Documentation

[0120] In one example, the document generation system 100 is used to generate detailed patent documents. The system 100 leverages the data structure 10, the set of terms 22, and the LLM 52 to generate patent documents that accurately and comprehensively describe a physical apparatus 200. The patent documents provide a detailed, comprehensive, and standardized description of the physical apparatus 200, including detailed descriptions of the physical components 210 and methods of operation 220, represented by specific terms 22 in the data structure 10.

[0121] 5.1.1. Generation of Detailed Patent Documents

[0122] In this example, the document generation system 100 generates detailed patent documents that accurately and comprehensively describe a physical apparatus 200. The patent documents include detailed descriptions of the physical components 210 and methods of operation 220 of the physical apparatus 200, represented by specific terms 22 in the data structure 10. The patent documents also include detailed specifications for each physical component 210, including dimensions, materials, and functionalities, as well as detailed descriptions of the various procedures or processes by which the physical apparatus 200 operates. By using the document generation system 100, users can generate high-quality patent documents that are readable, comprehensible, and accurate, enhancing the effectiveness and utility of the patent documents.

[0123] 5.1 .2. Standardization of Patent Terminology

[0124] In this example, the document generation system 100 standardizes the terminology used in patent documents. The system 100 uses the set of terms 22 to provide standardized names or labels for the physical components 210, methods of operation 220, or other attributes of the physical apparatus 200. By using standardized terms 22, the system 100 ensures that the information about the physical apparatus 200 is represented in a consistent and understandable manner across all patent documents. This enhances the readability and comprehensibility of the patent documents, making them more useful and effective for their intended audiences.

[0125] 5.2. Application in Technical Documentation

[0126] In another example, the document generation system 100 is used to generate detailed technical specifications. The system 100 leverages the data structure 10, the set of terms 22, and the LLM 52 to generate technical specifications that accurately and comprehensively describe a physical apparatus 200. The technical specifications provide a detailed, comprehensive, and standardized description of the physical apparatus 200, including detailed descriptions of the physical components 210 and methods of operation 220, represented by specific terms 22 in the data structure 10.

[0127] 5.2.1. Generation of Detailed Technical Specifications

[0128] In this example, the document generation system 100 generates detailed technical specifications that accurately and comprehensively describe a physical apparatus 200. The technical specifications include detailed descriptions of the physical components 210 and methods of operation 220 of the physical apparatus 200, represented by specific terms 22 in the data structure 10. The technical specifications also include detailed specifications for each physical component 210, including dimensions, materials, and functionalities, as well as detailed descriptions of the various procedures or processes by which the physical apparatus 200 operates. By using the document generation system 100, users can generate high-quality technical specifications that are readable, comprehensible, and accurate, enhancing the effectiveness and utility of the technical specifications.

[0129] 5.2.2. Standardization of Technical Terminology

[0130] In this example, the document generation system 100 standardizes the terminology used in technical specifications. The system 100 uses the set of terms 22 to provide standardized names or labels for the physical components 210, methods of operation 220, or other attributes of the physical apparatus 200. By using standardized terms 22, the system 100 ensures that the information about the physical apparatus 200 is represented in a consistent and understandable manner across all technical specifications. This enhances the readability and comprehensibility of the technical specifications, making them more useful and effective for their intended audiences.

[0131] 5.3. Application in User Manuals In yet another example, the document generation system 100 is used to generate detailed user manuals. The system 100 leverages the data structure 10, the set of terms 22, and the LLM 52 to generate user manuals that accurately and comprehensively describe a physical apparatus 200. The user manuals provide a detailed, comprehensive, and standardized description of the physical apparatus 200, including detailed descriptions of the physical components 210 and methods of operation 220, represented by specific terms 22 in the data structure 10.

[0132] 5.3.1 . Generation of Detailed User Manuals

[0133] In this example, the document generation system 100 generates detailed user manuals that accurately and comprehensively describe a physical apparatus 200. The user manuals include detailed descriptions of the physical components 210 and methods of operation 220 of the physical apparatus 200, represented by specific terms 22 in the data structure 10. The user manuals also include detailed specifications for each physical component 210, including dimensions, materials, and functionalities, as well as detailed descriptions of the various procedures or processes by which the physical apparatus 200 operates. By using the document generation system 100, users can generate high-quality user manuals that are readable, comprehensible, and accurate, enhancing the effectiveness and utility of the user manuals.

[0134] 5.3.2. Standardization of User Manual Terminology

[0135] In this example, the document generation system 100 standardizes the terminology used in user manuals. The system 100 uses the set of terms 22 to provide standardized names or labels for the physical components 210, methods of operation 220, or other attributes of the physical apparatus 200. By using standardized terms 22, the system 100 ensures that the information about the physical apparatus 200 is represented in a consistent and understandable manner across all user manuals. This enhances the readability and comprehensibility of the user manuals, making them more useful and effective for their intended audiences. The terminology used herein is for the purpose of describing particular aspects only and is not intended to be limiting of the disclosure. As used herein, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items. It will be further understood that the terms "comprises," "comprising," "includes," and / or "including" when used herein specify the presence of stated features, integers, actions, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, actions, steps, operations, elements, components, and / or groups thereof.

[0136] It will be understood that, although the terms first, second, etc., may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element without departing from the scope of the present disclosure.

[0137] Relative terms such as "below" or "above" or "upper" or "lower" or "horizontal" or "vertical" may be used herein to describe a relationship of one element to another element as illustrated in the Figures. It will be understood that these terms and those discussed above are intended to encompass different orientations of the device in addition to the orientation depicted in the Figures. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element, or intervening elements may be present. In contrast, when an element is referred to as being "directly connected" or "directly coupled" to another element, there are no intervening elements present.

[0138] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. It will be further understood that terms used herein should be interpreted as having a meaning consistent with their meaning in the context of this specification and the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein. The document generation system 100 further enhances efficiency and accuracy by seamlessly integrating with external systems such as Computer-Aided Design (CAD) tools and Product Lifecycle Management (PLM) systems. This integration is achieved through a standardized API (Application Programming Interface) allowing bidirectional data exchange. Specifically, the system 100 utilizes a JSON (JavaScript Object Notation) data format to import component information, including names, attributes, and relationships, from CAD and PLM systems. A dedicated data mapping module within the document generation system 100 automatically maps imported data elements to corresponding terms 22 in the register of terms 20. This automated mapping process eliminates manual data entry, reducing document creation time and minimizing errors associated with manual data transfer. Furthermore, changes made to component descriptions within the document generation system 100 can be exported back to the external systems via the API, ensuring consistency across all platforms.

[0139] A specific example, of the document generation system operation will now be described. For example, consider a physical apparatus 200 comprising a 'housing' and a 'circuit board'. The housing 210 may have attributes such as 'material' (e.g., 'aluminum') and 'dimensions'. The circuit board 210 may contain components like a 'processor' and 'memory'. The relationship between these components would be represented within the tree structure 12. In a scenario where the circuit board is inconsistently labeled as 'PCB' in some parts of an input document, the system 100 automatically maps both 'circuit board' and 'PCB' to the same term 22 in the register of terms 20, ensuring consistent labeling in the generated output document 54. The LLM 52, when prompted to describe the apparatus 200, would use the standardized term 'circuit board' throughout the generated text, resolving the initial inconsistency.

[0140] The LLM 52 interacts with the data structure 10 and register of terms 20 through a structured prompting mechanism. The text processing module 50 constructs prompts that include the term names 24, their descriptions 28, and their relationships as defined in the tree structure 12. For instance, a prompt might be structured as: 'Describe the function of [term name 24] given its relationship to [related term name 24] and considering its attributes [attributes of term 22]'. This structured approach guides the LLM 52 to generate text that is both accurate and consistent with the information stored in the data structure 10.

[0141] The document generation system 100 is implemented on a computing device comprising a processor, memory, and storage. The processor executes the software modules of the system 100, including the user interface 30, the document augmentor 40, and the text processing module 50. The memory stores the data structure 10, the register of terms 20, and the LLM 52 during operation. The storage provides persistent storage for data and software components. The computing device may be a general- purpose computer, a server, or a specialized device configured to execute the document generation system 100. The LLM 52 may be stored locally on the storage or accessed remotely via a network connection.

[0142] Fallback Mechanisms and Error Handling

[0143] In some examples, the document generation system 100 may incorporate robust error handling and fallback mechanisms to ensure reliable operation. For example, if the LLM 52 encounters an ambiguous prompt or is unable to generate suitable text based on the provided data, the system 100 can provide feedback to the user through the user interface 30, highlighting the potential issue. In some embodiments, the system may suggest alternative terms 22 or prompt modifications to resolve the ambiguity. Alternatively, the system 100 may default to a rule-based text generation module as a fallback mechanism, ensuring that a document can still be generated even if the LLM 52 encounters limitations.

[0144] Security Considerations

[0145] In certain examples, the document generation system 100 may include security features to protect sensitive information. Access control mechanisms, such as user authentication and authorization, restrict access to the system and its data based on predefined user roles and permissions. Data encryption techniques protect the confidentiality of the data stored in the data structure 10 and register of terms 20, both in transit and at rest. These security measures ensure that confidential information related to the physical apparatus 200 is protected from unauthorized access and disclosure. User Customization and Preferences

[0146] In some examples, the document generation system 100 may allows for user customization to accommodate individual preferences and specific requirements. Users can personalize the user interface 30, such as adjusting the layout, font sizes, and color schemes. Users can also configure the document generation process by selecting preferred output formats, specifying document templates, and defining custom rules for text generation. This flexibility enables the system 100 to adapt to diverse user needs and generate documents tailored to specific applications.

[0147] Batch Processing and Automation

[0148] In some examples, the system 100 may supports batch processing to automate the generation of multiple documents. Users can specify a set of input data structures 10 and generate corresponding output documents 54 in an automated fashion. This feature is particularly useful for generating large volumes of documentation, such as a set of patent applications for related inventions or a collection of technical manuals for different product configurations. Batch processing significantly improves efficiency compared to generating each document individually.

[0149] Version Control and Document History

[0150] In some examples, the document generation system 100 may incorporate version control mechanisms to track changes to the data structure 10, register of terms 20, and generated documents 54. Each modification is logged with a timestamp and user identifier, creating a comprehensive history of revisions. This allows users to revert to previous versions of documents if necessary and facilitates collaboration among multiple users working on the same project. The version control system ensures data integrity and provides an audit trail for document generation.

[0151] The present invention provides significant advantages over conventional document generation systems by addressing the technical challenges inherent in integrating LLMs for consistent and reliable technical document creation. Specifically, the system's structured data representation, coupled with a controlled vocabulary (register of terms), provides essential constraints and guidance to the LLM, mitigating the risk of unpredictable or irrelevant outputs. This structured approach ensures that the generated text adheres to predefined terminology and accurately reflects the relationships between different components of the described physical apparatus.

[0152] Furthermore, the invention tackles the persistent problem of inconsistent labeling in technical documentation, a challenge that previous template-based or rule-based systems struggled to overcome. The system's innovative approach, combining the strengths of LLMs with a structured data framework, offers a solution for generating accurate, consistent, and high-quality technical documents. This integration, while complex, is a key differentiator, enabling the system to achieve levels of consistency and automation not readily attainable previously.

[0153] It is to be understood that the present disclosure is not limited to the aspects described above and illustrated in the drawings; rather, the skilled person will recognize that many changes and modifications may be made within the scope of the present disclosure and appended claims. In the drawings and specification, there have been disclosed aspects for purposes of illustration only and not for purposes of limitation, the scope of the disclosure being set forth in the following claims.

Claims

Claims1. A document generation system (100) for generating an output document (54) describing a physical apparatus (200), the system comprising: a data structure (10) configured to store information about physical components (210) of the physical apparatus (200) in a standardized format using a set of terms (22), each term (22) having a term name (24) and a term number (26); and a text processing module (50) configured to generate the output document (54) using a large language model (LLM) (52) based on the data structure (10) and set of terms (22).

2. The document generation system (100) according to claim 1 , further comprising a register of terms (20) configured to provide the set of terms (22) for use in the data structure (10).

3. The document generation system (100) according to claim 2, wherein each term (22) further comprises a term description (28) providing a detailed description of the term (22).

4. The document generation system (100) according to any one of claims 1 to 3, wherein the data structure (10) comprises a tree structure (12) representing a hierarchical organization of the terms (22), attributes of the terms (22), and relationships between different terms (22).

5. The document generation system (100) according to claim 4, wherein the tree structure (12) comprises tree nodes (14) configured to contain the terms (22), attributes of the terms (22), and relationships between different terms (22).

6. The document generation system (100) according to any one of claims 1 to 5, further comprising a user interface (30) configured to allow a user to edit the data structure (10) and the set of terms (22).

7. The document generation system (100) according to claim 6, wherein the user interface (30) is configured to provide a visual interface for the user to add, modify, ordelete tree nodes (14) within the data structure (10), specify relationships between different tree nodes (14) within the data structure (10), add descriptions or additional attributes to the tree nodes (14) within the data structure (10), or organize and structure the tree nodes (14) within the data structure (10) through drag-and-drop or hierarchical tools.

8. The document generation system (100) according to claim 6 or claim 7, wherein the user interface (30) is configured to facilitate the user in searching and selecting existing terms (22) from the register of terms (20) to assign to the components within the data structure (10).

9. The document generation system (100) according to any one of claims 1 to 8, further comprising a document augmentor (40) configured to automatically generate new tree nodes (14) within the tree structure (12) describing the physical apparatus (200).

10. The document generation system (100) according to claim 9, wherein the document augmentor (40) is configured to automatically generate new tree nodes (14) comprising terms (22) describing the physical components (210) of the physical apparatus (200), including their names, attributes, and relationships with other components.

11. The document generation system (100) according to any one of claims 1 to 10, wherein the output document (54) is selected from the group consisting of a patent document, a technical specification document, a user manual, a design document, a manufacturing document, a maintenance guide, assembly instructions, a product description, documentation for regulatory compliance, training materials, and marketing materials.

12. The document generation system (100) according to any one of claims 1 to 11 , wherein the large language model (LLM) (52) is connected to a remote server or operated within the text processing module (50).

13. The document generation system (100) according to any of claims 9 to 12, wherein the document augmentor (40) utilizes a rule-based system or machine learning model to automatically generate new tree nodes (14).

14. The document generation system (100) according to any of claims 1 to 12, wherein the LLM (52) generates text by processing the terms (22) and their hierarchical relationships within the tree structure (12) using template-based prompting or context- aware prompting.

15. The document generation system (100) according to any of claims 6 to 13, wherein the user interface (30) provides a drag-and-drop interface for manipulating tree nodes (14) within the tree structure (12).

16. The document generation system (100) according to any of claims 6 to 13, wherein the system (100) integrates with on or more external systems via an API (60) using a standardized data format.

Images