system

Abstract

Provide a system. 【Solution means】 Means for collecting existing documents and converting them into text, Means for analyzing text data, classifying and tagging it, Means for detecting and integrating duplicate content, Means for generating a unified document, Means for collecting business result data and matching it with the document content, Means for automatically updating the document, Means for verifying user authentication and permissions, Means for allowing users to provide feedback, A system including means for revising the document based on the feedback.

Description

Technical Field

[0001] The technology of the present disclosure relates to a system.

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, the method including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a character of the chatbot, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Currently, many companies and organizations have the following problems regarding the management of business manuals. That is, problems such as "no manual", "in disarray", "not updated and it is unclear whether it is still valid", and "duplicate content". These problems are factors that reduce the efficiency of business and prevent accurate information sharing. Also, security may not be ensured in the handling of in-house confidential information. There is a need for a new manual management system to solve these problems.

Means for Solving the Problems

[0005] The present invention solves the aforementioned problems with a system that includes means for collecting existing documents and converting them into text, means for analyzing, classifying, and tagging text data, means for detecting and integrating duplicate content, means for generating a unified document, means for collecting business result data and comparing it with document content, means for automatically updating documents, means for verifying user authentication and permissions, means for users to provide feedback, and means for revising documents based on feedback.

[0006] Specifically, existing documents are converted into text using OCR (Optical Character Recognition) and PDF text extraction tools, and the text data is analyzed using natural language processing to classify and tag each section. This detects duplicate content and generates unified documents. In addition, business result data is collected periodically and compared with the document content to automatically update the documents. Furthermore, user authentication and permissions are verified, and documents are provided to users with appropriate permissions. Users can provide feedback, and documents can be revised based on that feedback. This system enables the centralization of business manuals, maintenance of up-to-date information, and secure handling of confidential information.

[0007] "Existing documents" refer to various manuals, guides, and other documents that have already been created within a company or organization.

[0008] "Collection" refers to the act of gathering electronic files from a specific folder or database.

[0009] "Converting to text" refers to the act of converting electronic files such as PDFs, Word documents, and Excel spreadsheets into text data using optical character recognition (OCR) or other text extraction tools.

[0010] "Text data" refers to the content of a document that has been converted into text.

[0011] "Analysis" refers to the act of using natural language processing (NLP) to thoroughly examine the content of text data and understand its meaning and structure.

[0012] "Classification" refers to the act of dividing analyzed text data into specific categories or sections.

[0013] "Tagging" refers to the act of assigning appropriate keywords or labels to text data.

[0014] "Duplicate content" refers to sections of a document that contain the same or very similar information.

[0015] "Integration" refers to the act of combining overlapping content into one and eliminating duplication.

[0016] A "unified document" refers to a single, complete document that eliminates redundant content and is optimized for clarity.

[0017] "Business results data" refers to data that shows the results of a company's or organization's business activities, such as sales performance and project progress.

[0018] "Verification" refers to the act of comparing business result data with document content to check for matches and discrepancies.

[0019] "Automatic updates" refer to the process where a system automatically modifies a document to reflect new information or changes.

[0020] "Authentication" refers to the process of verifying a user's identity and establishing access rights.

[0021] "Permissions" refer to the range of information a user can access and the operations they can perform.

[0022] "Feedback" refers to the opinions and suggestions for improvement provided by users.

[0023] "Revision" refers to the act of modifying or updating the content of a document based on feedback.

[0024] The above are the definition sentences of important words included in the claims of the patent.

Brief Description of the Drawings

[0025] [Figure 1] It is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] It is a conceptual diagram showing an example of the main functions of a data processing device and a smart device according to the first embodiment. [Figure 3] It is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] It is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] It is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] It is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] It is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] It is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] It shows an emotion map to which multiple emotions are mapped. [Figure 10] It shows an emotion map to which multiple emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when the emotion engine is combined. [Figure 14] This is a sequence diagram showing the processing flow of the data processing system in Application Example 2, which combines an emotion engine. [Modes for carrying out the invention]

[0026] Hereinafter, an example of an embodiment of the system relating to the technology of this disclosure will be described with reference to the attached drawings.

[0027] First, let's explain the terminology used in the following explanation.

[0028] In the following embodiments, the signed processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Furthermore, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include CPU (Central Processing Unit), GPU (Graphics Processing Unit), GPGPU (General-Purpose computing on Graphics Processing Units), and APU (Accelerated Processing Unit).

[0029] In the following embodiments, signed RAM (Random Access Memory) is a memory that temporarily stores information and is used as work memory by the processor.

[0030] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0031] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0032] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0033] [First Embodiment]

[0034] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0035] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0036] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0037] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0038] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0039] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0040] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0041] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0042] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0043] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0044] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0045] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0046] ---

[0047] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. This system includes multiple programs running on a server and terminals used by users. The system also features a function to automatically update documents based on user feedback and business result data.

[0048] System Overview

[0049] 1. Document collection and text conversion

[0050] Server: Collects existing documents (PDF, Word, Excel, etc.) from various folders and databases within the company. The collected documents are converted into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are also converted to text using a dedicated library.

[0051] 2. Analysis and Classification of Text Data

[0052] Server: The converted text data is analyzed using natural language processing (NLP) techniques. During the analysis process, the document is categorized into sections and assigned appropriate tags. For example, tags such as "project management" and "sales strategy" may be added.

[0053] 3. Duplicate detection and unification

[0054] Server: Text analysis detects duplication when the same content is found in multiple documents. The duplicate content is combined into a single, unified document. This eliminates unnecessary duplication and generates a well-organized manual.

[0055] 4. Verification with business result data and automatic updating

[0056] Server: The server periodically collects business performance data (e.g., sales figures and project progress). The collected data is compared with document content, and if discrepancies are found, the document is automatically updated. This ensures that manuals always reflect the latest information.

[0057] 5. User Authentication and Permission Management

[0058] Server: Users attempting to access the manual must first undergo identity verification through an authentication system. Authentication technologies such as LDAP are used to verify user access rights. This ensures that only users with appropriate permissions can access specific information.

[0059] 6. Feedback function and document revision

[0060] Users: Users can view the manual and provide feedback. For example, they can report if a particular procedure is outdated. The feedback is sent to the server, analyzed, and necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent access.

[0061] Specific example

[0062] For example, consider a company with two documents: "Sales Manual.pdf" and "Project Management.docx". The server collects these documents and converts them to text. Next, it analyzes each document using natural language processing, classifying and tagging them into sections such as "Sales Policy" and "Project Management Methods". Duplicate parts are detected from the analysis results and reconstructed into a single, unified document. Subsequently, the latest sales performance data is collected and reflected in the manual, automatically updating its content. When users access the manual, authentication is performed, ensuring that only users with appropriate permissions can access the information. When user feedback is provided, the server analyzes it and revises the document.

[0063] This allows for centralized management of internal business manuals, ensuring they are always up-to-date, and guaranteeing the secure handling of confidential information.

[0064] ---

[0065] The following describes the processing flow.

[0066] Step 1:

[0067] The server collects existing documents (PDFs, Word documents, Excel files, etc.) from specific folders or databases within the company. The collection script is run periodically to retrieve the most recent documents.

[0068] Step 2:

[0069] The server converts collected documents into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are converted to text using a dedicated library.

[0070] Step 3:

[0071] The server analyzes the converted text data using natural language processing (NLP) techniques. It analyzes the document content, categorizes it into sections, and applies appropriate tags.

[0072] Step 4:

[0073] The server detects duplicate content based on the analysis results. It compares similar sections and paragraphs and integrates them into a single, unified text.

[0074] Step 5:

[0075] The server formats the integrated documents and automatically generates a standardized table of contents and index. The unified documents are then stored in a database.

[0076] Step 6:

[0077] The server periodically collects business result data (sales performance, project progress, etc.) from various business systems within the company via APIs.

[0078] Step 7:

[0079] The server compares the collected work result data with the contents of the manual. If a discrepancy is detected, it automatically performs the necessary updates and updates the document.

[0080] Step 8:

[0081] When a user accesses a manual, the server authenticates the user using an authentication system such as LDAP. It verifies the user's permissions and grants access to the corresponding document only to users with the appropriate privileges.

[0082] Step 9:

[0083] Users view the manual and report areas for improvement or deficiencies using a feedback form. The feedback data is sent to the server.

[0084] Step 10:

[0085] The server analyzes the collected feedback and revises the document. The revised document is saved back to the database and the changes are reflected the next time it is accessed.

[0086] These steps ensure that internal operational manuals are efficiently collected and analyzed, and provided in a unified format free of duplication. Furthermore, they remain up-to-date by reflecting the latest operational results, and are quickly revised based on user feedback.

[0087] (Example 1)

[0088] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0089] The diverse range of documents generated within a company increases over time, making their management complex. The existence of multiple documents can lead to information duplication and inconsistencies, hindering efficient work execution. Furthermore, manual document updates are labor-intensive and can delay the reflection of the latest information. In addition, a lack of proper user authentication and access control increases the risk of confidential information leaks. Addressing these challenges is essential.

[0090] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0091] In this invention, the server includes means for collecting existing documents and converting them into text, means for analyzing, classifying, and tagging the text data, and means for detecting and integrating duplicate content. This streamlines document management within the company, eliminates duplication and inconsistencies, and generates a unified manual that always reflects the latest information. Furthermore, user authentication and permission verification ensure the secure handling of confidential information. In addition, document revisions are automatically performed based on feedback provided by users, ensuring that the latest information is always reflected.

[0092] "Document collection methods" refer to methods for collecting existing documents from folders and databases within a company.

[0093] "Text conversion means" refers to means of converting collected documents into text data. This includes optical character recognition and document text extraction tools.

[0094] "Text analysis means" refers to means for analyzing the converted text data, and this uses natural language processing technology.

[0095] "Classification and tagging means" refers to means of classifying the analyzed text data into sections and assigning appropriate tags.

[0096] A "duplicate detection means" is a means of calculating the similarity between analyzed text data and detecting duplicate content.

[0097] A "combination method" is a means of integrating duplicate text data into a single, unified document.

[0098] "Methods for collecting business data" refer to methods for periodically collecting business result data from a company, and this includes the use of various APIs.

[0099] A "document verification means" is a means of verifying collected business data against the contents of documents.

[0100] An "automatic update mechanism" is a means of automatically updating the document content based on the verification results.

[0101] "User authentication means" refers to methods for verifying a user's identity, and this includes using LDAP or authentication protocols.

[0102] A "permission verification method" is a means of verifying a user's access rights and ensuring that only users with appropriate permissions can access the information.

[0103] A "feedback provision method" is a means by which users can provide feedback on a document.

[0104] "Document revision methods" refer to methods for revising documents based on user feedback and reflecting the latest information.

[0105] ---

[0106] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. This system includes multiple programs running on a server and terminals used by users. The system also features a function to automatically update documents based on user feedback and business result data.

[0107] System Overview

[0108] Document collection and text conversion

[0109] The server collects existing documents (PDF, Word, Excel, etc.) from various folders and databases within the company. This collection process utilizes remote file transfer methods, specifically SFTP (Secure File Transfer Protocol). The collected documents are converted into text data using optical character recognition (OCR) and document text extraction tools. ABBYY FineReader is used for OCR, and the PDF library is used for PDF text extraction. Word and Excel files are converted to text using the Apache® POI library.

[0110] Text data analysis and classification

[0111] The server analyzes the converted text data using natural language processing (NLP) techniques. Software such as Google® Cloud Natural Language and SpaCy are used for analysis, including text tokenization and morphological analysis. As a result of the analysis, the documents are categorized into sections and tagged with terms such as "Project Management" and "Sales Strategy." Machine learning techniques (e.g., k-means clustering) are used for classification.

[0112] Duplicate detection and unification

[0113] The server detects duplicates when multiple documents contain the same content based on the analysis results. Text similarity calculation methods (such as Jaccard similarity or Cosine similarity) are used for duplicate detection. Duplicate content is combined into a single, unified document. The most reliable text and the most recent content are prioritized for unification as needed.

[0114] Matching with business result data and automatic updating

[0115] The server periodically collects various business result data from the company (e.g., sales performance data and project progress data). This data collection often utilizes APIs, specifically the Salesforce API and other business system APIs. The collected data is compared with document content, and if discrepancies are found, the documents are automatically updated. This update process uses PHP or Python scripts.

[0116] User authentication and access control

[0117] When a user attempts to access a manual, the server first verifies their identity through an authentication system. It uses LDAP (Lightweight Directory Access Protocol) or OAuth protocols to verify the user's access rights. Only users who successfully authenticate are allowed to access specific documents with the appropriate permissions.

[0118] Feedback function and document revisions

[0119] Users can provide feedback while browsing the manual. This can be done by filling out a feedback form and submitting it, noting that specific steps or information are outdated. The feedback is sent to the server, which uses natural language processing to analyze the content and determine if the document needs revision. If necessary, the document is revised and saved again in the database.

[0120] Specific example

[0121] For example, if a company has documents such as "Sales Manual.pdf" and "Project Management.docx," the server collects these documents and converts them into text data. Next, the converted text data is analyzed using Google Cloud Natural Language and tagged with terms such as "Sales Policy" and "Project Management Methods." Based on the analysis results, duplicate parts are detected using Jaccard similarity and reconstructed into a single, unified document. The latest sales performance data from Salesforce is collected and reflected in the manual. When users access the manual, authentication is performed using LDAP. If a user provides feedback that a particular procedure is outdated, the server analyzes the feedback, makes the necessary revisions, and saves them in the database.

[0122] Example of a prompt

[0123] Examples of prompts for a generative AI model are as follows:

[0124] Please describe in detail, step by step, a system for collecting existing documents within a company and generating and maintaining a unified manual. This system includes steps such as document collection via SFTP, OCR processing, NLP analysis, duplicate detection, LDAP authentication, and feedback analysis.

[0125] ---

[0126] The above is the description of the "Mode for Carrying Out the Invention" in the patent specification.

[0127] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0128] Step 1:

[0129] Collect existing documents.

[0130] The server accesses designated folders and databases within the company to collect documents.

[0131] The input consists of a folder path and database connection information.

[0132] Specifically, the system will perform the operation of downloading PDF, Word, and Excel files from a remote folder using SFTP.

[0133] The output includes the collected document files.

[0134] Step 2:

[0135] Convert a document into text data.

[0136] The server converts the collected documents into text data using optical character recognition (OCR) and document text extraction tools.

[0137] The document files collected in Step 1 are provided as input.

[0138] Specifically, the process involves using ABBYY FineReader to convert PDFs into text data, and then using the Apache POI library to extract text from Word and Excel files.

[0139] The output includes the converted text data.

[0140] Step 3:

[0141] Analyze text data and perform classification and tagging.

[0142] The server uses natural language processing (NLP) techniques to analyze the converted text data.

[0143] The text data obtained in step 2 is provided as input.

[0144] Specifically, the system uses the Google Cloud Natural Language API and SpaCy to perform tokenization and morphological analysis, and then analyzes the meaning and structure of the text.

[0145] The output includes analytical data categorized by section and tagged appropriately.

[0146] Step 4:

[0147] It detects duplicate content and generates a unified document.

[0148] The server detects duplicate content from the analysis results and reconstructs it into a unified document.

[0149] The analysis data obtained in step 3 is provided as input.

[0150] Specifically, the process involves calculating the degree of text duplication using Jaccard similarity or Cosine similarity, and then combining the duplicated parts into a single text.

[0151] The output includes a unified document with duplicates removed and organized.

[0152] Step 5:

[0153] Collect business result data and compare it with document content.

[0154] The server periodically collects business results data from the company and compares it with document content.

[0155] The input includes business result data collected using the API.

[0156] Specifically, the system will use business system APIs, such as the Salesforce API, to retrieve the latest sales performance and project progress data.

[0157] The output includes the results of comparing the collected business data with the document content.

[0158] Step 6:

[0159] Automatically update the document.

[0160] The server automatically updates the document content based on the matching results.

[0161] The input is provided with the matching result obtained in step 5.

[0162] Specifically, the system will use PHP or Python scripts to update document content and reflect the latest information.

[0163] The output will include the latest updated document.

[0164] Step 7:

[0165] Verify user authentication and permissions.

[0166] The server authenticates users when they access the manual.

[0167] The user's authentication information is provided as input.

[0168] Specifically, it performs user authentication and authorization verification using LDAP and OAuth protocols.

[0169] The output includes the authentication results and the user's access permissions.

[0170] Step 8:

[0171] Users provide feedback and revise the document.

[0172] Users provide feedback while viewing the manual.

[0173] The input is the feedback content.

[0174] Specifically, the process involves sending feedback to the server through a feedback form.

[0175] The server analyzes the feedback and revises the document.

[0176] User feedback is provided as input.

[0177] Specifically, this involves using NLP tools to understand the feedback and making necessary revisions.

[0178] The output will include the revised document.

[0179] (Application Example 1)

[0180] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0181] Work manuals and procedures used by robots operating in factories need to be updated frequently. However, manual updates are inefficient, making it difficult to maintain consistency and up-to-dateness of information. Furthermore, the lack of a mechanism to incorporate real-time feedback from robots during actual work can lead to decreased accuracy and efficiency. In addition, there is a need for a mechanism to quickly update robot operating procedures when they become outdated.

[0182] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0183] In this invention, the server includes means for collecting existing documents and converting them into text; means for analyzing, classifying, and tagging the text data; means for detecting and integrating duplicate content; means for generating a unified document; means for collecting work result data and comparing it with the document content; means for automatically updating the document; means for verifying user authentication and authorization; means for users to provide feedback; means for revising the document based on the feedback; means for robots to provide real-time feedback during work and automatically revising the document based on that data; and means for automatically updating the robot's operating procedures. This makes it possible for robots in the factory to always operate based on the latest work manuals, improving the accuracy and efficiency of their work.

[0184] "Methods for collecting existing documents and converting them into text" refers to the process of gathering documents of various formats that exist within a factory into one place and converting them into machine-readable text data using OCR technology and a dedicated library.

[0185] "Means for analyzing, classifying, and tagging text data" refers to the process of analyzing each piece of text data using natural language processing technology, classifying it into sections based on its content, and assigning appropriate labels and tags.

[0186] "Methods for detecting and integrating duplicate content" refers to the process of extracting and detecting identical or similar content from analyzed text data and combining them into a single, unified document.

[0187] "Methods for generating unified documents" refers to the process of integrating redundant content and then compiling the resulting organized documents into a single, complete document.

[0188] "Means for collecting business result data and comparing it with document content" refers to the process of regularly collecting performance data related to operations within the factory and comparing and verifying it with existing documents.

[0189] "Methods for automatically updating documents" refer to a process that automatically updates the content of documents as needed, based on the results of cross-referencing with business result data, to reflect the most accurate information.

[0190] "Means of verifying user authentication and permissions" refers to the process of verifying the identity of the user attempting to access a document and checking whether they have the appropriate access rights.

[0191] "Means for users to provide feedback" refers to a function that allows users to provide feedback when viewing a document, and a process to prompt corrections if the content is outdated.

[0192] "Means of revising documents based on feedback" refers to the process of analyzing feedback provided by users and modifying or revising the content of the document based on that feedback.

[0193] "A method for robots to provide real-time feedback during work and automatically revise documents based on that data" refers to a process in which robots record information and feedback obtained during actual work in real time and automatically revise documents based on that data.

[0194] "Means for automatically updating robot operating procedures" refers to a process that automatically updates robot operating procedures based on the latest collected data and feedback, thereby improving work efficiency and accuracy.

[0195] System Overview

[0196] This invention is a system for keeping the work manuals and operating procedures for robots operating in a factory constantly up-to-date. The system includes multiple programs running on a server and applications installed on the robots. The system also has the function of receiving real-time feedback from the robots and automatically updating the documents. This can improve the work efficiency and accuracy of the robots.

[0197] Hardware and software configuration

[0198] Hardware: Server, robot, camera (for OCR)

[0199] Software: OCR tool (e.g., Tesseract), NLP engine (e.g., SpaCy), database (MySQL®), authentication system (LDAP)

[0200] Document collection and text conversion

[0201] The server collects existing documents from various terminals and databases within the factory and uses OCR tools to convert the contents of PDFs and JPEGs into text data. The converted text data is further analyzed and processed using a dedicated library. For example, nouns and verbs are extracted and sentences are split.

[0202] Data analysis and classification

[0203] The server uses an NLP engine to analyze text data, classify it into sections, and assign tags to each section. This organizes the data by content, such as "operating procedures" or "maintenance procedures." Specifically, it uses a natural language processing engine (SpaCy) to analyze the parts of speech in the text data and automatically assigns important sections and labels.

[0204] Duplicate detection and merging

[0205] The system detects duplicate sections from the analyzed data and consolidates them into a single, unified document. This eliminates unnecessary duplication and generates a well-organized manual. This process utilizes a document consolidation algorithm. For example, it includes a process to combine identical operating procedures found in multiple documents into a single document.

[0206] Matching with business result data and automatic updating

[0207] The server periodically collects work result data and compares it with existing documents. If there are discrepancies, the documents are automatically updated. This ensures that the latest information is always reflected. Specifically, it retrieves work performance data from the work database and compares it with the analysis results.

[0208] User authentication and access management

[0209] When a user attempts to access a document, their identity is verified using an LDAP-based authentication system. Only users with appropriate permissions can access specific information. This ensures security.

[0210] Feedback function and document revision

[0211] The robot can provide real-time feedback during operation. For example, if there is an inconsistency in the operating procedure, it will provide feedback immediately. This feedback is sent to the server, analyzed, and any necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent access. This ensures that the procedure is up-to-date and reflects current practices.

[0212] Examples of usage and specific prompt statements

[0213] For example, if a robot starts performing a new work procedure and there is a contradiction with the existing work manual, the feedback function will report it as follows:

[0214] Prompt example:

[0215] "Please check the instructions in Step 2 of the manual. Since it does not match the actual work, it needs to be updated based on the latest work result data."

[0216] This allows robots in the factory to perform tasks efficiently based on the latest information at all times.

[0217] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0218] Step 1:

[0219] The server collects existing documents from various terminals and databases within the factory. Input documents include formats such as PDF, Word, and Excel. The server gathers these documents in one place via FTP or API. The output is a list of the collected document data.

[0220] Step 2:

[0221] The server converts the collected documents into text data using an OCR tool (e.g., Tesseract). The input is the collected document data. Using OCR or a dedicated library, this is converted into machine-readable text, and the output is text data.

[0222] Step 3:

[0223] The server analyzes text data using an NLP engine (e.g., SpaCy). The input is converted text data. Using natural language processing techniques, this data is analyzed, classified, and tagged by part of speech and section. The analyzed data is then output.

[0224] Step 4:

[0225] The server detects and integrates duplicate parts from the analyzed data. The input is the analyzed data. A document integration algorithm is used to combine identical or similar content into a single document. The integrated document is then output.

[0226] Step 5:

[0227] The server generates integrated documents as a single, unified document. The input is integrated data. Based on this, it generates documents in the most organized format. The output is a unified document.

[0228] Step 6:

[0229] The server collects business result data and compares it with existing document content. The input consists of business result data and a standardized document. It compares this data with work performance data retrieved from the business database, and if there are discrepancies, it automatically updates the content. The updated document is output.

[0230] Step 7:

[0231] The server verifies authentication and permissions when a user accesses a document. The input is the user's authentication information. The server uses an authentication system such as LDAP to verify the user's identity and permissions. The output is a list of users whose access permissions have been verified.

[0232] Step 8:

[0233] Users view documents and provide feedback. The input is user feedback information. The feedback is sent to the server, analyzed, and necessary revisions are made. The updated feedback information is recorded as output.

[0234] Step 9:

[0235] The server revises documents based on user feedback. Its inputs include feedback information and document data. Based on this data, it analyzes which parts of the document need revision and performs the revisions. The revised document is then output.

[0236] Step 10:

[0237] The robot provides real-time feedback during the task and automatically revises the document based on that data. The input is real-time work feedback data. The robot sends the feedback it provides to a server, which then revises the document based on that data. The output is a document reflecting the latest work procedure.

[0238] Step 11:

[0239] The server automatically updates the robot's operating procedures based on the latest collected data and feedback. The input is revised document data. Based on this, the robot's operating procedures are automatically updated, enabling operation based on the latest information. The updated operating procedures are obtained as output.

[0240] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0241] ---

[0242] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. Furthermore, the invention incorporates an emotion engine that recognizes user emotions and has a function to optimize the user experience based on those emotions. This system includes multiple programs running on a server and terminals used by users.

[0243] System Overview

[0244] 1. Document collection and text conversion

[0245] Server: Collects existing documents (PDF, Word, Excel, etc.) from various folders and databases within the company. It periodically runs collection scripts to retrieve the latest documents. The collected documents are converted into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are also converted to text using a dedicated library.

[0246] 2. Analysis and Classification of Text Data

[0247] Server: The converted documents are analyzed using natural language processing (NLP) techniques. The document content is categorized into sections and appropriately tagged. For example, tags such as "project management" and "sales strategy" may be applied.

[0248] 3. Duplicate detection and unification

[0249] Server: Detects duplicate content from the analysis results and combines it into a single file. Compares duplicate sections and paragraphs to generate a unified text. This eliminates unnecessary duplication and produces a well-organized manual.

[0250] 4. Verification with business result data and automatic updating

[0251] Server: This server periodically collects business performance data from companies (e.g., sales figures and project progress). The collected data is compared with document content, and if there are discrepancies, the document is automatically updated to reflect the latest information.

[0252] 5. User Authentication and Permission Management

[0253] Server: Users attempting to access the manual undergo identity verification through an authentication system such as LDAP. The server verifies user access rights and grants access to the document only to users with appropriate permissions.

[0254] 6. Feedback function and document revision

[0255] Users: Users can view the manual and provide feedback. For example, they can report if a particular procedure is outdated. The feedback is sent to the server, analyzed, and necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent access.

[0256] 7. Integration of the Emotional Engine

[0257] Server: The system incorporates an emotion engine that recognizes user emotions. It analyzes user emotions in real time based on user input and usage, and adjusts the user experience based on the recognized emotions. For example, if a user is feeling stressed, it can prioritize providing support information.

[0258] 8. Adjusting document content based on emotions

[0259] Server: Based on emotional feedback collected by the emotion engine, the document content is automatically adjusted. In this way, users can utilize the information more efficiently.

[0260] Specific example

[0261] For example, consider a company with two documents: "Sales Manual.pdf" and "Project Management.docx". The server collects these documents and converts them to text. Next, it analyzes each document using natural language processing, classifying and tagging them into sections such as "Sales Policy" and "Project Management Methods". Duplicate parts are detected from the analysis results and reconstructed into a single, unified document. Subsequently, the latest sales performance data is collected and reflected in the manual, automatically updating its content. When users access the manual, authentication is performed, ensuring that only users with appropriate permissions can access the information. When user feedback is provided, the server analyzes it and revises the document.

[0262] Furthermore, an emotion engine recognizes user emotions based on their input and usage. For example, if a user repeatedly refers to a specific section of a manual and expresses confusion, the system will offer additional support regarding that section. It also automatically adjusts the document content based on emotion data, making it easier to understand on subsequent accesses.

[0263] This allows for the efficient collection and analysis of internal business manuals, providing them in a unified format free of duplication, while also reflecting the latest business results and enabling optimized information delivery based on user sentiment.

[0264] ---

[0265] The following describes the processing flow.

[0266] Step 1:

[0267] The server collects existing documents (PDFs, Word documents, Excel files, etc.) from specific folders or databases within the company. The collection script is run periodically to retrieve the most recent documents.

[0268] Step 2:

[0269] The server converts collected documents into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are converted to text using a dedicated library.

[0270] Step 3:

[0271] The server analyzes the converted text data using natural language processing (NLP) techniques. It categorizes the document content into sections and assigns appropriate tags. For example, tags such as "project management" and "sales strategy" might be applied.

[0272] Step 4:

[0273] The server detects duplicate content based on the analysis results. It compares similar sections and paragraphs and combines them into one. By generating the integrated text, unnecessary duplicates are eliminated and an organized manual is created.

[0274] Step 5:

[0275] The server formats the integrated document and automatically generates a standardized table of contents and index. The unified document is saved in the database.

[0276] Step 6:

[0277] The server regularly collects business result data (such as sales performance, project progress, etc.) from various business systems within the company via the API.

[0278] Step 7:

[0279] The server matches the collected business result data with the content of the manual. Based on the matching results, if there are any discrepancies, the document is automatically updated to reflect the latest information.

[0280] Step 8:

[0281] When a user accesses the manual, the server authenticates the user using an authentication system such as LDAP. It checks the user's permissions and allows access to the document only for users with appropriate permissions.

[0282] Step 9:

[0283] The user views the manual and reports improvement points and deficiencies using the feedback form. The feedback content is sent to the server.

[0284] Step 10:

[0285] The server analyzes the collected feedback and makes necessary revisions. The revised document is saved back to the database and will be reflected during subsequent accesses.

[0286] Step 11:

[0287] The server operates an emotion engine that recognizes emotions in real time from the user's input content and usage situation. For example, it analyzes the user's emotion from the keyboard input speed, mouse movement, and specific keywords.

[0288] Step 12:

[0289] Based on the emotion recognized by the server, it displays a revision manual or support information to the user. For example, if the user is recognized as being confused, it presents additional help information.

[0290] Step 13:

[0291] Based on the emotion data collected by the server through the emotion engine, the server adjusts the document content. It revises the document content to be more understandable according to the user's emotion feedback.

[0292] Step 14:

[0293] The server accumulates the emotion data and uses it to provide optimal information tailored to specific users. For example, it improves the sections that were confusing in the past so that they can be easily understood, and ensures that appropriate information is displayed during the next access.

[0294] Through the above steps, the business manuals within the enterprise are efficiently collected and analyzed, and a unified document with duplicates eliminated is generated. Also, it can reflect the latest business data of the users and optimize the user experience using the emotion engine.

[0295] (Example 2)

[0296] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0297] When efficiently collecting existing documents within a company and managing them as a unified manual, it takes time and effort to address document duplication and ensure that the latest information is reflected. Furthermore, it is difficult to quickly incorporate user feedback and optimize information based on user sentiment. This leads to problems such as a decline in the quality of the user experience and a decrease in operational efficiency.

[0298] The identification processing performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting existing documents and converting them into text, means for analyzing, classifying, and tagging the text data, means for detecting and integrating duplicate content, means for recognizing user emotions and optimizing the user experience based on those emotions, and means for automatically adjusting document content based on emotional feedback. This enables efficient collection and management of documents, rapid reflection of feedback, and optimization of information based on user emotions.

[0299] "Existing documents" refer to various files such as PDFs, Word documents, and Excel spreadsheets that have already been created within the company.

[0300] "Means of converting to text" refers to the process of converting existing documents into text data using OCR, PDF text extraction tools, program libraries, and the tools used for that purpose.

[0301] "Means for analyzing, classifying, and tagging text data" refers to the process of analyzing text data using natural language processing technology and assigning classifications and tags to specific sections.

[0302] "Means for detecting and integrating duplicate content" refers to the process of detecting duplicate content contained within text data and integrating it into a single entry.

[0303] The means for generating a unified document is a process of organizing duplicate content and generating a document stored in a unified form.

[0304] The means for collecting business result data and comparing it with document content is a process of periodically collecting a company's business result data and comparing it with existing document content.

[0305] The means for automatically updating a document is a process of automatically updating an existing document based on business result data.

[0306] The means for authenticating a user and verifying their permissions is a process of authenticating a user using an authentication system such as LDAP or OAuth and verifying their permissions.

[0307] The means for allowing a user to provide feedback is a web form or interface through which a user can provide feedback on the content of a document.

[0308] The means for revising a document based on feedback is a process of revising a document based on feedback provided by a user.

[0309] The means for recognizing a user's emotion and optimizing the user experience based on that emotion is a process of using an emotion engine to analyze a user's emotion in real time and adjusting the use case based on that emotion.

[0310] The means for automatically adjusting document content based on emotion feedback is a process of automatically adjusting the content of a document based on a user's emotion feedback and providing optimized information when the user accesses it next time.

[0311] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. Furthermore, the invention incorporates an emotion engine that recognizes user emotions and has the function of optimizing the user experience based on those emotions. This system includes multiple programs running on a server and terminals used by users.

[0312] System Overview

[0313] 1. Document collection and text conversion

[0314] The server collects existing documents from folders and databases within the company. This collection uses Python's os module and cloud storage APIs (e.g., AWS® S3, Google Drive API). The collection script is run periodically to retrieve the latest documents. The collected documents are converted into text data using optical character recognition (OCR) or PDF text extraction tools (e.g., PyPDF2, python-docx, pandas, Tesseract OCR).

[0315] 2. Analysis and Classification of Text Data

[0316] The server analyzes the text data using natural language processing (NLP) tools (e.g., SpaCy, NLTK). It categorizes the document content into sections and assigns appropriate tags (e.g., "Project Management," "Sales Strategy").

[0317] 3. Duplicate detection and unification

[0318] The server detects duplicate content from the analysis results and combines it into a single file. The difflib module is used for text comparison. It generates a unified text file and saves it with unnecessary duplicates removed.

[0319] 4. Verification with business result data and automatic updating

[0320] The server periodically collects business performance data (e.g., sales figures and project progress) from SQL or NoSQL databases (e.g., MySQL, MongoDB). The collected data is compared with the document content, and if there are any discrepancies, the document is automatically updated based on the latest information.

[0321] 5. User Authentication and Permission Management

[0322] The server authenticates users using LDAP or OAuth when they access documents. Once authentication is complete, it checks the user's permissions and grants access to the documents only to users with the appropriate permissions.

[0323] 6. Feedback function and document revision

[0324] Users can view documents through the system and provide feedback. This feedback is sent to the server via web forms or other means. The server analyzes the feedback and revises the document accordingly.

[0325] 7. Integration of the Emotional Engine

[0326] The server uses an emotion engine (e.g., Azure® Text Analytics, IBM Watson®) to analyze user emotions in real time based on user input and usage. The user experience can then be tailored based on the recognized emotions.

[0327] 8. Adjusting document content based on emotions

[0328] The server automatically adjusts document content based on sentiment feedback, allowing users to access the information more efficiently the next time they visit.

[0329] Specific example

[0330] For example, consider a company where there are two documents: "Sales Manual.pdf" and "Project Management.docx". The server collects these documents and converts them to text using PyPDF2 or python-docx. Next, the converted text is analyzed using SpaCy, classifying each document into sections such as "Sales Policy" and "Project Management Methods," and tagging them. Duplicate parts are detected from the analysis results, and the documents are reconstructed as a single, unified document.

[0331] Subsequently, the latest sales performance data is collected from the MySQL database and reflected in the manual, automatically updating its content. When users access the manual, they are authenticated via LDAP, ensuring that only users with appropriate permissions can access the information. When user feedback is provided, the server analyzes it and revises the document accordingly.

[0332] Furthermore, the emotion engine analyzes user input and usage in real time, and based on the recognized emotions, if confusion or stress is detected regarding a particular section, the system will offer additional support for that section. It also automatically adjusts document content based on emotion data, making it easier to understand on subsequent accesses.

[0333] Example of a prompt

[0334] How can I use a generative AI model to detect user emotions and adjust the content of company manuals based on those emotions?

[0335] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0336] Step 1: Document collection and text conversion

[0337] The server collects existing documents from designated folders and databases within the company. Input is the folder path and database connection information, while output is the collected document files.

[0338] Specifically, the server retrieves files using Python's os module and cloud storage APIs.

[0339] Based on the entered folder path, the system searches for document files and uploads them to the database.

[0340] Step 2: Text conversion

[0341] The server converts collected documents into text data. Input is document files (e.g., PDF, Word, Excel), and output is the converted text data.

[0342] Tools such as PyPDF2, python-docx, pandas, and Tesseract OCR will be used.

[0343] PDF files are extracted using Optical Character Recognition (OCR), while Word and Excel files are converted to text using a dedicated library.

[0344] Step 3: Analysis and classification of text data

[0345] The server analyzes text data using natural language processing (NLP) tools. The input is the transformed text data, and the output is the analyzed section data and tags.

[0346] Specifically, the server uses SpaCy or NLTK to analyze the text, classify the document content into sections, and assign appropriate tags.

[0347] Based on the analyzed text, tags such as "Project Management" and "Sales Strategy" are added to each section.

[0348] Step 4: Duplicate detection and unification

[0349] The server detects duplicate content from the analysis results and combines it into a single file. The input is tagged section data, and the output is unified content with duplicates removed.

[0350] The server uses the difflib module to detect duplicate text.

[0351] Merge highly similar texts and remove unnecessary duplicates.

[0352] Step 5: Match with business result data and automatic update

[0353] The server periodically collects business results data from companies and compares it with document content. The input is information from the business results database, and the output is a document reflecting the latest information.

[0354] The server uses MySQL or MongoDB to retrieve the latest sales performance and project progress data from the database.

[0355] Compare the business results data with the documents, and update the document content if there are any discrepancies.

[0356] Step 6: User Authentication and Permission Management

[0357] The server authenticates the user and verifies their permissions. The input is the user's authentication information (e.g., user ID, password), and the output is the authentication and permission verification result.

[0358] The server uses LDAP or OAuth to authenticate users.

[0359] Once authentication is complete, the system verifies the user's access rights and grants access to the document only to users with the appropriate permissions.

[0360] Step 7: Feedback function and document revision

[0361] Users provide feedback on a document. The input is the content of the feedback, and the output is the document with the feedback reflected.

[0362] Feedback is sent to the server via web forms or similar means.

[0363] The server analyzes the feedback, identifies areas that need improvement, and revises the document accordingly.

[0364] Step 8: Integrating the Emotional Engine

[0365] The server uses an emotion engine to analyze the user's emotions. The input is the user's input data and feedback, and the output is the emotion analysis result.

[0366] The server utilizes sentiment analysis APIs (e.g., Azure Text Analytics, IBM Watson).

[0367] The user experience is adjusted based on recognized emotions, and additional support information is provided if negative emotions are detected.

[0368] Step 9: Adjusting document content based on emotions

[0369] The server automatically adjusts the document content based on emotional feedback. The input is emotional feedback data, and the output is the adjusted document content.

[0370] The server modifies the document content based on sentiment data, providing optimized information for the user's next access.

[0371] Specifically, this involves adding a QA section and simplifying the content of the documents.

[0372] This enables efficient document collection and management, rapid incorporation of feedback, and optimization of information based on user sentiment.

[0373] (Application Example 2)

[0374] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0375] Modern businesses are required to manage and appropriately utilize a wide variety of documents. Simultaneously, it is crucial to recognize user emotions in real time and provide information efficiently. However, existing systems struggle to integrate document management and user support, and real-time work instructions and emotion-based support are particularly lacking in manufacturing environments. This makes it difficult to optimize the user experience while simultaneously demanding document consistency, up-to-date information, and improved work efficiency. Therefore, this invention proposes a system that solves these problems, providing efficient and unified document management and real-time support tailored to user emotions.

[0376] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0377] In this invention, the server includes means for collecting existing documents and converting them to text; means for analyzing, classifying, and tagging the text data; means for detecting and integrating duplicate content; means for generating a unified document; means for collecting work result data and comparing it with document content; means for automatically updating documents; means for verifying user authentication and permissions; means for users to provide feedback; means for revising documents based on feedback; means for providing work instructions in real time; means for recognizing user emotions and optimizing the experience; and means for providing additional support information based on emotion recognition. This enables efficient document management within the company, real-time work support, and optimal information provision based on user emotions.

[0378] "Means of collecting documents and converting them into text" refers to a function that converts existing documents within a company, such as manuals, procedures, and reports, into analyzable text data.

[0379] "Means for analyzing, classifying, and tagging text data" refers to a function that analyzes the converted text data using natural language processing technology and assigns appropriate tags to each section.

[0380] "Means for detecting and integrating duplicate content" refers to a function that detects duplicate parts from analyzed text data and combines them into a single entity.

[0381] "Means for generating unified documents" refers to a function that generates consistent and unified manuals and reports based on text data that has been merged and whose duplicates have been eliminated.

[0382] "Means for collecting business result data and comparing it with document content" refers to a function that collects company performance and project progress data and compares and verifies it with existing document content.

[0383] "Methods for automatically updating documents" refers to a function that automatically updates documents to the latest content based on collected performance data and project progress data.

[0384] "Means of verifying user authentication and permissions" refers to functions that authenticate users attempting to access manuals and reports, and verify and manage appropriate permissions.

[0385] "Means for users to provide feedback" refers to features that allow users to provide feedback and comments on manuals and reports.

[0386] "Means for revising documents based on feedback" refers to a function that analyzes collected user feedback and revises documents as needed.

[0387] "Means of providing work instructions in real time" refers to a function that provides the instructions and procedures that workers need in real time.

[0388] "Means of recognizing user emotions and optimizing the experience" refers to a function that recognizes the emotional state of workers in real time and optimizes user support based on that information.

[0389] "Means of providing additional support information based on emotion recognition" refers to a function that provides necessary support information and additional instructions based on the user's emotional state.

[0390] Modes for carrying out the invention

[0391] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. Furthermore, the invention incorporates an emotion engine that recognizes user emotions and has a function to optimize the user experience based on those emotions. This system includes multiple software programs running on a server and terminals used by users.

[0392] System Overview

[0393] 1. Document collection and text conversion

[0394] The server collects existing documents (PDFs, Word documents, Excel files, etc.) from various folders and databases within the company. A collection script is run periodically to retrieve the latest documents. The collected documents are converted into text data using optical character recognition (OCR) tools and PDF text extraction tools. Word and Excel files are also converted to text data using dedicated libraries. Specifically, tools such as Tesseract OCR and PDFMiner can be used.

[0395] 2. Analysis and Classification of Text Data

[0396] The server analyzes the converted text data using natural language processing (NLP) techniques. It classifies the document content into sections and applies appropriate tags. For example, NLP libraries such as spaCy or NLTK can be used.

[0397] 3. Duplicate detection and unification

[0398] The server detects and merges duplicate content from the analysis results. It compares overlapping sections and paragraphs to generate a single, unified document. This eliminates unnecessary duplication, resulting in a more organized manual.

[0399] 4. Verification with business result data and automatic updating

[0400] The server periodically collects business performance data from the company (e.g., sales figures and project progress). The collected data is compared with the document content, and if there are any discrepancies, the document is automatically updated to reflect the latest information.

[0401] 5. User Authentication and Permission Management

[0402] The server authenticates users attempting to access manuals and reports. It uses authentication systems such as LDAP to verify identity and confirm user access rights. Only users with appropriate permissions are granted access to the documents.

[0403] 6. Feedback function and document revision

[0404] Users can view manuals and reports and provide feedback. For example, they can report if a particular procedure is outdated. The feedback is sent to the server, analyzed, and necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent accesses.

[0405] 7. Integration of the Emotional Engine

[0406] The server incorporates an emotion engine that recognizes user emotions. It analyzes emotions in real time based on user input and usage, and adjusts the user experience based on the recognized emotions. For example, if a user is experiencing stress, it can prioritize providing support information. Emotion recognition models such as TENSORFLOW® and OpenCV can be used.

[0407] 8. Adjusting document content based on emotions

[0408] The server automatically adjusts document content based on emotional feedback collected by the emotion engine. This allows users to utilize information more efficiently.

[0409] Specific example

[0410] When a new manufacturing process is introduced in a factory, the server collects relevant documents and generates standardized work instructions. When workers receive instructions using smart glasses, the emotion engine provides additional support information if it detects stress. The server uses TensorFlow to perform emotion recognition and provides optimal information in real time to reduce worker confusion.

[0411] Example of a prompt

[0412] Develop standard procedures for a new manufacturing process and design a real-time work guidance system that incorporates emotion recognition capabilities. This system should include features that allow workers to receive instructions in real time via a head-mounted display and receive additional support based on their emotional state.

[0413] Thus, the system of the present invention streamlines document management within a company and realizes real-time work support and optimized user experience.

[0414] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0415] Step 1: Document collection and text conversion

[0416] The server collects existing documents (PDF, Word, Excel, etc.) from various databases and folders within the company. The collected documents are converted into text data using OCR tools (e.g., Tesseract OCR) or PDF text extraction tools. The input is various document files, and the output is their corresponding text data.

[0417] Step 2: Analysis and classification of text data

[0418] The server analyzes the converted text data using natural language processing (NLP) techniques (e.g., spaCy, NLTK). It analyzes each section of the document and assigns appropriate tags to them. The input is text data, and the output is the analysis results and tagged data.

[0419] Step 3: Duplicate detection and merging

[0420] The server detects duplicate content from the parsed and tagged text data. It then compares and merges the duplicate sections to generate a single, unified document. The input is the parsed and tagged text data, and the output is the unified document. Specifically, it uses an algorithm that merges paragraphs and sections with high similarity.

[0421] Step 4: Matching with business result data and automatic updating

[0422] The server periodically collects corporate performance data and project progress data. It then compares this data with existing document content and automatically updates it to reflect the latest information. The input consists of performance data and existing documents, while the output is the updated documents. Specifically, it uses database queries and matching algorithms.

[0423] Step 5: User Authentication and Permission Management

[0424] The server authenticates users through an authentication system such as LDAP when they attempt to access manuals or reports. It verifies the user's access rights and allows access only to those with appropriate permissions. The input is the user's authentication information, and the output is the result of the access rights verification. Specifically, it performs LDAP queries and generates authentication tokens.

[0425] Step 6: Feedback function and document revision

[0426] Users can provide feedback while viewing manuals and reports. The feedback is sent to the server, analyzed, and necessary revisions are made. The input is user feedback, and the output is the revised document. Specifically, a feedback analysis algorithm is used to update the document content.

[0427] Step 7: Provide work instructions in real time

[0428] The server provides workers with real-time instructions and procedures through smart glasses or head-mounted displays. Input is data from the work environment, and output is real-time work instructions. Specifically, it performs tasks such as presenting predefined work procedures and analyzing environmental data.

[0429] Step 8: Emotion Recognition and Experience Optimization

[0430] The server recognizes user emotions in real time and optimizes the user experience based on that. TensorFlow is among the emotion recognition models used. Inputs include user facial images and usage data, while output is optimized user experience information. Specifically, it uses facial recognition algorithms and emotion recognition models (e.g., TensorFlow, OpenCV).

[0431] Step 9: Provide additional support information based on emotion recognition.

[0432] The server provides necessary support information and additional instructions based on the user's emotional state. Input consists of the results of emotion recognition and data regarding the user's confusion and stress; output is additional support information. Specifically, it searches for appropriate support information and presents it to the user.

[0433] In summary, this system streamlines document management within companies and provides real-time work support and optimized user experience.

[0434] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0435] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0436] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0437] [Second Embodiment]

[0438] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0439] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0440] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0441] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0442] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0443] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0444] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0445] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0446] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0447] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0448] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0449] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0450] ---

[0451] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. This system includes multiple programs running on a server and terminals used by users. The system also features a function to automatically update documents based on user feedback and business result data.

[0452] System Overview

[0453] 1. Document collection and text conversion

[0454] Server: Collects existing documents (PDF, Word, Excel, etc.) from various folders and databases within the company. The collected documents are converted into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are also converted to text using a dedicated library.

[0455] 2. Analysis and Classification of Text Data

[0456] Server: The converted text data is analyzed using natural language processing (NLP) techniques. During the analysis process, the document is categorized into sections and assigned appropriate tags. For example, tags such as "project management" and "sales strategy" may be added.

[0457] 3. Duplicate detection and unification

[0458] Server: Text analysis detects duplication when the same content is found in multiple documents. The duplicate content is combined into a single, unified document. This eliminates unnecessary duplication and generates a well-organized manual.

[0459] 4. Verification with business result data and automatic updating

[0460] Server: The server periodically collects business performance data (e.g., sales figures and project progress). The collected data is compared with document content, and if discrepancies are found, the document is automatically updated. This ensures that manuals always reflect the latest information.

[0461] 5. User Authentication and Permission Management

[0462] Server: Users attempting to access the manual must first undergo identity verification through an authentication system. Authentication technologies such as LDAP are used to verify user access rights. This ensures that only users with appropriate permissions can access specific information.

[0463] 6. Feedback function and document revision

[0464] Users: Users can view the manual and provide feedback. For example, they can report if a particular procedure is outdated. The feedback is sent to the server, analyzed, and necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent access.

[0465] Specific example

[0466] For example, consider a company with two documents: "Sales Manual.pdf" and "Project Management.docx". The server collects these documents and converts them to text. Next, it analyzes each document using natural language processing, classifying and tagging them into sections such as "Sales Policy" and "Project Management Methods". Duplicate parts are detected from the analysis results and reconstructed into a single, unified document. Subsequently, the latest sales performance data is collected and reflected in the manual, automatically updating its content. When users access the manual, authentication is performed, ensuring that only users with appropriate permissions can access the information. When user feedback is provided, the server analyzes it and revises the document.

[0467] This allows for centralized management of internal business manuals, ensuring they are always up-to-date, and guaranteeing the secure handling of confidential information.

[0468] ---

[0469] The following describes the processing flow.

[0470] Step 1:

[0471] The server collects existing documents (PDFs, Word documents, Excel files, etc.) from specific folders or databases within the company. The collection script is run periodically to retrieve the most recent documents.

[0472] Step 2:

[0473] The server converts collected documents into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are converted to text using a dedicated library.

[0474] Step 3:

[0475] The server analyzes the converted text data using natural language processing (NLP) techniques. It analyzes the document content, categorizes it into sections, and applies appropriate tags.

[0476] Step 4:

[0477] The server detects duplicate content based on the analysis results. It compares similar sections and paragraphs and integrates them into a single, unified text.

[0478] Step 5:

[0479] The server formats the integrated documents and automatically generates a standardized table of contents and index. The unified documents are then stored in a database.

[0480] Step 6:

[0481] The server periodically collects business result data (sales performance, project progress, etc.) from various business systems within the company via APIs.

[0482] Step 7:

[0483] The server compares the collected work result data with the contents of the manual. If a discrepancy is detected, it automatically performs the necessary updates and updates the document.

[0484] Step 8:

[0485] When a user accesses a manual, the server authenticates the user using an authentication system such as LDAP. It verifies the user's permissions and grants access to the corresponding document only to users with the appropriate privileges.

[0486] Step 9:

[0487] Users view the manual and report areas for improvement or deficiencies using a feedback form. The feedback data is sent to the server.

[0488] Step 10:

[0489] The server analyzes the collected feedback and revises the document. The revised document is saved back to the database and the changes are reflected the next time it is accessed.

[0490] These steps ensure that internal operational manuals are efficiently collected and analyzed, and provided in a unified format free of duplication. Furthermore, they remain up-to-date by reflecting the latest operational results, and are quickly revised based on user feedback.

[0491] (Example 1)

[0492] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0493] The diverse range of documents generated within a company increases over time, making their management complex. The existence of multiple documents can lead to information duplication and inconsistencies, hindering efficient work execution. Furthermore, manual document updates are labor-intensive and can delay the reflection of the latest information. In addition, a lack of proper user authentication and access control increases the risk of confidential information leaks. Addressing these challenges is essential.

[0494] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0495] In this invention, the server includes means for collecting existing documents and converting them into text, means for analyzing, classifying, and tagging the text data, and means for detecting and integrating duplicate content. This streamlines document management within the company, eliminates duplication and inconsistencies, and generates a unified manual that always reflects the latest information. Furthermore, user authentication and permission verification ensure the secure handling of confidential information. In addition, document revisions are automatically performed based on feedback provided by users, ensuring that the latest information is always reflected.

[0496] "Document collection methods" refer to methods for collecting existing documents from folders and databases within a company.

[0497] "Text conversion means" refers to means of converting collected documents into text data. This includes optical character recognition and document text extraction tools.

[0498] "Text analysis means" refers to means for analyzing the converted text data, and this uses natural language processing technology.

[0499] "Classification and tagging means" refers to means of classifying the analyzed text data into sections and assigning appropriate tags.

[0500] A "duplicate detection means" is a means of calculating the similarity between analyzed text data and detecting duplicate content.

[0501] A "combination method" is a means of integrating duplicate text data into a single, unified document.

[0502] "Methods for collecting business data" refer to methods for periodically collecting business result data from a company, and this includes the use of various APIs.

[0503] A "document verification means" is a means of verifying collected business data against the contents of documents.

[0504] An "automatic update mechanism" is a means of automatically updating the document content based on the verification results.

[0505] "User authentication means" refers to methods for verifying a user's identity, and this includes using LDAP or authentication protocols.

[0506] A "permission verification method" is a means of verifying a user's access rights and ensuring that only users with appropriate permissions can access the information.

[0507] A "feedback provision method" is a means by which users can provide feedback on a document.

[0508] "Document revision methods" refer to methods for revising documents based on user feedback and reflecting the latest information.

[0509] ---

[0510] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. This system includes multiple programs running on a server and terminals used by users. The system also features a function to automatically update documents based on user feedback and business result data.

[0511] System Overview

[0512] Document collection and text conversion

[0513] The server collects existing documents (PDF, Word, Excel, etc.) from various folders and databases within the company. This collection process utilizes remote file transfer methods, specifically SFTP (Secure File Transfer Protocol). The collected documents are converted into text data using optical character recognition (OCR) and document text extraction tools. ABBYY FineReader is used for OCR, and the PDF library is used for PDF text extraction. Word and Excel files are converted to text using the Apache POI library.

[0514] Text data analysis and classification

[0515] The server analyzes the converted text data using natural language processing (NLP) techniques. Software such as Google Cloud Natural Language and SpaCy are used for analysis, including text tokenization and morphological analysis. As a result of the analysis, the documents are categorized into sections and tagged with terms like "Project Management" and "Sales Strategy." Machine learning techniques (e.g., k-means clustering) are used for classification.

[0516] Duplicate detection and unification

[0517] The server detects duplicates when multiple documents contain the same content based on the analysis results. Text similarity calculation methods (such as Jaccard similarity or Cosine similarity) are used for duplicate detection. Duplicate content is combined into a single, unified document. The most reliable text and the most recent content are prioritized for unification as needed.

[0518] Matching with business result data and automatic updating

[0519] The server periodically collects various business result data from the company (e.g., sales performance data and project progress data). This data collection often utilizes APIs, specifically the Salesforce API and other business system APIs. The collected data is compared with document content, and if discrepancies are found, the documents are automatically updated. This update process uses PHP or Python scripts.

[0520] User authentication and access control

[0521] When a user attempts to access a manual, the server first verifies their identity through an authentication system. It uses LDAP (Lightweight Directory Access Protocol) or OAuth protocols to verify the user's access rights. Only users who successfully authenticate are allowed to access specific documents with the appropriate permissions.

[0522] Feedback function and document revisions

[0523] Users can provide feedback while browsing the manual. This can be done by filling out a feedback form and submitting it, noting that specific steps or information are outdated. The feedback is sent to the server, which uses natural language processing to analyze the content and determine if the document needs revision. If necessary, the document is revised and saved again in the database.

[0524] Specific example

[0525] For example, if a company has documents such as "Sales Manual.pdf" and "Project Management.docx," the server collects these documents and converts them into text data. Next, the converted text data is analyzed using Google Cloud Natural Language and tagged with terms such as "Sales Policy" and "Project Management Methods." Based on the analysis results, duplicate parts are detected using Jaccard similarity and reconstructed into a single, unified document. The latest sales performance data from Salesforce is collected and reflected in the manual. When users access the manual, authentication is performed using LDAP. If a user provides feedback that a particular procedure is outdated, the server analyzes the feedback, makes the necessary revisions, and saves them in the database.

[0526] Example of a prompt

[0527] Examples of prompts for a generative AI model are as follows:

[0528] Please describe in detail, step by step, a system for collecting existing documents within a company and generating and maintaining a unified manual. This system includes steps such as document collection via SFTP, OCR processing, NLP analysis, duplicate detection, LDAP authentication, and feedback analysis.

[0529] ---

[0530] The above is the description of the "Mode for Carrying Out the Invention" in the patent specification.

[0531] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0532] Step 1:

[0533] Collect existing documents.

[0534] The server accesses designated folders and databases within the company to collect documents.

[0535] The input consists of a folder path and database connection information.

[0536] Specifically, the system will perform the operation of downloading PDF, Word, and Excel files from a remote folder using SFTP.

[0537] The output includes the collected document files.

[0538] Step 2:

[0539] Convert a document into text data.

[0540] The server converts the collected documents into text data using optical character recognition (OCR) and document text extraction tools.

[0541] The document files collected in Step 1 are provided as input.

[0542] Specifically, the process involves using ABBYY FineReader to convert PDFs into text data, and then using the Apache POI library to extract text from Word and Excel files.

[0543] The output includes the converted text data.

[0544] Step 3:

[0545] Analyze text data and perform classification and tagging.

[0546] The server uses natural language processing (NLP) techniques to analyze the converted text data.

[0547] The text data obtained in step 2 is provided as input.

[0548] Specifically, the system uses the Google Cloud Natural Language API and SpaCy to perform tokenization and morphological analysis, and then analyzes the meaning and structure of the text.

[0549] The output includes analytical data categorized by section and tagged appropriately.

[0550] Step 4:

[0551] It detects duplicate content and generates a unified document.

[0552] The server detects duplicate content from the analysis results and reconstructs it into a unified document.

[0553] The analysis data obtained in step 3 is provided as input.

[0554] Specifically, the process involves calculating the degree of text duplication using Jaccard similarity or Cosine similarity, and then combining the duplicated parts into a single text.

[0555] The output includes a unified document with duplicates removed and organized.

[0556] Step 5:

[0557] Collect business result data and compare it with document content.

[0558] The server periodically collects business results data from the company and compares it with document content.

[0559] The input includes business result data collected using the API.

[0560] Specifically, the system will use business system APIs, such as the Salesforce API, to retrieve the latest sales performance and project progress data.

[0561] The output includes the results of comparing the collected business data with the document content.

[0562] Step 6:

[0563] Automatically update the document.

[0564] The server automatically updates the document content based on the matching results.

[0565] The input is provided with the matching result obtained in step 5.

[0566] Specifically, the system will use PHP or Python scripts to update document content and reflect the latest information.

[0567] The output will include the latest updated document.

[0568] Step 7:

[0569] Verify user authentication and permissions.

[0570] The server authenticates users when they access the manual.

[0571] The user's authentication information is provided as input.

[0572] Specifically, it performs user authentication and authorization verification using LDAP and OAuth protocols.

[0573] The output includes the authentication results and the user's access permissions.

[0574] Step 8:

[0575] Users provide feedback and revise the document.

[0576] Users provide feedback while viewing the manual.

[0577] The input is the feedback content.

[0578] Specifically, the process involves sending feedback to the server through a feedback form.

[0579] The server analyzes the feedback and revises the document.

[0580] User feedback is provided as input.

[0581] Specifically, this involves using NLP tools to understand the feedback and making necessary revisions.

[0582] The output will include the revised document.

[0583] (Application Example 1)

[0584] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0585] Work manuals and procedures used by robots operating in factories need to be updated frequently. However, manual updates are inefficient, making it difficult to maintain consistency and up-to-dateness of information. Furthermore, the lack of a mechanism to incorporate real-time feedback from robots during actual work can lead to decreased accuracy and efficiency. In addition, there is a need for a mechanism to quickly update robot operating procedures when they become outdated.

[0586] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0587] In this invention, the server includes means for collecting existing documents and converting them into text; means for analyzing, classifying, and tagging the text data; means for detecting and integrating duplicate content; means for generating a unified document; means for collecting work result data and comparing it with the document content; means for automatically updating the document; means for verifying user authentication and authorization; means for users to provide feedback; means for revising the document based on the feedback; means for robots to provide real-time feedback during work and automatically revising the document based on that data; and means for automatically updating the robot's operating procedures. This makes it possible for robots in the factory to always operate based on the latest work manuals, improving the accuracy and efficiency of their work.

[0588] "Methods for collecting existing documents and converting them into text" refers to the process of gathering documents of various formats that exist within a factory into one place and converting them into machine-readable text data using OCR technology and a dedicated library.

[0589] "Means for analyzing, classifying, and tagging text data" refers to the process of analyzing each piece of text data using natural language processing technology, classifying it into sections based on its content, and assigning appropriate labels and tags.

[0590] "Methods for detecting and integrating duplicate content" refers to the process of extracting and detecting identical or similar content from analyzed text data and combining them into a single, unified document.

[0591] "Methods for generating unified documents" refers to the process of integrating redundant content and then compiling the resulting organized documents into a single, complete document.

[0592] "Means for collecting business result data and comparing it with document content" refers to the process of regularly collecting performance data related to operations within the factory and comparing and verifying it with existing documents.

[0593] "Methods for automatically updating documents" refer to a process that automatically updates the content of documents as needed, based on the results of cross-referencing with business result data, to reflect the most accurate information.

[0594] "Means of verifying user authentication and permissions" refers to the process of verifying the identity of the user attempting to access a document and checking whether they have the appropriate access rights.

[0595] "Means for users to provide feedback" refers to a function that allows users to provide feedback when viewing a document, and a process to prompt corrections if the content is outdated.

[0596] "Means of revising documents based on feedback" refers to the process of analyzing feedback provided by users and modifying or revising the content of the document based on that feedback.

[0597] "A method for robots to provide real-time feedback during work and automatically revise documents based on that data" refers to a process in which robots record information and feedback obtained during actual work in real time and automatically revise documents based on that data.

[0598] "Means for automatically updating robot operating procedures" refers to a process that automatically updates robot operating procedures based on the latest collected data and feedback, thereby improving work efficiency and accuracy.

[0599] System Overview

[0600] This invention is a system for keeping the work manuals and operating procedures for robots operating in a factory constantly up-to-date. The system includes multiple programs running on a server and applications installed on the robots. The system also has the function of receiving real-time feedback from the robots and automatically updating the documents. This can improve the work efficiency and accuracy of the robots.

[0601] Hardware and software configuration

[0602] Hardware: Server, robot, camera (for OCR)

[0603] Software: OCR tool (e.g., Tesseract), NLP engine (e.g., SpaCy), database (MySQL), authentication system (LDAP)

[0604] Document collection and text conversion

[0605] The server collects existing documents from various terminals and databases within the factory and uses OCR tools to convert the contents of PDFs and JPEGs into text data. The converted text data is further analyzed and processed using a dedicated library. For example, nouns and verbs are extracted and sentences are split.

[0606] Data analysis and classification

[0607] The server uses an NLP engine to analyze text data, classify it into sections, and assign tags to each section. This organizes the data by content, such as "operating procedures" or "maintenance procedures." Specifically, it uses a natural language processing engine (SpaCy) to analyze the parts of speech in the text data and automatically assigns important sections and labels.

[0608] Duplicate detection and merging

[0609] The system detects duplicate sections from the analyzed data and consolidates them into a single, unified document. This eliminates unnecessary duplication and generates a well-organized manual. This process utilizes a document consolidation algorithm. For example, it includes a process to combine identical operating procedures found in multiple documents into a single document.

[0610] Matching with business result data and automatic updating

[0611] The server periodically collects work result data and compares it with existing documents. If there are discrepancies, the documents are automatically updated. This ensures that the latest information is always reflected. Specifically, it retrieves work performance data from the work database and compares it with the analysis results.

[0612] User authentication and access management

[0613] When a user attempts to access a document, their identity is verified using an LDAP-based authentication system. Only users with appropriate permissions can access specific information. This ensures security.

[0614] Feedback function and document revision

[0615] The robot can provide real-time feedback during operation. For example, if there is an inconsistency in the operating procedure, it will provide feedback immediately. This feedback is sent to the server, analyzed, and any necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent access. This ensures that the procedure is up-to-date and reflects current practices.

[0616] Examples of usage and specific prompt statements

[0617] For example, if a robot starts performing a new work procedure and there is a contradiction with the existing work manual, the feedback function will report it as follows:

[0618] Prompt example:

[0619] "Please check the instructions in Step 2 of the manual. Since it does not match the actual work, it needs to be updated based on the latest work result data."

[0620] This allows robots in the factory to perform tasks efficiently based on the latest information at all times.

[0621] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0622] Step 1:

[0623] The server collects existing documents from various terminals and databases within the factory. Input documents include formats such as PDF, Word, and Excel. The server gathers these documents in one place via FTP or API. The output is a list of the collected document data.

[0624] Step 2:

[0625] The server converts the collected documents into text data using an OCR tool (e.g., Tesseract). The input is the collected document data. Using OCR or a dedicated library, this is converted into machine-readable text, and the output is text data.

[0626] Step 3:

[0627] The server analyzes text data using an NLP engine (e.g., SpaCy). The input is converted text data. Using natural language processing techniques, this data is analyzed, classified, and tagged by part of speech and section. The analyzed data is then output.

[0628] Step 4:

[0629] The server detects and integrates duplicate parts from the analyzed data. The input is the analyzed data. A document integration algorithm is used to combine identical or similar content into a single document. The integrated document is then output.

[0630] Step 5:

[0631] The server generates integrated documents as a single, unified document. The input is integrated data. Based on this, it generates documents in the most organized format. The output is a unified document.

[0632] Step 6:

[0633] The server collects business result data and compares it with existing document content. The input consists of business result data and a standardized document. It compares this data with work performance data retrieved from the business database, and if there are discrepancies, it automatically updates the content. The updated document is output.

[0634] Step 7:

[0635] The server verifies authentication and permissions when a user accesses a document. The input is the user's authentication information. The server uses an authentication system such as LDAP to verify the user's identity and permissions. The output is a list of users whose access permissions have been verified.

[0636] Step 8:

[0637] Users view documents and provide feedback. The input is user feedback information. The feedback is sent to the server, analyzed, and necessary revisions are made. The updated feedback information is recorded as output.

[0638] Step 9:

[0639] The server revises documents based on user feedback. Its inputs include feedback information and document data. Based on this data, it analyzes which parts of the document need revision and performs the revisions. The revised document is then output.

[0640] Step 10:

[0641] The robot provides real-time feedback during the task and automatically revises the document based on that data. The input is real-time work feedback data. The robot sends the feedback it provides to a server, which then revises the document based on that data. The output is a document reflecting the latest work procedure.

[0642] Step 11:

[0643] The server automatically updates the robot's operating procedures based on the latest collected data and feedback. The input is revised document data. Based on this, the robot's operating procedures are automatically updated, enabling operation based on the latest information. The updated operating procedures are obtained as output.

[0644] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0645] ---

[0646] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. Furthermore, the invention incorporates an emotion engine that recognizes user emotions and has a function to optimize the user experience based on those emotions. This system includes multiple programs running on a server and terminals used by users.

[0647] System Overview

[0648] 1. Document collection and text conversion

[0649] Server: Collects existing documents (PDF, Word, Excel, etc.) from various folders and databases within the company. It periodically runs collection scripts to retrieve the latest documents. The collected documents are converted into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are also converted to text using a dedicated library.

[0650] 2. Analysis and Classification of Text Data

[0651] Server: The converted documents are analyzed using natural language processing (NLP) techniques. The document content is categorized into sections and appropriately tagged. For example, tags such as "project management" and "sales strategy" may be applied.

[0652] 3. Duplicate detection and unification

[0653] Server: Detects duplicate content from the analysis results and combines it into a single file. Compares duplicate sections and paragraphs to generate a unified text. This eliminates unnecessary duplication and produces a well-organized manual.

[0654] 4. Verification with business result data and automatic updating

[0655] Server: This server periodically collects business performance data from companies (e.g., sales figures and project progress). The collected data is compared with document content, and if there are discrepancies, the document is automatically updated to reflect the latest information.

[0656] 5. User Authentication and Permission Management

[0657] Server: Users attempting to access the manual undergo identity verification through an authentication system such as LDAP. The server verifies user access rights and grants access to the document only to users with appropriate permissions.

[0658] 6. Feedback function and document revision

[0659] Users: Users can view the manual and provide feedback. For example, they can report if a particular procedure is outdated. The feedback is sent to the server, analyzed, and necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent access.

[0660] 7. Integration of the Emotional Engine

[0661] Server: The system incorporates an emotion engine that recognizes user emotions. It analyzes user emotions in real time based on user input and usage, and adjusts the user experience based on the recognized emotions. For example, if a user is feeling stressed, it can prioritize providing support information.

[0662] 8. Adjusting document content based on emotions

[0663] Server: Based on emotional feedback collected by the emotion engine, the document content is automatically adjusted. In this way, users can utilize the information more efficiently.

[0664] Specific example

[0665] For example, consider a company with two documents: "Sales Manual.pdf" and "Project Management.docx". The server collects these documents and converts them to text. Next, it analyzes each document using natural language processing, classifying and tagging them into sections such as "Sales Policy" and "Project Management Methods". Duplicate parts are detected from the analysis results and reconstructed into a single, unified document. Subsequently, the latest sales performance data is collected and reflected in the manual, automatically updating its content. When users access the manual, authentication is performed, ensuring that only users with appropriate permissions can access the information. When user feedback is provided, the server analyzes it and revises the document.

[0666] Furthermore, an emotion engine recognizes user emotions based on their input and usage. For example, if a user repeatedly refers to a specific section of a manual and expresses confusion, the system will offer additional support regarding that section. It also automatically adjusts the document content based on emotion data, making it easier to understand on subsequent accesses.

[0667] This allows for the efficient collection and analysis of internal business manuals, providing them in a unified format free of duplication, while also reflecting the latest business results and enabling optimized information delivery based on user sentiment.

[0668] ---

[0669] The following describes the processing flow.

[0670] Step 1:

[0671] The server collects existing documents (PDFs, Word documents, Excel files, etc.) from specific folders or databases within the company. The collection script is run periodically to retrieve the most recent documents.

[0672] Step 2:

[0673] The server converts collected documents into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are converted to text using a dedicated library.

[0674] Step 3:

[0675] The server analyzes the converted text data using natural language processing (NLP) techniques. It categorizes the document content into sections and assigns appropriate tags. For example, tags such as "project management" and "sales strategy" might be applied.

[0676] Step 4:

[0677] The server detects duplicate content based on the analysis results. It compares similar sections and paragraphs and combines them into one. By generating integrated text, unnecessary duplication is eliminated, resulting in a well-organized manual.

[0678] Step 5:

[0679] The server formats the integrated documents and automatically generates a standardized table of contents and index. The unified documents are then stored in a database.

[0680] Step 6:

[0681] The server periodically collects business result data (sales performance, project progress, etc.) from various business systems within the company via APIs.

[0682] Step 7:

[0683] The server compares the collected work result data with the contents of the manual. Based on the comparison results, if there are any discrepancies, the document is automatically updated to reflect the latest information.

[0684] Step 8:

[0685] When a user accesses the manual, the server authenticates the user using an authentication system such as LDAP. It verifies the user's permissions and grants access to the document only to users with appropriate privileges.

[0686] Step 9:

[0687] Users view the manual and report areas for improvement or deficiencies using the feedback form. The feedback is then sent to the server.

[0688] Step 10:

[0689] The server analyzes the collected feedback and makes necessary revisions. The revised document is saved back to the database and reflected in subsequent accesses.

[0690] Step 11:

[0691] The server operates an emotion engine that recognizes user emotions in real time based on user input and usage patterns. For example, it analyzes user emotions based on keyboard input speed, mouse movements, and specific keywords.

[0692] Step 12:

[0693] Based on the emotions the server perceives, it displays revised manuals and support information to the user. For example, if the server detects that the user is confused, it will provide additional help information.

[0694] Step 13:

[0695] The server adjusts the document content based on sentiment data collected by the sentiment engine. The document content is revised to be more understandable based on user sentiment feedback.

[0696] Step 14:

[0697] The server collects emotional data and uses it to provide optimal information tailored to specific users. For example, it can improve sections that previously confused users, ensuring that appropriate information is displayed on subsequent visits.

[0698] Through the steps described above, internal business manuals are efficiently collected and analyzed, generating unified documents with no duplication. Furthermore, the system can reflect the latest user business data and optimize the user experience using an emotion engine.

[0699] (Example 2)

[0700] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0701] When efficiently collecting existing documents within a company and managing them as a unified manual, it takes time and effort to address document duplication and ensure that the latest information is reflected. Furthermore, it is difficult to quickly incorporate user feedback and optimize information based on user sentiment. This leads to problems such as a decline in the quality of the user experience and a decrease in operational efficiency.

[0702] The identification processing performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting existing documents and converting them into text, means for analyzing, classifying, and tagging the text data, means for detecting and integrating duplicate content, means for recognizing user emotions and optimizing the user experience based on those emotions, and means for automatically adjusting document content based on emotional feedback. This enables efficient collection and management of documents, rapid reflection of feedback, and optimization of information based on user emotions.

[0703] "Existing documents" refer to various files such as PDFs, Word documents, and Excel spreadsheets that have already been created within the company.

[0704] "Means of converting to text" refers to the process of converting existing documents into text data using OCR, PDF text extraction tools, program libraries, and the tools used for that purpose.

[0705] "Means for analyzing, classifying, and tagging text data" refers to the process of analyzing text data using natural language processing technology and assigning classifications and tags to specific sections.

[0706] "Means for detecting and integrating duplicate content" refers to the process of detecting duplicate content contained within text data and integrating it into a single entry.

[0707] "Means for generating unified documents" refers to the process of organizing redundant content and generating documents that are stored in a unified format.

[0708] "Means of collecting business result data and comparing it with document content" refers to the process of regularly collecting business result data for a company and comparing it with the content of existing documents.

[0709] "Methods for automatically updating documents" refer to processes that automatically update existing documents based on business result data.

[0710] "Means of verifying user authentication and authority" refers to the process of authenticating a user using authentication systems such as LDAP or OAuth, and verifying their authority.

[0711] "Means for users to provide feedback" refers to web forms or interfaces that allow users to provide feedback on the content of a document.

[0712] "Means of revising documents based on feedback" refers to the process of revising documents based on feedback provided by users.

[0713] "A means of recognizing user emotions and optimizing the user experience based on those emotions" refers to the process of using an emotion engine to analyze user emotions in real time and adjusting use cases based on those emotions.

[0714] "Methods for automatically adjusting document content based on emotional feedback" refers to a process that automatically adjusts the content of a document based on the user's emotional feedback, providing optimized information for the user's next access.

[0715] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. Furthermore, the invention incorporates an emotion engine that recognizes user emotions and has the function of optimizing the user experience based on those emotions. This system includes multiple programs running on a server and terminals used by users.

[0716] System Overview

[0717] 1. Document collection and text conversion

[0718] The server collects existing documents from folders and databases within the company. This collection uses Python's os module and cloud storage APIs (e.g., AWS S3, Google Drive API). The collection script is run periodically to retrieve the latest documents. The collected documents are converted into text data using optical character recognition (OCR) or PDF text extraction tools (e.g., PyPDF2, python-docx, pandas, Tesseract OCR).

[0719] 2. Analysis and Classification of Text Data

[0720] The server analyzes the text data using natural language processing (NLP) tools (e.g., SpaCy, NLTK). It categorizes the document content into sections and assigns appropriate tags (e.g., "Project Management," "Sales Strategy").

[0721] 3. Duplicate detection and unification

[0722] The server detects duplicate content from the analysis results and combines it into a single file. The difflib module is used for text comparison. It generates a unified text file and saves it with unnecessary duplicates removed.

[0723] 4. Verification with business result data and automatic updating

[0724] The server periodically collects business performance data (e.g., sales figures and project progress) from SQL or NoSQL databases (e.g., MySQL, MongoDB). The collected data is compared with the document content, and if there are any discrepancies, the document is automatically updated based on the latest information.

[0725] 5. User Authentication and Permission Management

[0726] The server authenticates users using LDAP or OAuth when they access documents. Once authentication is complete, it checks the user's permissions and grants access to the documents only to users with the appropriate permissions.

[0727] 6. Feedback function and document revision

[0728] Users can view documents through the system and provide feedback. This feedback is sent to the server via web forms or other means. The server analyzes the feedback and revises the document accordingly.

[0729] 7. Integration of the Emotional Engine

[0730] The server uses an emotion engine (e.g., Azure Text Analytics, IBM Watson) to analyze user sentiment in real time based on user input and usage. Based on the recognized sentiment, the user experience can be tailored accordingly.

[0731] 8. Adjusting document content based on emotions

[0732] The server automatically adjusts document content based on sentiment feedback, allowing users to access the information more efficiently the next time they visit.

[0733] Specific example

[0734] For example, consider a company where there are two documents: "Sales Manual.pdf" and "Project Management.docx". The server collects these documents and converts them to text using PyPDF2 or python-docx. Next, the converted text is analyzed using SpaCy, classifying each document into sections such as "Sales Policy" and "Project Management Methods," and tagging them. Duplicate parts are detected from the analysis results, and the documents are reconstructed as a single, unified document.

[0735] Subsequently, the latest sales performance data is collected from the MySQL database and reflected in the manual, automatically updating its content. When users access the manual, they are authenticated via LDAP, ensuring that only users with appropriate permissions can access the information. When user feedback is provided, the server analyzes it and revises the document accordingly.

[0736] Furthermore, the emotion engine analyzes user input and usage in real time, and based on the recognized emotions, if confusion or stress is detected regarding a particular section, the system will offer additional support for that section. It also automatically adjusts document content based on emotion data, making it easier to understand on subsequent accesses.

[0737] Example of a prompt

[0738] How can I use a generative AI model to detect user emotions and adjust the content of company manuals based on those emotions?

[0739] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0740] Step 1: Document collection and text conversion

[0741] The server collects existing documents from designated folders and databases within the company. Input is the folder path and database connection information, while output is the collected document files.

[0742] Specifically, the server retrieves files using Python's os module and cloud storage APIs.

[0743] Based on the entered folder path, the system searches for document files and uploads them to the database.

[0744] Step 2: Text conversion

[0745] The server converts collected documents into text data. Input is document files (e.g., PDF, Word, Excel), and output is the converted text data.

[0746] Tools such as PyPDF2, python-docx, pandas, and Tesseract OCR will be used.

[0747] PDF files are extracted using Optical Character Recognition (OCR), while Word and Excel files are converted to text using a dedicated library.

[0748] Step 3: Analysis and classification of text data

[0749] The server analyzes text data using natural language processing (NLP) tools. The input is the transformed text data, and the output is the analyzed section data and tags.

[0750] Specifically, the server uses SpaCy or NLTK to analyze the text, classify the document content into sections, and assign appropriate tags.

[0751] Based on the analyzed text, tags such as "Project Management" and "Sales Strategy" are added to each section.

[0752] Step 4: Duplicate detection and unification

[0753] The server detects duplicate content from the analysis results and combines it into a single file. The input is tagged section data, and the output is unified content with duplicates removed.

[0754] The server uses the difflib module to detect duplicate text.

[0755] Merge highly similar texts and remove unnecessary duplicates.

[0756] Step 5: Match with business result data and automatic update

[0757] The server periodically collects business results data from companies and compares it with document content. The input is information from the business results database, and the output is a document reflecting the latest information.

[0758] The server uses MySQL or MongoDB to retrieve the latest sales performance and project progress data from the database.

[0759] Compare the business results data with the documents, and update the document content if there are any discrepancies.

[0760] Step 6: User Authentication and Permission Management

[0761] The server authenticates the user and verifies their permissions. The input is the user's authentication information (e.g., user ID, password), and the output is the authentication and permission verification result.

[0762] The server uses LDAP or OAuth to authenticate users.

[0763] Once authentication is complete, the system verifies the user's access rights and grants access to the document only to users with the appropriate permissions.

[0764] Step 7: Feedback function and document revision

[0765] Users provide feedback on a document. The input is the content of the feedback, and the output is the document with the feedback reflected.

[0766] Feedback is sent to the server via web forms or similar means.

[0767] The server analyzes the feedback, identifies areas that need improvement, and revises the document accordingly.

[0768] Step 8: Integrating the Emotional Engine

[0769] The server uses an emotion engine to analyze the user's emotions. The input is the user's input data and feedback, and the output is the emotion analysis result.

[0770] The server utilizes sentiment analysis APIs (e.g., Azure Text Analytics, IBM Watson).

[0771] The user experience is adjusted based on recognized emotions, and additional support information is provided if negative emotions are detected.

[0772] Step 9: Adjusting document content based on emotions

[0773] The server automatically adjusts the document content based on emotional feedback. The input is emotional feedback data, and the output is the adjusted document content.

[0774] The server modifies the document content based on sentiment data, providing optimized information for the user's next access.

[0775] Specifically, this involves adding a QA section and simplifying the content of the documents.

[0776] This enables efficient document collection and management, rapid incorporation of feedback, and optimization of information based on user sentiment.

[0777] (Application Example 2)

[0778] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0779] Modern businesses are required to manage and appropriately utilize a wide variety of documents. Simultaneously, it is crucial to recognize user emotions in real time and provide information efficiently. However, existing systems struggle to integrate document management and user support, and real-time work instructions and emotion-based support are particularly lacking in manufacturing environments. This makes it difficult to optimize the user experience while simultaneously demanding document consistency, up-to-date information, and improved work efficiency. Therefore, this invention proposes a system that solves these problems, providing efficient and unified document management and real-time support tailored to user emotions.

[0780] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0781] In this invention, the server includes means for collecting existing documents and converting them to text; means for analyzing, classifying, and tagging the text data; means for detecting and integrating duplicate content; means for generating a unified document; means for collecting work result data and comparing it with document content; means for automatically updating documents; means for verifying user authentication and permissions; means for users to provide feedback; means for revising documents based on feedback; means for providing work instructions in real time; means for recognizing user emotions and optimizing the experience; and means for providing additional support information based on emotion recognition. This enables efficient document management within the company, real-time work support, and optimal information provision based on user emotions.

[0782] "Means of collecting documents and converting them into text" refers to a function that converts existing documents within a company, such as manuals, procedures, and reports, into analyzable text data.

[0783] "Means for analyzing, classifying, and tagging text data" refers to a function that analyzes the converted text data using natural language processing technology and assigns appropriate tags to each section.

[0784] "Means for detecting and integrating duplicate content" refers to a function that detects duplicate parts from analyzed text data and combines them into a single entity.

[0785] "Means for generating unified documents" refers to a function that generates consistent and unified manuals and reports based on text data that has been merged and whose duplicates have been eliminated.

[0786] "Means for collecting business result data and comparing it with document content" refers to a function that collects company performance and project progress data and compares and verifies it with existing document content.

[0787] "Methods for automatically updating documents" refers to a function that automatically updates documents to the latest content based on collected performance data and project progress data.

[0788] "Means of verifying user authentication and permissions" refers to functions that authenticate users attempting to access manuals and reports, and verify and manage appropriate permissions.

[0789] "Means for users to provide feedback" refers to features that allow users to provide feedback and comments on manuals and reports.

[0790] "Means for revising documents based on feedback" refers to a function that analyzes collected user feedback and revises documents as needed.

[0791] "Means of providing work instructions in real time" refers to a function that provides the instructions and procedures that workers need in real time.

[0792] "Means of recognizing user emotions and optimizing the experience" refers to a function that recognizes the emotional state of workers in real time and optimizes user support based on that information.

[0793] "Means of providing additional support information based on emotion recognition" refers to a function that provides necessary support information and additional instructions based on the user's emotional state.

[0794] Modes for carrying out the invention

[0795] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. Furthermore, the invention incorporates an emotion engine that recognizes user emotions and has a function to optimize the user experience based on those emotions. This system includes multiple software programs running on a server and terminals used by users.

[0796] System Overview

[0797] 1. Document collection and text conversion

[0798] The server collects existing documents (PDFs, Word documents, Excel files, etc.) from various folders and databases within the company. A collection script is run periodically to retrieve the latest documents. The collected documents are converted into text data using optical character recognition (OCR) tools and PDF text extraction tools. Word and Excel files are also converted to text data using dedicated libraries. Specifically, tools such as Tesseract OCR and PDFMiner can be used.

[0799] 2. Analysis and Classification of Text Data

[0800] The server analyzes the converted text data using natural language processing (NLP) techniques. It classifies the document content into sections and applies appropriate tags. For example, NLP libraries such as spaCy or NLTK can be used.

[0801] 3. Duplicate detection and unification

[0802] The server detects and merges duplicate content from the analysis results. It compares overlapping sections and paragraphs to generate a single, unified document. This eliminates unnecessary duplication, resulting in a more organized manual.

[0803] 4. Verification with business result data and automatic updating

[0804] The server periodically collects business performance data from the company (e.g., sales figures and project progress). The collected data is compared with the document content, and if there are any discrepancies, the document is automatically updated to reflect the latest information.

[0805] 5. User Authentication and Permission Management

[0806] The server authenticates users attempting to access manuals and reports. It uses authentication systems such as LDAP to verify identity and confirm user access rights. Only users with appropriate permissions are granted access to the documents.

[0807] 6. Feedback function and document revision

[0808] Users can view manuals and reports and provide feedback. For example, they can report if a particular procedure is outdated. The feedback is sent to the server, analyzed, and necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent accesses.

[0809] 7. Integration of the Emotional Engine

[0810] The server incorporates an emotion engine that recognizes user emotions. It analyzes emotions in real time based on user input and usage, and adjusts the user experience based on the recognized emotions. For example, if a user is experiencing stress, it can prioritize providing supportive information. TensorFlow and OpenCV can be used as emotion recognition models.

[0811] 8. Adjusting document content based on emotions

[0812] The server automatically adjusts document content based on emotional feedback collected by the emotion engine. This allows users to utilize information more efficiently.

[0813] Specific example

[0814] When a new manufacturing process is introduced in a factory, the server collects relevant documents and generates standardized work instructions. When workers receive instructions using smart glasses, the emotion engine provides additional support information if it detects stress. The server uses TensorFlow to perform emotion recognition and provides optimal information in real time to reduce worker confusion.

[0815] Example of a prompt

[0816] Develop standard procedures for a new manufacturing process and design a real-time work guidance system that incorporates emotion recognition capabilities. This system should include features that allow workers to receive instructions in real time via a head-mounted display and receive additional support based on their emotional state.

[0817] Thus, the system of the present invention streamlines document management within a company and realizes real-time work support and optimized user experience.

[0818] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0819] Step 1: Document collection and text conversion

[0820] The server collects existing documents (PDF, Word, Excel, etc.) from various databases and folders within the company. The collected documents are converted into text data using OCR tools (e.g., Tesseract OCR) or PDF text extraction tools. The input is various document files, and the output is their corresponding text data.

[0821] Step 2: Analysis and classification of text data

[0822] The server analyzes the converted text data using natural language processing (NLP) techniques (e.g., spaCy, NLTK). It analyzes each section of the document and assigns appropriate tags to them. The input is text data, and the output is the analysis results and tagged data.

[0823] Step 3: Duplicate detection and merging

[0824] The server detects duplicate content from the parsed and tagged text data. It then compares and merges the duplicate sections to generate a single, unified document. The input is the parsed and tagged text data, and the output is the unified document. Specifically, it uses an algorithm that merges paragraphs and sections with high similarity.

[0825] Step 4: Matching with business result data and automatic updating

[0826] The server periodically collects corporate performance data and project progress data. It then compares this data with existing document content and automatically updates it to reflect the latest information. The input consists of performance data and existing documents, while the output is the updated documents. Specifically, it uses database queries and matching algorithms.

[0827] Step 5: User Authentication and Permission Management

[0828] The server authenticates users through an authentication system such as LDAP when they attempt to access manuals or reports. It verifies the user's access rights and allows access only to those with appropriate permissions. The input is the user's authentication information, and the output is the result of the access rights verification. Specifically, it performs LDAP queries and generates authentication tokens.

[0829] Step 6: Feedback function and document revision

[0830] Users can provide feedback while viewing manuals and reports. The feedback is sent to the server, analyzed, and necessary revisions are made. The input is user feedback, and the output is the revised document. Specifically, a feedback analysis algorithm is used to update the document content.

[0831] Step 7: Provide work instructions in real time

[0832] The server provides workers with real-time instructions and procedures through smart glasses or head-mounted displays. Input is data from the work environment, and output is real-time work instructions. Specifically, it performs tasks such as presenting predefined work procedures and analyzing environmental data.

[0833] Step 8: Emotion Recognition and Experience Optimization

[0834] The server recognizes user emotions in real time and optimizes the user experience based on that. TensorFlow is among the emotion recognition models used. Inputs include user facial images and usage data, while output is optimized user experience information. Specifically, it uses facial recognition algorithms and emotion recognition models (e.g., TensorFlow, OpenCV).

[0835] Step 9: Provide additional support information based on emotion recognition.

[0836] The server provides necessary support information and additional instructions based on the user's emotional state. Input consists of the results of emotion recognition and data regarding the user's confusion and stress; output is additional support information. Specifically, it searches for appropriate support information and presents it to the user.

[0837] In summary, this system streamlines document management within companies and provides real-time work support and optimized user experience.

[0838] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0839] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0840] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0841] [Third Embodiment]

[0842] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0843] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0844] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0845] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0846] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0847] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0848] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0849] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0850] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0851] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0852] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0853] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0854] ---

[0855] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. This system includes multiple programs running on a server and terminals used by users. The system also features a function to automatically update documents based on user feedback and business result data.

[0856] System Overview

[0857] 1. Document collection and text conversion

[0858] Server: Collects existing documents (PDF, Word, Excel, etc.) from various folders and databases within the company. The collected documents are converted into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are also converted to text using a dedicated library.

[0859] 2. Analysis and Classification of Text Data

[0860] Server: The converted text data is analyzed using natural language processing (NLP) techniques. During the analysis process, the document is categorized into sections and assigned appropriate tags. For example, tags such as "project management" and "sales strategy" may be added.

[0861] 3. Duplicate detection and unification

[0862] Server: Text analysis detects duplication when the same content is found in multiple documents. The duplicate content is combined into a single, unified document. This eliminates unnecessary duplication and generates a well-organized manual.

[0863] 4. Verification with business result data and automatic updating

[0864] Server: The server periodically collects business performance data (e.g., sales figures and project progress). The collected data is compared with document content, and if discrepancies are found, the document is automatically updated. This ensures that manuals always reflect the latest information.

[0865] 5. User Authentication and Permission Management

[0866] Server: Users attempting to access the manual must first undergo identity verification through an authentication system. Authentication technologies such as LDAP are used to verify user access rights. This ensures that only users with appropriate permissions can access specific information.

[0867] 6. Feedback function and document revision

[0868] Users: Users can view the manual and provide feedback. For example, they can report if a particular procedure is outdated. The feedback is sent to the server, analyzed, and necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent access.

[0869] Specific example

[0870] For example, consider a company with two documents: "Sales Manual.pdf" and "Project Management.docx". The server collects these documents and converts them to text. Next, it analyzes each document using natural language processing, classifying and tagging them into sections such as "Sales Policy" and "Project Management Methods". Duplicate parts are detected from the analysis results and reconstructed into a single, unified document. Subsequently, the latest sales performance data is collected and reflected in the manual, automatically updating its content. When users access the manual, authentication is performed, ensuring that only users with appropriate permissions can access the information. When user feedback is provided, the server analyzes it and revises the document.

[0871] This allows for centralized management of internal business manuals, ensuring they are always up-to-date, and guaranteeing the secure handling of confidential information.

[0872] ---

[0873] The following describes the processing flow.

[0874] Step 1:

[0875] The server collects existing documents (PDFs, Word documents, Excel files, etc.) from specific folders or databases within the company. The collection script is run periodically to retrieve the most recent documents.

[0876] Step 2:

[0877] The server converts collected documents into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are converted to text using a dedicated library.

[0878] Step 3:

[0879] The server analyzes the converted text data using natural language processing (NLP) techniques. It analyzes the document content, categorizes it into sections, and applies appropriate tags.

[0880] Step 4:

[0881] The server detects duplicate content based on the analysis results. It compares similar sections and paragraphs and integrates them into a single, unified text.

[0882] Step 5:

[0883] The server formats the integrated documents and automatically generates a standardized table of contents and index. The unified documents are then stored in a database.

[0884] Step 6:

[0885] The server periodically collects business result data (sales performance, project progress, etc.) from various business systems within the company via APIs.

[0886] Step 7:

[0887] The server compares the collected work result data with the contents of the manual. If a discrepancy is detected, it automatically performs the necessary updates and updates the document.

[0888] Step 8:

[0889] When a user accesses a manual, the server authenticates the user using an authentication system such as LDAP. It verifies the user's permissions and grants access to the corresponding document only to users with the appropriate privileges.

[0890] Step 9:

[0891] Users view the manual and report areas for improvement or deficiencies using a feedback form. The feedback data is sent to the server.

[0892] Step 10:

[0893] The server analyzes the collected feedback and revises the document. The revised document is saved back to the database and the changes are reflected the next time it is accessed.

[0894] These steps ensure that internal operational manuals are efficiently collected and analyzed, and provided in a unified format free of duplication. Furthermore, they remain up-to-date by reflecting the latest operational results, and are quickly revised based on user feedback.

[0895] (Example 1)

[0896] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0897] The diverse range of documents generated within a company increases over time, making their management complex. The existence of multiple documents can lead to information duplication and inconsistencies, hindering efficient work execution. Furthermore, manual document updates are labor-intensive and can delay the reflection of the latest information. In addition, a lack of proper user authentication and access control increases the risk of confidential information leaks. Addressing these challenges is essential.

[0898] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0899] In this invention, the server includes means for collecting existing documents and converting them into text, means for analyzing, classifying, and tagging the text data, and means for detecting and integrating duplicate content. This streamlines document management within the company, eliminates duplication and inconsistencies, and generates a unified manual that always reflects the latest information. Furthermore, user authentication and permission verification ensure the secure handling of confidential information. In addition, document revisions are automatically performed based on feedback provided by users, ensuring that the latest information is always reflected.

[0900] "Document collection methods" refer to methods for collecting existing documents from folders and databases within a company.

[0901] "Text conversion means" refers to means of converting collected documents into text data. This includes optical character recognition and document text extraction tools.

[0902] "Text analysis means" refers to means for analyzing the converted text data, and this uses natural language processing technology.

[0903] "Classification and tagging means" refers to means of classifying the analyzed text data into sections and assigning appropriate tags.

[0904] A "duplicate detection means" is a means of calculating the similarity between analyzed text data and detecting duplicate content.

[0905] A "combination method" is a means of integrating duplicate text data into a single, unified document.

[0906] "Methods for collecting business data" refer to methods for periodically collecting business result data from a company, and this includes the use of various APIs.

[0907] A "document verification means" is a means of verifying collected business data against the contents of documents.

[0908] An "automatic update mechanism" is a means of automatically updating the document content based on the verification results.

[0909] "User authentication means" refers to methods for verifying a user's identity, and this includes using LDAP or authentication protocols.

[0910] A "permission verification method" is a means of verifying a user's access rights and ensuring that only users with appropriate permissions can access the information.

[0911] A "feedback provision method" is a means by which users can provide feedback on a document.

[0912] "Document revision methods" refer to methods for revising documents based on user feedback and reflecting the latest information.

[0913] ---

[0914] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. This system includes multiple programs running on a server and terminals used by users. The system also features a function to automatically update documents based on user feedback and business result data.

[0915] System Overview

[0916] Document collection and text conversion

[0917] The server collects existing documents (PDF, Word, Excel, etc.) from various folders and databases within the company. This collection process utilizes remote file transfer methods, specifically SFTP (Secure File Transfer Protocol). The collected documents are converted into text data using optical character recognition (OCR) and document text extraction tools. ABBYY FineReader is used for OCR, and the PDF library is used for PDF text extraction. Word and Excel files are converted to text using the Apache POI library.

[0918] Text data analysis and classification

[0919] The server analyzes the converted text data using natural language processing (NLP) techniques. Software such as Google Cloud Natural Language and SpaCy are used for analysis, including text tokenization and morphological analysis. As a result of the analysis, the documents are categorized into sections and tagged with terms like "Project Management" and "Sales Strategy." Machine learning techniques (e.g., k-means clustering) are used for classification.

[0920] Duplicate detection and unification

[0921] The server detects duplicates when multiple documents contain the same content based on the analysis results. Text similarity calculation methods (such as Jaccard similarity or Cosine similarity) are used for duplicate detection. Duplicate content is combined into a single, unified document. The most reliable text and the most recent content are prioritized for unification as needed.

[0922] Matching with business result data and automatic updating

[0923] The server periodically collects various business result data from the company (e.g., sales performance data and project progress data). This data collection often utilizes APIs, specifically the Salesforce API and other business system APIs. The collected data is compared with document content, and if discrepancies are found, the documents are automatically updated. This update process uses PHP or Python scripts.

[0924] User authentication and access control

[0925] When a user attempts to access a manual, the server first verifies their identity through an authentication system. It uses LDAP (Lightweight Directory Access Protocol) or OAuth protocols to verify the user's access rights. Only users who successfully authenticate are allowed to access specific documents with the appropriate permissions.

[0926] Feedback function and document revisions

[0927] Users can provide feedback while browsing the manual. This can be done by filling out a feedback form and submitting it, noting that specific steps or information are outdated. The feedback is sent to the server, which uses natural language processing to analyze the content and determine if the document needs revision. If necessary, the document is revised and saved again in the database.

[0928] Specific example

[0929] For example, if a company has documents such as "Sales Manual.pdf" and "Project Management.docx," the server collects these documents and converts them into text data. Next, the converted text data is analyzed using Google Cloud Natural Language and tagged with terms such as "Sales Policy" and "Project Management Methods." Based on the analysis results, duplicate parts are detected using Jaccard similarity and reconstructed into a single, unified document. The latest sales performance data from Salesforce is collected and reflected in the manual. When users access the manual, authentication is performed using LDAP. If a user provides feedback that a particular procedure is outdated, the server analyzes the feedback, makes the necessary revisions, and saves them in the database.

[0930] Example of a prompt

[0931] Examples of prompts for a generative AI model are as follows:

[0932] Please describe in detail, step by step, a system for collecting existing documents within a company and generating and maintaining a unified manual. This system includes steps such as document collection via SFTP, OCR processing, NLP analysis, duplicate detection, LDAP authentication, and feedback analysis.

[0933] ---

[0934] The above is the description of the "Mode for Carrying Out the Invention" in the patent specification.

[0935] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0936] Step 1:

[0937] Collect existing documents.

[0938] The server accesses designated folders and databases within the company to collect documents.

[0939] The input consists of a folder path and database connection information.

[0940] Specifically, the system will perform the operation of downloading PDF, Word, and Excel files from a remote folder using SFTP.

[0941] The output includes the collected document files.

[0942] Step 2:

[0943] Convert a document into text data.

[0944] The server converts the collected documents into text data using optical character recognition (OCR) and document text extraction tools.

[0945] The document files collected in Step 1 are provided as input.

[0946] Specifically, the process involves using ABBYY FineReader to convert PDFs into text data, and then using the Apache POI library to extract text from Word and Excel files.

[0947] The output includes the converted text data.

[0948] Step 3:

[0949] Analyze text data and perform classification and tagging.

[0950] The server uses natural language processing (NLP) techniques to analyze the converted text data.

[0951] The text data obtained in step 2 is provided as input.

[0952] Specifically, the system uses the Google Cloud Natural Language API and SpaCy to perform tokenization and morphological analysis, and then analyzes the meaning and structure of the text.

[0953] The output includes analytical data categorized by section and tagged appropriately.

[0954] Step 4:

[0955] It detects duplicate content and generates a unified document.

[0956] The server detects duplicate content from the analysis results and reconstructs it into a unified document.

[0957] The analysis data obtained in step 3 is provided as input.

[0958] Specifically, the process involves calculating the degree of text duplication using Jaccard similarity or Cosine similarity, and then combining the duplicated parts into a single text.

[0959] The output includes a unified document with duplicates removed and organized.

[0960] Step 5:

[0961] Collect business result data and compare it with document content.

[0962] The server periodically collects business results data from the company and compares it with document content.

[0963] The input includes business result data collected using the API.

[0964] Specifically, the system will use business system APIs, such as the Salesforce API, to retrieve the latest sales performance and project progress data.

[0965] The output includes the results of comparing the collected business data with the document content.

[0966] Step 6:

[0967] Automatically update the document.

[0968] The server automatically updates the document content based on the matching results.

[0969] The input is provided with the matching result obtained in step 5.

[0970] Specifically, the system will use PHP or Python scripts to update document content and reflect the latest information.

[0971] The output will include the latest updated document.

[0972] Step 7:

[0973] Verify user authentication and permissions.

[0974] The server authenticates users when they access the manual.

[0975] The user's authentication information is provided as input.

[0976] Specifically, it performs user authentication and authorization verification using LDAP and OAuth protocols.

[0977] The output includes the authentication results and the user's access permissions.

[0978] Step 8:

[0979] Users provide feedback and revise the document.

[0980] Users provide feedback while viewing the manual.

[0981] The input is the feedback content.

[0982] Specifically, the process involves sending feedback to the server through a feedback form.

[0983] The server analyzes the feedback and revises the document.

[0984] User feedback is provided as input.

[0985] Specifically, this involves using NLP tools to understand the feedback and making necessary revisions.

[0986] The output will include the revised document.

[0987] (Application Example 1)

[0988] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0989] Work manuals and procedures used by robots operating in factories need to be updated frequently. However, manual updates are inefficient, making it difficult to maintain consistency and up-to-dateness of information. Furthermore, the lack of a mechanism to incorporate real-time feedback from robots during actual work can lead to decreased accuracy and efficiency. In addition, there is a need for a mechanism to quickly update robot operating procedures when they become outdated.

[0990] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0991] In this invention, the server includes means for collecting existing documents and converting them into text; means for analyzing, classifying, and tagging the text data; means for detecting and integrating duplicate content; means for generating a unified document; means for collecting work result data and comparing it with the document content; means for automatically updating the document; means for verifying user authentication and authorization; means for users to provide feedback; means for revising the document based on the feedback; means for robots to provide real-time feedback during work and automatically revising the document based on that data; and means for automatically updating the robot's operating procedures. This makes it possible for robots in the factory to always operate based on the latest work manuals, improving the accuracy and efficiency of their work.

[0992] "Methods for collecting existing documents and converting them into text" refers to the process of gathering documents of various formats that exist within a factory into one place and converting them into machine-readable text data using OCR technology and a dedicated library.

[0993] "Means for analyzing, classifying, and tagging text data" refers to the process of analyzing each piece of text data using natural language processing technology, classifying it into sections based on its content, and assigning appropriate labels and tags.

[0994] "Methods for detecting and integrating duplicate content" refers to the process of extracting and detecting identical or similar content from analyzed text data and combining them into a single, unified document.

[0995] "Methods for generating unified documents" refers to the process of integrating redundant content and then compiling the resulting organized documents into a single, complete document.

[0996] "Means for collecting business result data and comparing it with document content" refers to the process of regularly collecting performance data related to operations within the factory and comparing and verifying it with existing documents.

[0997] "Methods for automatically updating documents" refer to a process that automatically updates the content of documents as needed, based on the results of cross-referencing with business result data, to reflect the most accurate information.

[0998] "Means of verifying user authentication and permissions" refers to the process of verifying the identity of the user attempting to access a document and checking whether they have the appropriate access rights.

[0999] "Means for users to provide feedback" refers to a function that allows users to provide feedback when viewing a document, and a process to prompt corrections if the content is outdated.

[1000] "Means of revising documents based on feedback" refers to the process of analyzing feedback provided by users and modifying or revising the content of the document based on that feedback.

[1001] "A method for robots to provide real-time feedback during work and automatically revise documents based on that data" refers to a process in which robots record information and feedback obtained during actual work in real time and automatically revise documents based on that data.

[1002] "Means for automatically updating robot operating procedures" refers to a process that automatically updates robot operating procedures based on the latest collected data and feedback, thereby improving work efficiency and accuracy.

[1003] System Overview

[1004] This invention is a system for keeping the work manuals and operating procedures for robots operating in a factory constantly up-to-date. The system includes multiple programs running on a server and applications installed on the robots. The system also has the function of receiving real-time feedback from the robots and automatically updating the documents. This can improve the work efficiency and accuracy of the robots.

[1005] Hardware and software configuration

[1006] Hardware: Server, robot, camera (for OCR)

[1007] Software: OCR tool (e.g., Tesseract), NLP engine (e.g., SpaCy), database (MySQL), authentication system (LDAP)

[1008] Document collection and text conversion

[1009] The server collects existing documents from various terminals and databases within the factory and uses OCR tools to convert the contents of PDFs and JPEGs into text data. The converted text data is further analyzed and processed using a dedicated library. For example, nouns and verbs are extracted and sentences are split.

[1010] Data analysis and classification

[1011] The server uses an NLP engine to analyze text data, classify it into sections, and assign tags to each section. This organizes the data by content, such as "operating procedures" or "maintenance procedures." Specifically, it uses a natural language processing engine (SpaCy) to analyze the parts of speech in the text data and automatically assigns important sections and labels.

[1012] Duplicate detection and merging

[1013] The system detects duplicate sections from the analyzed data and consolidates them into a single, unified document. This eliminates unnecessary duplication and generates a well-organized manual. This process utilizes a document consolidation algorithm. For example, it includes a process to combine identical operating procedures found in multiple documents into a single document.

[1014] Matching with business result data and automatic updating

[1015] The server periodically collects work result data and compares it with existing documents. If there are discrepancies, the documents are automatically updated. This ensures that the latest information is always reflected. Specifically, it retrieves work performance data from the work database and compares it with the analysis results.

[1016] User authentication and access management

[1017] When a user attempts to access a document, their identity is verified using an LDAP-based authentication system. Only users with appropriate permissions can access specific information. This ensures security.

[1018] Feedback function and document revision

[1019] The robot can provide real-time feedback during operation. For example, if there is an inconsistency in the operating procedure, it will provide feedback immediately. This feedback is sent to the server, analyzed, and any necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent access. This ensures that the procedure is up-to-date and reflects current practices.

[1020] Examples of usage and specific prompt statements

[1021] For example, if a robot starts performing a new work procedure and there is a contradiction with the existing work manual, the feedback function will report it as follows:

[1022] Prompt example:

[1023] "Please check the instructions in Step 2 of the manual. Since it does not match the actual work, it needs to be updated based on the latest work result data."

[1024] This allows robots in the factory to perform tasks efficiently based on the latest information at all times.

[1025] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1026] Step 1:

[1027] The server collects existing documents from various terminals and databases within the factory. Input documents include formats such as PDF, Word, and Excel. The server gathers these documents in one place via FTP or API. The output is a list of the collected document data.

[1028] Step 2:

[1029] The server converts the collected documents into text data using an OCR tool (e.g., Tesseract). The input is the collected document data. Using OCR or a dedicated library, this is converted into machine-readable text, and the output is text data.

[1030] Step 3:

[1031] The server analyzes text data using an NLP engine (e.g., SpaCy). The input is converted text data. Using natural language processing techniques, this data is analyzed, classified, and tagged by part of speech and section. The analyzed data is then output.

[1032] Step 4:

[1033] The server detects and integrates duplicate parts from the analyzed data. The input is the analyzed data. A document integration algorithm is used to combine identical or similar content into a single document. The integrated document is then output.

[1034] Step 5:

[1035] The server generates integrated documents as a single, unified document. The input is integrated data. Based on this, it generates documents in the most organized format. The output is a unified document.

[1036] Step 6:

[1037] The server collects business result data and compares it with existing document content. The input consists of business result data and a standardized document. It compares this data with work performance data retrieved from the business database, and if there are discrepancies, it automatically updates the content. The updated document is output.

[1038] Step 7:

[1039] The server verifies authentication and permissions when a user accesses a document. The input is the user's authentication information. The server uses an authentication system such as LDAP to verify the user's identity and permissions. The output is a list of users whose access permissions have been verified.

[1040] Step 8:

[1041] Users view documents and provide feedback. The input is user feedback information. The feedback is sent to the server, analyzed, and necessary revisions are made. The updated feedback information is recorded as output.

[1042] Step 9:

[1043] The server revises documents based on user feedback. Its inputs include feedback information and document data. Based on this data, it analyzes which parts of the document need revision and performs the revisions. The revised document is then output.

[1044] Step 10:

[1045] The robot provides real-time feedback during the task and automatically revises the document based on that data. The input is real-time work feedback data. The robot sends the feedback it provides to a server, which then revises the document based on that data. The output is a document reflecting the latest work procedure.

[1046] Step 11:

[1047] The server automatically updates the robot's operating procedures based on the latest collected data and feedback. The input is revised document data. Based on this, the robot's operating procedures are automatically updated, enabling operation based on the latest information. The updated operating procedures are obtained as output.

[1048] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1049] ---

[1050] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. Furthermore, the invention incorporates an emotion engine that recognizes user emotions and has a function to optimize the user experience based on those emotions. This system includes multiple programs running on a server and terminals used by users.

[1051] System Overview

[1052] 1. Document collection and text conversion

[1053] Server: Collects existing documents (PDF, Word, Excel, etc.) from various folders and databases within the company. It periodically runs collection scripts to retrieve the latest documents. The collected documents are converted into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are also converted to text using a dedicated library.

[1054] 2. Analysis and Classification of Text Data

[1055] Server: The converted documents are analyzed using natural language processing (NLP) techniques. The document content is categorized into sections and appropriately tagged. For example, tags such as "project management" and "sales strategy" may be applied.

[1056] 3. Duplicate detection and unification

[1057] Server: Detects duplicate content from the analysis results and combines it into a single file. Compares duplicate sections and paragraphs to generate a unified text. This eliminates unnecessary duplication and produces a well-organized manual.

[1058] 4. Verification with business result data and automatic updating

[1059] Server: This server periodically collects business performance data from companies (e.g., sales figures and project progress). The collected data is compared with document content, and if there are discrepancies, the document is automatically updated to reflect the latest information.

[1060] 5. User Authentication and Permission Management

[1061] Server: Users attempting to access the manual undergo identity verification through an authentication system such as LDAP. The server verifies user access rights and grants access to the document only to users with appropriate permissions.

[1062] 6. Feedback function and document revision

[1063] Users: Users can view the manual and provide feedback. For example, they can report if a particular procedure is outdated. The feedback is sent to the server, analyzed, and necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent access.

[1064] 7. Integration of the Emotional Engine

[1065] Server: The system incorporates an emotion engine that recognizes user emotions. It analyzes user emotions in real time based on user input and usage, and adjusts the user experience based on the recognized emotions. For example, if a user is feeling stressed, it can prioritize providing support information.

[1066] 8. Adjusting document content based on emotions

[1067] Server: Based on emotional feedback collected by the emotion engine, the document content is automatically adjusted. In this way, users can utilize the information more efficiently.

[1068] Specific example

[1069] For example, consider a company with two documents: "Sales Manual.pdf" and "Project Management.docx". The server collects these documents and converts them to text. Next, it analyzes each document using natural language processing, classifying and tagging them into sections such as "Sales Policy" and "Project Management Methods". Duplicate parts are detected from the analysis results and reconstructed into a single, unified document. Subsequently, the latest sales performance data is collected and reflected in the manual, automatically updating its content. When users access the manual, authentication is performed, ensuring that only users with appropriate permissions can access the information. When user feedback is provided, the server analyzes it and revises the document.

[1070] Furthermore, an emotion engine recognizes user emotions based on their input and usage. For example, if a user repeatedly refers to a specific section of a manual and expresses confusion, the system will offer additional support regarding that section. It also automatically adjusts the document content based on emotion data, making it easier to understand on subsequent accesses.

[1071] This allows for the efficient collection and analysis of internal business manuals, providing them in a unified format free of duplication, while also reflecting the latest business results and enabling optimized information delivery based on user sentiment.

[1072] ---

[1073] The following describes the processing flow.

[1074] Step 1:

[1075] The server collects existing documents (PDFs, Word documents, Excel files, etc.) from specific folders or databases within the company. The collection script is run periodically to retrieve the most recent documents.

[1076] Step 2:

[1077] The server converts collected documents into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are converted to text using a dedicated library.

[1078] Step 3:

[1079] The server analyzes the converted text data using natural language processing (NLP) techniques. It categorizes the document content into sections and assigns appropriate tags. For example, tags such as "project management" and "sales strategy" might be applied.

[1080] Step 4:

[1081] The server detects duplicate content based on the analysis results. It compares similar sections and paragraphs and combines them into one. By generating integrated text, unnecessary duplication is eliminated, resulting in a well-organized manual.

[1082] Step 5:

[1083] The server formats the integrated documents and automatically generates a standardized table of contents and index. The unified documents are then stored in a database.

[1084] Step 6:

[1085] The server periodically collects business result data (sales performance, project progress, etc.) from various business systems within the company via APIs.

[1086] Step 7:

[1087] The server compares the collected work result data with the contents of the manual. Based on the comparison results, if there are any discrepancies, the document is automatically updated to reflect the latest information.

[1088] Step 8:

[1089] When a user accesses the manual, the server authenticates the user using an authentication system such as LDAP. It verifies the user's permissions and grants access to the document only to users with appropriate privileges.

[1090] Step 9:

[1091] Users view the manual and report areas for improvement or deficiencies using the feedback form. The feedback is then sent to the server.

[1092] Step 10:

[1093] The server analyzes the collected feedback and makes necessary revisions. The revised document is saved back to the database and reflected in subsequent accesses.

[1094] Step 11:

[1095] The server operates an emotion engine that recognizes user emotions in real time based on user input and usage patterns. For example, it analyzes user emotions based on keyboard input speed, mouse movements, and specific keywords.

[1096] Step 12:

[1097] Based on the emotions the server perceives, it displays revised manuals and support information to the user. For example, if the server detects that the user is confused, it will provide additional help information.

[1098] Step 13:

[1099] The server adjusts the document content based on sentiment data collected by the sentiment engine. The document content is revised to be more understandable based on user sentiment feedback.

[1100] Step 14:

[1101] The server collects emotional data and uses it to provide optimal information tailored to specific users. For example, it can improve sections that previously confused users, ensuring that appropriate information is displayed on subsequent visits.

[1102] Through the steps described above, internal business manuals are efficiently collected and analyzed, generating unified documents with no duplication. Furthermore, the system can reflect the latest user business data and optimize the user experience using an emotion engine.

[1103] (Example 2)

[1104] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1105] When efficiently collecting existing documents within a company and managing them as a unified manual, it takes time and effort to address document duplication and ensure that the latest information is reflected. Furthermore, it is difficult to quickly incorporate user feedback and optimize information based on user sentiment. This leads to problems such as a decline in the quality of the user experience and a decrease in operational efficiency.

[1106] The identification processing performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting existing documents and converting them into text, means for analyzing, classifying, and tagging the text data, means for detecting and integrating duplicate content, means for recognizing user emotions and optimizing the user experience based on those emotions, and means for automatically adjusting document content based on emotional feedback. This enables efficient collection and management of documents, rapid reflection of feedback, and optimization of information based on user emotions.

[1107] "Existing documents" refer to various files such as PDFs, Word documents, and Excel spreadsheets that have already been created within the company.

[1108] "Means of converting to text" refers to the process of converting existing documents into text data using OCR, PDF text extraction tools, program libraries, and the tools used for that purpose.

[1109] "Means for analyzing, classifying, and tagging text data" refers to the process of analyzing text data using natural language processing technology and assigning classifications and tags to specific sections.

[1110] "Means for detecting and integrating duplicate content" refers to the process of detecting duplicate content contained within text data and integrating it into a single entry.

[1111] "Means for generating unified documents" refers to the process of organizing redundant content and generating documents that are stored in a unified format.

[1112] "Means of collecting business result data and comparing it with document content" refers to the process of regularly collecting business result data for a company and comparing it with the content of existing documents.

[1113] "Methods for automatically updating documents" refer to processes that automatically update existing documents based on business result data.

[1114] "Means of verifying user authentication and authority" refers to the process of authenticating a user using authentication systems such as LDAP or OAuth, and verifying their authority.

[1115] "Means for users to provide feedback" refers to web forms or interfaces that allow users to provide feedback on the content of a document.

[1116] "Means of revising documents based on feedback" refers to the process of revising documents based on feedback provided by users.

[1117] "A means of recognizing user emotions and optimizing the user experience based on those emotions" refers to the process of using an emotion engine to analyze user emotions in real time and adjusting use cases based on those emotions.

[1118] "Methods for automatically adjusting document content based on emotional feedback" refers to a process that automatically adjusts the content of a document based on the user's emotional feedback, providing optimized information for the user's next access.

[1119] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. Furthermore, the invention incorporates an emotion engine that recognizes user emotions and has the function of optimizing the user experience based on those emotions. This system includes multiple programs running on a server and terminals used by users.

[1120] System Overview

[1121] 1. Document collection and text conversion

[1122] The server collects existing documents from folders and databases within the company. This collection uses Python's os module and cloud storage APIs (e.g., AWS S3, Google Drive API). The collection script is run periodically to retrieve the latest documents. The collected documents are converted into text data using optical character recognition (OCR) or PDF text extraction tools (e.g., PyPDF2, python-docx, pandas, Tesseract OCR).

[1123] 2. Analysis and Classification of Text Data

[1124] The server analyzes the text data using natural language processing (NLP) tools (e.g., SpaCy, NLTK). It categorizes the document content into sections and assigns appropriate tags (e.g., "Project Management," "Sales Strategy").

[1125] 3. Duplicate detection and unification

[1126] The server detects duplicate content from the analysis results and combines it into a single file. The difflib module is used for text comparison. It generates a unified text file and saves it with unnecessary duplicates removed.

[1127] 4. Verification with business result data and automatic updating

[1128] The server periodically collects business performance data (e.g., sales figures and project progress) from SQL or NoSQL databases (e.g., MySQL, MongoDB). The collected data is compared with the document content, and if there are any discrepancies, the document is automatically updated based on the latest information.

[1129] 5. User Authentication and Permission Management

[1130] The server authenticates users using LDAP or OAuth when they access documents. Once authentication is complete, it checks the user's permissions and grants access to the documents only to users with the appropriate permissions.

[1131] 6. Feedback function and document revision

[1132] Users can view documents through the system and provide feedback. This feedback is sent to the server via web forms or other means. The server analyzes the feedback and revises the document accordingly.

[1133] 7. Integration of the Emotional Engine

[1134] The server uses an emotion engine (e.g., Azure Text Analytics, IBM Watson) to analyze user sentiment in real time based on user input and usage. Based on the recognized sentiment, the user experience can be tailored accordingly.

[1135] 8. Adjusting document content based on emotions

[1136] The server automatically adjusts document content based on sentiment feedback, allowing users to access the information more efficiently the next time they visit.

[1137] Specific example

[1138] For example, consider a company where there are two documents: "Sales Manual.pdf" and "Project Management.docx". The server collects these documents and converts them to text using PyPDF2 or python-docx. Next, the converted text is analyzed using SpaCy, classifying each document into sections such as "Sales Policy" and "Project Management Methods," and tagging them. Duplicate parts are detected from the analysis results, and the documents are reconstructed as a single, unified document.

[1139] Subsequently, the latest sales performance data is collected from the MySQL database and reflected in the manual, automatically updating its content. When users access the manual, they are authenticated via LDAP, ensuring that only users with appropriate permissions can access the information. When user feedback is provided, the server analyzes it and revises the document accordingly.

[1140] Furthermore, the emotion engine analyzes user input and usage in real time, and based on the recognized emotions, if confusion or stress is detected regarding a particular section, the system will offer additional support for that section. It also automatically adjusts document content based on emotion data, making it easier to understand on subsequent accesses.

[1141] Example of a prompt

[1142] How can I use a generative AI model to detect user emotions and adjust the content of company manuals based on those emotions?

[1143] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1144] Step 1: Document collection and text conversion

[1145] The server collects existing documents from designated folders and databases within the company. Input is the folder path and database connection information, while output is the collected document files.

[1146] Specifically, the server retrieves files using Python's os module and cloud storage APIs.

[1147] Based on the entered folder path, the system searches for document files and uploads them to the database.

[1148] Step 2: Text conversion

[1149] The server converts collected documents into text data. Input is document files (e.g., PDF, Word, Excel), and output is the converted text data.

[1150] Tools such as PyPDF2, python-docx, pandas, and Tesseract OCR will be used.

[1151] PDF files are extracted using Optical Character Recognition (OCR), while Word and Excel files are converted to text using a dedicated library.

[1152] Step 3: Analysis and classification of text data

[1153] The server analyzes text data using natural language processing (NLP) tools. The input is the transformed text data, and the output is the analyzed section data and tags.

[1154] Specifically, the server uses SpaCy or NLTK to analyze the text, classify the document content into sections, and assign appropriate tags.

[1155] Based on the analyzed text, tags such as "Project Management" and "Sales Strategy" are added to each section.

[1156] Step 4: Duplicate detection and unification

[1157] The server detects duplicate content from the analysis results and combines it into a single file. The input is tagged section data, and the output is unified content with duplicates removed.

[1158] The server uses the difflib module to detect duplicate text.

[1159] Merge highly similar texts and remove unnecessary duplicates.

[1160] Step 5: Match with business result data and automatic update

[1161] The server periodically collects business results data from companies and compares it with document content. The input is information from the business results database, and the output is a document reflecting the latest information.

[1162] The server uses MySQL or MongoDB to retrieve the latest sales performance and project progress data from the database.

[1163] Compare the business results data with the documents, and update the document content if there are any discrepancies.

[1164] Step 6: User Authentication and Permission Management

[1165] The server authenticates the user and verifies their permissions. The input is the user's authentication information (e.g., user ID, password), and the output is the authentication and permission verification result.

[1166] The server uses LDAP or OAuth to authenticate users.

[1167] Once authentication is complete, the system verifies the user's access rights and grants access to the document only to users with the appropriate permissions.

[1168] Step 7: Feedback function and document revision

[1169] Users provide feedback on a document. The input is the content of the feedback, and the output is the document with the feedback reflected.

[1170] Feedback is sent to the server via web forms or similar means.

[1171] The server analyzes the feedback, identifies areas that need improvement, and revises the document accordingly.

[1172] Step 8: Integrating the Emotional Engine

[1173] The server uses an emotion engine to analyze the user's emotions. The input is the user's input data and feedback, and the output is the emotion analysis result.

[1174] The server utilizes sentiment analysis APIs (e.g., Azure Text Analytics, IBM Watson).

[1175] The user experience is adjusted based on recognized emotions, and additional support information is provided if negative emotions are detected.

[1176] Step 9: Adjusting document content based on emotions

[1177] The server automatically adjusts the document content based on emotional feedback. The input is emotional feedback data, and the output is the adjusted document content.

[1178] The server modifies the document content based on sentiment data, providing optimized information for the user's next access.

[1179] Specifically, this involves adding a QA section and simplifying the content of the documents.

[1180] This enables efficient document collection and management, rapid incorporation of feedback, and optimization of information based on user sentiment.

[1181] (Application Example 2)

[1182] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1183] Modern businesses are required to manage and appropriately utilize a wide variety of documents. Simultaneously, it is crucial to recognize user emotions in real time and provide information efficiently. However, existing systems struggle to integrate document management and user support, and real-time work instructions and emotion-based support are particularly lacking in manufacturing environments. This makes it difficult to optimize the user experience while simultaneously demanding document consistency, up-to-date information, and improved work efficiency. Therefore, this invention proposes a system that solves these problems, providing efficient and unified document management and real-time support tailored to user emotions.

[1184] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[1185] In this invention, the server includes means for collecting existing documents and converting them to text; means for analyzing, classifying, and tagging the text data; means for detecting and integrating duplicate content; means for generating a unified document; means for collecting work result data and comparing it with document content; means for automatically updating documents; means for verifying user authentication and permissions; means for users to provide feedback; means for revising documents based on feedback; means for providing work instructions in real time; means for recognizing user emotions and optimizing the experience; and means for providing additional support information based on emotion recognition. This enables efficient document management within the company, real-time work support, and optimal information provision based on user emotions.

[1186] "Means of collecting documents and converting them into text" refers to a function that converts existing documents within a company, such as manuals, procedures, and reports, into analyzable text data.

[1187] "Means for analyzing, classifying, and tagging text data" refers to a function that analyzes the converted text data using natural language processing technology and assigns appropriate tags to each section.

[1188] "Means for detecting and integrating duplicate content" refers to a function that detects duplicate parts from analyzed text data and combines them into a single entity.

[1189] "Means for generating unified documents" refers to a function that generates consistent and unified manuals and reports based on text data that has been merged and whose duplicates have been eliminated.

[1190] "Means for collecting business result data and comparing it with document content" refers to a function that collects company performance and project progress data and compares and verifies it with existing document content.

[1191] "Methods for automatically updating documents" refers to a function that automatically updates documents to the latest content based on collected performance data and project progress data.

[1192] "Means of verifying user authentication and permissions" refers to functions that authenticate users attempting to access manuals and reports, and verify and manage appropriate permissions.

[1193] "Means for users to provide feedback" refers to features that allow users to provide feedback and comments on manuals and reports.

[1194] "Means for revising documents based on feedback" refers to a function that analyzes collected user feedback and revises documents as needed.

[1195] "Means of providing work instructions in real time" refers to a function that provides the instructions and procedures that workers need in real time.

[1196] "Means of recognizing user emotions and optimizing the experience" refers to a function that recognizes the emotional state of workers in real time and optimizes user support based on that information.

[1197] "Means of providing additional support information based on emotion recognition" refers to a function that provides necessary support information and additional instructions based on the user's emotional state.

[1198] Modes for carrying out the invention

[1199] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. Furthermore, the invention incorporates an emotion engine that recognizes user emotions and has a function to optimize the user experience based on those emotions. This system includes multiple software programs running on a server and terminals used by users.

[1200] System Overview

[1201] 1. Document collection and text conversion

[1202] The server collects existing documents (PDFs, Word documents, Excel files, etc.) from various folders and databases within the company. A collection script is run periodically to retrieve the latest documents. The collected documents are converted into text data using optical character recognition (OCR) tools and PDF text extraction tools. Word and Excel files are also converted to text data using dedicated libraries. Specifically, tools such as Tesseract OCR and PDFMiner can be used.

[1203] 2. Analysis and Classification of Text Data

[1204] The server analyzes the converted text data using natural language processing (NLP) techniques. It classifies the document content into sections and applies appropriate tags. For example, NLP libraries such as spaCy or NLTK can be used.

[1205] 3. Duplicate detection and unification

[1206] The server detects and merges duplicate content from the analysis results. It compares overlapping sections and paragraphs to generate a single, unified document. This eliminates unnecessary duplication, resulting in a more organized manual.

[1207] 4. Verification with business result data and automatic updating

[1208] The server periodically collects business performance data from the company (e.g., sales figures and project progress). The collected data is compared with the document content, and if there are any discrepancies, the document is automatically updated to reflect the latest information.

[1209] 5. User Authentication and Permission Management

[1210] The server authenticates users attempting to access manuals and reports. It uses authentication systems such as LDAP to verify identity and confirm user access rights. Only users with appropriate permissions are granted access to the documents.

[1211] 6. Feedback function and document revision

[1212] Users can view manuals and reports and provide feedback. For example, they can report if a particular procedure is outdated. The feedback is sent to the server, analyzed, and necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent accesses.

[1213] 7. Integration of the Emotional Engine

[1214] The server incorporates an emotion engine that recognizes user emotions. It analyzes emotions in real time based on user input and usage, and adjusts the user experience based on the recognized emotions. For example, if a user is experiencing stress, it can prioritize providing supportive information. TensorFlow and OpenCV can be used as emotion recognition models.

[1215] 8. Adjusting document content based on emotions

[1216] The server automatically adjusts document content based on emotional feedback collected by the emotion engine. This allows users to utilize information more efficiently.

[1217] Specific example

[1218] When a new manufacturing process is introduced in a factory, the server collects relevant documents and generates standardized work instructions. When workers receive instructions using smart glasses, the emotion engine provides additional support information if it detects stress. The server uses TensorFlow to perform emotion recognition and provides optimal information in real time to reduce worker confusion.

[1219] Example of a prompt

[1220] Develop standard procedures for a new manufacturing process and design a real-time work guidance system that incorporates emotion recognition capabilities. This system should include features that allow workers to receive instructions in real time via a head-mounted display and receive additional support based on their emotional state.

[1221] Thus, the system of the present invention streamlines document management within a company and realizes real-time work support and optimized user experience.

[1222] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1223] Step 1: Document collection and text conversion

[1224] The server collects existing documents (PDF, Word, Excel, etc.) from various databases and folders within the company. The collected documents are converted into text data using OCR tools (e.g., Tesseract OCR) or PDF text extraction tools. The input is various document files, and the output is their corresponding text data.

[1225] Step 2: Analysis and classification of text data

[1226] The server analyzes the converted text data using natural language processing (NLP) techniques (e.g., spaCy, NLTK). It analyzes each section of the document and assigns appropriate tags to them. The input is text data, and the output is the analysis results and tagged data.

[1227] Step 3: Duplicate detection and merging

[1228] The server detects duplicate content from the parsed and tagged text data. It then compares and merges the duplicate sections to generate a single, unified document. The input is the parsed and tagged text data, and the output is the unified document. Specifically, it uses an algorithm that merges paragraphs and sections with high similarity.

[1229] Step 4: Matching with business result data and automatic updating

[1230] The server periodically collects corporate performance data and project progress data. It then compares this data with existing document content and automatically updates it to reflect the latest information. The input consists of performance data and existing documents, while the output is the updated documents. Specifically, it uses database queries and matching algorithms.

[1231] Step 5: User Authentication and Permission Management

[1232] The server authenticates users through an authentication system such as LDAP when they attempt to access manuals or reports. It verifies the user's access rights and allows access only to those with appropriate permissions. The input is the user's authentication information, and the output is the result of the access rights verification. Specifically, it performs LDAP queries and generates authentication tokens.

[1233] Step 6: Feedback function and document revision

[1234] Users can provide feedback while viewing manuals and reports. The feedback is sent to the server, analyzed, and necessary revisions are made. The input is user feedback, and the output is the revised document. Specifically, a feedback analysis algorithm is used to update the document content.

[1235] Step 7: Provide work instructions in real time

[1236] The server provides workers with real-time instructions and procedures through smart glasses or head-mounted displays. Input is data from the work environment, and output is real-time work instructions. Specifically, it performs tasks such as presenting predefined work procedures and analyzing environmental data.

[1237] Step 8: Emotion Recognition and Experience Optimization

[1238] The server recognizes user emotions in real time and optimizes the user experience based on that. TensorFlow is among the emotion recognition models used. Inputs include user facial images and usage data, while output is optimized user experience information. Specifically, it uses facial recognition algorithms and emotion recognition models (e.g., TensorFlow, OpenCV).

[1239] Step 9: Provide additional support information based on emotion recognition.

[1240] The server provides necessary support information and additional instructions based on the user's emotional state. Input consists of the results of emotion recognition and data regarding the user's confusion and stress; output is additional support information. Specifically, it searches for appropriate support information and presents it to the user.

[1241] In summary, this system streamlines document management within companies and provides real-time work support and optimized user experience.

[1242] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1243] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1244] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[1245] [Fourth Embodiment]

[1246] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1247] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1248] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1249] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[1250] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1251] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1252] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1253] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[1254] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1255] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1256] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1257] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1258] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1259] ---

[1260] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. This system includes multiple programs running on a server and terminals used by users. The system also features a function to automatically update documents based on user feedback and business result data.

[1261] System Overview

[1262] 1. Document collection and text conversion

[1263] Server: Collects existing documents (PDF, Word, Excel, etc.) from various folders and databases within the company. The collected documents are converted into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are also converted to text using a dedicated library.

[1264] 2. Analysis and Classification of Text Data

[1265] Server: The converted text data is analyzed using natural language processing (NLP) techniques. During the analysis process, the document is categorized into sections and assigned appropriate tags. For example, tags such as "project management" and "sales strategy" may be added.

[1266] 3. Duplicate detection and unification

[1267] Server: Text analysis detects duplication when the same content is found in multiple documents. The duplicate content is combined into a single, unified document. This eliminates unnecessary duplication and generates a well-organized manual.

[1268] 4. Verification with business result data and automatic updating

[1269] Server: The server periodically collects business performance data (e.g., sales figures and project progress). The collected data is compared with document content, and if discrepancies are found, the document is automatically updated. This ensures that manuals always reflect the latest information.

[1270] 5. User Authentication and Permission Management

[1271] Server: Users attempting to access the manual must first undergo identity verification through an authentication system. Authentication technologies such as LDAP are used to verify user access rights. This ensures that only users with appropriate permissions can access specific information.

[1272] 6. Feedback function and document revision

[1273] Users: Users can view the manual and provide feedback. For example, they can report if a particular procedure is outdated. The feedback is sent to the server, analyzed, and necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent access.

[1274] Specific example

[1275] For example, consider a company with two documents: "Sales Manual.pdf" and "Project Management.docx". The server collects these documents and converts them to text. Next, it analyzes each document using natural language processing, classifying and tagging them into sections such as "Sales Policy" and "Project Management Methods". Duplicate parts are detected from the analysis results and reconstructed into a single, unified document. Subsequently, the latest sales performance data is collected and reflected in the manual, automatically updating its content. When users access the manual, authentication is performed, ensuring that only users with appropriate permissions can access the information. When user feedback is provided, the server analyzes it and revises the document.

[1276] This allows for centralized management of internal business manuals, ensuring they are always up-to-date, and guaranteeing the secure handling of confidential information.

[1277] ---

[1278] The following describes the processing flow.

[1279] Step 1:

[1280] The server collects existing documents (PDFs, Word documents, Excel files, etc.) from specific folders or databases within the company. The collection script is run periodically to retrieve the most recent documents.

[1281] Step 2:

[1282] The server converts collected documents into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are converted to text using a dedicated library.

[1283] Step 3:

[1284] The server analyzes the converted text data using natural language processing (NLP) techniques. It analyzes the document content, categorizes it into sections, and applies appropriate tags.

[1285] Step 4:

[1286] The server detects duplicate content based on the analysis results. It compares similar sections and paragraphs and integrates them into a single, unified text.

[1287] Step 5:

[1288] The server formats the integrated documents and automatically generates a standardized table of contents and index. The unified documents are then stored in a database.

[1289] Step 6:

[1290] The server periodically collects business result data (sales performance, project progress, etc.) from various business systems within the company via APIs.

[1291] Step 7:

[1292] The server compares the collected work result data with the contents of the manual. If a discrepancy is detected, it automatically performs the necessary updates and updates the document.

[1293] Step 8:

[1294] When a user accesses a manual, the server authenticates the user using an authentication system such as LDAP. It verifies the user's permissions and grants access to the corresponding document only to users with the appropriate privileges.

[1295] Step 9:

[1296] Users view the manual and report areas for improvement or deficiencies using a feedback form. The feedback data is sent to the server.

[1297] Step 10:

[1298] The server analyzes the collected feedback and revises the document. The revised document is saved back to the database and the changes are reflected the next time it is accessed.

[1299] These steps ensure that internal operational manuals are efficiently collected and analyzed, and provided in a unified format free of duplication. Furthermore, they remain up-to-date by reflecting the latest operational results, and are quickly revised based on user feedback.

[1300] (Example 1)

[1301] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1302] The diverse range of documents generated within a company increases over time, making their management complex. The existence of multiple documents can lead to information duplication and inconsistencies, hindering efficient work execution. Furthermore, manual document updates are labor-intensive and can delay the reflection of the latest information. In addition, a lack of proper user authentication and access control increases the risk of confidential information leaks. Addressing these challenges is essential.

[1303] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1304] In this invention, the server includes means for collecting existing documents and converting them into text, means for analyzing, classifying, and tagging the text data, and means for detecting and integrating duplicate content. This streamlines document management within the company, eliminates duplication and inconsistencies, and generates a unified manual that always reflects the latest information. Furthermore, user authentication and permission verification ensure the secure handling of confidential information. In addition, document revisions are automatically performed based on feedback provided by users, ensuring that the latest information is always reflected.

[1305] "Document collection methods" refer to methods for collecting existing documents from folders and databases within a company.

[1306] "Text conversion means" refers to means of converting collected documents into text data. This includes optical character recognition and document text extraction tools.

[1307] "Text analysis means" refers to means for analyzing the converted text data, and this uses natural language processing technology.

[1308] "Classification and tagging means" refers to means of classifying the analyzed text data into sections and assigning appropriate tags.

[1309] A "duplicate detection means" is a means of calculating the similarity between analyzed text data and detecting duplicate content.

[1310] A "combination method" is a means of integrating duplicate text data into a single, unified document.

[1311] "Methods for collecting business data" refer to methods for periodically collecting business result data from a company, and this includes the use of various APIs.

[1312] A "document verification means" is a means of verifying collected business data against the contents of documents.

[1313] An "automatic update mechanism" is a means of automatically updating the document content based on the verification results.

[1314] "User authentication means" refers to methods for verifying a user's identity, and this includes using LDAP or authentication protocols.

[1315] A "permission verification method" is a means of verifying a user's access rights and ensuring that only users with appropriate permissions can access the information.

[1316] A "feedback provision method" is a means by which users can provide feedback on a document.

[1317] "Document revision methods" refer to methods for revising documents based on user feedback and reflecting the latest information.

[1318] ---

[1319] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. This system includes multiple programs running on a server and terminals used by users. The system also features a function to automatically update documents based on user feedback and business result data.

[1320] System Overview

[1321] Document collection and text conversion

[1322] The server collects existing documents (PDF, Word, Excel, etc.) from various folders and databases within the company. This collection process utilizes remote file transfer methods, specifically SFTP (Secure File Transfer Protocol). The collected documents are converted into text data using optical character recognition (OCR) and document text extraction tools. ABBYY FineReader is used for OCR, and the PDF library is used for PDF text extraction. Word and Excel files are converted to text using the Apache POI library.

[1323] Text data analysis and classification

[1324] The server analyzes the converted text data using natural language processing (NLP) techniques. Software such as Google Cloud Natural Language and SpaCy are used for analysis, including text tokenization and morphological analysis. As a result of the analysis, the documents are categorized into sections and tagged with terms like "Project Management" and "Sales Strategy." Machine learning techniques (e.g., k-means clustering) are used for classification.

[1325] Duplicate detection and unification

[1326] The server detects duplicates when multiple documents contain the same content based on the analysis results. Text similarity calculation methods (such as Jaccard similarity or Cosine similarity) are used for duplicate detection. Duplicate content is combined into a single, unified document. The most reliable text and the most recent content are prioritized for unification as needed.

[1327] Matching with business result data and automatic updating

[1328] The server periodically collects various business result data from the company (e.g., sales performance data and project progress data). This data collection often utilizes APIs, specifically the Salesforce API and other business system APIs. The collected data is compared with document content, and if discrepancies are found, the documents are automatically updated. This update process uses PHP or Python scripts.

[1329] User authentication and access control

[1330] When a user attempts to access a manual, the server first verifies their identity through an authentication system. It uses LDAP (Lightweight Directory Access Protocol) or OAuth protocols to verify the user's access rights. Only users who successfully authenticate are allowed to access specific documents with the appropriate permissions.

[1331] Feedback function and document revisions

[1332] Users can provide feedback while browsing the manual. This can be done by filling out a feedback form and submitting it, noting that specific steps or information are outdated. The feedback is sent to the server, which uses natural language processing to analyze the content and determine if the document needs revision. If necessary, the document is revised and saved again in the database.

[1333] Specific example

[1334] For example, if a company has documents such as "Sales Manual.pdf" and "Project Management.docx," the server collects these documents and converts them into text data. Next, the converted text data is analyzed using Google Cloud Natural Language and tagged with terms such as "Sales Policy" and "Project Management Methods." Based on the analysis results, duplicate parts are detected using Jaccard similarity and reconstructed into a single, unified document. The latest sales performance data from Salesforce is collected and reflected in the manual. When users access the manual, authentication is performed using LDAP. If a user provides feedback that a particular procedure is outdated, the server analyzes the feedback, makes the necessary revisions, and saves them in the database.

[1335] Example of a prompt

[1336] Examples of prompts for a generative AI model are as follows:

[1337] Please describe in detail, step by step, a system for collecting existing documents within a company and generating and maintaining a unified manual. This system includes steps such as document collection via SFTP, OCR processing, NLP analysis, duplicate detection, LDAP authentication, and feedback analysis.

[1338] ---

[1339] The above is the description of the "Mode for Carrying Out the Invention" in the patent specification.

[1340] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1341] Step 1:

[1342] Collect existing documents.

[1343] The server accesses designated folders and databases within the company to collect documents.

[1344] The input consists of a folder path and database connection information.

[1345] Specifically, the system will perform the operation of downloading PDF, Word, and Excel files from a remote folder using SFTP.

[1346] The output includes the collected document files.

[1347] Step 2:

[1348] Convert a document into text data.

[1349] The server converts the collected documents into text data using optical character recognition (OCR) and document text extraction tools.

[1350] The document files collected in Step 1 are provided as input.

[1351] Specifically, the process involves using ABBYY FineReader to convert PDFs into text data, and then using the Apache POI library to extract text from Word and Excel files.

[1352] The output includes the converted text data.

[1353] Step 3:

[1354] Analyze text data and perform classification and tagging.

[1355] The server uses natural language processing (NLP) techniques to analyze the converted text data.

[1356] The text data obtained in step 2 is provided as input.

[1357] Specifically, the system uses the Google Cloud Natural Language API and SpaCy to perform tokenization and morphological analysis, and then analyzes the meaning and structure of the text.

[1358] The output includes analytical data categorized by section and tagged appropriately.

[1359] Step 4:

[1360] It detects duplicate content and generates a unified document.

[1361] The server detects duplicate content from the analysis results and reconstructs it into a unified document.

[1362] The analysis data obtained in step 3 is provided as input.

[1363] Specifically, the process involves calculating the degree of text duplication using Jaccard similarity or Cosine similarity, and then combining the duplicated parts into a single text.

[1364] The output includes a unified document with duplicates removed and organized.

[1365] Step 5:

[1366] Collect business result data and compare it with document content.

[1367] The server periodically collects business results data from the company and compares it with document content.

[1368] The input includes business result data collected using the API.

[1369] Specifically, the system will use business system APIs, such as the Salesforce API, to retrieve the latest sales performance and project progress data.

[1370] The output includes the results of comparing the collected business data with the document content.

[1371] Step 6:

[1372] Automatically update the document.

[1373] The server automatically updates the document content based on the matching results.

[1374] The input is provided with the matching result obtained in step 5.

[1375] Specifically, the system will use PHP or Python scripts to update document content and reflect the latest information.

[1376] The output will include the latest updated document.

[1377] Step 7:

[1378] Verify user authentication and permissions.

[1379] The server authenticates users when they access the manual.

[1380] The user's authentication information is provided as input.

[1381] Specifically, it performs user authentication and authorization verification using LDAP and OAuth protocols.

[1382] The output includes the authentication results and the user's access permissions.

[1383] Step 8:

[1384] Users provide feedback and revise the document.

[1385] Users provide feedback while viewing the manual.

[1386] The input is the feedback content.

[1387] Specifically, the process involves sending feedback to the server through a feedback form.

[1388] The server analyzes the feedback and revises the document.

[1389] User feedback is provided as input.

[1390] Specifically, this involves using NLP tools to understand the feedback and making necessary revisions.

[1391] The output will include the revised document.

[1392] (Application Example 1)

[1393] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1394] Work manuals and procedures used by robots operating in factories need to be updated frequently. However, manual updates are inefficient, making it difficult to maintain consistency and up-to-dateness of information. Furthermore, the lack of a mechanism to incorporate real-time feedback from robots during actual work can lead to decreased accuracy and efficiency. In addition, there is a need for a mechanism to quickly update robot operating procedures when they become outdated.

[1395] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1396] In this invention, the server includes means for collecting existing documents and converting them into text; means for analyzing, classifying, and tagging the text data; means for detecting and integrating duplicate content; means for generating a unified document; means for collecting work result data and comparing it with the document content; means for automatically updating the document; means for verifying user authentication and authorization; means for users to provide feedback; means for revising the document based on the feedback; means for robots to provide real-time feedback during work and automatically revising the document based on that data; and means for automatically updating the robot's operating procedures. This makes it possible for robots in the factory to always operate based on the latest work manuals, improving the accuracy and efficiency of their work.

[1397] "Methods for collecting existing documents and converting them into text" refers to the process of gathering documents of various formats that exist within a factory into one place and converting them into machine-readable text data using OCR technology and a dedicated library.

[1398] "Means for analyzing, classifying, and tagging text data" refers to the process of analyzing each piece of text data using natural language processing technology, classifying it into sections based on its content, and assigning appropriate labels and tags.

[1399] "Methods for detecting and integrating duplicate content" refers to the process of extracting and detecting identical or similar content from analyzed text data and combining them into a single, unified document.

[1400] "Methods for generating unified documents" refers to the process of integrating redundant content and then compiling the resulting organized documents into a single, complete document.

[1401] "Means for collecting business result data and comparing it with document content" refers to the process of regularly collecting performance data related to operations within the factory and comparing and verifying it with existing documents.

[1402] "Methods for automatically updating documents" refer to a process that automatically updates the content of documents as needed, based on the results of cross-referencing with business result data, to reflect the most accurate information.

[1403] "Means of verifying user authentication and permissions" refers to the process of verifying the identity of the user attempting to access a document and checking whether they have the appropriate access rights.

[1404] "Means for users to provide feedback" refers to a function that allows users to provide feedback when viewing a document, and a process to prompt corrections if the content is outdated.

[1405] "Means of revising documents based on feedback" refers to the process of analyzing feedback provided by users and modifying or revising the content of the document based on that feedback.

[1406] "A method for robots to provide real-time feedback during work and automatically revise documents based on that data" refers to a process in which robots record information and feedback obtained during actual work in real time and automatically revise documents based on that data.

[1407] "Means for automatically updating robot operating procedures" refers to a process that automatically updates robot operating procedures based on the latest collected data and feedback, thereby improving work efficiency and accuracy.

[1408] System Overview

[1409] This invention is a system for keeping the work manuals and operating procedures for robots operating in a factory constantly up-to-date. The system includes multiple programs running on a server and applications installed on the robots. The system also has the function of receiving real-time feedback from the robots and automatically updating the documents. This can improve the work efficiency and accuracy of the robots.

[1410] Hardware and software configuration

[1411] Hardware: Server, robot, camera (for OCR)

[1412] Software: OCR tool (e.g., Tesseract), NLP engine (e.g., SpaCy), database (MySQL), authentication system (LDAP)

[1413] Document collection and text conversion

[1414] The server collects existing documents from various terminals and databases within the factory and uses OCR tools to convert the contents of PDFs and JPEGs into text data. The converted text data is further analyzed and processed using a dedicated library. For example, nouns and verbs are extracted and sentences are split.

[1415] Data analysis and classification

[1416] The server uses an NLP engine to analyze text data, classify it into sections, and assign tags to each section. This organizes the data by content, such as "operating procedures" or "maintenance procedures." Specifically, it uses a natural language processing engine (SpaCy) to analyze the parts of speech in the text data and automatically assigns important sections and labels.

[1417] Duplicate detection and merging

[1418] The system detects duplicate sections from the analyzed data and consolidates them into a single, unified document. This eliminates unnecessary duplication and generates a well-organized manual. This process utilizes a document consolidation algorithm. For example, it includes a process to combine identical operating procedures found in multiple documents into a single document.

[1419] Matching with business result data and automatic updating

[1420] The server periodically collects work result data and compares it with existing documents. If there are discrepancies, the documents are automatically updated. This ensures that the latest information is always reflected. Specifically, it retrieves work performance data from the work database and compares it with the analysis results.

[1421] User authentication and access management

[1422] When a user attempts to access a document, their identity is verified using an LDAP-based authentication system. Only users with appropriate permissions can access specific information. This ensures security.

[1423] Feedback function and document revision

[1424] The robot can provide real-time feedback during operation. For example, if there is an inconsistency in the operating procedure, it will provide feedback immediately. This feedback is sent to the server, analyzed, and any necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent access. This ensures that the procedure is up-to-date and reflects current practices.

[1425] Examples of usage and specific prompt statements

[1426] For example, if a robot starts performing a new work procedure and there is a contradiction with the existing work manual, the feedback function will report it as follows:

[1427] Prompt example:

[1428] "Please check the instructions in Step 2 of the manual. Since it does not match the actual work, it needs to be updated based on the latest work result data."

[1429] This allows robots in the factory to perform tasks efficiently based on the latest information at all times.

[1430] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1431] Step 1:

[1432] The server collects existing documents from various terminals and databases within the factory. Input documents include formats such as PDF, Word, and Excel. The server gathers these documents in one place via FTP or API. The output is a list of the collected document data.

[1433] Step 2:

[1434] The server converts the collected documents into text data using an OCR tool (e.g., Tesseract). The input is the collected document data. Using OCR or a dedicated library, this is converted into machine-readable text, and the output is text data.

[1435] Step 3:

[1436] The server analyzes text data using an NLP engine (e.g., SpaCy). The input is converted text data. Using natural language processing techniques, this data is analyzed, classified, and tagged by part of speech and section. The analyzed data is then output.

[1437] Step 4:

[1438] The server detects and integrates duplicate parts from the analyzed data. The input is the analyzed data. A document integration algorithm is used to combine identical or similar content into a single document. The integrated document is then output.

[1439] Step 5:

[1440] The server generates integrated documents as a single, unified document. The input is integrated data. Based on this, it generates documents in the most organized format. The output is a unified document.

[1441] Step 6:

[1442] The server collects business result data and compares it with existing document content. The input consists of business result data and a standardized document. It compares this data with work performance data retrieved from the business database, and if there are discrepancies, it automatically updates the content. The updated document is output.

[1443] Step 7:

[1444] The server verifies authentication and permissions when a user accesses a document. The input is the user's authentication information. The server uses an authentication system such as LDAP to verify the user's identity and permissions. The output is a list of users whose access permissions have been verified.

[1445] Step 8:

[1446] Users view documents and provide feedback. The input is user feedback information. The feedback is sent to the server, analyzed, and necessary revisions are made. The updated feedback information is recorded as output.

[1447] Step 9:

[1448] The server revises documents based on user feedback. Its inputs include feedback information and document data. Based on this data, it analyzes which parts of the document need revision and performs the revisions. The revised document is then output.

[1449] Step 10:

[1450] The robot provides real-time feedback during the task and automatically revises the document based on that data. The input is real-time work feedback data. The robot sends the feedback it provides to a server, which then revises the document based on that data. The output is a document reflecting the latest work procedure.

[1451] Step 11:

[1452] The server automatically updates the robot's operating procedures based on the latest collected data and feedback. The input is revised document data. Based on this, the robot's operating procedures are automatically updated, enabling operation based on the latest information. The updated operating procedures are obtained as output.

[1453] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1454] ---

[1455] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. Furthermore, the invention incorporates an emotion engine that recognizes user emotions and has a function to optimize the user experience based on those emotions. This system includes multiple programs running on a server and terminals used by users.

[1456] System Overview

[1457] 1. Document collection and text conversion

[1458] Server: Collects existing documents (PDF, Word, Excel, etc.) from various folders and databases within the company. It periodically runs collection scripts to retrieve the latest documents. The collected documents are converted into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are also converted to text using a dedicated library.

[1459] 2. Analysis and Classification of Text Data

[1460] Server: The converted documents are analyzed using natural language processing (NLP) techniques. The document content is categorized into sections and appropriately tagged. For example, tags such as "project management" and "sales strategy" may be applied.

[1461] 3. Duplicate detection and unification

[1462] Server: Detects duplicate content from the analysis results and combines it into a single file. Compares duplicate sections and paragraphs to generate a unified text. This eliminates unnecessary duplication and produces a well-organized manual.

[1463] 4. Verification with business result data and automatic updating

[1464] Server: This server periodically collects business performance data from companies (e.g., sales figures and project progress). The collected data is compared with document content, and if there are discrepancies, the document is automatically updated to reflect the latest information.

[1465] 5. User Authentication and Permission Management

[1466] Server: Users attempting to access the manual undergo identity verification through an authentication system such as LDAP. The server verifies user access rights and grants access to the document only to users with appropriate permissions.

[1467] 6. Feedback function and document revision

[1468] Users: Users can view the manual and provide feedback. For example, they can report if a particular procedure is outdated. The feedback is sent to the server, analyzed, and necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent access.

[1469] 7. Integration of the Emotional Engine

[1470] Server: The system incorporates an emotion engine that recognizes user emotions. It analyzes user emotions in real time based on user input and usage, and adjusts the user experience based on the recognized emotions. For example, if a user is feeling stressed, it can prioritize providing support information.

[1471] 8. Adjusting document content based on emotions

[1472] Server: Based on emotional feedback collected by the emotion engine, the document content is automatically adjusted. In this way, users can utilize the information more efficiently.

[1473] Specific example

[1474] For example, consider a company with two documents: "Sales Manual.pdf" and "Project Management.docx". The server collects these documents and converts them to text. Next, it analyzes each document using natural language processing, classifying and tagging them into sections such as "Sales Policy" and "Project Management Methods". Duplicate parts are detected from the analysis results and reconstructed into a single, unified document. Subsequently, the latest sales performance data is collected and reflected in the manual, automatically updating its content. When users access the manual, authentication is performed, ensuring that only users with appropriate permissions can access the information. When user feedback is provided, the server analyzes it and revises the document.

[1475] Furthermore, an emotion engine recognizes user emotions based on their input and usage. For example, if a user repeatedly refers to a specific section of a manual and expresses confusion, the system will offer additional support regarding that section. It also automatically adjusts the document content based on emotion data, making it easier to understand on subsequent accesses.

[1476] This allows for the efficient collection and analysis of internal business manuals, providing them in a unified format free of duplication, while also reflecting the latest business results and enabling optimized information delivery based on user sentiment.

[1477] ---

[1478] The following describes the processing flow.

[1479] Step 1:

[1480] The server collects existing documents (PDFs, Word documents, Excel files, etc.) from specific folders or databases within the company. The collection script is run periodically to retrieve the most recent documents.

[1481] Step 2:

[1482] The server converts collected documents into text data using optical character recognition (OCR) and PDF text extraction tools. Word and Excel files are converted to text using a dedicated library.

[1483] Step 3:

[1484] The server analyzes the converted text data using natural language processing (NLP) techniques. It categorizes the document content into sections and assigns appropriate tags. For example, tags such as "project management" and "sales strategy" might be applied.

[1485] Step 4:

[1486] The server detects duplicate content based on the analysis results. It compares similar sections and paragraphs and combines them into one. By generating integrated text, unnecessary duplication is eliminated, resulting in a well-organized manual.

[1487] Step 5:

[1488] The server formats the integrated documents and automatically generates a standardized table of contents and index. The unified documents are then stored in a database.

[1489] Step 6:

[1490] The server periodically collects business result data (sales performance, project progress, etc.) from various business systems within the company via APIs.

[1491] Step 7:

[1492] The server compares the collected work result data with the contents of the manual. Based on the comparison results, if there are any discrepancies, the document is automatically updated to reflect the latest information.

[1493] Step 8:

[1494] When a user accesses the manual, the server authenticates the user using an authentication system such as LDAP. It verifies the user's permissions and grants access to the document only to users with appropriate privileges.

[1495] Step 9:

[1496] Users view the manual and report areas for improvement or deficiencies using the feedback form. The feedback is then sent to the server.

[1497] Step 10:

[1498] The server analyzes the collected feedback and makes necessary revisions. The revised document is saved back to the database and reflected in subsequent accesses.

[1499] Step 11:

[1500] The server operates an emotion engine that recognizes user emotions in real time based on user input and usage patterns. For example, it analyzes user emotions based on keyboard input speed, mouse movements, and specific keywords.

[1501] Step 12:

[1502] Based on the emotions the server perceives, it displays revised manuals and support information to the user. For example, if the server detects that the user is confused, it will provide additional help information.

[1503] Step 13:

[1504] The server adjusts the document content based on sentiment data collected by the sentiment engine. The document content is revised to be more understandable based on user sentiment feedback.

[1505] Step 14:

[1506] The server collects emotional data and uses it to provide optimal information tailored to specific users. For example, it can improve sections that previously confused users, ensuring that appropriate information is displayed on subsequent visits.

[1507] Through the steps described above, internal business manuals are efficiently collected and analyzed, generating unified documents with no duplication. Furthermore, the system can reflect the latest user business data and optimize the user experience using an emotion engine.

[1508] (Example 2)

[1509] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1510] When efficiently collecting existing documents within a company and managing them as a unified manual, it takes time and effort to address document duplication and ensure that the latest information is reflected. Furthermore, it is difficult to quickly incorporate user feedback and optimize information based on user sentiment. This leads to problems such as a decline in the quality of the user experience and a decrease in operational efficiency.

[1511] The identification processing performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting existing documents and converting them into text, means for analyzing, classifying, and tagging the text data, means for detecting and integrating duplicate content, means for recognizing user emotions and optimizing the user experience based on those emotions, and means for automatically adjusting document content based on emotional feedback. This enables efficient collection and management of documents, rapid reflection of feedback, and optimization of information based on user emotions.

[1512] "Existing documents" refer to various files such as PDFs, Word documents, and Excel spreadsheets that have already been created within the company.

[1513] "Means of converting to text" refers to the process of converting existing documents into text data using OCR, PDF text extraction tools, program libraries, and the tools used for that purpose.

[1514] "Means for analyzing, classifying, and tagging text data" refers to the process of analyzing text data using natural language processing technology and assigning classifications and tags to specific sections.

[1515] "Means for detecting and integrating duplicate content" refers to the process of detecting duplicate content contained within text data and integrating it into a single entry.

[1516] "Means for generating unified documents" refers to the process of organizing redundant content and generating documents that are stored in a unified format.

[1517] "Means of collecting business result data and comparing it with document content" refers to the process of regularly collecting business result data for a company and comparing it with the content of existing documents.

[1518] "Methods for automatically updating documents" refer to processes that automatically update existing documents based on business result data.

[1519] "Means of verifying user authentication and authority" refers to the process of authenticating a user using authentication systems such as LDAP or OAuth, and verifying their authority.

[1520] "Means for users to provide feedback" refers to web forms or interfaces that allow users to provide feedback on the content of a document.

[1521] "Means of revising documents based on feedback" refers to the process of revising documents based on feedback provided by users.

[1522] "A means of recognizing user emotions and optimizing the user experience based on those emotions" refers to the process of using an emotion engine to analyze user emotions in real time and adjusting use cases based on those emotions.

[1523] "Methods for automatically adjusting document content based on emotional feedback" refers to a process that automatically adjusts the content of a document based on the user's emotional feedback, providing optimized information for the user's next access.

[1524] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. Furthermore, the invention incorporates an emotion engine that recognizes user emotions and has the function of optimizing the user experience based on those emotions. This system includes multiple programs running on a server and terminals used by users.

[1525] System Overview

[1526] 1. Document collection and text conversion

[1527] The server collects existing documents from folders and databases within the company. This collection uses Python's os module and cloud storage APIs (e.g., AWS S3, Google Drive API). The collection script is run periodically to retrieve the latest documents. The collected documents are converted into text data using optical character recognition (OCR) or PDF text extraction tools (e.g., PyPDF2, python-docx, pandas, Tesseract OCR).

[1528] 2. Analysis and Classification of Text Data

[1529] The server analyzes the text data using natural language processing (NLP) tools (e.g., SpaCy, NLTK). It categorizes the document content into sections and assigns appropriate tags (e.g., "Project Management," "Sales Strategy").

[1530] 3. Duplicate detection and unification

[1531] The server detects duplicate content from the analysis results and combines it into a single file. The difflib module is used for text comparison. It generates a unified text file and saves it with unnecessary duplicates removed.

[1532] 4. Verification with business result data and automatic updating

[1533] The server periodically collects business performance data (e.g., sales figures and project progress) from SQL or NoSQL databases (e.g., MySQL, MongoDB). The collected data is compared with the document content, and if there are any discrepancies, the document is automatically updated based on the latest information.

[1534] 5. User Authentication and Permission Management

[1535] The server authenticates users using LDAP or OAuth when they access documents. Once authentication is complete, it checks the user's permissions and grants access to the documents only to users with the appropriate permissions.

[1536] 6. Feedback function and document revision

[1537] Users can view documents through the system and provide feedback. This feedback is sent to the server via web forms or other means. The server analyzes the feedback and revises the document accordingly.

[1538] 7. Integration of the Emotional Engine

[1539] The server uses an emotion engine (e.g., Azure Text Analytics, IBM Watson) to analyze user sentiment in real time based on user input and usage. Based on the recognized sentiment, the user experience can be tailored accordingly.

[1540] 8. Adjusting document content based on emotions

[1541] The server automatically adjusts document content based on sentiment feedback, allowing users to access the information more efficiently the next time they visit.

[1542] Specific example

[1543] For example, consider a company where there are two documents: "Sales Manual.pdf" and "Project Management.docx". The server collects these documents and converts them to text using PyPDF2 or python-docx. Next, the converted text is analyzed using SpaCy, classifying each document into sections such as "Sales Policy" and "Project Management Methods," and tagging them. Duplicate parts are detected from the analysis results, and the documents are reconstructed as a single, unified document.

[1544] Subsequently, the latest sales performance data is collected from the MySQL database and reflected in the manual, automatically updating its content. When users access the manual, they are authenticated via LDAP, ensuring that only users with appropriate permissions can access the information. When user feedback is provided, the server analyzes it and revises the document accordingly.

[1545] Furthermore, the emotion engine analyzes user input and usage in real time, and based on the recognized emotions, if confusion or stress is detected regarding a particular section, the system will offer additional support for that section. It also automatically adjusts document content based on emotion data, making it easier to understand on subsequent accesses.

[1546] Example of a prompt

[1547] How can I use a generative AI model to detect user emotions and adjust the content of company manuals based on those emotions?

[1548] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1549] Step 1: Document collection and text conversion

[1550] The server collects existing documents from designated folders and databases within the company. Input is the folder path and database connection information, while output is the collected document files.

[1551] Specifically, the server retrieves files using Python's os module and cloud storage APIs.

[1552] Based on the entered folder path, the system searches for document files and uploads them to the database.

[1553] Step 2: Text conversion

[1554] The server converts collected documents into text data. Input is document files (e.g., PDF, Word, Excel), and output is the converted text data.

[1555] Tools such as PyPDF2, python-docx, pandas, and Tesseract OCR will be used.

[1556] PDF files are extracted using Optical Character Recognition (OCR), while Word and Excel files are converted to text using a dedicated library.

[1557] Step 3: Analysis and classification of text data

[1558] The server analyzes text data using natural language processing (NLP) tools. The input is the transformed text data, and the output is the analyzed section data and tags.

[1559] Specifically, the server uses SpaCy or NLTK to analyze the text, classify the document content into sections, and assign appropriate tags.

[1560] Based on the analyzed text, tags such as "Project Management" and "Sales Strategy" are added to each section.

[1561] Step 4: Duplicate detection and unification

[1562] The server detects duplicate content from the analysis results and combines it into a single file. The input is tagged section data, and the output is unified content with duplicates removed.

[1563] The server uses the difflib module to detect duplicate text.

[1564] Merge highly similar texts and remove unnecessary duplicates.

[1565] Step 5: Match with business result data and automatic update

[1566] The server periodically collects business results data from companies and compares it with document content. The input is information from the business results database, and the output is a document reflecting the latest information.

[1567] The server uses MySQL or MongoDB to retrieve the latest sales performance and project progress data from the database.

[1568] Compare the business results data with the documents, and update the document content if there are any discrepancies.

[1569] Step 6: User Authentication and Permission Management

[1570] The server authenticates the user and verifies their permissions. The input is the user's authentication information (e.g., user ID, password), and the output is the authentication and permission verification result.

[1571] The server uses LDAP or OAuth to authenticate users.

[1572] Once authentication is complete, the system verifies the user's access rights and grants access to the document only to users with the appropriate permissions.

[1573] Step 7: Feedback function and document revision

[1574] Users provide feedback on a document. The input is the content of the feedback, and the output is the document with the feedback reflected.

[1575] Feedback is sent to the server via web forms or similar means.

[1576] The server analyzes the feedback, identifies areas that need improvement, and revises the document accordingly.

[1577] Step 8: Integrating the Emotional Engine

[1578] The server uses an emotion engine to analyze the user's emotions. The input is the user's input data and feedback, and the output is the emotion analysis result.

[1579] The server utilizes sentiment analysis APIs (e.g., Azure Text Analytics, IBM Watson).

[1580] The user experience is adjusted based on recognized emotions, and additional support information is provided if negative emotions are detected.

[1581] Step 9: Adjusting document content based on emotions

[1582] The server automatically adjusts the document content based on emotional feedback. The input is emotional feedback data, and the output is the adjusted document content.

[1583] The server modifies the document content based on sentiment data, providing optimized information for the user's next access.

[1584] Specifically, this involves adding a QA section and simplifying the content of the documents.

[1585] This enables efficient document collection and management, rapid incorporation of feedback, and optimization of information based on user sentiment.

[1586] (Application Example 2)

[1587] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1588] Modern businesses are required to manage and appropriately utilize a wide variety of documents. Simultaneously, it is crucial to recognize user emotions in real time and provide information efficiently. However, existing systems struggle to integrate document management and user support, and real-time work instructions and emotion-based support are particularly lacking in manufacturing environments. This makes it difficult to optimize the user experience while simultaneously demanding document consistency, up-to-date information, and improved work efficiency. Therefore, this invention proposes a system that solves these problems, providing efficient and unified document management and real-time support tailored to user emotions.

[1589] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[1590] In this invention, the server includes means for collecting existing documents and converting them to text; means for analyzing, classifying, and tagging the text data; means for detecting and integrating duplicate content; means for generating a unified document; means for collecting work result data and comparing it with document content; means for automatically updating documents; means for verifying user authentication and permissions; means for users to provide feedback; means for revising documents based on feedback; means for providing work instructions in real time; means for recognizing user emotions and optimizing the experience; and means for providing additional support information based on emotion recognition. This enables efficient document management within the company, real-time work support, and optimal information provision based on user emotions.

[1591] "Means of collecting documents and converting them into text" refers to a function that converts existing documents within a company, such as manuals, procedures, and reports, into analyzable text data.

[1592] "Means for analyzing, classifying, and tagging text data" refers to a function that analyzes the converted text data using natural language processing technology and assigns appropriate tags to each section.

[1593] "Means for detecting and integrating duplicate content" refers to a function that detects duplicate parts from analyzed text data and combines them into a single entity.

[1594] "Means for generating unified documents" refers to a function that generates consistent and unified manuals and reports based on text data that has been merged and whose duplicates have been eliminated.

[1595] "Means for collecting business result data and comparing it with document content" refers to a function that collects company performance and project progress data and compares and verifies it with existing document content.

[1596] "Methods for automatically updating documents" refers to a function that automatically updates documents to the latest content based on collected performance data and project progress data.

[1597] "Means of verifying user authentication and permissions" refers to functions that authenticate users attempting to access manuals and reports, and verify and manage appropriate permissions.

[1598] "Means for users to provide feedback" refers to features that allow users to provide feedback and comments on manuals and reports.

[1599] "Means for revising documents based on feedback" refers to a function that analyzes collected user feedback and revises documents as needed.

[1600] "Means of providing work instructions in real time" refers to a function that provides the instructions and procedures that workers need in real time.

[1601] "Means of recognizing user emotions and optimizing the experience" refers to a function that recognizes the emotional state of workers in real time and optimizes user support based on that information.

[1602] "Means of providing additional support information based on emotion recognition" refers to a function that provides necessary support information and additional instructions based on the user's emotional state.

[1603] Modes for carrying out the invention

[1604] This invention is a system for collecting existing documents within a company and generating and maintaining a unified manual. Furthermore, the invention incorporates an emotion engine that recognizes user emotions and has a function to optimize the user experience based on those emotions. This system includes multiple software programs running on a server and terminals used by users.

[1605] System Overview

[1606] 1. Document collection and text conversion

[1607] The server collects existing documents (PDFs, Word documents, Excel files, etc.) from various folders and databases within the company. A collection script is run periodically to retrieve the latest documents. The collected documents are converted into text data using optical character recognition (OCR) tools and PDF text extraction tools. Word and Excel files are also converted to text data using dedicated libraries. Specifically, tools such as Tesseract OCR and PDFMiner can be used.

[1608] 2. Analysis and Classification of Text Data

[1609] The server analyzes the converted text data using natural language processing (NLP) techniques. It classifies the document content into sections and applies appropriate tags. For example, NLP libraries such as spaCy or NLTK can be used.

[1610] 3. Duplicate detection and unification

[1611] The server detects and merges duplicate content from the analysis results. It compares overlapping sections and paragraphs to generate a single, unified document. This eliminates unnecessary duplication, resulting in a more organized manual.

[1612] 4. Verification with business result data and automatic updating

[1613] The server periodically collects business performance data from the company (e.g., sales figures and project progress). The collected data is compared with the document content, and if there are any discrepancies, the document is automatically updated to reflect the latest information.

[1614] 5. User Authentication and Permission Management

[1615] The server authenticates users attempting to access manuals and reports. It uses authentication systems such as LDAP to verify identity and confirm user access rights. Only users with appropriate permissions are granted access to the documents.

[1616] 6. Feedback function and document revision

[1617] Users can view manuals and reports and provide feedback. For example, they can report if a particular procedure is outdated. The feedback is sent to the server, analyzed, and necessary revisions are made. The revised document is then saved back to the database and reflected in subsequent accesses.

[1618] 7. Integration of the Emotional Engine

[1619] The server incorporates an emotion engine that recognizes user emotions. It analyzes emotions in real time based on user input and usage, and adjusts the user experience based on the recognized emotions. For example, if a user is experiencing stress, it can prioritize providing supportive information. TensorFlow and OpenCV can be used as emotion recognition models.

[1620] 8. Adjusting document content based on emotions

[1621] The server automatically adjusts document content based on emotional feedback collected by the emotion engine. This allows users to utilize information more efficiently.

[1622] Specific example

[1623] When a new manufacturing process is introduced in a factory, the server collects relevant documents and generates standardized work instructions. When workers receive instructions using smart glasses, the emotion engine provides additional support information if it detects stress. The server uses TensorFlow to perform emotion recognition and provides optimal information in real time to reduce worker confusion.

[1624] Example of a prompt

[1625] Develop standard procedures for a new manufacturing process and design a real-time work guidance system that incorporates emotion recognition capabilities. This system should include features that allow workers to receive instructions in real time via a head-mounted display and receive additional support based on their emotional state.

[1626] Thus, the system of the present invention streamlines document management within a company and realizes real-time work support and optimized user experience.

[1627] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1628] Step 1: Document collection and text conversion

[1629] The server collects existing documents (PDF, Word, Excel, etc.) from various databases and folders within the company. The collected documents are converted into text data using OCR tools (e.g., Tesseract OCR) or PDF text extraction tools. The input is various document files, and the output is their corresponding text data.

[1630] Step 2: Analysis and classification of text data

[1631] The server analyzes the converted text data using natural language processing (NLP) techniques (e.g., spaCy, NLTK). It analyzes each section of the document and assigns appropriate tags to them. The input is text data, and the output is the analysis results and tagged data.

[1632] Step 3: Duplicate detection and merging

[1633] The server detects duplicate content from the parsed and tagged text data. It then compares and merges the duplicate sections to generate a single, unified document. The input is the parsed and tagged text data, and the output is the unified document. Specifically, it uses an algorithm that merges paragraphs and sections with high similarity.

[1634] Step 4: Matching with business result data and automatic updating

[1635] The server periodically collects corporate performance data and project progress data. It then compares this data with existing document content and automatically updates it to reflect the latest information. The input consists of performance data and existing documents, while the output is the updated documents. Specifically, it uses database queries and matching algorithms.

[1636] Step 5: User Authentication and Permission Management

[1637] The server authenticates users through an authentication system such as LDAP when they attempt to access manuals or reports. It verifies the user's access rights and allows access only to those with appropriate permissions. The input is the user's authentication information, and the output is the result of the access rights verification. Specifically, it performs LDAP queries and generates authentication tokens.

[1638] Step 6: Feedback function and document revision

[1639] Users can provide feedback while viewing manuals and reports. The feedback is sent to the server, analyzed, and necessary revisions are made. The input is user feedback, and the output is the revised document. Specifically, a feedback analysis algorithm is used to update the document content.

[1640] Step 7: Provide work instructions in real time

[1641] The server provides workers with real-time instructions and procedures through smart glasses or head-mounted displays. Input is data from the work environment, and output is real-time work instructions. Specifically, it performs tasks such as presenting predefined work procedures and analyzing environmental data.

[1642] Step 8: Emotion Recognition and Experience Optimization

[1643] The server recognizes user emotions in real time and optimizes the user experience based on that. TensorFlow is among the emotion recognition models used. Inputs include user facial images and usage data, while output is optimized user experience information. Specifically, it uses facial recognition algorithms and emotion recognition models (e.g., TensorFlow, OpenCV).

[1644] Step 9: Provide additional support information based on emotion recognition.

[1645] The server provides necessary support information and additional instructions based on the user's emotional state. Input consists of the results of emotion recognition and data regarding the user's confusion and stress; output is additional support information. Specifically, it searches for appropriate support information and presents it to the user.

[1646] In summary, this system streamlines document management within companies and provides real-time work support and optimized user experience.

[1647] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1648] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1649] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1650] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1651] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1652] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1653] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1654] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1655] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1656] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values. 【165...

Claims

1. A means of collecting existing documents and converting them into text, A means for analyzing, classifying, and tagging text data, A means for detecting and merging duplicate content, A means of generating a unified document, A means of collecting business result data and comparing it with document content, A means of automatically updating documents, Means for verifying user authentication and permissions, Means by which users can provide feedback, A system that includes means for revising documents based on feedback.

2. The system according to claim 1, comprising means for converting an existing document into text using OCR or a PDF text extraction tool.

3. The system according to claim 1, comprising means for analyzing text data using natural language processing and classifying and tagging each section.

Images