System

Abstract

A system is provided.SOLUTION: The system includes a means for collecting past similar documents and analyzing the contents by using a natural language processing technology, a means for inputting the kind of material desired to be generated and necessary information through a user interface, a means for performing similarity calculation and specifying the most relevant document, a means for automatically generating a draft of new material based on the specified document, a means for performing grammatical check and content check of the generated draft, and a means for generating feedback based on the result of the check and providing it to the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, 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] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] The traditional document creation and review process is laborious, time-consuming, and prone to human error. Furthermore, there is no system in place for efficiently utilizing similar documents from the past, and individual documents are scattered, making reuse difficult. As a result, creating new documents requires a lot of resources, resulting in reduced work efficiency. [Means for solving the problem]

[0005] The present invention includes a means for collecting similar documents from the past and analyzing them using natural language processing technology. This allows for efficient summarization and classification of the documents. Next, the user inputs the type of material they wish to generate and the necessary information via a user interface, and the system performs a similarity calculation to identify the most relevant documents.

[0006] The system then provides a means for automatically generating a draft of new material based on the identified documents. It also includes a means for checking the grammar and content of the generated draft, and generates and provides feedback to the user based on the results of the check. This allows users to quickly and efficiently create appropriate materials, improving work efficiency.

[0007] "Similar documents" are documents that were created in the past and have similar content and format that can be used as references for creating current materials.

[0008] "Natural language processing technology" is a technology that allows computers to understand, analyze, and process human language.

[0009] A "user interface" is an interface (screen or input means) that allows a user to interact with and operate a system.

[0010] "Type of document to be generated" refers to the category or purpose of the document the user wishes to create (for example, a business plan or project plan).

[0011] "Similarity calculation" is a mathematical method for measuring the relevance of content between multiple documents.

[0012] "Cosine similarity" is a method for calculating the similarity between two vectors using the angle between them.

[0013] A "draft of new material" is an initial version of a new document automatically generated by the system based on specified criteria.

[0014] "Grammar check and content check" is a process for detecting grammatical errors, ambiguous expressions, missing information, etc. in a document and suggesting corrections.

[0015] "Feedback" refers to comments and suggestions for corrections provided to the user based on the check results. [Brief explanation of the drawings]

[0016] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0017] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0018] First, the terms used in the following description will be explained.

[0019] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0020] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0021] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0022] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0023] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0024] [First embodiment]

[0025] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0026] 1, a 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.

[0027] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0028] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0029] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the 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.

[0030] 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 of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0031] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0032] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] 2, in the data processing device 12, a specific process 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" according to the technology of the present 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 process 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.

[0034] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0035] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0036] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0037] The present invention provides a system that automatically generates new materials based on similar projects and business documents accumulated in the past, and automatically provides a chat-style discussion before the approver reviews them. Specific embodiments of this system are described below.

[0038] System configuration

[0039] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents.

[0040] Document collection and analysis

[0041] server

[0042] Access the folder path specified by the user (e.g., " / projects / past_docs / ") and retrieve all files in the folder.

[0043] The collected document files (PDF, Word, Excel, PowerPoint, etc.) are opened one by one and the contents (text data) are extracted.

[0044] Using natural language processing techniques, the extracted text content is analyzed and summaries and main topics are extracted.

[0045] Classify document types (business plans, project plans, etc.) and store them in a database.

[0046] Automatic generation of new materials

[0047] User

[0048] Through the chat interface, you input the type of document you want to generate (e.g., "Business Proposal") and the topic (e.g., "New Product X").

[0049] server

[0050] Search a parsed document database for documents most relevant to the input topic.

[0051] Perform a similarity calculation (e.g., cosine similarity) to identify the most relevant documents.

[0052] Based on the identified documents, the system automatically generates a draft of the new material according to a template, including basic elements such as an introduction, objectives, and a project overview.

[0053] Automatic wall hitting and feedback

[0054] User

[0055] Request a review of the generated draft through the chat interface.

[0056] server

[0057] Analyze the text content of the generated draft using the automatic wall-writing function. Use grammar checking software (e.g., Grammarly, LanguageTool, etc.) to detect grammatical errors.

[0058] It also analyzes ambiguous expressions and missing information, and generates feedback based on the detection results.

[0059] The generated feedback is provided to the user through a chat interface, including comments such as "Item A is missing a specific number" or "There are three grammatical errors."

[0060] Specific examples

[0061] For example, if a user types "I want to create a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant document. It then automatically generates a draft of a new business proposal based on the identified documents and presents it to the user. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[0062] In this way, the present invention provides a system that allows users to efficiently create new materials and quickly review them.

[0063] The processing flow will be explained below.

[0064] Step 1:

[0065] User

[0066] A user accesses the system via a chat interface and inputs the path of the folder to be analyzed.

[0067] User: Specify the " / projects / past_docs / " folder.

[0068] Step 2:

[0069] server

[0070] The server collects all document files (PDF, Word, Excel, PowerPoint) from the specified folder path and keeps a list of file names and paths.

[0071] The server retrieves all files in the " / projects / past_docs / " folder and lists their paths.

[0072] Step 3:

[0073] server

[0074] The server opens each collected file and extracts the text content using the appropriate library (e.g. PyPDF2, python-docx) depending on the document format (PDF, Word, etc.).

[0075] The server opens PDF files with PyPDF2 and Word files with python-docx and extracts the text.

[0076] Step 4:

[0077] server

[0078] The server analyzes the extracted text using natural language processing (NLP) techniques (e.g., spaCy, NLTK) to summarize the content and extract key topics and keywords.

[0079] The server uses the spaCy library to parse the text and extract summaries and keywords.

[0080] Step 5:

[0081] server

[0082] Based on the analysis results, the server classifies the documents into categories (e.g., business proposals, project plans) and stores them in a database.

[0083] The server stores the parsed text along with the classification information in a database.

[0084] Step 6:

[0085] User

[0086] The user inputs the type and theme of the material they wish to generate through the chat interface.

[0087] User: Enter "Create a business proposal on the theme of new product X."

[0088] Step 7:

[0089] server

[0090] The server searches the database for relevant parsed documents based on the themes entered by the user, and performs similarity calculations (e.g., cosine similarity) to identify the most relevant documents.

[0091] The server searches the database for keywords related to "business proposal" and "new product X" and calculates cosine similarity to identify highly relevant documents.

[0092] Step 8:

[0093] server

[0094] The server automatically generates a draft of the new material based on the template, including basic elements such as an introduction, objectives, and a project outline, based on the identified related documents.

[0095] The server generates a new draft based on a business proposal template.

[0096] Step 9:

[0097] User

[0098] The user requests a review of the generated draft through a chat interface.

[0099] User: Enter "Please review the business proposal I created."

[0100] Step 10:

[0101] server

[0102] The server uses an automatic writing-on-the-wall function to analyze the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool) to detect errors and ambiguous expressions.

[0103] The server uses Grammarly to detect grammatical errors and ambiguous expressions in the draft.

[0104] Step 11:

[0105] server

[0106] The server generates a summary of the detected problems and suggestions for improvement, and provides feedback to the user through a chat interface.

[0107] The server generates error information and improvement suggestions as feedback and provides them to the user.

[0108] In this way, by performing specific processing at each step, the user can efficiently create new materials and have them reviewed quickly.

[0109] Example 1

[0110] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0111] Conventional document creation systems have the problem of requiring a great deal of time and effort to create and review new documents. Furthermore, there is a lack of effective means for utilizing similar documents from the past, making efficient document creation difficult. Furthermore, document content and grammar checks are not fully automated, placing a burden on users.

[0112] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0113] In this invention, the server includes means for collecting similar documents from the past and analyzing their contents using natural language processing technology, means for inputting the type and theme of the document to be created via a user interface, means for calculating similarity and identifying the most relevant document, means for automatically generating a draft of a new document based on the identified document, means for performing grammar and content checks on the content of the generated draft, and means for generating feedback based on the check results and providing it to the user, thereby enabling the user to quickly and accurately create new documents and review them effectively.

[0114] "Past similar documents" refers to documents that were previously created and are the subject of collection and analysis by the system.

[0115] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language.

[0116] "User interface" refers to an interface through which a user inputs information into a system and performs operations.

[0117] "Similarity calculation" refers to a calculation method that evaluates the relevance between text data and expresses it numerically.

[0118] "Cosine similarity" refers to a method of measuring the similarity between two pieces of text data using angles in a vector space.

[0119] "Feedback" refers to information that indicates reviews and improvements to the generated document.

[0120] "Grammar checking" refers to the process of detecting and correcting grammatical errors in a piece of writing.

[0121] "Content check" refers to the process of checking whether the content of a document is appropriate and whether it contains the necessary information.

[0122] The present invention provides a system that automatically generates new materials based on similar projects and business documents accumulated in the past, and automatically provides a chat-style discussion before the approver reviews them. Specific embodiments of this system are described below.

[0123] System configuration

[0124] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents.

[0125] Document collection and analysis

[0126] server

[0127] The server accesses the folder path specified by the user (e.g., " / projects / past_docs / ") and retrieves all files in the folder.

[0128] The server sequentially opens the collected document files (PDF, Word, Excel, PowerPoint, etc.) and extracts the contents (text data). To do this, it uses a library such as Apache (registered trademark) Tika.

[0129] The server uses natural language processing techniques to analyze the extracted text content and extract summaries and main topics, specifically using libraries such as NLTK and SpaCy.

[0130] The server classifies the documents by type (business proposal, project plan, etc.) and stores them in a database. The database used here is a general database management system (e.g., MySQL (registered trademark), PostgreSQL).

[0131] Automatic generation of new materials

[0132] User

[0133] Through the chat interface, users input the type of document they want to generate (e.g., "business proposal") and theme (e.g., "new product X") as keywords.

[0134] server

[0135] The server searches the parsed document database for documents that are most relevant to the input topic, using a search engine such as ElasticSearch (registered trademark).

[0136] The server performs a similarity calculation (e.g., cosine similarity) to identify the most relevant documents, using a machine learning library such as scikit-learn.

[0137] The server automatically generates a draft of the new material based on the template, using OpenAI's (registered trademark) generative AI model (e.g., GPT-3 (registered trademark)). The generated draft includes basic components such as an introduction, purpose, and outline of the plan.

[0138] Automatic wall hitting and feedback

[0139] User

[0140] The user requests a review of the generated draft through a chat interface.

[0141] server

[0142] The server analyzes the text content of the generated draft using an automated writing-on-the-wall function and detects grammar errors using grammar checking software (e.g., Grammarly or LanguageTool).

[0143] The server also analyzes ambiguous expressions and missing information and generates feedback based on the results of its detection.

[0144] The server provides the generated feedback to the user through a chat interface, including comments such as "Item A is missing a specific number" or "There are three grammatical errors."

[0145] Examples and prompts

[0146] For example, if a user types "I want to create a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant document. It then automatically generates a draft of a new business proposal based on the identified documents and presents it to the user. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[0147] An example prompt is, "Write a business proposal for new product X. Include the use, challenges, and solutions."

[0148] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0149] Step 1: Gather documentation

[0150] server

[0151] Input: A user-specified folder path (e.g. " / projects / past_docs / ").

[0152] Specific operation: The server accesses the specified folder and lists all files in the folder using the os.listdir() function.

[0153] Data processing: Open each file one by one and extract the file contents using Apache Tika, for example, extract text data from PDF, Word, Excel, and PowerPoint files.

[0154] Output: A list of extracted text data.

[0155] Step 2: Parsing the document

[0156] server

[0157] Input: A list of extracted text data.

[0158] Specific operation: The server performs natural language processing on each text data, specifically tokenization, POS tagging, summarization, and major topic extraction using SpaCy.

[0159] Data Calculation: Calculate TF-IDF scores to identify major topics and keywords in the text.

[0160] Output: A list of analysis results, including summary, major topics, and classification information.

[0161] Step 3: Saving to the database

[0162] server

[0163] Input: A list of analysis results.

[0164] Specific operation: The server stores the analysis results in a database such as MySQL or PostgreSQL.

[0165] Data processing: Classify the data based on the type of document (e.g., business plan, project plan), and store the relevant information for each document in a database.

[0166] Output: Parsed document data stored in a database.

[0167] Step 4: Generate request from user

[0168] User

[0169] Input: Enter the type and topic of the document you want to generate (e.g., "Business Proposal" or "New Product X") via the chat interface.

[0170] Specific operation: The user accesses the chat interface and inputs the necessary information, which causes the user's request to be sent to the server.

[0171] Output: Generated request information.

[0172] Step 5: Finding Similar Documents

[0173] server

[0174] Input: Generation request information (document type and subject).

[0175] What it does: The server uses ElasticSearch to search the database for relevant documents.

[0176] Data calculation: Calculate cosine similarity and identify the documents most relevant to the request information. Similarity calculation is performed using scikit-learn.

[0177] Output: A list of the most relevant documents.

[0178] Step 6: Draft a new document

[0179] server

[0180] Input: A list of the most relevant documents.

[0181] Specific operation: The server uses OpenAI's generative AI model (e.g., GPT-3) to automatically generate a draft of a new document based on a template.

[0182] Data processing: Using a generative AI model, a draft is created that includes the document's basic structure, such as its introduction, purpose, and outline.

[0183] Output: A draft of the new document.

[0184] Step 7: Provide the draft to users

[0185] server

[0186] Input: New document draft.

[0187] Specific operation: The server provides the generated draft to the user through a chat interface.

[0188] Output: A draft of the new document sent to the user.

[0189] Step 8: Automatic wall-hitting and feedback generation

[0190] User

[0191] Input: Review request (via chat interface).

[0192] What happens: A user requests a draft for review through the chat interface. This information is sent to the server.

[0193] Output: Review request information.

[0194] server

[0195] Input: Review request information and new document draft.

[0196] What it does: The server uses LanguageTool and Grammarly to detect grammatical errors in the draft, as well as analyzes ambiguous expressions and missing information.

[0197] Data operations: Detecting grammatical errors, checking content, and generating feedback.

[0198] Output: Feedback information.

[0199] Step 9: Provide feedback

[0200] server

[0201] Input: Feedback information.

[0202] Specific Operation: The server provides the generated feedback to the user through a chat interface.

[0203] Output: Feedback sent to the user.

[0204] (Application example 1)

[0205] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0206] In production management and quality control within factories, document creation and review work is a time-consuming and labor-intensive issue. In particular, when launching a new product, it is necessary to extract the necessary information from a huge amount of related past documents and quickly create and review new documents. However, the current manual process is inefficient and prone to human error. There is a need to solve these problems and streamline and automate document creation and review.

[0207] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0208] In this invention, the server includes a means for collecting similar documents from the past and analyzing their contents using natural language processing technology, a means for inputting the type of document to be generated and necessary information via a user interface installed in the factory robot, and a means for calculating similarity and identifying the most relevant document. This makes it possible to efficiently create new documents based on past documents and to generate high-quality documents and quickly review them by using a generative AI model.

[0209] "Past similar documents" are documents that were created previously and have content and formats related to the current project or work.

[0210] "Natural language processing technology" is a technology that allows computers to understand text data written in human language and perform operations such as analysis, generation, and translation.

[0211] A "user interface mounted on a factory robot" is an interface that is installed on a factory robot and allows the user to input data and perform operations.

[0212] "Similarity calculation" is a calculation method for numerically evaluating the degree of similarity between different documents or data sets.

[0213] "Cosine similarity" is a method for assessing the similarity of documents or datasets by calculating the angle between two vectors.

[0214] A "generative AI model" is a learning model that uses artificial intelligence to generate new text or information based on specific input prompts.

[0215] "Grammar checking" is the process of detecting grammatical errors in a document and suggesting corrections.

[0216] "Content check" is a process for verifying whether the contents of a document are logically correct, consistent, and whether there is any missing information.

[0217] "Feedback" means suggestions for correction or improvement provided as a result of analysis or review.

[0218] This invention is a system for automatically generating new documents based on similar past documents and then reviewing their contents. The purpose of this system is to improve the efficiency of creating and reviewing production management and quality control documents within a factory.

[0219] System configuration

[0220] The system consists of three main components: a server, a terminal mounted on the factory robot, and a user. The user interacts with the system through a user interface mounted on the factory robot, while the server consistently collects, analyzes, generates, and reviews documents.

[0221] Document collection and analysis

[0222] server

[0223] Accesses a folder path specified by the user (e.g. " / projects / past_docs / ") and retrieves all files in the folder. Formats include PDF, Word, Excel, PowerPoint, etc.

[0224] The text content of the collected document files is extracted and analyzed using natural language processing techniques, which makes it possible to extract summaries and main topics.

[0225] Classify the document type (e.g. production control report, quality control report, etc.) and store it in a database.

[0226] Automatic generation of new materials

[0227] User

[0228] Through the user interface installed on the factory robot, the type of document to be generated (e.g., "Quality Control Report") and topic (e.g., "New Product X") are entered.

[0229] server

[0230] Search a parsed document database for documents most relevant to the input topic.

[0231] Based on similarity calculations (e.g., cosine similarity) and the most relevant documents, a generative AI model is used to automatically generate a draft of new material, including quality check items, inspection results, and improvement suggestions.

[0232] Automatic wall hitting and feedback

[0233] User

[0234] A user interface on the factory robot allows a review of the generated draft to be requested.

[0235] server

[0236] Analyze the text content of the generated draft using the automatic wall-writing function. Use grammar checking software (e.g., Grammarly, LanguageTool, etc.) to detect grammatical errors.

[0237] It also analyzes ambiguous expressions and missing information and generates feedback, such as "Item A is missing a specific number" or "There are three grammatical errors."

[0238] The generated feedback is provided to the user through a user interface of the factory robot.

[0239] Specific examples

[0240] For example, if a user inputs "I want to create a quality control report" into a factory robot, the system will search the database for past documents related to "quality control report" and "new product X" to identify the most relevant documents. It then uses a generative AI model to automatically generate a draft of a new quality control report based on the identified documents. The generated draft is then checked for grammar and content, and feedback is provided to the user.

[0241] Prompt Sentence Examples

[0242] I want to create a quality control report

[0243] Please prepare a quality control report for new product X.

[0244] In this manner, the present invention provides a system that allows users to efficiently create new materials and quickly review them.

[0245] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0246] Step 1:

[0247] The user inputs the type of new material (e.g., "Quality Control Report") and topic (e.g., "New Product X") via a user interface installed on the factory robot. The input information is sent to the server.

[0248] Input: Type of material and subject

[0249] Output: Request data sent to the server

[0250] Step 2:

[0251] The server accesses the specified folder path (e.g., " / projects / past_docs / "), collects all document files (PDF, Word, Excel, PowerPoint, etc.) in the folder, and opens them to extract text data.

[0252] Input: the required folder path

[0253] Output: Text data of collected documents

[0254] Step 3:

[0255] The server analyzes the collected text data using natural language processing technology, extracts summaries and main topics from the analysis results, classifies them by document type, and stores them in a database.

[0256] Input: Text data of collected documents

[0257] Output: Summary of the analyzed document, main topics, and classification information

[0258] Step 4:

[0259] The server searches through the documents stored in the database to find the documents that are most relevant to the topic entered by the user. It identifies the relevant documents using similarity calculations (e.g., cosine similarity).

[0260] Input: User-entered themes, documents from the database

[0261] Output: Most relevant documents

[0262] Step 5:

[0263] Based on the identified documents, the server uses a generative AI model to automatically generate a draft of the new material, including basic components such as an introduction, purpose, and outline of the project.

[0264] Input: Most relevant documents

[0265] Output: Automatically generated draft of new material

[0266] Step 6:

[0267] When a user requests a review of a generated draft, the server uses an automated text-based grammar and content checker to detect errors.

[0268] Input: Draft of new material, user review request

[0269] Output: Grammar and content error detection results

[0270] Step 7:

[0271] The server generates feedback based on the detected errors and provides it to the user with specific suggestions for correction, such as "Item A is missing a specific number" or "There are three grammatical errors."

[0272] Input: Error detection result

[0273] Output: Generated feedback, suggested corrections

[0274] By the above steps, the present invention makes it possible to efficiently and automatically create and review new materials in a factory.

[0275] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0276] The present invention combines an emotion engine with a system that automatically generates new materials based on similar projects and business documents accumulated in the past and provides automatic chat-style discussion before the approver reviews them. Specific embodiments of this system are described below.

[0277] System configuration

[0278] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents. The emotion engine also recognizes the user's emotions and reflects them in the document creation process.

[0279] Document collection and analysis

[0280] server

[0281] Access the folder path specified by the user (e.g., " / projects / past_docs / ") and collect all files in the folder.

[0282] The collected document files (PDF, Word, Excel, PowerPoint, etc.) are opened one by one and the contents (text data) are extracted.

[0283] Using natural language processing techniques, the extracted text content is analyzed and summaries and main topics are extracted.

[0284] Classify document types (business plans, project plans, etc.) and store them in a database.

[0285] Automatic generation of new materials

[0286] User

[0287] Through the chat interface, you input the type of document you want to generate (e.g., "Business Proposal") and the topic (e.g., "New Product X").

[0288] server

[0289] An emotion engine is used to recognize the user's emotions from the input text data.

[0290] Search a parsed document database for documents most relevant to the input topic.

[0291] Perform a similarity calculation (e.g., cosine similarity) to identify the most relevant documents.

[0292] Based on the identified documents, the system automatically generates a template-based draft of the new material, including basic elements such as an introduction, objectives, and project outline, while adjusting the tone and style of the material based on the user's perceived emotions.

[0293] Automatic wall hitting and feedback

[0294] User

[0295] Request a review of the generated draft through the chat interface.

[0296] server

[0297] The text content of the generated draft is analyzed using grammar checking software (e.g., Grammarly, LanguageTool) to detect errors and ambiguous expressions.

[0298] It also analyzes ambiguous expressions and missing information, and generates feedback based on the detection results.

[0299] The generated feedback is provided to the user through a chat interface, including comments such as "Item A is missing specific numbers," "There are three grammatical errors," and "Please adjust your tone based on sentiment analysis."

[0300] Specific examples

[0301] For example, if a user types "I want to write a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant documents. It then automatically generates a draft of the new business proposal based on the identified documents. The emotion engine recognizes the user's emotions based on the text entered by the user and adjusts the tone and style of the document accordingly. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[0302] In this way, the present invention not only enables users to efficiently create new materials and quickly review them, but also provides a system that enables document creation that takes the user's emotions into consideration.

[0303] The processing flow will be explained below.

[0304] Step 1:

[0305] User

[0306] A user accesses the system via a chat interface and inputs the path of the folder to be analyzed.

[0307] User: Specify the " / projects / past_docs / " folder.

[0308] Step 2:

[0309] server

[0310] The server collects all document files (PDF, Word, Excel, PowerPoint) from the specified folder path and keeps a list of file names and paths.

[0311] The server retrieves all files in the " / projects / past_docs / " folder and lists their paths.

[0312] Step 3:

[0313] server

[0314] The server opens each collected file and extracts the text content using the appropriate library (e.g. PyPDF2, python-docx) depending on the document format (PDF, Word, etc.).

[0315] The server opens PDF files with PyPDF2 and Word files with python-docx and extracts the text.

[0316] Step 4:

[0317] server

[0318] The server analyzes the extracted text using natural language processing (NLP) techniques (e.g., spaCy, NLTK) to summarize the content and extract key topics and keywords.

[0319] The server uses the spaCy library to parse the text and extract summaries and keywords.

[0320] Step 5:

[0321] server

[0322] Based on the analysis results, the server classifies the documents into categories (e.g., business proposals, project plans) and stores them in a database.

[0323] The server stores the parsed text along with the classification information in a database.

[0324] Step 6:

[0325] User

[0326] The user inputs the type and theme of the material they wish to generate through the chat interface.

[0327] User: Enter "Create a business proposal on the theme of new product X."

[0328] Step 7:

[0329] server

[0330] The server analyzes input data from the user interface using an emotion engine based on the theme entered by the user to recognize the emotion of the user.

[0331] The server uses an emotion engine to analyze the user's input text and recognize emotions such as "positive" and "negative."

[0332] Step 8:

[0333] server

[0334] The server searches the analyzed document database for documents that are most relevant to the input topic, taking into account the emotion recognition results. It performs similarity calculations (e.g., cosine similarity) to identify the most relevant documents.

[0335] The server searches the database for keywords related to "business proposal" and "new product X" and calculates cosine similarity to identify highly relevant documents.

[0336] Step 9:

[0337] server

[0338] The server automatically generates a template-based draft of the new document based on the identified related documents. The draft includes basic elements such as an introduction, objectives, and a project outline. The tone and style of the document are adjusted based on the user's perceived emotions.

[0339] The server generates a new draft based on a business proposal template, adapting the tone and style to match the user's emotional needs.

[0340] Step 10:

[0341] User

[0342] The user requests a review of the generated draft through a chat interface.

[0343] User: Enter "Please review the business proposal I created."

[0344] Step 11:

[0345] server

[0346] The server uses an automatic writing-on-the-wall function to analyze the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool) to detect errors and ambiguous expressions.

[0347] The server uses Grammarly to detect grammatical errors and ambiguous expressions in the draft.

[0348] Step 12:

[0349] server

[0350] The server generates a summary of the detected issues and suggestions for improving tone and style based on sentiment analysis, and provides feedback to the user via a chat interface.

[0351] The server generates feedback including error information, suggestions for improvement, and comments based on sentiment analysis, and provides it to the user.

[0352] In this way, by performing specific processing at each step, the user can efficiently create new materials, have them reviewed quickly, and create documents that take the user's feelings into consideration.

[0353] Example 2

[0354] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0355] In conventional document generation systems, the manual search for similar documents, analysis of their contents, and creation of new materials required time and effort, and it was difficult to create documents that took the user's feelings into consideration. Furthermore, the grammar and content of the generated drafts had to be checked manually, preventing automation. This made it difficult for users to create documents efficiently and with high quality.

[0356] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes: means for collecting past similar documents and analyzing their contents using natural language processing technology; means for inputting the type of material to be generated and necessary information via a user interface; means for recognizing the user's sentiment from the input text data; means for calculating similarity and identifying the most relevant document; means for automatically generating a draft of a new material based on the identified document; means for adjusting the tone and style of the generated draft based on the user's sentiment; means for checking the grammar and content of the generated draft; and means for generating feedback based on the check results and providing it to the user. This enables the automation of document generation work, allowing users to efficiently create high-quality documents and quickly review them.

[0357] "Previous similar documents" refers to previously created, highly relevant materials and records.

[0358] "Natural language processing technology" refers to technology that enables computers to understand, interpret, and manipulate human language.

[0359] "User interface" refers to the components that provide the means for a user to interact with a system.

[0360] "Means of recognizing emotions" refers to technologies and algorithms that analyze emotions from user input and understand those emotions.

[0361] "Similarity calculation" refers to a calculation method for measuring the similarity between multiple documents and evaluating their relevance.

[0362] "Document" means any electronic or physical document containing certain information.

[0363] A "draft" is a rough draft of a document that is not yet complete, but has a certain degree of content.

[0364] "Tone and style adjustment" refers to the process of changing the wording and writing style of a document based on the user's feelings.

[0365] "Grammar Check" refers to the functionality that detects grammatical errors in text and suggests corrections.

[0366] "Content check" refers to the function that checks the content of the text for consistency and accuracy and suggests corrections if necessary.

[0367] "Feedback" refers to evaluations and suggestions for improvement provided to users by the system.

[0368] "Chat interface" refers to an interface that allows users and systems to communicate text-based.

[0369] The present invention is a system that automatically generates new materials based on similar documents from the past and adjusts the tone and style according to the user's emotions. An embodiment of the system will be described in detail below.

[0370] System configuration

[0371] This system consists of three main components: a server, a terminal, and a user. The server is responsible for the collection, analysis, generation, and review of documents, while the terminal provides an interface for users to interact with the system. Users access the system through their terminal and perform the necessary operations.

[0372] Hardware and software used

[0373] 1. Server

[0374] Hardware: General server machine (CPU, memory, storage)

[0375] software:

[0376] Document processing: PyPDF2 (analyzing PDF files), python-docx (analyzing Word files)

[0377] Natural Language Processing: spaCy, NLTK

[0378] Emotion Recognition: A publicly available sentiment analysis API

[0379] Similarity calculation: Similarity calculation algorithm (e.g., cosine similarity)

[0380] Grammar Check: Grammarly, LanguageTool

[0381] 2. Terminal

[0382] Hardware: General purpose computers, tablets, smartphones

[0383] Software: Browser or dedicated chat interface application

[0384] Process Overview

[0385] 1. Document collection and analysis

[0386] The server accesses the folder path specified by the user and collects all files found there. It then opens the collected document files using a library corresponding to their format (e.g., PyPDF2 or python-docx) and extracts the text content. The extracted text is then analyzed using natural language processing techniques (spaCy or NLTK) to extract summaries and main topics. The results of these analyses are categorized based on document type and stored in a database.

[0387] 2. Automatic generation of new materials

[0388] Users input the type and theme of the document they wish to generate through a chat interface. The server uses an emotion engine to recognize the user's emotions from the input text data. It then searches the analyzed data stored in the database for documents most relevant to the input theme and performs a similarity calculation. Once the most relevant document is identified, a draft of the new document is automatically generated based on a template. The generated draft includes basic components such as an introduction, purpose, and outline of the plan, and the tone and style are adjusted based on the user's emotions.

[0389] 3. Automatic wall hitting and feedback

[0390] The user requests a review of the generated draft. The server then uses an automated wall-hitting function to analyze the text content of the draft using grammar checking software (Grammarly, LanguageTool) to detect grammatical errors, ambiguous expressions, and missing information. Based on the results, feedback is generated and provided to the user through a chat interface. The feedback includes specific error indications and suggestions for improvement.

[0391] Specific examples

[0392] For example, if a user types "I want to write a business proposal for new product X" into the chat interface, the server collects past business proposal documents from the specified folder and performs the necessary analysis. It then searches for documents related to the user's theme, "new product X," and automatically generates a draft of a new business proposal based on the most relevant documents after similarity calculations. This draft is adjusted to reflect the user's sentiment in tone and style, and based on the user's review request, it undergoes grammar and content checks and provides specific feedback.

[0393] Example prompt: "I would like to create a business proposal for new product X. Please refer to the previous project materials to generate a draft and review it."

[0394] This series of processes enables users to efficiently create high-quality documents and review them quickly.

[0395] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0396] Step 1:

[0397] Document file collection

[0398] The server accesses the folder path specified by the user (e.g. " / projects / past_docs / ") and lists all the files in the folder.

[0399] Input: Folder path

[0400] Specific behavior: Scans the file system and generates a list of files in the specified folder.

[0401] Output: File list

[0402] Step 2:

[0403] Document Content Extraction

[0404] The server filters document files (PDF, Word, Excel, PowerPoint, etc.) from the listed files.

[0405] Each document file is opened using a library corresponding to the format and the text content is extracted (e.g. PyPDF2 for PDF, python-docx for Word).

[0406] Input: File list

[0407] Specific operation: Opens a document file and extracts the text.

[0408] Output: Extracted text data

[0409] Step 3:

[0410] Text Analysis and Classification

[0411] The server analyzes the extracted text using natural language processing techniques (e.g., spaCy, NLTK) to extract summaries and main topics.

[0412] It automatically classifies documents into types (e.g., business proposals, project plans, etc.) based on their content.

[0413] The analysis results are stored in a database.

[0414] Input: Extracted text data

[0415] What it does: Summarizes text, extracts major topics, and classifies document types.

[0416] Output: Abstract, main topics, classification information

[0417] Step 4:

[0418] Acceptance of document generation requests

[0419] Users simply input the type of document they want to generate (e.g., business proposal) and topic (e.g., new product X) through a chat interface.

[0420] Input: Material type and subject

[0421] Specific actions: Enter a request for document generation into the chat interface.

[0422] Output: Document generation request

[0423] Step 5:

[0424] Emotion Recognition and Document Retrieval

[0425] The server uses an emotion engine to recognize the user's emotions from the input text data.

[0426] It searches a database of analyzed documents for the documents most relevant to the input topic.

[0427] Input: Document generation request, existing parsed document

[0428] Specific behavior: Recognizes emotions from input text and searches for related documents.

[0429] Output: Emotion recognition results, related documents

[0430] Step 6:

[0431] Automatic draft generation

[0432] The server performs a similarity calculation (e.g., cosine similarity) to identify the most relevant documents.

[0433] Based on the identified documents, it automatically generates drafts of new materials based on templates (e.g., introduction, objectives, project overview, etc.).

[0434] It adjusts the tone and style of the material based on the user's emotion recognition results.

[0435] Input: Emotion recognition results, related documents

[0436] What it does: Selects the most relevant documents based on similarity calculations, generates drafts, and adjusts tone and style.

[0437] Output: Draft of new material

[0438] Step 7:

[0439] Request a draft review

[0440] Users can request a review of the generated draft through a chat interface.

[0441] Input: Draft of new material

[0442] What it does: Sends a review request in the chat interface.

[0443] Output: Review request

[0444] Step 8:

[0445] Checking and parsing the draft

[0446] The server uses an automatic wall-hitting function and analyzes the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool).

[0447] It detects grammatical errors, ambiguous expressions, and missing information.

[0448] Input: Draft of new material, grammar checking software

[0449] Specific actions: Analyze the draft text with grammar checking software to detect errors and omissions.

[0450] Output: Syntax error detection results

[0451] Step 9:

[0452] Generating and Providing Feedback

[0453] The server generates feedback based on the detection results and adds specific improvement suggestions (e.g., "Item A lacks specific numbers," "There are three grammatical errors," "Please adjust your tone based on sentiment analysis").

[0454] Feedback is provided to users through a chat interface.

[0455] Input: Syntax error detection result

[0456] Specific behavior: Generate feedback from the detection results and provide it in a chat interface.

[0457] Output: Feedback

[0458] Through the above steps, the user can efficiently create new materials and quickly review them.

[0459] (Application example 2)

[0460] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0461] In logistics centers, creating new documents requires a lot of time and effort, and the process of extracting necessary information from past documents to create new plans and reports is particularly cumbersome. Furthermore, traditional methods tend to delay review and feedback on created documents, hindering rapid business processes. Furthermore, documents created without considering user sentiment may be difficult for users to understand and may be ineffective.

[0462] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0463] In this invention, the server includes means for collecting similar documents from the past and analyzing their contents using natural language processing technology, means for inputting the type of material to be generated and necessary information via a user interface, and means for calculating similarity and identifying the most relevant document. This makes it possible to efficiently automatically generate new materials based on past documents and provide real-time feedback in a document style that reflects the user's feelings.

[0464] "Past similar documents" are documents or plans created in the past, which have similar content and format to the newly generated document.

[0465] "Natural language processing technology" is a technology for analyzing human language and automatically understanding, processing, and generating it, and includes functions such as extracting, summarizing, and classifying text data.

[0466] A "user interface" is an interface that allows a system and a user to interact with each other, and provides operation means using voice input, text input, touch gestures, etc.

[0467] "Similarity calculation" is a method for calculating the similarity between multiple documents, and measures the similarity using a statistical method such as cosine similarity.

[0468] "Draft of new material" refers to an early stage material that is automatically generated based on the theme and format specified by the user.

[0469] "Generated Draft Content" refers to the text and information contained in the automatically generated early stage material.

[0470] "Grammar checking" is a process of checking whether the generated text content contains any grammatical errors, and is carried out using natural language processing technology.

[0471] "Content checking" is the process of verifying that the generated text content is appropriate and consistent and that no important information is missing.

[0472] An "emotion engine" is a system that analyzes emotions from user input and adjusts the style and tone of the generated document based on those emotions.

[0473] "Feedback" is information including review results and improvement suggestions for the generated draft, and is provided in response to a user's request.

[0474] "Providing feedback in real time" refers to a process in which the review results of the generated materials are immediately provided to the user, allowing the user to quickly make any necessary corrections or confirmations.

[0475] The present invention is a system that automatically generates new materials based on similar projects and business documents accumulated in the past, and combines them with an emotion engine to automatically bounce ideas off each other in a chat format. Specific embodiments of the present invention are described below.

[0476] System configuration

[0477] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents. The emotion engine also recognizes the user's emotions and reflects them in the document creation process.

[0478] Document collection and analysis

[0479] The server accesses the folder path specified by the user (e.g., " / projects / past_docs / ") and collects all files in the folder. It opens the collected document files (PDF, Word, Excel, PowerPoint, etc.) and extracts the content (text data). It uses natural language processing technology to analyze the extracted text content and extract summaries and main topics. It classifies the document type (business proposal, project plan, etc.) and stores it in a database.

[0480] Automatic generation of new materials

[0481] The user inputs the type of document they wish to generate (e.g., "business proposal") and theme (e.g., "new product X") through the user interface on their device. The server uses an emotion engine to recognize the user's emotions from the input text data. It then searches the analyzed document database for documents most relevant to the input theme. It performs similarity calculations (e.g., cosine similarity) to identify the most relevant documents. Based on the identified documents, it automatically generates a draft of the new document in accordance with a template. The generated draft includes basic components such as an introduction, purpose, and plan outline. The tone and style of the document are then adjusted based on the user's recognized emotions.

[0482] Automatic wall hitting and feedback

[0483] The user requests a review of the generated draft through the chat interface. The server uses an automated back-and-forth function to analyze the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool). Errors and ambiguous expressions are detected. Ambiguous expressions and missing information are also analyzed, and feedback is generated based on the detection results. The generated feedback is provided to the user through the chat interface. The feedback includes comments such as "Item A is missing a specific number," "There are three grammatical errors," and "Please adjust the tone based on sentiment analysis."

[0484] Specific examples

[0485] For example, if a user types "I want to write a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant documents. It then automatically generates a draft of the new business proposal based on the identified documents. The emotion engine recognizes the user's emotions based on the text entered by the user and adjusts the tone and style of the document accordingly. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[0486] An example of a prompt sentence to input to the generative AI model is as follows:

[0487] "I want to create a shift plan for next week."

[0488] "We want to create a plan for implementing a new warehouse management system."

[0489] This invention enables logistics center managers and staff to efficiently create new documents and quickly review them. It also enables document creation that takes user emotions into consideration, reducing the burden on workers.

[0490] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0491] Step 1:

[0492] The server accesses the folder path specified by the user (e.g., " / projects / past_docs / ") and collects all files in the folder. The collected files include different formats such as PDF, Word, Excel, and PowerPoint. These files are opened sequentially and text data is extracted. This is done using natural language processing technology to extract summaries and main topics of the text content. The extracted data is classified by document type and stored in a database.

[0493] Input: Folder path

[0494] Output: Extracted text data, classified document data

[0495] Step 2:

[0496] The user inputs the type of document (e.g., "business proposal") and theme (e.g., "new product X") they want to generate through the user interface of their terminal. This input information is sent to the server.

[0497] Input: Type of material, theme

[0498] Output: Sending input information to the server

[0499] Step 3:

[0500] The server uses an emotion engine to recognize emotions from the text data entered by the user, searches the analyzed document database for documents that are most relevant to the entered topic, and uses similarity calculations, such as cosine similarity, to identify the most relevant documents.

[0501] Input: User-entered text data

[0502] Output: Identifying relevant documents

[0503] Step 4:

[0504] The server automatically generates a draft of the new document based on the template, including basic elements such as an introduction, objectives, and a project outline, and adjusts the tone and style of the document based on the user's perceived emotions.

[0505] Input: Identified documents, sentiment data

[0506] Output: Draft of new material

[0507] Step 5:

[0508] Through the chat interface, the user requests a review of the generated draft, which is then sent to the server.

[0509] Input: Review request

[0510] Output: Request sent to server

[0511] Step 6:

[0512] The server uses an automated writing function to analyze the text content of the generated draft using grammar checking software, such as Grammarly or LanguageTool, to detect errors and ambiguous expressions, as well as to analyze missing information and generate feedback based on the results.

[0513] Input: Generated draft

[0514] Output: Grammar check results, feedback

[0515] Step 7:

[0516] The server provides the generated feedback to the user through a chat interface, including comments such as "Item A is missing a specific number," "There are three grammatical errors," and "Please adjust your tone based on sentiment analysis."

[0517] Input: Feedback

[0518] Output: Providing feedback to the user

[0519] This series of steps allows the distribution center manager and staff to efficiently create new documents and quickly review them. Specific examples of prompts are as follows:

[0520] "I want to create a shift plan for next week."

[0521] "We want to create a plan for implementing a new warehouse management system."

[0522] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the 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.

[0523] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0524] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0525] [Second embodiment]

[0526] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0527] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0528] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0529] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0530] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0531] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0532] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0533] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0534] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[0535] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0536] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0537] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. 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."

[0538] The present invention provides a system that automatically generates new materials based on similar projects and business documents accumulated in the past, and automatically provides a chat-style discussion before the approver reviews them. Specific embodiments of this system are described below.

[0539] System configuration

[0540] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents.

[0541] Document collection and analysis

[0542] server

[0543] Access the folder path specified by the user (e.g., " / projects / past_docs / ") and retrieve all files in the folder.

[0544] The collected document files (PDF, Word, Excel, PowerPoint, etc.) are opened one by one and the contents (text data) are extracted.

[0545] Using natural language processing techniques, the extracted text content is analyzed and summaries and main topics are extracted.

[0546] Classify document types (business plans, project plans, etc.) and store them in a database.

[0547] Automatic generation of new materials

[0548] User

[0549] Through the chat interface, you input the type of document you want to generate (e.g., "Business Proposal") and the topic (e.g., "New Product X").

[0550] server

[0551] Search a parsed document database for documents most relevant to the input topic.

[0552] Perform a similarity calculation (e.g., cosine similarity) to identify the most relevant documents.

[0553] Based on the identified documents, the system automatically generates a draft of the new material according to a template, including basic elements such as an introduction, objectives, and a project overview.

[0554] Automatic wall hitting and feedback

[0555] User

[0556] Request a review of the generated draft through the chat interface.

[0557] server

[0558] Analyze the text content of the generated draft using the automatic wall-writing function. Use grammar checking software (e.g., Grammarly, LanguageTool, etc.) to detect grammatical errors.

[0559] It also analyzes ambiguous expressions and missing information, and generates feedback based on the detection results.

[0560] The generated feedback is provided to the user through a chat interface, including comments such as "Item A is missing a specific number" or "There are three grammatical errors."

[0561] Specific examples

[0562] For example, if a user types "I want to create a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant document. It then automatically generates a draft of a new business proposal based on the identified documents and presents it to the user. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[0563] In this way, the present invention provides a system that allows users to efficiently create new materials and quickly review them.

[0564] The processing flow will be explained below.

[0565] Step 1:

[0566] User

[0567] A user accesses the system via a chat interface and inputs the path of the folder to be analyzed.

[0568] User: Specify the " / projects / past_docs / " folder.

[0569] Step 2:

[0570] server

[0571] The server collects all document files (PDF, Word, Excel, PowerPoint) from the specified folder path and keeps a list of file names and paths.

[0572] The server retrieves all files in the " / projects / past_docs / " folder and lists their paths.

[0573] Step 3:

[0574] server

[0575] The server opens each collected file and extracts the text content using the appropriate library (e.g. PyPDF2, python-docx) depending on the document format (PDF, Word, etc.).

[0576] The server opens PDF files with PyPDF2 and Word files with python-docx and extracts the text.

[0577] Step 4:

[0578] server

[0579] The server analyzes the extracted text using natural language processing (NLP) techniques (e.g., spaCy, NLTK) to summarize the content and extract key topics and keywords.

[0580] The server uses the spaCy library to parse the text and extract summaries and keywords.

[0581] Step 5:

[0582] server

[0583] Based on the analysis results, the server classifies the documents into categories (e.g., business proposals, project plans) and stores them in a database.

[0584] The server stores the parsed text along with the classification information in a database.

[0585] Step 6:

[0586] User

[0587] The user inputs the type and theme of the material they wish to generate through the chat interface.

[0588] User: Enter "Create a business proposal on the theme of new product X."

[0589] Step 7:

[0590] server

[0591] The server searches the database for relevant parsed documents based on the themes entered by the user, and performs similarity calculations (e.g., cosine similarity) to identify the most relevant documents.

[0592] The server searches the database for keywords related to "business proposal" and "new product X" and calculates cosine similarity to identify highly relevant documents.

[0593] Step 8:

[0594] server

[0595] The server automatically generates a draft of the new material based on the template, including basic elements such as an introduction, objectives, and a project outline, based on the identified related documents.

[0596] The server generates a new draft based on a business proposal template.

[0597] Step 9:

[0598] User

[0599] The user requests a review of the generated draft through a chat interface.

[0600] User: Enter "Please review the business proposal I created."

[0601] Step 10:

[0602] server

[0603] The server uses an automatic writing-on-the-wall function to analyze the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool) to detect errors and ambiguous expressions.

[0604] The server uses Grammarly to detect grammatical errors and ambiguous expressions in the draft.

[0605] Step 11:

[0606] server

[0607] The server generates a summary of the detected problems and suggestions for improvement, and provides feedback to the user through a chat interface.

[0608] The server generates error information and improvement suggestions as feedback and provides them to the user.

[0609] In this way, by performing specific processing at each step, the user can efficiently create new materials and have them reviewed quickly.

[0610] Example 1

[0611] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0612] Conventional document creation systems have the problem of requiring a great deal of time and effort to create and review new documents. Furthermore, there is a lack of effective means for utilizing similar documents from the past, making efficient document creation difficult. Furthermore, document content and grammar checks are not fully automated, placing a burden on users.

[0613] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0614] In this invention, the server includes means for collecting similar documents from the past and analyzing their contents using natural language processing technology, means for inputting the type and theme of the document to be created via a user interface, means for calculating similarity and identifying the most relevant document, means for automatically generating a draft of a new document based on the identified document, means for performing grammar and content checks on the content of the generated draft, and means for generating feedback based on the check results and providing it to the user, thereby enabling the user to quickly and accurately create new documents and review them effectively.

[0615] "Past similar documents" refers to documents that were previously created and are the subject of collection and analysis by the system.

[0616] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language.

[0617] "User interface" refers to an interface through which a user inputs information into a system and performs operations.

[0618] "Similarity calculation" refers to a calculation method that evaluates the relevance between text data and expresses it numerically.

[0619] "Cosine similarity" refers to a method of measuring the similarity between two pieces of text data using angles in a vector space.

[0620] "Feedback" refers to information that indicates reviews and improvements to the generated document.

[0621] "Grammar checking" refers to the process of detecting and correcting grammatical errors in a piece of writing.

[0622] "Content check" refers to the process of checking whether the content of a document is appropriate and whether it contains the necessary information.

[0623] The present invention provides a system that automatically generates new materials based on similar projects and business documents accumulated in the past, and automatically provides a chat-style discussion before the approver reviews them. Specific embodiments of this system are described below.

[0624] System configuration

[0625] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents.

[0626] Document collection and analysis

[0627] server

[0628] The server accesses the folder path specified by the user (e.g., " / projects / past_docs / ") and retrieves all files in the folder.

[0629] The server sequentially opens the collected document files (PDF, Word, Excel, PowerPoint, etc.) and extracts the contents (text data). To do this, it uses libraries such as Apache Tika.

[0630] The server uses natural language processing techniques to analyze the extracted text content and extract summaries and main topics, specifically using libraries such as NLTK and SpaCy.

[0631] The server classifies the documents by type (business proposal, project plan, etc.) and stores them in a database. The database used here is a general database management system (e.g., MySQL, PostgreSQL).

[0632] Automatic generation of new materials

[0633] User

[0634] Through the chat interface, users input the type of document they want to generate (e.g., "business proposal") and theme (e.g., "new product X") as keywords.

[0635] server

[0636] The server searches the parsed document database for documents that are most relevant to the input topic, using a search engine such as ElasticSearch.

[0637] The server performs a similarity calculation (e.g., cosine similarity) to identify the most relevant documents, using a machine learning library such as scikit-learn.

[0638] The server automatically generates a draft of the new material based on the identified document, following a template, using OpenAI's generative AI model (e.g., GPT-3). The generated draft includes basic components such as an introduction, purpose, and outline of the plan.

[0639] Automatic wall hitting and feedback

[0640] User

[0641] The user requests a review of the generated draft through a chat interface.

[0642] server

[0643] The server analyzes the text content of the generated draft using an automated writing-on-the-wall function and detects grammar errors using grammar checking software (e.g., Grammarly or LanguageTool).

[0644] The server also analyzes ambiguous expressions and missing information and generates feedback based on the results of its detection.

[0645] The server provides the generated feedback to the user through a chat interface, including comments such as "Item A is missing a specific number" or "There are three grammatical errors."

[0646] Examples and prompts

[0647] For example, if a user types "I want to create a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant document. It then automatically generates a draft of a new business proposal based on the identified documents and presents it to the user. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[0648] An example prompt is, "Write a business proposal for new product X. Include the use, challenges, and solutions."

[0649] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0650] Step 1: Gather documentation

[0651] server

[0652] Input: A user-specified folder path (e.g. " / projects / past_docs / ").

[0653] Specific operation: The server accesses the specified folder and lists all files in the folder using the os.listdir() function.

[0654] Data processing: Open each file one by one and extract the file contents using Apache Tika, for example, extract text data from PDF, Word, Excel, and PowerPoint files.

[0655] Output: A list of extracted text data.

[0656] Step 2: Parsing the document

[0657] server

[0658] Input: A list of extracted text data.

[0659] Specific operation: The server performs natural language processing on each text data, specifically tokenization, POS tagging, summarization, and major topic extraction using SpaCy.

[0660] Data Calculation: Calculate TF-IDF scores to identify major topics and keywords in the text.

[0661] Output: A list of analysis results, including summary, major topics, and classification information.

[0662] Step 3: Saving to the database

[0663] server

[0664] Input: A list of analysis results.

[0665] Specific operation: The server stores the analysis results in a database such as MySQL or PostgreSQL.

[0666] Data processing: Classify the data based on the type of document (e.g., business plan, project plan), and store the relevant information for each document in a database.

[0667] Output: Parsed document data stored in a database.

[0668] Step 4: Generate request from user

[0669] User

[0670] Input: Enter the type and topic of the document you want to generate (e.g., "Business Proposal" or "New Product X") via the chat interface.

[0671] Specific operation: The user accesses the chat interface and inputs the necessary information, which causes the user's request to be sent to the server.

[0672] Output: Generated request information.

[0673] Step 5: Finding Similar Documents

[0674] server

[0675] Input: Generation request information (document type and subject).

[0676] What it does: The server uses ElasticSearch to search the database for relevant documents.

[0677] Data calculation: Calculate cosine similarity and identify the documents most relevant to the request information. Similarity calculation is performed using scikit-learn.

[0678] Output: A list of the most relevant documents.

[0679] Step 6: Draft a new document

[0680] server

[0681] Input: A list of the most relevant documents.

[0682] Specific operation: The server uses OpenAI's generative AI model (e.g., GPT-3) to automatically generate a draft of a new document based on a template.

[0683] Data processing: Using a generative AI model, a draft is created that includes the document's basic structure, such as its introduction, purpose, and outline.

[0684] Output: A draft of the new document.

[0685] Step 7: Provide the draft to users

[0686] server

[0687] Input: New document draft.

[0688] Specific operation: The server provides the generated draft to the user through a chat interface.

[0689] Output: A draft of the new document sent to the user.

[0690] Step 8: Automatic wall-hitting and feedback generation

[0691] User

[0692] Input: Review request (via chat interface).

[0693] What happens: A user requests a draft for review through the chat interface. This information is sent to the server.

[0694] Output: Review request information.

[0695] server

[0696] Input: Review request information and new document draft.

[0697] What it does: The server uses LanguageTool and Grammarly to detect grammatical errors in the draft, as well as analyzes ambiguous expressions and missing information.

[0698] Data operations: Detecting grammatical errors, checking content, and generating feedback.

[0699] Output: Feedback information.

[0700] Step 9: Provide feedback

[0701] server

[0702] Input: Feedback information.

[0703] Specific Operation: The server provides the generated feedback to the user through a chat interface.

[0704] Output: Feedback sent to the user.

[0705] (Application example 1)

[0706] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0707] In production management and quality control within factories, document creation and review work is a time-consuming and labor-intensive issue. In particular, when launching a new product, it is necessary to extract the necessary information from a huge amount of related past documents and quickly create and review new documents. However, the current manual process is inefficient and prone to human error. There is a need to solve these problems and streamline and automate document creation and review.

[0708] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0709] In this invention, the server includes a means for collecting similar documents from the past and analyzing their contents using natural language processing technology, a means for inputting the type of document to be generated and necessary information via a user interface installed in the factory robot, and a means for calculating similarity and identifying the most relevant document. This makes it possible to efficiently create new documents based on past documents and to generate high-quality documents and quickly review them by using a generative AI model.

[0710] "Past similar documents" are documents that were created previously and have content and formats related to the current project or work.

[0711] "Natural language processing technology" is a technology that allows computers to understand text data written in human language and perform operations such as analysis, generation, and translation.

[0712] A "user interface mounted on a factory robot" is an interface that is installed on a factory robot and allows the user to input data and perform operations.

[0713] "Similarity calculation" is a calculation method for numerically evaluating the degree of similarity between different documents or data sets.

[0714] "Cosine similarity" is a method for assessing the similarity of documents or datasets by calculating the angle between two vectors.

[0715] A "generative AI model" is a learning model that uses artificial intelligence to generate new text or information based on specific input prompts.

[0716] "Grammar checking" is the process of detecting grammatical errors in a document and suggesting corrections.

[0717] "Content check" is a process for verifying whether the contents of a document are logically correct, consistent, and whether there is any missing information.

[0718] "Feedback" means suggestions for correction or improvement provided as a result of analysis or review.

[0719] This invention is a system for automatically generating new documents based on similar past documents and then reviewing their contents. The purpose of this system is to improve the efficiency of creating and reviewing production management and quality control documents within a factory.

[0720] System configuration

[0721] The system consists of three main components: a server, a terminal mounted on the factory robot, and a user. The user interacts with the system through a user interface mounted on the factory robot, while the server consistently collects, analyzes, generates, and reviews documents.

[0722] Document collection and analysis

[0723] server

[0724] Accesses a folder path specified by the user (e.g. " / projects / past_docs / ") and retrieves all files in the folder. Formats include PDF, Word, Excel, PowerPoint, etc.

[0725] The text content of the collected document files is extracted and analyzed using natural language processing techniques, which makes it possible to extract summaries and main topics.

[0726] Classify the document type (e.g. production control report, quality control report, etc.) and store it in a database.

[0727] Automatic generation of new materials

[0728] User

[0729] Through the user interface installed on the factory robot, the type of document to be generated (e.g., "Quality Control Report") and topic (e.g., "New Product X") are entered.

[0730] server

[0731] Search a parsed document database for documents most relevant to the input topic.

[0732] Based on similarity calculations (e.g., cosine similarity) and the most relevant documents, a generative AI model is used to automatically generate a draft of new material, including quality check items, inspection results, and improvement suggestions.

[0733] Automatic wall hitting and feedback

[0734] User

[0735] A user interface on the factory robot allows a review of the generated draft to be requested.

[0736] server

[0737] Analyze the text content of the generated draft using the automatic wall-writing function. Use grammar checking software (e.g., Grammarly, LanguageTool, etc.) to detect grammatical errors.

[0738] It also analyzes ambiguous expressions and missing information and generates feedback, such as "Item A is missing a specific number" or "There are three grammatical errors."

[0739] The generated feedback is provided to the user through a user interface of the factory robot.

[0740] Specific examples

[0741] For example, if a user inputs "I want to create a quality control report" into a factory robot, the system will search the database for past documents related to "quality control report" and "new product X" to identify the most relevant documents. It then uses a generative AI model to automatically generate a draft of a new quality control report based on the identified documents. The generated draft is then checked for grammar and content, and feedback is provided to the user.

[0742] Prompt Sentence Examples

[0743] I want to create a quality control report

[0744] Please prepare a quality control report for new product X.

[0745] In this manner, the present invention provides a system that allows users to efficiently create new materials and quickly review them.

[0746] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0747] Step 1:

[0748] The user inputs the type of new material (e.g., "Quality Control Report") and topic (e.g., "New Product X") via a user interface installed on the factory robot. The input information is sent to the server.

[0749] Input: Type of material and subject

[0750] Output: Request data sent to the server

[0751] Step 2:

[0752] The server accesses the specified folder path (e.g., " / projects / past_docs / "), collects all document files (PDF, Word, Excel, PowerPoint, etc.) in the folder, and opens them to extract text data.

[0753] Input: the required folder path

[0754] Output: Text data of collected documents

[0755] Step 3:

[0756] The server analyzes the collected text data using natural language processing technology, extracts summaries and main topics from the analysis results, classifies them by document type, and stores them in a database.

[0757] Input: Text data of collected documents

[0758] Output: Summary of the analyzed document, main topics, and classification information

[0759] Step 4:

[0760] The server searches through the documents stored in the database to find the documents that are most relevant to the topic entered by the user. It identifies the relevant documents using similarity calculations (e.g., cosine similarity).

[0761] Input: User-entered themes, documents from the database

[0762] Output: Most relevant documents

[0763] Step 5:

[0764] Based on the identified documents, the server uses a generative AI model to automatically generate a draft of the new material, including basic components such as an introduction, purpose, and outline of the project.

[0765] Input: Most relevant documents

[0766] Output: Automatically generated draft of new material

[0767] Step 6:

[0768] When a user requests a review of a generated draft, the server uses an automated text-based grammar and content checker to detect errors.

[0769] Input: Draft of new material, user review request

[0770] Output: Grammar and content error detection results

[0771] Step 7:

[0772] The server generates feedback based on the detected errors and provides it to the user with specific suggestions for correction, such as "Item A is missing a specific number" or "There are three grammatical errors."

[0773] Input: Error detection result

[0774] Output: Generated feedback, suggested corrections

[0775] By the above steps, the present invention makes it possible to efficiently and automatically create and review new materials in a factory.

[0776] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0777] The present invention combines an emotion engine with a system that automatically generates new materials based on similar projects and business documents accumulated in the past and provides automatic chat-style discussion before the approver reviews them. Specific embodiments of this system are described below.

[0778] System configuration

[0779] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents. The emotion engine also recognizes the user's emotions and reflects them in the document creation process.

[0780] Document collection and analysis

[0781] server

[0782] Access the folder path specified by the user (e.g., " / projects / past_docs / ") and collect all files in the folder.

[0783] The collected document files (PDF, Word, Excel, PowerPoint, etc.) are opened one by one and the contents (text data) are extracted.

[0784] Using natural language processing techniques, the extracted text content is analyzed and summaries and main topics are extracted.

[0785] Classify document types (business plans, project plans, etc.) and store them in a database.

[0786] Automatic generation of new materials

[0787] User

[0788] Through the chat interface, you input the type of document you want to generate (e.g., "Business Proposal") and the topic (e.g., "New Product X").

[0789] server

[0790] An emotion engine is used to recognize the user's emotions from the input text data.

[0791] Search a parsed document database for documents most relevant to the input topic.

[0792] Perform a similarity calculation (e.g., cosine similarity) to identify the most relevant documents.

[0793] Based on the identified documents, the system automatically generates a template-based draft of the new material, including basic elements such as an introduction, objectives, and project outline, while adjusting the tone and style of the material based on the user's perceived emotions.

[0794] Automatic wall hitting and feedback

[0795] User

[0796] Request a review of the generated draft through the chat interface.

[0797] server

[0798] The text content of the generated draft is analyzed using grammar checking software (e.g., Grammarly, LanguageTool) to detect errors and ambiguous expressions.

[0799] It also analyzes ambiguous expressions and missing information, and generates feedback based on the detection results.

[0800] The generated feedback is provided to the user through a chat interface, including comments such as "Item A is missing specific numbers," "There are three grammatical errors," and "Please adjust your tone based on sentiment analysis."

[0801] Specific examples

[0802] For example, if a user types "I want to write a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant documents. It then automatically generates a draft of the new business proposal based on the identified documents. The emotion engine recognizes the user's emotions based on the text entered by the user and adjusts the tone and style of the document accordingly. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[0803] In this way, the present invention not only enables users to efficiently create new materials and quickly review them, but also provides a system that enables document creation that takes the user's emotions into consideration.

[0804] The processing flow will be explained below.

[0805] Step 1:

[0806] User

[0807] A user accesses the system via a chat interface and inputs the path of the folder to be analyzed.

[0808] User: Specify the " / projects / past_docs / " folder.

[0809] Step 2:

[0810] server

[0811] The server collects all document files (PDF, Word, Excel, PowerPoint) from the specified folder path and keeps a list of file names and paths.

[0812] The server retrieves all files in the " / projects / past_docs / " folder and lists their paths.

[0813] Step 3:

[0814] server

[0815] The server opens each collected file and extracts the text content using the appropriate library (e.g. PyPDF2, python-docx) depending on the document format (PDF, Word, etc.).

[0816] The server opens PDF files with PyPDF2 and Word files with python-docx and extracts the text.

[0817] Step 4:

[0818] server

[0819] The server analyzes the extracted text using natural language processing (NLP) techniques (e.g., spaCy, NLTK) to summarize the content and extract key topics and keywords.

[0820] The server uses the spaCy library to parse the text and extract summaries and keywords.

[0821] Step 5:

[0822] server

[0823] Based on the analysis results, the server classifies the documents into categories (e.g., business proposals, project plans) and stores them in a database.

[0824] The server stores the parsed text along with the classification information in a database.

[0825] Step 6:

[0826] User

[0827] The user inputs the type and theme of the material they wish to generate through the chat interface.

[0828] User: Enter "Create a business proposal on the theme of new product X."

[0829] Step 7:

[0830] server

[0831] The server analyzes input data from the user interface using an emotion engine based on the theme entered by the user to recognize the emotion of the user.

[0832] The server uses an emotion engine to analyze the user's input text and recognize emotions such as "positive" and "negative."

[0833] Step 8:

[0834] server

[0835] The server searches the analyzed document database for documents that are most relevant to the input topic, taking into account the emotion recognition results. It performs similarity calculations (e.g., cosine similarity) to identify the most relevant documents.

[0836] The server searches the database for keywords related to "business proposal" and "new product X" and calculates cosine similarity to identify highly relevant documents.

[0837] Step 9:

[0838] server

[0839] The server automatically generates a template-based draft of the new document based on the identified related documents. The draft includes basic elements such as an introduction, objectives, and a project outline. The tone and style of the document are adjusted based on the user's perceived emotions.

[0840] The server generates a new draft based on a business proposal template, adapting the tone and style to match the user's emotional needs.

[0841] Step 10:

[0842] User

[0843] The user requests a review of the generated draft through a chat interface.

[0844] User: Enter "Please review the business proposal I created."

[0845] Step 11:

[0846] server

[0847] The server uses an automatic writing-on-the-wall function to analyze the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool) to detect errors and ambiguous expressions.

[0848] The server uses Grammarly to detect grammatical errors and ambiguous expressions in the draft.

[0849] Step 12:

[0850] server

[0851] The server generates a summary of the detected issues and suggestions for improving tone and style based on sentiment analysis, and provides feedback to the user via a chat interface.

[0852] The server generates feedback including error information, suggestions for improvement, and comments based on sentiment analysis, and provides it to the user.

[0853] In this way, by performing specific processing at each step, the user can efficiently create new materials, have them reviewed quickly, and create documents that take the user's feelings into consideration.

[0854] Example 2

[0855] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0856] In conventional document generation systems, the manual search for similar documents, analysis of their contents, and creation of new materials required time and effort, and it was difficult to create documents that took the user's feelings into consideration. Furthermore, the grammar and content of the generated drafts had to be checked manually, preventing automation. This made it difficult for users to create documents efficiently and with high quality.

[0857] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes: means for collecting past similar documents and analyzing their contents using natural language processing technology; means for inputting the type of material to be generated and necessary information via a user interface; means for recognizing the user's sentiment from the input text data; means for calculating similarity and identifying the most relevant document; means for automatically generating a draft of a new material based on the identified document; means for adjusting the tone and style of the generated draft based on the user's sentiment; means for checking the grammar and content of the generated draft; and means for generating feedback based on the check results and providing it to the user. This enables the automation of document generation work, allowing users to efficiently create high-quality documents and quickly review them.

[0858] "Previous similar documents" refers to previously created, highly relevant materials and records.

[0859] "Natural language processing technology" refers to technology that enables computers to understand, interpret, and manipulate human language.

[0860] "User interface" refers to the components that provide the means for a user to interact with a system.

[0861] "Means of recognizing emotions" refers to technologies and algorithms that analyze emotions from user input and understand those emotions.

[0862] "Similarity calculation" refers to a calculation method for measuring the similarity between multiple documents and evaluating their relevance.

[0863] "Document" means any electronic or physical document containing certain information.

[0864] A "draft" is a rough draft of a document that is not yet complete, but has a certain degree of content.

[0865] "Tone and style adjustment" refers to the process of changing the wording and writing style of a document based on the user's feelings.

[0866] "Grammar Check" refers to the functionality that detects grammatical errors in text and suggests corrections.

[0867] "Content check" refers to the function that checks the content of the text for consistency and accuracy and suggests corrections if necessary.

[0868] "Feedback" refers to evaluations and suggestions for improvement provided to users by the system.

[0869] "Chat interface" refers to an interface that allows users and systems to communicate text-based.

[0870] The present invention is a system that automatically generates new materials based on similar documents from the past and adjusts the tone and style according to the user's emotions. An embodiment of the system will be described in detail below.

[0871] System configuration

[0872] This system consists of three main components: a server, a terminal, and a user. The server is responsible for the collection, analysis, generation, and review of documents, while the terminal provides an interface for users to interact with the system. Users access the system through their terminal and perform the necessary operations.

[0873] Hardware and software used

[0874] 1. Server

[0875] Hardware: General server machine (CPU, memory, storage)

[0876] software:

[0877] Document processing: PyPDF2 (analyzing PDF files), python-docx (analyzing Word files)

[0878] Natural Language Processing: spaCy, NLTK

[0879] Emotion Recognition: A publicly available sentiment analysis API

[0880] Similarity calculation: Similarity calculation algorithm (e.g., cosine similarity)

[0881] Grammar Check: Grammarly, LanguageTool

[0882] 2. Terminal

[0883] Hardware: General purpose computers, tablets, smartphones

[0884] Software: Browser or dedicated chat interface application

[0885] Process Overview

[0886] 1. Document collection and analysis

[0887] The server accesses the folder path specified by the user and collects all files found there. It then opens the collected document files using a library corresponding to their format (e.g., PyPDF2 or python-docx) and extracts the text content. The extracted text is then analyzed using natural language processing techniques (spaCy or NLTK) to extract summaries and main topics. The results of these analyses are categorized based on document type and stored in a database.

[0888] 2. Automatic generation of new materials

[0889] Users input the type and theme of the document they wish to generate through a chat interface. The server uses an emotion engine to recognize the user's emotions from the input text data. It then searches the analyzed data stored in the database for documents most relevant to the input theme and performs a similarity calculation. Once the most relevant document is identified, a draft of the new document is automatically generated based on a template. The generated draft includes basic components such as an introduction, purpose, and outline of the plan, and the tone and style are adjusted based on the user's emotions.

[0890] 3. Automatic wall hitting and feedback

[0891] The user requests a review of the generated draft. The server then uses an automated wall-hitting function to analyze the text content of the draft using grammar checking software (Grammarly, LanguageTool) to detect grammatical errors, ambiguous expressions, and missing information. Based on the results, feedback is generated and provided to the user through a chat interface. The feedback includes specific error indications and suggestions for improvement.

[0892] Specific examples

[0893] For example, if a user types "I want to write a business proposal for new product X" into the chat interface, the server collects past business proposal documents from the specified folder and performs the necessary analysis. It then searches for documents related to the user's theme, "new product X," and automatically generates a draft of a new business proposal based on the most relevant documents after similarity calculations. This draft is adjusted to reflect the user's sentiment in tone and style, and based on the user's review request, it undergoes grammar and content checks and provides specific feedback.

[0894] Example prompt: "I would like to create a business proposal for new product X. Please refer to the previous project materials to generate a draft and review it."

[0895] This series of processes enables users to efficiently create high-quality documents and review them quickly.

[0896] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0897] Step 1:

[0898] Document file collection

[0899] The server accesses the folder path specified by the user (e.g. " / projects / past_docs / ") and lists all the files in the folder.

[0900] Input: Folder path

[0901] Specific behavior: Scans the file system and generates a list of files in the specified folder.

[0902] Output: File list

[0903] Step 2:

[0904] Document Content Extraction

[0905] The server filters document files (PDF, Word, Excel, PowerPoint, etc.) from the listed files.

[0906] Each document file is opened using a library corresponding to the format and the text content is extracted (e.g. PyPDF2 for PDF, python-docx for Word).

[0907] Input: File list

[0908] Specific operation: Opens a document file and extracts the text.

[0909] Output: Extracted text data

[0910] Step 3:

[0911] Text Analysis and Classification

[0912] The server analyzes the extracted text using natural language processing techniques (e.g., spaCy, NLTK) to extract summaries and main topics.

[0913] It automatically classifies documents into types (e.g., business proposals, project plans, etc.) based on their content.

[0914] The analysis results are stored in a database.

[0915] Input: Extracted text data

[0916] What it does: Summarizes text, extracts major topics, and classifies document types.

[0917] Output: Abstract, main topics, classification information

[0918] Step 4:

[0919] Acceptance of document generation requests

[0920] Users simply input the type of document they want to generate (e.g., business proposal) and topic (e.g., new product X) through a chat interface.

[0921] Input: Material type and subject

[0922] Specific actions: Enter a request for document generation into the chat interface.

[0923] Output: Document generation request

[0924] Step 5:

[0925] Emotion Recognition and Document Retrieval

[0926] The server uses an emotion engine to recognize the user's emotions from the input text data.

[0927] It searches a database of analyzed documents for the documents most relevant to the input topic.

[0928] Input: Document generation request, existing parsed document

[0929] Specific behavior: Recognizes emotions from input text and searches for related documents.

[0930] Output: Emotion recognition results, related documents

[0931] Step 6:

[0932] Automatic draft generation

[0933] The server performs a similarity calculation (e.g., cosine similarity) to identify the most relevant documents.

[0934] Based on the identified documents, it automatically generates drafts of new materials based on templates (e.g., introduction, objectives, project overview, etc.).

[0935] It adjusts the tone and style of the material based on the user's emotion recognition results.

[0936] Input: Emotion recognition results, related documents

[0937] What it does: Selects the most relevant documents based on similarity calculations, generates drafts, and adjusts tone and style.

[0938] Output: Draft of new material

[0939] Step 7:

[0940] Request a draft review

[0941] Users can request a review of the generated draft through a chat interface.

[0942] Input: Draft of new material

[0943] What it does: Sends a review request in the chat interface.

[0944] Output: Review request

[0945] Step 8:

[0946] Checking and parsing the draft

[0947] The server uses an automatic wall-hitting function and analyzes the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool).

[0948] It detects grammatical errors, ambiguous expressions, and missing information.

[0949] Input: Draft of new material, grammar checking software

[0950] Specific actions: Analyze the draft text with grammar checking software to detect errors and omissions.

[0951] Output: Syntax error detection results

[0952] Step 9:

[0953] Generating and Providing Feedback

[0954] The server generates feedback based on the detection results and adds specific improvement suggestions (e.g., "Item A lacks specific numbers," "There are three grammatical errors," "Please adjust your tone based on sentiment analysis").

[0955] Feedback is provided to users through a chat interface.

[0956] Input: Syntax error detection result

[0957] Specific behavior: Generate feedback from the detection results and provide it in a chat interface.

[0958] Output: Feedback

[0959] Through the above steps, the user can efficiently create new materials and quickly review them.

[0960] (Application example 2)

[0961] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0962] In logistics centers, creating new documents requires a lot of time and effort, and the process of extracting necessary information from past documents to create new plans and reports is particularly cumbersome. Furthermore, traditional methods tend to delay review and feedback on created documents, hindering rapid business processes. Furthermore, documents created without considering user sentiment may be difficult for users to understand and may be ineffective.

[0963] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0964] In this invention, the server includes means for collecting similar documents from the past and analyzing their contents using natural language processing technology, means for inputting the type of material to be generated and necessary information via a user interface, and means for calculating similarity and identifying the most relevant document. This makes it possible to efficiently automatically generate new materials based on past documents and provide real-time feedback in a document style that reflects the user's feelings.

[0965] "Past similar documents" are documents or plans created in the past, which have similar content and format to the newly generated document.

[0966] "Natural language processing technology" is a technology for analyzing human language and automatically understanding, processing, and generating it, and includes functions such as extracting, summarizing, and classifying text data.

[0967] A "user interface" is an interface that allows a system and a user to interact with each other, and provides operation means using voice input, text input, touch gestures, etc.

[0968] "Similarity calculation" is a method for calculating the similarity between multiple documents, and measures the similarity using a statistical method such as cosine similarity.

[0969] "Draft of new material" refers to an early stage material that is automatically generated based on the theme and format specified by the user.

[0970] "Generated Draft Content" refers to the text and information contained in the automatically generated early stage material.

[0971] "Grammar checking" is a process of checking whether the generated text content contains any grammatical errors, and is carried out using natural language processing technology.

[0972] "Content checking" is the process of verifying that the generated text content is appropriate and consistent and that no important information is missing.

[0973] An "emotion engine" is a system that analyzes emotions from user input and adjusts the style and tone of the generated document based on those emotions.

[0974] "Feedback" is information including review results and improvement suggestions for the generated draft, and is provided in response to a user's request.

[0975] "Providing feedback in real time" refers to a process in which the review results of the generated materials are immediately provided to the user, allowing the user to quickly make any necessary corrections or confirmations.

[0976] The present invention is a system that automatically generates new materials based on similar projects and business documents accumulated in the past, and combines them with an emotion engine to automatically bounce ideas off each other in a chat format. Specific embodiments of the present invention are described below.

[0977] System configuration

[0978] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents. The emotion engine also recognizes the user's emotions and reflects them in the document creation process.

[0979] Document collection and analysis

[0980] The server accesses the folder path specified by the user (e.g., " / projects / past_docs / ") and collects all files in the folder. It opens the collected document files (PDF, Word, Excel, PowerPoint, etc.) and extracts the content (text data). It uses natural language processing technology to analyze the extracted text content and extract summaries and main topics. It classifies the document type (business proposal, project plan, etc.) and stores it in a database.

[0981] Automatic generation of new materials

[0982] The user inputs the type of document they wish to generate (e.g., "business proposal") and theme (e.g., "new product X") through the user interface on their device. The server uses an emotion engine to recognize the user's emotions from the input text data. It then searches the analyzed document database for documents most relevant to the input theme. It performs similarity calculations (e.g., cosine similarity) to identify the most relevant documents. Based on the identified documents, it automatically generates a draft of the new document in accordance with a template. The generated draft includes basic components such as an introduction, purpose, and plan outline. The tone and style of the document are then adjusted based on the user's recognized emotions.

[0983] Automatic wall hitting and feedback

[0984] The user requests a review of the generated draft through the chat interface. The server uses an automated back-and-forth function to analyze the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool). Errors and ambiguous expressions are detected. Ambiguous expressions and missing information are also analyzed, and feedback is generated based on the detection results. The generated feedback is provided to the user through the chat interface. The feedback includes comments such as "Item A is missing a specific number," "There are three grammatical errors," and "Please adjust the tone based on sentiment analysis."

[0985] Specific examples

[0986] For example, if a user types "I want to write a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant documents. It then automatically generates a draft of the new business proposal based on the identified documents. The emotion engine recognizes the user's emotions based on the text entered by the user and adjusts the tone and style of the document accordingly. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[0987] An example of a prompt sentence to input to the generative AI model is as follows:

[0988] "I want to create a shift plan for next week."

[0989] "We want to create a plan for implementing a new warehouse management system."

[0990] This invention enables logistics center managers and staff to efficiently create new documents and quickly review them. It also enables document creation that takes user emotions into consideration, reducing the burden on workers.

[0991] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0992] Step 1:

[0993] The server accesses the folder path specified by the user (e.g., " / projects / past_docs / ") and collects all files in the folder. The collected files include different formats such as PDF, Word, Excel, and PowerPoint. These files are opened sequentially and text data is extracted. This is done using natural language processing technology to extract summaries and main topics of the text content. The extracted data is classified by document type and stored in a database.

[0994] Input: Folder path

[0995] Output: Extracted text data, classified document data

[0996] Step 2:

[0997] The user inputs the type of document (e.g., "business proposal") and theme (e.g., "new product X") they want to generate through the user interface of their terminal. This input information is sent to the server.

[0998] Input: Type of material, theme

[0999] Output: Sending input information to the server

[1000] Step 3:

[1001] The server uses an emotion engine to recognize emotions from the text data entered by the user, searches the analyzed document database for documents that are most relevant to the entered topic, and uses similarity calculations, such as cosine similarity, to identify the most relevant documents.

[1002] Input: User-entered text data

[1003] Output: Identifying relevant documents

[1004] Step 4:

[1005] The server automatically generates a draft of the new document based on the template, including basic elements such as an introduction, objectives, and a project outline, and adjusts the tone and style of the document based on the user's perceived emotions.

[1006] Input: Identified documents, sentiment data

[1007] Output: Draft of new material

[1008] Step 5:

[1009] Through the chat interface, the user requests a review of the generated draft, which is then sent to the server.

[1010] Input: Review request

[1011] Output: Request sent to server

[1012] Step 6:

[1013] The server uses an automated writing function to analyze the text content of the generated draft using grammar checking software, such as Grammarly or LanguageTool, to detect errors and ambiguous expressions, as well as to analyze missing information and generate feedback based on the results.

[1014] Input: Generated draft

[1015] Output: Grammar check results, feedback

[1016] Step 7:

[1017] The server provides the generated feedback to the user through a chat interface, including comments such as "Item A is missing a specific number," "There are three grammatical errors," and "Please adjust your tone based on sentiment analysis."

[1018] Input: Feedback

[1019] Output: Providing feedback to the user

[1020] This series of steps allows the distribution center manager and staff to efficiently create new documents and quickly review them. Specific examples of prompts are as follows:

[1021] "I want to create a shift plan for next week."

[1022] "We want to create a plan for implementing a new warehouse management system."

[1023] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1024] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1025] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1026] [Third embodiment]

[1027] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1028] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[1029] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1030] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[1031] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1032] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1033] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1034] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1035] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[1036] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1037] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1038] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. 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."

[1039] The present invention provides a system that automatically generates new materials based on similar projects and business documents accumulated in the past, and automatically provides a chat-style discussion before the approver reviews them. Specific embodiments of this system are described below.

[1040] System configuration

[1041] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents.

[1042] Document collection and analysis

[1043] server

[1044] Access the folder path specified by the user (e.g., " / projects / past_docs / ") and retrieve all files in the folder.

[1045] The collected document files (PDF, Word, Excel, PowerPoint, etc.) are opened one by one and the contents (text data) are extracted.

[1046] Using natural language processing techniques, the extracted text content is analyzed and summaries and main topics are extracted.

[1047] Classify document types (business plans, project plans, etc.) and store them in a database.

[1048] Automatic generation of new materials

[1049] User

[1050] Through the chat interface, you input the type of document you want to generate (e.g., "Business Proposal") and the topic (e.g., "New Product X").

[1051] server

[1052] Search a parsed document database for documents most relevant to the input topic.

[1053] Perform a similarity calculation (e.g., cosine similarity) to identify the most relevant documents.

[1054] Based on the identified documents, the system automatically generates a draft of the new material according to a template, including basic elements such as an introduction, objectives, and a project overview.

[1055] Automatic wall hitting and feedback

[1056] User

[1057] Request a review of the generated draft through the chat interface.

[1058] server

[1059] Analyze the text content of the generated draft using the automatic wall-writing function. Use grammar checking software (e.g., Grammarly, LanguageTool, etc.) to detect grammatical errors.

[1060] It also analyzes ambiguous expressions and missing information, and generates feedback based on the detection results.

[1061] The generated feedback is provided to the user through a chat interface, including comments such as "Item A is missing a specific number" or "There are three grammatical errors."

[1062] Specific examples

[1063] For example, if a user types "I want to create a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant document. It then automatically generates a draft of a new business proposal based on the identified documents and presents it to the user. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[1064] In this way, the present invention provides a system that allows users to efficiently create new materials and quickly review them.

[1065] The processing flow will be explained below.

[1066] Step 1:

[1067] User

[1068] A user accesses the system via a chat interface and inputs the path of the folder to be analyzed.

[1069] User: Specify the " / projects / past_docs / " folder.

[1070] Step 2:

[1071] server

[1072] The server collects all document files (PDF, Word, Excel, PowerPoint) from the specified folder path and keeps a list of file names and paths.

[1073] The server retrieves all files in the " / projects / past_docs / " folder and lists their paths.

[1074] Step 3:

[1075] server

[1076] The server opens each collected file and extracts the text content using the appropriate library (e.g. PyPDF2, python-docx) depending on the document format (PDF, Word, etc.).

[1077] The server opens PDF files with PyPDF2 and Word files with python-docx and extracts the text.

[1078] Step 4:

[1079] server

[1080] The server analyzes the extracted text using natural language processing (NLP) techniques (e.g., spaCy, NLTK) to summarize the content and extract key topics and keywords.

[1081] The server uses the spaCy library to parse the text and extract summaries and keywords.

[1082] Step 5:

[1083] server

[1084] Based on the analysis results, the server classifies the documents into categories (e.g., business proposals, project plans) and stores them in a database.

[1085] The server stores the parsed text along with the classification information in a database.

[1086] Step 6:

[1087] User

[1088] The user inputs the type and theme of the material they wish to generate through the chat interface.

[1089] User: Enter "Create a business proposal on the theme of new product X."

[1090] Step 7:

[1091] server

[1092] The server searches the database for relevant parsed documents based on the themes entered by the user, and performs similarity calculations (e.g., cosine similarity) to identify the most relevant documents.

[1093] The server searches the database for keywords related to "business proposal" and "new product X" and calculates cosine similarity to identify highly relevant documents.

[1094] Step 8:

[1095] server

[1096] The server automatically generates a draft of the new material based on the template, including basic elements such as an introduction, objectives, and a project outline, based on the identified related documents.

[1097] The server generates a new draft based on a business proposal template.

[1098] Step 9:

[1099] User

[1100] The user requests a review of the generated draft through a chat interface.

[1101] User: Enter "Please review the business proposal I created."

[1102] Step 10:

[1103] server

[1104] The server uses an automatic writing-on-the-wall function to analyze the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool) to detect errors and ambiguous expressions.

[1105] The server uses Grammarly to detect grammatical errors and ambiguous expressions in the draft.

[1106] Step 11:

[1107] server

[1108] The server generates a summary of the detected problems and suggestions for improvement, and provides feedback to the user through a chat interface.

[1109] The server generates error information and improvement suggestions as feedback and provides them to the user.

[1110] In this way, by performing specific processing at each step, the user can efficiently create new materials and have them reviewed quickly.

[1111] Example 1

[1112] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1113] Conventional document creation systems have the problem of requiring a great deal of time and effort to create and review new documents. Furthermore, there is a lack of effective means for utilizing similar documents from the past, making efficient document creation difficult. Furthermore, document content and grammar checks are not fully automated, placing a burden on users.

[1114] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1115] In this invention, the server includes means for collecting similar documents from the past and analyzing their contents using natural language processing technology, means for inputting the type and theme of the document to be created via a user interface, means for calculating similarity and identifying the most relevant document, means for automatically generating a draft of a new document based on the identified document, means for performing grammar and content checks on the content of the generated draft, and means for generating feedback based on the check results and providing it to the user, thereby enabling the user to quickly and accurately create new documents and review them effectively.

[1116] "Past similar documents" refers to documents that were previously created and are the subject of collection and analysis by the system.

[1117] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language.

[1118] "User interface" refers to an interface through which a user inputs information into a system and performs operations.

[1119] "Similarity calculation" refers to a calculation method that evaluates the relevance between text data and expresses it numerically.

[1120] "Cosine similarity" refers to a method of measuring the similarity between two pieces of text data using angles in a vector space.

[1121] "Feedback" refers to information that indicates reviews and improvements to the generated document.

[1122] "Grammar checking" refers to the process of detecting and correcting grammatical errors in a piece of writing.

[1123] "Content check" refers to the process of checking whether the content of a document is appropriate and whether it contains the necessary information.

[1124] The present invention provides a system that automatically generates new materials based on similar projects and business documents accumulated in the past, and automatically provides a chat-style discussion before the approver reviews them. Specific embodiments of this system are described below.

[1125] System configuration

[1126] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents.

[1127] Document collection and analysis

[1128] server

[1129] The server accesses the folder path specified by the user (e.g., " / projects / past_docs / ") and retrieves all files in the folder.

[1130] The server sequentially opens the collected document files (PDF, Word, Excel, PowerPoint, etc.) and extracts the contents (text data). To do this, it uses libraries such as Apache Tika.

[1131] The server uses natural language processing techniques to analyze the extracted text content and extract summaries and main topics, specifically using libraries such as NLTK and SpaCy.

[1132] The server classifies the documents by type (business proposal, project plan, etc.) and stores them in a database. The database used here is a general database management system (e.g., MySQL, PostgreSQL).

[1133] Automatic generation of new materials

[1134] User

[1135] Through the chat interface, users input the type of document they want to generate (e.g., "business proposal") and theme (e.g., "new product X") as keywords.

[1136] server

[1137] The server searches the parsed document database for documents that are most relevant to the input topic, using a search engine such as ElasticSearch.

[1138] The server performs a similarity calculation (e.g., cosine similarity) to identify the most relevant documents, using a machine learning library such as scikit-learn.

[1139] The server automatically generates a draft of the new material based on the identified document, following a template, using OpenAI's generative AI model (e.g., GPT-3). The generated draft includes basic components such as an introduction, purpose, and outline of the plan.

[1140] Automatic wall hitting and feedback

[1141] User

[1142] The user requests a review of the generated draft through a chat interface.

[1143] server

[1144] The server analyzes the text content of the generated draft using an automated writing-on-the-wall function and detects grammar errors using grammar checking software (e.g., Grammarly or LanguageTool).

[1145] The server also analyzes ambiguous expressions and missing information and generates feedback based on the results of its detection.

[1146] The server provides the generated feedback to the user through a chat interface, including comments such as "Item A is missing a specific number" or "There are three grammatical errors."

[1147] Examples and prompts

[1148] For example, if a user types "I want to create a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant document. It then automatically generates a draft of a new business proposal based on the identified documents and presents it to the user. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[1149] An example prompt is, "Write a business proposal for new product X. Include the use, challenges, and solutions."

[1150] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1151] Step 1: Gather documentation

[1152] server

[1153] Input: A user-specified folder path (e.g. " / projects / past_docs / ").

[1154] Specific operation: The server accesses the specified folder and lists all files in the folder using the os.listdir() function.

[1155] Data processing: Open each file one by one and extract the file contents using Apache Tika, for example, extract text data from PDF, Word, Excel, and PowerPoint files.

[1156] Output: A list of extracted text data.

[1157] Step 2: Parsing the document

[1158] server

[1159] Input: A list of extracted text data.

[1160] Specific operation: The server performs natural language processing on each text data, specifically tokenization, POS tagging, summarization, and major topic extraction using SpaCy.

[1161] Data Calculation: Calculate TF-IDF scores to identify major topics and keywords in the text.

[1162] Output: A list of analysis results, including summary, major topics, and classification information.

[1163] Step 3: Saving to the database

[1164] server

[1165] Input: A list of analysis results.

[1166] Specific operation: The server stores the analysis results in a database such as MySQL or PostgreSQL.

[1167] Data processing: Classify the data based on the type of document (e.g., business plan, project plan), and store the relevant information for each document in a database.

[1168] Output: Parsed document data stored in a database.

[1169] Step 4: Generate request from user

[1170] User

[1171] Input: Enter the type and topic of the document you want to generate (e.g., "Business Proposal" or "New Product X") via the chat interface.

[1172] Specific operation: The user accesses the chat interface and inputs the necessary information, which causes the user's request to be sent to the server.

[1173] Output: Generated request information.

[1174] Step 5: Finding Similar Documents

[1175] server

[1176] Input: Generation request information (document type and subject).

[1177] What it does: The server uses ElasticSearch to search the database for relevant documents.

[1178] Data calculation: Calculate cosine similarity and identify the documents most relevant to the request information. Similarity calculation is performed using scikit-learn.

[1179] Output: A list of the most relevant documents.

[1180] Step 6: Draft a new document

[1181] server

[1182] Input: A list of the most relevant documents.

[1183] Specific operation: The server uses OpenAI's generative AI model (e.g., GPT-3) to automatically generate a draft of a new document based on a template.

[1184] Data processing: Using a generative AI model, a draft is created that includes the document's basic structure, such as its introduction, purpose, and outline.

[1185] Output: A draft of the new document.

[1186] Step 7: Provide the draft to users

[1187] server

[1188] Input: New document draft.

[1189] Specific operation: The server provides the generated draft to the user through a chat interface.

[1190] Output: A draft of the new document sent to the user.

[1191] Step 8: Automatic wall-hitting and feedback generation

[1192] User

[1193] Input: Review request (via chat interface).

[1194] What happens: A user requests a draft for review through the chat interface. This information is sent to the server.

[1195] Output: Review request information.

[1196] server

[1197] Input: Review request information and new document draft.

[1198] What it does: The server uses LanguageTool and Grammarly to detect grammatical errors in the draft, as well as analyzes ambiguous expressions and missing information.

[1199] Data operations: Detecting grammatical errors, checking content, and generating feedback.

[1200] Output: Feedback information.

[1201] Step 9: Provide feedback

[1202] server

[1203] Input: Feedback information.

[1204] Specific Operation: The server provides the generated feedback to the user through a chat interface.

[1205] Output: Feedback sent to the user.

[1206] (Application example 1)

[1207] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1208] In production management and quality control within factories, document creation and review work is a time-consuming and labor-intensive issue. In particular, when launching a new product, it is necessary to extract the necessary information from a huge amount of related past documents and quickly create and review new documents. However, the current manual process is inefficient and prone to human error. There is a need to solve these problems and streamline and automate document creation and review.

[1209] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1210] In this invention, the server includes a means for collecting similar documents from the past and analyzing their contents using natural language processing technology, a means for inputting the type of document to be generated and necessary information via a user interface installed in the factory robot, and a means for calculating similarity and identifying the most relevant document. This makes it possible to efficiently create new documents based on past documents and to generate high-quality documents and quickly review them by using a generative AI model.

[1211] "Past similar documents" are documents that were created previously and have content and formats related to the current project or work.

[1212] "Natural language processing technology" is a technology that allows computers to understand text data written in human language and perform operations such as analysis, generation, and translation.

[1213] A "user interface mounted on a factory robot" is an interface that is installed on a factory robot and allows the user to input data and perform operations.

[1214] "Similarity calculation" is a calculation method for numerically evaluating the degree of similarity between different documents or data sets.

[1215] "Cosine similarity" is a method for assessing the similarity of documents or datasets by calculating the angle between two vectors.

[1216] A "generative AI model" is a learning model that uses artificial intelligence to generate new text or information based on specific input prompts.

[1217] "Grammar checking" is the process of detecting grammatical errors in a document and suggesting corrections.

[1218] "Content check" is a process for verifying whether the contents of a document are logically correct, consistent, and whether there is any missing information.

[1219] "Feedback" means suggestions for correction or improvement provided as a result of analysis or review.

[1220] This invention is a system for automatically generating new documents based on similar past documents and then reviewing their contents. The purpose of this system is to improve the efficiency of creating and reviewing production management and quality control documents within a factory.

[1221] System configuration

[1222] The system consists of three main components: a server, a terminal mounted on the factory robot, and a user. The user interacts with the system through a user interface mounted on the factory robot, while the server consistently collects, analyzes, generates, and reviews documents.

[1223] Document collection and analysis

[1224] server

[1225] Accesses a folder path specified by the user (e.g. " / projects / past_docs / ") and retrieves all files in the folder. Formats include PDF, Word, Excel, PowerPoint, etc.

[1226] The text content of the collected document files is extracted and analyzed using natural language processing techniques, which makes it possible to extract summaries and main topics.

[1227] Classify the document type (e.g. production control report, quality control report, etc.) and store it in a database.

[1228] Automatic generation of new materials

[1229] User

[1230] Through the user interface installed on the factory robot, the type of document to be generated (e.g., "Quality Control Report") and topic (e.g., "New Product X") are entered.

[1231] server

[1232] Search a parsed document database for documents most relevant to the input topic.

[1233] Based on similarity calculations (e.g., cosine similarity) and the most relevant documents, a generative AI model is used to automatically generate a draft of new material, including quality check items, inspection results, and improvement suggestions.

[1234] Automatic wall hitting and feedback

[1235] User

[1236] A user interface on the factory robot allows a review of the generated draft to be requested.

[1237] server

[1238] Analyze the text content of the generated draft using the automatic wall-writing function. Use grammar checking software (e.g., Grammarly, LanguageTool, etc.) to detect grammatical errors.

[1239] It also analyzes ambiguous expressions and missing information and generates feedback, such as "Item A is missing a specific number" or "There are three grammatical errors."

[1240] The generated feedback is provided to the user through a user interface of the factory robot.

[1241] Specific examples

[1242] For example, if a user inputs "I want to create a quality control report" into a factory robot, the system will search the database for past documents related to "quality control report" and "new product X" to identify the most relevant documents. It then uses a generative AI model to automatically generate a draft of a new quality control report based on the identified documents. The generated draft is then checked for grammar and content, and feedback is provided to the user.

[1243] Prompt Sentence Examples

[1244] I want to create a quality control report

[1245] Please prepare a quality control report for new product X.

[1246] In this manner, the present invention provides a system that allows users to efficiently create new materials and quickly review them.

[1247] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1248] Step 1:

[1249] The user inputs the type of new material (e.g., "Quality Control Report") and topic (e.g., "New Product X") via a user interface installed on the factory robot. The input information is sent to the server.

[1250] Input: Type of material and subject

[1251] Output: Request data sent to the server

[1252] Step 2:

[1253] The server accesses the specified folder path (e.g., " / projects / past_docs / "), collects all document files (PDF, Word, Excel, PowerPoint, etc.) in the folder, and opens them to extract text data.

[1254] Input: the required folder path

[1255] Output: Text data of collected documents

[1256] Step 3:

[1257] The server analyzes the collected text data using natural language processing technology, extracts summaries and main topics from the analysis results, classifies them by document type, and stores them in a database.

[1258] Input: Text data of collected documents

[1259] Output: Summary of the analyzed document, main topics, and classification information

[1260] Step 4:

[1261] The server searches through the documents stored in the database to find the documents that are most relevant to the topic entered by the user. It identifies the relevant documents using similarity calculations (e.g., cosine similarity).

[1262] Input: User-entered themes, documents from the database

[1263] Output: Most relevant documents

[1264] Step 5:

[1265] Based on the identified documents, the server uses a generative AI model to automatically generate a draft of the new material, including basic components such as an introduction, purpose, and outline of the project.

[1266] Input: Most relevant documents

[1267] Output: Automatically generated draft of new material

[1268] Step 6:

[1269] When a user requests a review of a generated draft, the server uses an automated text-based grammar and content checker to detect errors.

[1270] Input: Draft of new material, user review request

[1271] Output: Grammar and content error detection results

[1272] Step 7:

[1273] The server generates feedback based on the detected errors and provides it to the user with specific suggestions for correction, such as "Item A is missing a specific number" or "There are three grammatical errors."

[1274] Input: Error detection result

[1275] Output: Generated feedback, suggested corrections

[1276] By the above steps, the present invention makes it possible to efficiently and automatically create and review new materials in a factory.

[1277] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1278] The present invention combines an emotion engine with a system that automatically generates new materials based on similar projects and business documents accumulated in the past and provides automatic chat-style discussion before the approver reviews them. Specific embodiments of this system are described below.

[1279] System configuration

[1280] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents. The emotion engine also recognizes the user's emotions and reflects them in the document creation process.

[1281] Document collection and analysis

[1282] server

[1283] Access the folder path specified by the user (e.g., " / projects / past_docs / ") and collect all files in the folder.

[1284] The collected document files (PDF, Word, Excel, PowerPoint, etc.) are opened one by one and the contents (text data) are extracted.

[1285] Using natural language processing techniques, the extracted text content is analyzed and summaries and main topics are extracted.

[1286] Classify document types (business plans, project plans, etc.) and store them in a database.

[1287] Automatic generation of new materials

[1288] User

[1289] Through the chat interface, you input the type of document you want to generate (e.g., "Business Proposal") and the topic (e.g., "New Product X").

[1290] server

[1291] An emotion engine is used to recognize the user's emotions from the input text data.

[1292] Search a parsed document database for documents most relevant to the input topic.

[1293] Perform a similarity calculation (e.g., cosine similarity) to identify the most relevant documents.

[1294] Based on the identified documents, the system automatically generates a template-based draft of the new material, including basic elements such as an introduction, objectives, and project outline, while adjusting the tone and style of the material based on the user's perceived emotions.

[1295] Automatic wall hitting and feedback

[1296] User

[1297] Request a review of the generated draft through the chat interface.

[1298] server

[1299] The text content of the generated draft is analyzed using grammar checking software (e.g., Grammarly, LanguageTool) to detect errors and ambiguous expressions.

[1300] It also analyzes ambiguous expressions and missing information, and generates feedback based on the detection results.

[1301] The generated feedback is provided to the user through a chat interface, including comments such as "Item A is missing specific numbers," "There are three grammatical errors," and "Please adjust your tone based on sentiment analysis."

[1302] Specific examples

[1303] For example, if a user types "I want to write a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant documents. It then automatically generates a draft of the new business proposal based on the identified documents. The emotion engine recognizes the user's emotions based on the text entered by the user and adjusts the tone and style of the document accordingly. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[1304] In this way, the present invention not only enables users to efficiently create new materials and quickly review them, but also provides a system that enables document creation that takes the user's emotions into consideration.

[1305] The processing flow will be explained below.

[1306] Step 1:

[1307] User

[1308] A user accesses the system via a chat interface and inputs the path of the folder to be analyzed.

[1309] User: Specify the " / projects / past_docs / " folder.

[1310] Step 2:

[1311] server

[1312] The server collects all document files (PDF, Word, Excel, PowerPoint) from the specified folder path and keeps a list of file names and paths.

[1313] The server retrieves all files in the " / projects / past_docs / " folder and lists their paths.

[1314] Step 3:

[1315] server

[1316] The server opens each collected file and extracts the text content using the appropriate library (e.g. PyPDF2, python-docx) depending on the document format (PDF, Word, etc.).

[1317] The server opens PDF files with PyPDF2 and Word files with python-docx and extracts the text.

[1318] Step 4:

[1319] server

[1320] The server analyzes the extracted text using natural language processing (NLP) techniques (e.g., spaCy, NLTK) to summarize the content and extract key topics and keywords.

[1321] The server uses the spaCy library to parse the text and extract summaries and keywords.

[1322] Step 5:

[1323] server

[1324] Based on the analysis results, the server classifies the documents into categories (e.g., business proposals, project plans) and stores them in a database.

[1325] The server stores the parsed text along with the classification information in a database.

[1326] Step 6:

[1327] User

[1328] The user inputs the type and theme of the material they wish to generate through the chat interface.

[1329] User: Enter "Create a business proposal on the theme of new product X."

[1330] Step 7:

[1331] server

[1332] The server analyzes input data from the user interface using an emotion engine based on the theme entered by the user to recognize the emotion of the user.

[1333] The server uses an emotion engine to analyze the user's input text and recognize emotions such as "positive" and "negative."

[1334] Step 8:

[1335] server

[1336] The server searches the analyzed document database for documents that are most relevant to the input topic, taking into account the emotion recognition results. It performs similarity calculations (e.g., cosine similarity) to identify the most relevant documents.

[1337] The server searches the database for keywords related to "business proposal" and "new product X" and calculates cosine similarity to identify highly relevant documents.

[1338] Step 9:

[1339] server

[1340] The server automatically generates a template-based draft of the new document based on the identified related documents. The draft includes basic elements such as an introduction, objectives, and a project outline. The tone and style of the document are adjusted based on the user's perceived emotions.

[1341] The server generates a new draft based on a business proposal template, adapting the tone and style to match the user's emotional needs.

[1342] Step 10:

[1343] User

[1344] The user requests a review of the generated draft through a chat interface.

[1345] User: Enter "Please review the business proposal I created."

[1346] Step 11:

[1347] server

[1348] The server uses an automatic writing-on-the-wall function to analyze the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool) to detect errors and ambiguous expressions.

[1349] The server uses Grammarly to detect grammatical errors and ambiguous expressions in the draft.

[1350] Step 12:

[1351] server

[1352] The server generates a summary of the detected issues and suggestions for improving tone and style based on sentiment analysis, and provides feedback to the user via a chat interface.

[1353] The server generates feedback including error information, suggestions for improvement, and comments based on sentiment analysis, and provides it to the user.

[1354] In this way, by performing specific processing at each step, the user can efficiently create new materials, have them reviewed quickly, and create documents that take the user's feelings into consideration.

[1355] Example 2

[1356] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1357] In conventional document generation systems, the manual search for similar documents, analysis of their contents, and creation of new materials required time and effort, and it was difficult to create documents that took the user's feelings into consideration. Furthermore, the grammar and content of the generated drafts had to be checked manually, preventing automation. This made it difficult for users to create documents efficiently and with high quality.

[1358] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes: means for collecting past similar documents and analyzing their contents using natural language processing technology; means for inputting the type of material to be generated and necessary information via a user interface; means for recognizing the user's sentiment from the input text data; means for calculating similarity and identifying the most relevant document; means for automatically generating a draft of a new material based on the identified document; means for adjusting the tone and style of the generated draft based on the user's sentiment; means for checking the grammar and content of the generated draft; and means for generating feedback based on the check results and providing it to the user. This enables the automation of document generation work, allowing users to efficiently create high-quality documents and quickly review them.

[1359] "Previous similar documents" refers to previously created, highly relevant materials and records.

[1360] "Natural language processing technology" refers to technology that enables computers to understand, interpret, and manipulate human language.

[1361] "User interface" refers to the components that provide the means for a user to interact with a system.

[1362] "Means of recognizing emotions" refers to technologies and algorithms that analyze emotions from user input and understand those emotions.

[1363] "Similarity calculation" refers to a calculation method for measuring the similarity between multiple documents and evaluating their relevance.

[1364] "Document" means any electronic or physical document containing certain information.

[1365] A "draft" is a rough draft of a document that is not yet complete, but has a certain degree of content.

[1366] "Tone and style adjustment" refers to the process of changing the wording and writing style of a document based on the user's feelings.

[1367] "Grammar Check" refers to the functionality that detects grammatical errors in text and suggests corrections.

[1368] "Content check" refers to the function that checks the content of the text for consistency and accuracy and suggests corrections if necessary.

[1369] "Feedback" refers to evaluations and suggestions for improvement provided to users by the system.

[1370] "Chat interface" refers to an interface that allows users and systems to communicate text-based.

[1371] The present invention is a system that automatically generates new materials based on similar documents from the past and adjusts the tone and style according to the user's emotions. An embodiment of the system will be described in detail below.

[1372] System configuration

[1373] This system consists of three main components: a server, a terminal, and a user. The server is responsible for the collection, analysis, generation, and review of documents, while the terminal provides an interface for users to interact with the system. Users access the system through their terminal and perform the necessary operations.

[1374] Hardware and software used

[1375] 1. Server

[1376] Hardware: General server machine (CPU, memory, storage)

[1377] software:

[1378] Document processing: PyPDF2 (analyzing PDF files), python-docx (analyzing Word files)

[1379] Natural Language Processing: spaCy, NLTK

[1380] Emotion Recognition: A publicly available sentiment analysis API

[1381] Similarity calculation: Similarity calculation algorithm (e.g., cosine similarity)

[1382] Grammar Check: Grammarly, LanguageTool

[1383] 2. Terminal

[1384] Hardware: General purpose computers, tablets, smartphones

[1385] Software: Browser or dedicated chat interface application

[1386] Process Overview

[1387] 1. Document collection and analysis

[1388] The server accesses the folder path specified by the user and collects all files found there. It then opens the collected document files using a library corresponding to their format (e.g., PyPDF2 or python-docx) and extracts the text content. The extracted text is then analyzed using natural language processing techniques (spaCy or NLTK) to extract summaries and main topics. The results of these analyses are categorized based on document type and stored in a database.

[1389] 2. Automatic generation of new materials

[1390] Users input the type and theme of the document they wish to generate through a chat interface. The server uses an emotion engine to recognize the user's emotions from the input text data. It then searches the analyzed data stored in the database for documents most relevant to the input theme and performs a similarity calculation. Once the most relevant document is identified, a draft of the new document is automatically generated based on a template. The generated draft includes basic components such as an introduction, purpose, and outline of the plan, and the tone and style are adjusted based on the user's emotions.

[1391] 3. Automatic wall hitting and feedback

[1392] The user requests a review of the generated draft. The server then uses an automated wall-hitting function to analyze the text content of the draft using grammar checking software (Grammarly, LanguageTool) to detect grammatical errors, ambiguous expressions, and missing information. Based on the results, feedback is generated and provided to the user through a chat interface. The feedback includes specific error indications and suggestions for improvement.

[1393] Specific examples

[1394] For example, if a user types "I want to write a business proposal for new product X" into the chat interface, the server collects past business proposal documents from the specified folder and performs the necessary analysis. It then searches for documents related to the user's theme, "new product X," and automatically generates a draft of a new business proposal based on the most relevant documents after similarity calculations. This draft is adjusted to reflect the user's sentiment in tone and style, and based on the user's review request, it undergoes grammar and content checks and provides specific feedback.

[1395] Example prompt: "I would like to create a business proposal for new product X. Please refer to the previous project materials to generate a draft and review it."

[1396] This series of processes enables users to efficiently create high-quality documents and review them quickly.

[1397] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1398] Step 1:

[1399] Document file collection

[1400] The server accesses the folder path specified by the user (e.g. " / projects / past_docs / ") and lists all the files in the folder.

[1401] Input: Folder path

[1402] Specific behavior: Scans the file system and generates a list of files in the specified folder.

[1403] Output: File list

[1404] Step 2:

[1405] Document Content Extraction

[1406] The server filters document files (PDF, Word, Excel, PowerPoint, etc.) from the listed files.

[1407] Each document file is opened using a library corresponding to the format and the text content is extracted (e.g. PyPDF2 for PDF, python-docx for Word).

[1408] Input: File list

[1409] Specific operation: Opens a document file and extracts the text.

[1410] Output: Extracted text data

[1411] Step 3:

[1412] Text Analysis and Classification

[1413] The server analyzes the extracted text using natural language processing techniques (e.g., spaCy, NLTK) to extract summaries and main topics.

[1414] It automatically classifies documents into types (e.g., business proposals, project plans, etc.) based on their content.

[1415] The analysis results are stored in a database.

[1416] Input: Extracted text data

[1417] What it does: Summarizes text, extracts major topics, and classifies document types.

[1418] Output: Abstract, main topics, classification information

[1419] Step 4:

[1420] Acceptance of document generation requests

[1421] Users simply input the type of document they want to generate (e.g., business proposal) and topic (e.g., new product X) through a chat interface.

[1422] Input: Material type and subject

[1423] Specific actions: Enter a request for document generation into the chat interface.

[1424] Output: Document generation request

[1425] Step 5:

[1426] Emotion Recognition and Document Retrieval

[1427] The server uses an emotion engine to recognize the user's emotions from the input text data.

[1428] It searches a database of analyzed documents for the documents most relevant to the input topic.

[1429] Input: Document generation request, existing parsed document

[1430] Specific behavior: Recognizes emotions from input text and searches for related documents.

[1431] Output: Emotion recognition results, related documents

[1432] Step 6:

[1433] Automatic draft generation

[1434] The server performs a similarity calculation (e.g., cosine similarity) to identify the most relevant documents.

[1435] Based on the identified documents, it automatically generates drafts of new materials based on templates (e.g., introduction, objectives, project overview, etc.).

[1436] It adjusts the tone and style of the material based on the user's emotion recognition results.

[1437] Input: Emotion recognition results, related documents

[1438] What it does: Selects the most relevant documents based on similarity calculations, generates drafts, and adjusts tone and style.

[1439] Output: Draft of new material

[1440] Step 7:

[1441] Request a draft review

[1442] Users can request a review of the generated draft through a chat interface.

[1443] Input: Draft of new material

[1444] What it does: Sends a review request in the chat interface.

[1445] Output: Review request

[1446] Step 8:

[1447] Checking and parsing the draft

[1448] The server uses an automatic wall-hitting function and analyzes the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool).

[1449] It detects grammatical errors, ambiguous expressions, and missing information.

[1450] Input: Draft of new material, grammar checking software

[1451] Specific actions: Analyze the draft text with grammar checking software to detect errors and omissions.

[1452] Output: Syntax error detection results

[1453] Step 9:

[1454] Generating and Providing Feedback

[1455] The server generates feedback based on the detection results and adds specific improvement suggestions (e.g., "Item A lacks specific numbers," "There are three grammatical errors," "Please adjust your tone based on sentiment analysis").

[1456] Feedback is provided to users through a chat interface.

[1457] Input: Syntax error detection result

[1458] Specific behavior: Generate feedback from the detection results and provide it in a chat interface.

[1459] Output: Feedback

[1460] Through the above steps, the user can efficiently create new materials and quickly review them.

[1461] (Application example 2)

[1462] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1463] In logistics centers, creating new documents requires a lot of time and effort, and the process of extracting necessary information from past documents to create new plans and reports is particularly cumbersome. Furthermore, traditional methods tend to delay review and feedback on created documents, hindering rapid business processes. Furthermore, documents created without considering user sentiment may be difficult for users to understand and may be ineffective.

[1464] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1465] In this invention, the server includes means for collecting similar documents from the past and analyzing their contents using natural language processing technology, means for inputting the type of material to be generated and necessary information via a user interface, and means for calculating similarity and identifying the most relevant document. This makes it possible to efficiently automatically generate new materials based on past documents and provide real-time feedback in a document style that reflects the user's feelings.

[1466] "Past similar documents" are documents or plans created in the past, which have similar content and format to the newly generated document.

[1467] "Natural language processing technology" is a technology for analyzing human language and automatically understanding, processing, and generating it, and includes functions such as extracting, summarizing, and classifying text data.

[1468] A "user interface" is an interface that allows a system and a user to interact with each other, and provides operation means using voice input, text input, touch gestures, etc.

[1469] "Similarity calculation" is a method for calculating the similarity between multiple documents, and measures the similarity using a statistical method such as cosine similarity.

[1470] "Draft of new material" refers to an early stage material that is automatically generated based on the theme and format specified by the user.

[1471] "Generated Draft Content" refers to the text and information contained in the automatically generated early stage material.

[1472] "Grammar checking" is a process of checking whether the generated text content contains any grammatical errors, and is carried out using natural language processing technology.

[1473] "Content checking" is the process of verifying that the generated text content is appropriate and consistent and that no important information is missing.

[1474] An "emotion engine" is a system that analyzes emotions from user input and adjusts the style and tone of the generated document based on those emotions.

[1475] "Feedback" is information including review results and improvement suggestions for the generated draft, and is provided in response to a user's request.

[1476] "Providing feedback in real time" refers to a process in which the review results of the generated materials are immediately provided to the user, allowing the user to quickly make any necessary corrections or confirmations.

[1477] The present invention is a system that automatically generates new materials based on similar projects and business documents accumulated in the past, and combines them with an emotion engine to automatically bounce ideas off each other in a chat format. Specific embodiments of the present invention are described below.

[1478] System configuration

[1479] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents. The emotion engine also recognizes the user's emotions and reflects them in the document creation process.

[1480] Document collection and analysis

[1481] The server accesses the folder path specified by the user (e.g., " / projects / past_docs / ") and collects all files in the folder. It opens the collected document files (PDF, Word, Excel, PowerPoint, etc.) and extracts the content (text data). It uses natural language processing technology to analyze the extracted text content and extract summaries and main topics. It classifies the document type (business proposal, project plan, etc.) and stores it in a database.

[1482] Automatic generation of new materials

[1483] The user inputs the type of document they wish to generate (e.g., "business proposal") and theme (e.g., "new product X") through the user interface on their device. The server uses an emotion engine to recognize the user's emotions from the input text data. It then searches the analyzed document database for documents most relevant to the input theme. It performs similarity calculations (e.g., cosine similarity) to identify the most relevant documents. Based on the identified documents, it automatically generates a draft of the new document in accordance with a template. The generated draft includes basic components such as an introduction, purpose, and plan outline. The tone and style of the document are then adjusted based on the user's recognized emotions.

[1484] Automatic wall hitting and feedback

[1485] The user requests a review of the generated draft through the chat interface. The server uses an automated back-and-forth function to analyze the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool). Errors and ambiguous expressions are detected. Ambiguous expressions and missing information are also analyzed, and feedback is generated based on the detection results. The generated feedback is provided to the user through the chat interface. The feedback includes comments such as "Item A is missing a specific number," "There are three grammatical errors," and "Please adjust the tone based on sentiment analysis."

[1486] Specific examples

[1487] For example, if a user types "I want to write a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant documents. It then automatically generates a draft of the new business proposal based on the identified documents. The emotion engine recognizes the user's emotions based on the text entered by the user and adjusts the tone and style of the document accordingly. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[1488] An example of a prompt sentence to input to the generative AI model is as follows:

[1489] "I want to create a shift plan for next week."

[1490] "We want to create a plan for implementing a new warehouse management system."

[1491] This invention enables logistics center managers and staff to efficiently create new documents and quickly review them. It also enables document creation that takes user emotions into consideration, reducing the burden on workers.

[1492] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1493] Step 1:

[1494] The server accesses the folder path specified by the user (e.g., " / projects / past_docs / ") and collects all files in the folder. The collected files include different formats such as PDF, Word, Excel, and PowerPoint. These files are opened sequentially and text data is extracted. This is done using natural language processing technology to extract summaries and main topics of the text content. The extracted data is classified by document type and stored in a database.

[1495] Input: Folder path

[1496] Output: Extracted text data, classified document data

[1497] Step 2:

[1498] The user inputs the type of document (e.g., "business proposal") and theme (e.g., "new product X") they want to generate through the user interface of their terminal. This input information is sent to the server.

[1499] Input: Type of material, theme

[1500] Output: Sending input information to the server

[1501] Step 3:

[1502] The server uses an emotion engine to recognize emotions from the text data entered by the user, searches the analyzed document database for documents that are most relevant to the entered topic, and uses similarity calculations, such as cosine similarity, to identify the most relevant documents.

[1503] Input: User-entered text data

[1504] Output: Identifying relevant documents

[1505] Step 4:

[1506] The server automatically generates a draft of the new document based on the template, including basic elements such as an introduction, objectives, and a project outline, and adjusts the tone and style of the document based on the user's perceived emotions.

[1507] Input: Identified documents, sentiment data

[1508] Output: Draft of new material

[1509] Step 5:

[1510] Through the chat interface, the user requests a review of the generated draft, which is then sent to the server.

[1511] Input: Review request

[1512] Output: Request sent to server

[1513] Step 6:

[1514] The server uses an automated writing function to analyze the text content of the generated draft using grammar checking software, such as Grammarly or LanguageTool, to detect errors and ambiguous expressions, as well as to analyze missing information and generate feedback based on the results.

[1515] Input: Generated draft

[1516] Output: Grammar check results, feedback

[1517] Step 7:

[1518] The server provides the generated feedback to the user through a chat interface, including comments such as "Item A is missing a specific number," "There are three grammatical errors," and "Please adjust your tone based on sentiment analysis."

[1519] Input: Feedback

[1520] Output: Providing feedback to the user

[1521] This series of steps allows the distribution center manager and staff to efficiently create new documents and quickly review them. Specific examples of prompts are as follows:

[1522] "I want to create a shift plan for next week."

[1523] "We want to create a plan for implementing a new warehouse management system."

[1524] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1525] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1526] 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 the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1527] [Fourth embodiment]

[1528] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1529] 7, a 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.

[1530] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1531] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1532] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1533] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1534] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1535] The control object 443 includes a display device, LEDs in the eyes, and motors for driving 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 emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1536] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1537] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[1538] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1539] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1540] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1541] The present invention provides a system that automatically generates new materials based on similar projects and business documents accumulated in the past, and automatically provides a chat-style discussion before the approver reviews them. Specific embodiments of this system are described below.

[1542] System configuration

[1543] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents.

[1544] Document collection and analysis

[1545] server

[1546] Access the folder path specified by the user (e.g., " / projects / past_docs / ") and retrieve all files in the folder.

[1547] The collected document files (PDF, Word, Excel, PowerPoint, etc.) are opened one by one and the contents (text data) are extracted.

[1548] Using natural language processing techniques, the extracted text content is analyzed and summaries and main topics are extracted.

[1549] Classify document types (business plans, project plans, etc.) and store them in a database.

[1550] Automatic generation of new materials

[1551] User

[1552] Through the chat interface, you input the type of document you want to generate (e.g., "Business Proposal") and the topic (e.g., "New Product X").

[1553] server

[1554] Search a parsed document database for documents most relevant to the input topic.

[1555] Perform a similarity calculation (e.g., cosine similarity) to identify the most relevant documents.

[1556] Based on the identified documents, the system automatically generates a draft of the new material according to a template, including basic elements such as an introduction, objectives, and a project overview.

[1557] Automatic wall hitting and feedback

[1558] User

[1559] Request a review of the generated draft through the chat interface.

[1560] server

[1561] Analyze the text content of the generated draft using the automatic wall-writing function. Use grammar checking software (e.g., Grammarly, LanguageTool, etc.) to detect grammatical errors.

[1562] It also analyzes ambiguous expressions and missing information, and generates feedback based on the detection results.

[1563] The generated feedback is provided to the user through a chat interface, including comments such as "Item A is missing a specific number" or "There are three grammatical errors."

[1564] Specific examples

[1565] For example, if a user types "I want to create a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant document. It then automatically generates a draft of a new business proposal based on the identified documents and presents it to the user. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[1566] In this way, the present invention provides a system that allows users to efficiently create new materials and quickly review them.

[1567] The processing flow will be explained below.

[1568] Step 1:

[1569] User

[1570] A user accesses the system via a chat interface and inputs the path of the folder to be analyzed.

[1571] User: Specify the " / projects / past_docs / " folder.

[1572] Step 2:

[1573] server

[1574] The server collects all document files (PDF, Word, Excel, PowerPoint) from the specified folder path and keeps a list of file names and paths.

[1575] The server retrieves all files in the " / projects / past_docs / " folder and lists their paths.

[1576] Step 3:

[1577] server

[1578] The server opens each collected file and extracts the text content using the appropriate library (e.g. PyPDF2, python-docx) depending on the document format (PDF, Word, etc.).

[1579] The server opens PDF files with PyPDF2 and Word files with python-docx and extracts the text.

[1580] Step 4:

[1581] server

[1582] The server analyzes the extracted text using natural language processing (NLP) techniques (e.g., spaCy, NLTK) to summarize the content and extract key topics and keywords.

[1583] The server uses the spaCy library to parse the text and extract summaries and keywords.

[1584] Step 5:

[1585] server

[1586] Based on the analysis results, the server classifies the documents into categories (e.g., business proposals, project plans) and stores them in a database.

[1587] The server stores the parsed text along with the classification information in a database.

[1588] Step 6:

[1589] User

[1590] The user inputs the type and theme of the material they wish to generate through the chat interface.

[1591] User: Enter "Create a business proposal on the theme of new product X."

[1592] Step 7:

[1593] server

[1594] The server searches the database for relevant parsed documents based on the themes entered by the user, and performs similarity calculations (e.g., cosine similarity) to identify the most relevant documents.

[1595] The server searches the database for keywords related to "business proposal" and "new product X" and calculates cosine similarity to identify highly relevant documents.

[1596] Step 8:

[1597] server

[1598] The server automatically generates a draft of the new material based on the template, including basic elements such as an introduction, objectives, and a project outline, based on the identified related documents.

[1599] The server generates a new draft based on a business proposal template.

[1600] Step 9:

[1601] User

[1602] The user requests a review of the generated draft through a chat interface.

[1603] User: Enter "Please review the business proposal I created."

[1604] Step 10:

[1605] server

[1606] The server uses an automatic writing-on-the-wall function to analyze the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool) to detect errors and ambiguous expressions.

[1607] The server uses Grammarly to detect grammatical errors and ambiguous expressions in the draft.

[1608] Step 11:

[1609] server

[1610] The server generates a summary of the detected problems and suggestions for improvement, and provides feedback to the user through a chat interface.

[1611] The server generates error information and improvement suggestions as feedback and provides them to the user.

[1612] In this way, by performing specific processing at each step, the user can efficiently create new materials and have them reviewed quickly.

[1613] Example 1

[1614] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1615] Conventional document creation systems have the problem of requiring a great deal of time and effort to create and review new documents. Furthermore, there is a lack of effective means for utilizing similar documents from the past, making efficient document creation difficult. Furthermore, document content and grammar checks are not fully automated, placing a burden on users.

[1616] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1617] In this invention, the server includes means for collecting similar documents from the past and analyzing their contents using natural language processing technology, means for inputting the type and theme of the document to be created via a user interface, means for calculating similarity and identifying the most relevant document, means for automatically generating a draft of a new document based on the identified document, means for performing grammar and content checks on the content of the generated draft, and means for generating feedback based on the check results and providing it to the user, thereby enabling the user to quickly and accurately create new documents and review them effectively.

[1618] "Past similar documents" refers to documents that were previously created and are the subject of collection and analysis by the system.

[1619] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language.

[1620] "User interface" refers to an interface through which a user inputs information into a system and performs operations.

[1621] "Similarity calculation" refers to a calculation method that evaluates the relevance between text data and expresses it numerically.

[1622] "Cosine similarity" refers to a method of measuring the similarity between two pieces of text data using angles in a vector space.

[1623] "Feedback" refers to information that indicates reviews and improvements to the generated document.

[1624] "Grammar checking" refers to the process of detecting and correcting grammatical errors in a piece of writing.

[1625] "Content check" refers to the process of checking whether the content of a document is appropriate and whether it contains the necessary information.

[1626] The present invention provides a system that automatically generates new materials based on similar projects and business documents accumulated in the past, and automatically provides a chat-style discussion before the approver reviews them. Specific embodiments of this system are described below.

[1627] System configuration

[1628] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents.

[1629] Document collection and analysis

[1630] server

[1631] The server accesses the folder path specified by the user (e.g., " / projects / past_docs / ") and retrieves all files in the folder.

[1632] The server sequentially opens the collected document files (PDF, Word, Excel, PowerPoint, etc.) and extracts the contents (text data). To do this, it uses libraries such as Apache Tika.

[1633] The server uses natural language processing techniques to analyze the extracted text content and extract summaries and main topics, specifically using libraries such as NLTK and SpaCy.

[1634] The server classifies the documents by type (business proposal, project plan, etc.) and stores them in a database. The database used here is a general database management system (e.g., MySQL, PostgreSQL).

[1635] Automatic generation of new materials

[1636] User

[1637] Through the chat interface, users input the type of document they want to generate (e.g., "business proposal") and theme (e.g., "new product X") as keywords.

[1638] server

[1639] The server searches the parsed document database for documents that are most relevant to the input topic, using a search engine such as ElasticSearch.

[1640] The server performs a similarity calculation (e.g., cosine similarity) to identify the most relevant documents, using a machine learning library such as scikit-learn.

[1641] The server automatically generates a draft of the new material based on the identified document, following a template, using OpenAI's generative AI model (e.g., GPT-3). The generated draft includes basic components such as an introduction, purpose, and outline of the plan.

[1642] Automatic wall hitting and feedback

[1643] User

[1644] The user requests a review of the generated draft through a chat interface.

[1645] server

[1646] The server analyzes the text content of the generated draft using an automated writing-on-the-wall function and detects grammar errors using grammar checking software (e.g., Grammarly or LanguageTool).

[1647] The server also analyzes ambiguous expressions and missing information and generates feedback based on the results of its detection.

[1648] The server provides the generated feedback to the user through a chat interface, including comments such as "Item A is missing a specific number" or "There are three grammatical errors."

[1649] Examples and prompts

[1650] For example, if a user types "I want to create a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant document. It then automatically generates a draft of a new business proposal based on the identified documents and presents it to the user. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[1651] An example prompt is, "Write a business proposal for new product X. Include the use, challenges, and solutions."

[1652] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1653] Step 1: Gather documentation

[1654] server

[1655] Input: A user-specified folder path (e.g. " / projects / past_docs / ").

[1656] Specific operation: The server accesses the specified folder and lists all files in the folder using the os.listdir() function.

[1657] Data processing: Open each file one by one and extract the file contents using Apache Tika, for example, extract text data from PDF, Word, Excel, and PowerPoint files.

[1658] Output: A list of extracted text data.

[1659] Step 2: Parsing the document

[1660] server

[1661] Input: A list of extracted text data.

[1662] Specific operation: The server performs natural language processing on each text data, specifically tokenization, POS tagging, summarization, and major topic extraction using SpaCy.

[1663] Data Calculation: Calculate TF-IDF scores to identify major topics and keywords in the text.

[1664] Output: A list of analysis results, including summary, major topics, and classification information.

[1665] Step 3: Saving to the database

[1666] server

[1667] Input: A list of analysis results.

[1668] Specific operation: The server stores the analysis results in a database such as MySQL or PostgreSQL.

[1669] Data processing: Classify the data based on the type of document (e.g., business plan, project plan), and store the relevant information for each document in a database.

[1670] Output: Parsed document data stored in a database.

[1671] Step 4: Generate request from user

[1672] User

[1673] Input: Enter the type and topic of the document you want to generate (e.g., "Business Proposal" or "New Product X") via the chat interface.

[1674] Specific operation: The user accesses the chat interface and inputs the necessary information, which causes the user's request to be sent to the server.

[1675] Output: Generated request information.

[1676] Step 5: Finding Similar Documents

[1677] server

[1678] Input: Generation request information (document type and subject).

[1679] What it does: The server uses ElasticSearch to search the database for relevant documents.

[1680] Data calculation: Calculate cosine similarity and identify the documents most relevant to the request information. Similarity calculation is performed using scikit-learn.

[1681] Output: A list of the most relevant documents.

[1682] Step 6: Draft a new document

[1683] server

[1684] Input: A list of the most relevant documents.

[1685] Specific operation: The server uses OpenAI's generative AI model (e.g., GPT-3) to automatically generate a draft of a new document based on a template.

[1686] Data processing: Using a generative AI model, a draft is created that includes the document's basic structure, such as its introduction, purpose, and outline.

[1687] Output: A draft of the new document.

[1688] Step 7: Provide the draft to users

[1689] server

[1690] Input: New document draft.

[1691] Specific operation: The server provides the generated draft to the user through a chat interface.

[1692] Output: A draft of the new document sent to the user.

[1693] Step 8: Automatic wall-hitting and feedback generation

[1694] User

[1695] Input: Review request (via chat interface).

[1696] What happens: A user requests a draft for review through the chat interface. This information is sent to the server.

[1697] Output: Review request information.

[1698] server

[1699] Input: Review request information and new document draft.

[1700] What it does: The server uses LanguageTool and Grammarly to detect grammatical errors in the draft, as well as analyzes ambiguous expressions and missing information.

[1701] Data operations: Detecting grammatical errors, checking content, and generating feedback.

[1702] Output: Feedback information.

[1703] Step 9: Provide feedback

[1704] server

[1705] Input: Feedback information.

[1706] Specific Operation: The server provides the generated feedback to the user through a chat interface.

[1707] Output: Feedback sent to the user.

[1708] (Application example 1)

[1709] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1710] In production management and quality control within factories, document creation and review work is a time-consuming and labor-intensive issue. In particular, when launching a new product, it is necessary to extract the necessary information from a huge amount of related past documents and quickly create and review new documents. However, the current manual process is inefficient and prone to human error. There is a need to solve these problems and streamline and automate document creation and review.

[1711] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1712] In this invention, the server includes a means for collecting similar documents from the past and analyzing their contents using natural language processing technology, a means for inputting the type of document to be generated and necessary information via a user interface installed in the factory robot, and a means for calculating similarity and identifying the most relevant document. This makes it possible to efficiently create new documents based on past documents and to generate high-quality documents and quickly review them by using a generative AI model.

[1713] "Past similar documents" are documents that were created previously and have content and formats related to the current project or work.

[1714] "Natural language processing technology" is a technology that allows computers to understand text data written in human language and perform operations such as analysis, generation, and translation.

[1715] A "user interface mounted on a factory robot" is an interface that is installed on a factory robot and allows the user to input data and perform operations.

[1716] "Similarity calculation" is a calculation method for numerically evaluating the degree of similarity between different documents or data sets.

[1717] "Cosine similarity" is a method for assessing the similarity of documents or datasets by calculating the angle between two vectors.

[1718] A "generative AI model" is a learning model that uses artificial intelligence to generate new text or information based on specific input prompts.

[1719] "Grammar checking" is the process of detecting grammatical errors in a document and suggesting corrections.

[1720] "Content check" is a process for verifying whether the contents of a document are logically correct, consistent, and whether there is any missing information.

[1721] "Feedback" means suggestions for correction or improvement provided as a result of analysis or review.

[1722] This invention is a system for automatically generating new documents based on similar past documents and then reviewing their contents. The purpose of this system is to improve the efficiency of creating and reviewing production management and quality control documents within a factory.

[1723] System configuration

[1724] The system consists of three main components: a server, a terminal mounted on the factory robot, and a user. The user interacts with the system through a user interface mounted on the factory robot, while the server consistently collects, analyzes, generates, and reviews documents.

[1725] Document collection and analysis

[1726] server

[1727] Accesses a folder path specified by the user (e.g. " / projects / past_docs / ") and retrieves all files in the folder. Formats include PDF, Word, Excel, PowerPoint, etc.

[1728] The text content of the collected document files is extracted and analyzed using natural language processing techniques, which makes it possible to extract summaries and main topics.

[1729] Classify the document type (e.g. production control report, quality control report, etc.) and store it in a database.

[1730] Automatic generation of new materials

[1731] User

[1732] Through the user interface installed on the factory robot, the type of document to be generated (e.g., "Quality Control Report") and topic (e.g., "New Product X") are entered.

[1733] server

[1734] Search a parsed document database for documents most relevant to the input topic.

[1735] Based on similarity calculations (e.g., cosine similarity) and the most relevant documents, a generative AI model is used to automatically generate a draft of new material, including quality check items, inspection results, and improvement suggestions.

[1736] Automatic wall hitting and feedback

[1737] User

[1738] A user interface on the factory robot allows a review of the generated draft to be requested.

[1739] server

[1740] Analyze the text content of the generated draft using the automatic wall-writing function. Use grammar checking software (e.g., Grammarly, LanguageTool, etc.) to detect grammatical errors.

[1741] It also analyzes ambiguous expressions and missing information and generates feedback, such as "Item A is missing a specific number" or "There are three grammatical errors."

[1742] The generated feedback is provided to the user through a user interface of the factory robot.

[1743] Specific examples

[1744] For example, if a user inputs "I want to create a quality control report" into a factory robot, the system will search the database for past documents related to "quality control report" and "new product X" to identify the most relevant documents. It then uses a generative AI model to automatically generate a draft of a new quality control report based on the identified documents. The generated draft is then checked for grammar and content, and feedback is provided to the user.

[1745] Prompt Sentence Examples

[1746] I want to create a quality control report

[1747] Please prepare a quality control report for new product X.

[1748] In this manner, the present invention provides a system that allows users to efficiently create new materials and quickly review them.

[1749] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1750] Step 1:

[1751] The user inputs the type of new material (e.g., "Quality Control Report") and topic (e.g., "New Product X") via a user interface installed on the factory robot. The input information is sent to the server.

[1752] Input: Type of material and subject

[1753] Output: Request data sent to the server

[1754] Step 2:

[1755] The server accesses the specified folder path (e.g., " / projects / past_docs / "), collects all document files (PDF, Word, Excel, PowerPoint, etc.) in the folder, and opens them to extract text data.

[1756] Input: the required folder path

[1757] Output: Text data of collected documents

[1758] Step 3:

[1759] The server analyzes the collected text data using natural language processing technology, extracts summaries and main topics from the analysis results, classifies them by document type, and stores them in a database.

[1760] Input: Text data of collected documents

[1761] Output: Summary of the analyzed document, main topics, and classification information

[1762] Step 4:

[1763] The server searches through the documents stored in the database to find the documents that are most relevant to the topic entered by the user. It identifies the relevant documents using similarity calculations (e.g., cosine similarity).

[1764] Input: User-entered themes, documents from the database

[1765] Output: Most relevant documents

[1766] Step 5:

[1767] Based on the identified documents, the server uses a generative AI model to automatically generate a draft of the new material, including basic components such as an introduction, purpose, and outline of the project.

[1768] Input: Most relevant documents

[1769] Output: Automatically generated draft of new material

[1770] Step 6:

[1771] When a user requests a review of a generated draft, the server uses an automated text-based grammar and content checker to detect errors.

[1772] Input: Draft of new material, user review request

[1773] Output: Grammar and content error detection results

[1774] Step 7:

[1775] The server generates feedback based on the detected errors and provides it to the user with specific suggestions for correction, such as "Item A is missing a specific number" or "There are three grammatical errors."

[1776] Input: Error detection result

[1777] Output: Generated feedback, suggested corrections

[1778] By the above steps, the present invention makes it possible to efficiently and automatically create and review new materials in a factory.

[1779] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1780] The present invention combines an emotion engine with a system that automatically generates new materials based on similar projects and business documents accumulated in the past and provides automatic chat-style discussion before the approver reviews them. Specific embodiments of this system are described below.

[1781] System configuration

[1782] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents. The emotion engine also recognizes the user's emotions and reflects them in the document creation process.

[1783] Document collection and analysis

[1784] server

[1785] Access the folder path specified by the user (e.g., " / projects / past_docs / ") and collect all files in the folder.

[1786] The collected document files (PDF, Word, Excel, PowerPoint, etc.) are opened one by one and the contents (text data) are extracted.

[1787] Using natural language processing techniques, the extracted text content is analyzed and summaries and main topics are extracted.

[1788] Classify document types (business plans, project plans, etc.) and store them in a database.

[1789] Automatic generation of new materials

[1790] User

[1791] Through the chat interface, you input the type of document you want to generate (e.g., "Business Proposal") and the topic (e.g., "New Product X").

[1792] server

[1793] An emotion engine is used to recognize the user's emotions from the input text data.

[1794] Search a parsed document database for documents most relevant to the input topic.

[1795] Perform a similarity calculation (e.g., cosine similarity) to identify the most relevant documents.

[1796] Based on the identified documents, the system automatically generates a template-based draft of the new material, including basic elements such as an introduction, objectives, and project outline, while adjusting the tone and style of the material based on the user's perceived emotions.

[1797] Automatic wall hitting and feedback

[1798] User

[1799] Request a review of the generated draft through the chat interface.

[1800] server

[1801] The text content of the generated draft is analyzed using grammar checking software (e.g., Grammarly, LanguageTool) to detect errors and ambiguous expressions.

[1802] It also analyzes ambiguous expressions and missing information, and generates feedback based on the detection results.

[1803] The generated feedback is provided to the user through a chat interface, including comments such as "Item A is missing specific numbers," "There are three grammatical errors," and "Please adjust your tone based on sentiment analysis."

[1804] Specific examples

[1805] For example, if a user types "I want to write a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant documents. It then automatically generates a draft of the new business proposal based on the identified documents. The emotion engine recognizes the user's emotions based on the text entered by the user and adjusts the tone and style of the document accordingly. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[1806] In this way, the present invention not only enables users to efficiently create new materials and quickly review them, but also provides a system that enables document creation that takes the user's emotions into consideration.

[1807] The processing flow will be explained below.

[1808] Step 1:

[1809] User

[1810] A user accesses the system via a chat interface and inputs the path of the folder to be analyzed.

[1811] User: Specify the " / projects / past_docs / " folder.

[1812] Step 2:

[1813] server

[1814] The server collects all document files (PDF, Word, Excel, PowerPoint) from the specified folder path and keeps a list of file names and paths.

[1815] The server retrieves all files in the " / projects / past_docs / " folder and lists their paths.

[1816] Step 3:

[1817] server

[1818] The server opens each collected file and extracts the text content using the appropriate library (e.g. PyPDF2, python-docx) depending on the document format (PDF, Word, etc.).

[1819] The server opens PDF files with PyPDF2 and Word files with python-docx and extracts the text.

[1820] Step 4:

[1821] server

[1822] The server analyzes the extracted text using natural language processing (NLP) techniques (e.g., spaCy, NLTK) to summarize the content and extract key topics and keywords.

[1823] The server uses the spaCy library to parse the text and extract summaries and keywords.

[1824] Step 5:

[1825] server

[1826] Based on the analysis results, the server classifies the documents into categories (e.g., business proposals, project plans) and stores them in a database.

[1827] The server stores the parsed text along with the classification information in a database.

[1828] Step 6:

[1829] User

[1830] The user inputs the type and theme of the material they wish to generate through the chat interface.

[1831] User: Enter "Create a business proposal on the theme of new product X."

[1832] Step 7:

[1833] server

[1834] The server analyzes input data from the user interface using an emotion engine based on the theme entered by the user to recognize the emotion of the user.

[1835] The server uses an emotion engine to analyze the user's input text and recognize emotions such as "positive" and "negative."

[1836] Step 8:

[1837] server

[1838] The server searches the analyzed document database for documents that are most relevant to the input topic, taking into account the emotion recognition results. It performs similarity calculations (e.g., cosine similarity) to identify the most relevant documents.

[1839] The server searches the database for keywords related to "business proposal" and "new product X" and calculates cosine similarity to identify highly relevant documents.

[1840] Step 9:

[1841] server

[1842] The server automatically generates a template-based draft of the new document based on the identified related documents. The draft includes basic elements such as an introduction, objectives, and a project outline. The tone and style of the document are adjusted based on the user's perceived emotions.

[1843] The server generates a new draft based on a business proposal template, adapting the tone and style to match the user's emotional needs.

[1844] Step 10:

[1845] User

[1846] The user requests a review of the generated draft through a chat interface.

[1847] User: Enter "Please review the business proposal I created."

[1848] Step 11:

[1849] server

[1850] The server uses an automatic writing-on-the-wall function to analyze the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool) to detect errors and ambiguous expressions.

[1851] The server uses Grammarly to detect grammatical errors and ambiguous expressions in the draft.

[1852] Step 12:

[1853] server

[1854] The server generates a summary of the detected issues and suggestions for improving tone and style based on sentiment analysis, and provides feedback to the user via a chat interface.

[1855] The server generates feedback including error information, suggestions for improvement, and comments based on sentiment analysis, and provides it to the user.

[1856] In this way, by performing specific processing at each step, the user can efficiently create new materials, have them reviewed quickly, and create documents that take the user's feelings into consideration.

[1857] Example 2

[1858] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1859] In conventional document generation systems, the manual search for similar documents, analysis of their contents, and creation of new materials required time and effort, and it was difficult to create documents that took the user's feelings into consideration. Furthermore, the grammar and content of the generated drafts had to be checked manually, preventing automation. This made it difficult for users to create documents efficiently and with high quality.

[1860] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes: means for collecting past similar documents and analyzing their contents using natural language processing technology; means for inputting the type of material to be generated and necessary information via a user interface; means for recognizing the user's sentiment from the input text data; means for calculating similarity and identifying the most relevant document; means for automatically generating a draft of a new material based on the identified document; means for adjusting the tone and style of the generated draft based on the user's sentiment; means for checking the grammar and content of the generated draft; and means for generating feedback based on the check results and providing it to the user. This enables the automation of document generation work, allowing users to efficiently create high-quality documents and quickly review them.

[1861] "Previous similar documents" refers to previously created, highly relevant materials and records.

[1862] "Natural language processing technology" refers to technology that enables computers to understand, interpret, and manipulate human language.

[1863] "User interface" refers to the components that provide the means for a user to interact with a system.

[1864] "Means of recognizing emotions" refers to technologies and algorithms that analyze emotions from user input and understand those emotions.

[1865] "Similarity calculation" refers to a calculation method for measuring the similarity between multiple documents and evaluating their relevance.

[1866] "Document" means any electronic or physical document containing certain information.

[1867] A "draft" is a rough draft of a document that is not yet complete, but has a certain degree of content.

[1868] "Tone and style adjustment" refers to the process of changing the wording and writing style of a document based on the user's feelings.

[1869] "Grammar Check" refers to the functionality that detects grammatical errors in text and suggests corrections.

[1870] "Content check" refers to the function that checks the content of the text for consistency and accuracy and suggests corrections if necessary.

[1871] "Feedback" refers to evaluations and suggestions for improvement provided to users by the system.

[1872] "Chat interface" refers to an interface that allows users and systems to communicate text-based.

[1873] The present invention is a system that automatically generates new materials based on similar documents from the past and adjusts the tone and style according to the user's emotions. An embodiment of the system will be described in detail below.

[1874] System configuration

[1875] This system consists of three main components: a server, a terminal, and a user. The server is responsible for the collection, analysis, generation, and review of documents, while the terminal provides an interface for users to interact with the system. Users access the system through their terminal and perform the necessary operations.

[1876] Hardware and software used

[1877] 1. Server

[1878] Hardware: General server machine (CPU, memory, storage)

[1879] software:

[1880] Document processing: PyPDF2 (analyzing PDF files), python-docx (analyzing Word files)

[1881] Natural Language Processing: spaCy, NLTK

[1882] Emotion Recognition: A publicly available sentiment analysis API

[1883] Similarity calculation: Similarity calculation algorithm (e.g., cosine similarity)

[1884] Grammar Check: Grammarly, LanguageTool

[1885] 2. Terminal

[1886] Hardware: General purpose computers, tablets, smartphones

[1887] Software: Browser or dedicated chat interface application

[1888] Process Overview

[1889] 1. Document collection and analysis

[1890] The server accesses the folder path specified by the user and collects all files found there. It then opens the collected document files using a library corresponding to their format (e.g., PyPDF2 or python-docx) and extracts the text content. The extracted text is then analyzed using natural language processing techniques (spaCy or NLTK) to extract summaries and main topics. The results of these analyses are categorized based on document type and stored in a database.

[1891] 2. Automatic generation of new materials

[1892] Users input the type and theme of the document they wish to generate through a chat interface. The server uses an emotion engine to recognize the user's emotions from the input text data. It then searches the analyzed data stored in the database for documents most relevant to the input theme and performs a similarity calculation. Once the most relevant document is identified, a draft of the new document is automatically generated based on a template. The generated draft includes basic components such as an introduction, purpose, and outline of the plan, and the tone and style are adjusted based on the user's emotions.

[1893] 3. Automatic wall hitting and feedback

[1894] The user requests a review of the generated draft. The server then uses an automated wall-hitting function to analyze the text content of the draft using grammar checking software (Grammarly, LanguageTool) to detect grammatical errors, ambiguous expressions, and missing information. Based on the results, feedback is generated and provided to the user through a chat interface. The feedback includes specific error indications and suggestions for improvement.

[1895] Specific examples

[1896] For example, if a user types "I want to write a business proposal for new product X" into the chat interface, the server collects past business proposal documents from the specified folder and performs the necessary analysis. It then searches for documents related to the user's theme, "new product X," and automatically generates a draft of a new business proposal based on the most relevant documents after similarity calculations. This draft is adjusted to reflect the user's sentiment in tone and style, and based on the user's review request, it undergoes grammar and content checks and provides specific feedback.

[1897] Example prompt: "I would like to create a business proposal for new product X. Please refer to the previous project materials to generate a draft and review it."

[1898] This series of processes enables users to efficiently create high-quality documents and review them quickly.

[1899] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1900] Step 1:

[1901] Document file collection

[1902] The server accesses the folder path specified by the user (e.g. " / projects / past_docs / ") and lists all the files in the folder.

[1903] Input: Folder path

[1904] Specific behavior: Scans the file system and generates a list of files in the specified folder.

[1905] Output: File list

[1906] Step 2:

[1907] Document Content Extraction

[1908] The server filters document files (PDF, Word, Excel, PowerPoint, etc.) from the listed files.

[1909] Each document file is opened using a library corresponding to the format and the text content is extracted (e.g. PyPDF2 for PDF, python-docx for Word).

[1910] Input: File list

[1911] Specific operation: Opens a document file and extracts the text.

[1912] Output: Extracted text data

[1913] Step 3:

[1914] Text Analysis and Classification

[1915] The server analyzes the extracted text using natural language processing techniques (e.g., spaCy, NLTK) to extract summaries and main topics.

[1916] It automatically classifies documents into types (e.g., business proposals, project plans, etc.) based on their content.

[1917] The analysis results are stored in a database.

[1918] Input: Extracted text data

[1919] What it does: Summarizes text, extracts major topics, and classifies document types.

[1920] Output: Abstract, main topics, classification information

[1921] Step 4:

[1922] Acceptance of document generation requests

[1923] Users simply input the type of document they want to generate (e.g., business proposal) and topic (e.g., new product X) through a chat interface.

[1924] Input: Material type and subject

[1925] Specific actions: Enter a request for document generation into the chat interface.

[1926] Output: Document generation request

[1927] Step 5:

[1928] Emotion Recognition and Document Retrieval

[1929] The server uses an emotion engine to recognize the user's emotions from the input text data.

[1930] It searches a database of analyzed documents for the documents most relevant to the input topic.

[1931] Input: Document generation request, existing parsed document

[1932] Specific behavior: Recognizes emotions from input text and searches for related documents.

[1933] Output: Emotion recognition results, related documents

[1934] Step 6:

[1935] Automatic draft generation

[1936] The server performs a similarity calculation (e.g., cosine similarity) to identify the most relevant documents.

[1937] Based on the identified documents, it automatically generates drafts of new materials based on templates (e.g., introduction, objectives, project overview, etc.).

[1938] It adjusts the tone and style of the material based on the user's emotion recognition results.

[1939] Input: Emotion recognition results, related documents

[1940] What it does: Selects the most relevant documents based on similarity calculations, generates drafts, and adjusts tone and style.

[1941] Output: Draft of new material

[1942] Step 7:

[1943] Request a draft review

[1944] Users can request a review of the generated draft through a chat interface.

[1945] Input: Draft of new material

[1946] What it does: Sends a review request in the chat interface.

[1947] Output: Review request

[1948] Step 8:

[1949] Checking and parsing the draft

[1950] The server uses an automatic wall-hitting function and analyzes the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool).

[1951] It detects grammatical errors, ambiguous expressions, and missing information.

[1952] Input: Draft of new material, grammar checking software

[1953] Specific actions: Analyze the draft text with grammar checking software to detect errors and omissions.

[1954] Output: Syntax error detection results

[1955] Step 9:

[1956] Generating and Providing Feedback

[1957] The server generates feedback based on the detection results and adds specific improvement suggestions (e.g., "Item A lacks specific numbers," "There are three grammatical errors," "Please adjust your tone based on sentiment analysis").

[1958] Feedback is provided to users through a chat interface.

[1959] Input: Syntax error detection result

[1960] Specific behavior: Generate feedback from the detection results and provide it in a chat interface.

[1961] Output: Feedback

[1962] Through the above steps, the user can efficiently create new materials and quickly review them.

[1963] (Application example 2)

[1964] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1965] In logistics centers, creating new documents requires a lot of time and effort, and the process of extracting necessary information from past documents to create new plans and reports is particularly cumbersome. Furthermore, traditional methods tend to delay review and feedback on created documents, hindering rapid business processes. Furthermore, documents created without considering user sentiment may be difficult for users to understand and may be ineffective.

[1966] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1967] In this invention, the server includes means for collecting similar documents from the past and analyzing their contents using natural language processing technology, means for inputting the type of material to be generated and necessary information via a user interface, and means for calculating similarity and identifying the most relevant document. This makes it possible to efficiently automatically generate new materials based on past documents and provide real-time feedback in a document style that reflects the user's feelings.

[1968] "Past similar documents" are documents or plans created in the past, which have similar content and format to the newly generated document.

[1969] "Natural language processing technology" is a technology for analyzing human language and automatically understanding, processing, and generating it, and includes functions such as extracting, summarizing, and classifying text data.

[1970] A "user interface" is an interface that allows a system and a user to interact with each other, and provides operation means using voice input, text input, touch gestures, etc.

[1971] "Similarity calculation" is a method for calculating the similarity between multiple documents, and measures the similarity using a statistical method such as cosine similarity.

[1972] "Draft of new material" refers to an early stage material that is automatically generated based on the theme and format specified by the user.

[1973] "Generated Draft Content" refers to the text and information contained in the automatically generated early stage material.

[1974] "Grammar checking" is a process of checking whether the generated text content contains any grammatical errors, and is carried out using natural language processing technology.

[1975] "Content checking" is the process of verifying that the generated text content is appropriate and consistent and that no important information is missing.

[1976] An "emotion engine" is a system that analyzes emotions from user input and adjusts the style and tone of the generated document based on those emotions.

[1977] "Feedback" is information including review results and improvement suggestions for the generated draft, and is provided in response to a user's request.

[1978] "Providing feedback in real time" refers to a process in which the review results of the generated materials are immediately provided to the user, allowing the user to quickly make any necessary corrections or confirmations.

[1979] The present invention is a system that automatically generates new materials based on similar projects and business documents accumulated in the past, and combines them with an emotion engine to automatically bounce ideas off each other in a chat format. Specific embodiments of the present invention are described below.

[1980] System configuration

[1981] This system consists of three main components: a server, a terminal, and a user. Users interact with the system through their terminals, and the server consistently collects, analyzes, generates, and reviews documents. The emotion engine also recognizes the user's emotions and reflects them in the document creation process.

[1982] Document collection and analysis

[1983] The server accesses the folder path specified by the user (e.g., " / projects / past_docs / ") and collects all files in the folder. It opens the collected document files (PDF, Word, Excel, PowerPoint, etc.) and extracts the content (text data). It uses natural language processing technology to analyze the extracted text content and extract summaries and main topics. It classifies the document type (business proposal, project plan, etc.) and stores it in a database.

[1984] Automatic generation of new materials

[1985] The user inputs the type of document they wish to generate (e.g., "business proposal") and theme (e.g., "new product X") through the user interface on their device. The server uses an emotion engine to recognize the user's emotions from the input text data. It then searches the analyzed document database for documents most relevant to the input theme. It performs similarity calculations (e.g., cosine similarity) to identify the most relevant documents. Based on the identified documents, it automatically generates a draft of the new document in accordance with a template. The generated draft includes basic components such as an introduction, purpose, and plan outline. The tone and style of the document are then adjusted based on the user's recognized emotions.

[1986] Automatic wall hitting and feedback

[1987] The user requests a review of the generated draft through the chat interface. The server uses an automated back-and-forth function to analyze the text content of the generated draft using grammar checking software (e.g., Grammarly, LanguageTool). Errors and ambiguous expressions are detected. Ambiguous expressions and missing information are also analyzed, and feedback is generated based on the detection results. The generated feedback is provided to the user through the chat interface. The feedback includes comments such as "Item A is missing a specific number," "There are three grammatical errors," and "Please adjust the tone based on sentiment analysis."

[1988] Specific examples

[1989] For example, if a user types "I want to write a new business proposal" into the chat interface, the server searches the database for past documents related to "business proposals" and "new product X" to identify the most relevant documents. It then automatically generates a draft of the new business proposal based on the identified documents. The emotion engine recognizes the user's emotions based on the text entered by the user and adjusts the tone and style of the document accordingly. When the user requests a review, the server checks the grammar and content of the generated draft and provides feedback on any problems detected and suggestions for improvement.

[1990] An example of a prompt sentence to input to the generative AI model is as follows:

[1991] "I want to create a shift plan for next week."

[1992] "We want to create a plan for implementing a new warehouse management system."

[1993] This invention enables logistics center managers and staff to efficiently create new documents and quickly review them. It also enables document creation that takes user emotions into consideration, reducing the burden on workers.

[1994] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1995] Step 1:

[1996] The server accesses the folder path specified by the user (e.g., " / projects / past_docs / ") and collects all files in the folder. The collected files include different formats such as PDF, Word, Excel, and PowerPoint. These files are opened sequentially and text data is extracted. This is done using natural language processing technology to extract summaries and main topics of the text content. The extracted data is classified by document type and stored in a database.

[1997] Input: Folder path

[1998] Output: Extracted text data, classified document data

[1999] Step 2:

[2000] The user inputs the type of document (e.g., "business proposal") and theme (e.g., "new product X") they want to generate through the user interface of their terminal. This input information is sent to the server.

[2001] Input: Type of material, theme

[2002] Output: Sending input information to the server

[2003] Step 3:

[2004] The server uses an emotion engine to recognize emotions from the text data entered by the user, searches the analyzed document database for documents that are most relevant to the entered topic, and uses similarity calculations, such as cosine similarity, to identify the most relevant documents.

[2005] Input: User-entered text data

[2006] Output: Identifying relevant documents

[2007] Step 4:

[2008] The server automatically generates a draft of the new document based on the template, including basic elements such as an introduction, objectives, and a project outline, and adjusts the tone and style of the document based on the user's perceived emotions.

[2009] Input: Identified documents, sentiment data

[2010] Output: Draft of new material

[2011] Step 5:

[2012] Through the chat interface, the user requests a review of the generated draft, which is then sent to the server.

[2013] Input: Review request

[2014] Output: Request sent to server

[2015] Step 6:

[2016] The server uses an automated writing function to analyze the text content of the generated draft using grammar checking software, such as Grammarly or LanguageTool, to detect errors and ambiguous expressions, as well as to analyze missing information and generate feedback based on the results.

[2017] Input: Generated draft

[2018] Output: Grammar check results, feedback

[2019] Step 7:

[2020] The server provides the generated feedback to the user through a chat interface, including comments such as "Item A is missing a specific number," "There are three grammatical errors," and "Please adjust your tone based on sentiment analysis."

[2021] Input: Feedback

[2022] Output: Providing feedback to the user

[2023] This series of steps allows the distribution center manager and staff to efficiently create new documents and quickly review them. Specific examples of prompts are as follows:

[2024] "I want to create a shift plan for next week."

[2025] "We want to create a plan for implementing a new warehouse management system."

[2026] 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 control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[2027] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[2028] 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 the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[2029] The emotion identification model 59 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 an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[2030] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[2031] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[2032] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[2033] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[2034] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs 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 a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[2035] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[2036] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[2037] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[2038] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[2039] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[2040] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[2041] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[2042] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[2043] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[2044] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[2045] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[2046] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[2047] The following is further disclosed regarding the above embodiment.

[2048] (Claim 1)

[2049] A means of collecting similar documents from the past and analyzing their contents using natural language processing technology;

[2050] A means for inputting the type of material to be generated and necessary information via a user interface;

[2051] a means for performing similarity calculations to identify the most relevant documents;

[2052] a means for automatically generating a draft of new material based on the identified documents;

[2053] means for performing grammar and content checks on the content of the generated draft;

[2054] means for generating and providing feedback to the user based on the results of the check;

[2055] A system including:

[2056] (Claim 2)

[2057] 2. The system according to claim 1, wherein the means for performing the grammar check and content check uses natural language processing technology.

[2058] (Claim 3)

[2059] 2. The system according to claim 1, wherein the means for calculating the similarity uses cosine similarity.

[2060] "Example 1"

[2061] (Claim 1)

[2062] A means of collecting similar documents from the past and analyzing their contents using natural language processing technology;

[2063] means for inputting the type and subject of the document to be generated via a user interface;

[2064] a means for performing similarity calculations to identify the most relevant documents;

[2065] means for automatically generating a draft of a new document based on the identified document;

[2066] means for performing grammar and content checks on the content of the generated draft;

[2067] means for generating and providing feedback to the user based on the results of the check;

[2068] A system including:

[2069] (Claim 2)

[2070] 2. The system according to claim 1, wherein the means for performing the grammar check and content check uses natural language processing technology.

[2071] (Claim 3)

[2072] 2. The system according to claim 1, wherein the means for calculating the similarity uses cosine similarity.

[2073] "Application Example 1"

[2074] (Claim 1)

[2075] A means of collecting similar documents from the past and analyzing their contents using natural language processing technology;

[2076] A means for inputting the type of material to be generated and necessary information via a user interface mounted on the factory robot;

[2077] a means for performing similarity calculations to identify the most relevant documents;

[2078] A means of automatically generating a draft of new material based on the identified documents using an AI model;

[2079] means for performing grammar and content checks on the content of the generated draft;

[2080] means for generating and providing feedback to the user based on the results of the check;

[2081] A system including:

[2082] (Claim 2)

[2083] 2. The system according to claim 1, wherein the means for performing the grammar check and content check uses natural language processing technology.

[2084] (Claim 3)

[2085] 2. The system according to claim 1, wherein the means for calculating the similarity uses cosine similarity.

[2086] "Example 2: Combining Emotion Engines"

[2087] (Claim 1)

[2088] A means of collecting similar documents from the past and analyzing their contents using natural language processing technology;

[2089] A means for inputting the type of material to be generated and necessary information via a user interface;

[2090] means for recognizing a user's emotion from input text data;

[2091] a means for performing similarity calculations to identify the most relevant documents;

[2092] a means for automatically generating a draft of new material based on the identified documents;

[2093] means for adjusting the tone and style of the generated draft based on the user's emotions;

[2094] means for performing grammar and content checks on the content of the generated draft;

[2095] means for generating and providing feedback to the user based on the results of the check;

[2096] A system including:

[2097] (Claim 2)

[2098] 2. The system according to claim 1, wherein the means for performing the grammar check and content check uses natural language processing technology.

[2099] (Claim 3)

[2100] 2. The system according to claim 1, wherein the means for calculating the similarity uses cosine similarity.

[2101] "Application example 2 when combining emotion engines"

[2102] (Claim 1)

[2103] A means of collecting similar documents from the past and analyzing their contents using natural language processing technology;

[2104] A means for inputting the type of material to be generated and necessary information via a user interface;

[2105] a means for performing similarity calculations to identify the most relevant documents;

[2106] a means for automatically generating a draft of new material based on the identified documents;

[2107] means for performing grammar and content checks on the content of the generated draft;

[2108] A means of analyzing user emotions using an emotion engine and adjusting the tone and style of the material;

[2109] a means for generating and providing feedback to the user based on the review results;

[2110] a means for providing real-time feedback through a user interface;

[2111] A system including:

[2112] (Claim 2)

[2113] 2. The system according to claim 1, wherein the means for performing grammar check and content check uses natural language processing technology.

[2114] (Claim 3)

[2115] 2. The system according to claim 1, wherein the means for performing similarity calculation uses cosine similarity. [Explanation of symbols]

[2116] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of collecting similar documents from the past and analyzing their contents using natural language processing technology; A means for inputting the type of material to be generated and necessary information via a user interface; a means for performing similarity calculations to identify the most relevant documents; a means for automatically generating a draft of new material based on the identified documents; means for performing grammar and content checks on the content of the generated draft; means for generating and providing feedback to the user based on the results of the check; A system including:

2. 2. The system of claim 1, wherein said means for performing grammar check and content check uses natural language processing technology.

3. 2. The system according to claim 1, wherein said means for calculating similarity uses cosine similarity.

Images