System

Abstract

A system is provided.SOLUTION: A system, comprising: means for collecting and storing historical test questions, entrance information, tutorials, and departmental activity tutorial manuals; means for preprocessing and converting the collected data into a unified format; means for training a generative artificial intelligence model using the preprocessed data; means for receiving requests from users to create mini-tests, test scores, and generate departmental activity tutorials; means for generating mini-tests, test scores, and departmental activity tutorials based on the requests using the trained generative artificial intelligence model; and means for delivering the generated tests, score results, or tutorials to the users' terminals.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] Elementary and junior high school teachers have a wide range of tasks beyond their teaching activities, which is the main reason for unpaid overtime and working on holidays. Furthermore, creating and grading tests, coaching club activities, and other tasks are particularly time-consuming, so the excessive workload of teachers has become a social issue. The purpose of this invention is to improve the efficiency of these tasks and reduce the workload of teachers. [Means for solving the problem]

[0005] The present invention provides a system that includes a means for collecting and storing past test questions, entrance exam information, curriculum guidelines, and extracurricular activity coaching manuals; a means for preprocessing the collected data and converting it into a unified format; a means for training a generative artificial intelligence model using the preprocessed data; a means for receiving requests from users for creating quizzes, grading tests, and generating extracurricular activity curriculum guidelines; a means for generating quizzes, grading tests, and extracurricular activity curriculum guidelines based on the requests using the trained generative artificial intelligence model; and a means for delivering the generated tests, grading results, or curriculum guidelines to the user's device. This system significantly reduces the workload for creating quizzes and regular tests and automates the grading of multiple-choice and essay-type questions, significantly reducing the workload of teachers. Furthermore, even inexperienced teachers can easily generate standard curriculum guidelines for extracurricular activities.

[0006] "Data collection methods" are means of collecting and storing education-related data such as past test questions, entrance examination information, curriculum guidelines, and club activity instruction manuals from teachers and educational institutions.

[0007] The "data preprocessing means" is a means for converting collected data into a unified format and for removing noise from and normalizing text data.

[0008] A "generative artificial intelligence model" is an AI model that learns from input education-related data and is used to create quizzes, grade tests, generate club activity curriculum guidelines, etc.

[0009] The "user request receiving means" is a means for receiving requests for creating quizzes, grading tests, and creating club activity teaching guidelines sent from the teacher's terminal.

[0010] A "test generation means" is a means for automatically generating quizzes based on user requests using a trained generative artificial intelligence model.

[0011] The "scoring means" is a means for receiving student response data sent from the user's device and using an AI model to score multiple-choice and essay questions.

[0012] The "teaching guideline generation means" is a means by which even inexperienced teachers can generate standard teaching guidelines based on club activity teaching guidelines or sports manuals.

[0013] The "distribution means" is a means for converting the generated quiz, marking results, or club activity instruction guidelines into a PDF file or HTML format and sending it to the user's terminal. [Brief explanation of the drawings]

[0014] [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

[0015] 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.

[0016] First, the terms used in the following description will be explained.

[0017] 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).

[0018] 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.

[0019] 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.

[0020] 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.

[0021] 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."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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."

[0035] System Overview

[0036] This invention is a system that uses generative artificial intelligence (AI) to reduce the workload of elementary and junior high school teachers. Specifically, by learning from past test questions, entrance exam information, and teaching guidelines, it automatically creates quizzes, automates grading, and generates teaching guidelines for club activities.

[0037] System configuration

[0038] 1. Data collection methods:

[0039] The server collects data provided by teachers and educational institutions, such as past test questions, entrance exam information, teaching guidelines, and club activity instruction manuals. This data is stored in the system's database.

[0040] 2. Data preprocessing methods:

[0041] The server cleans the collected data to standardize it into an appropriate format, removing noise from the text data and normalizing it, making it usable as education-related data.

[0042] 3. Generative AI Models:

[0043] The server uses the preprocessed data to train generative artificial intelligence (AI) models that are specialized to perform education-related tasks (quiz creation, grading, curriculum generation).

[0044] 4. User request receiving means:

[0045] Teachers using the devices can request the creation and grading of quizzes and the generation of club activity guidelines. The requests are sent to the server and analyzed.

[0046] 5. Test generation methods:

[0047] Based on user requests, the server automatically generates quizzes using AI models, which are structured according to the user's specified subject matter and number of questions.

[0048] 6. Scoring Procedure:

[0049] The server receives the student's response data sent by the user and automatically scores multiple-choice and essay questions using an AI model. The results are automatically compiled and feedback is provided.

[0050] 7. Curriculum syllabus generation means:

[0051] For club activity teaching guidelines, the server uses a trained AI model to generate standard teaching guidelines based on user requests, and these guidelines are provided in a format that is easy to use even for inexperienced teachers.

[0052] 8. Delivery Method:

[0053] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's terminal.

[0054] Program processing and specific examples

[0055] Example of creating a quiz

[0056] Request Submission (User):

[0057] A teacher sends a request from a terminal saying, "Create a 10-question quiz on linear equations in second-year junior high school mathematics."

[0058] Receiving and processing the request (server):

[0059] The server receives this request and analyzes the contents.

[0060] Quiz generation (server):

[0061] The server uses a trained AI model to generate quizzes based on the specified parameters.

[0062] Distribution (server):

[0063] The generated quiz is converted into PDF format and sent to the teacher's device.

[0064] Specific examples of automatic scoring

[0065] Submit Answer (User):

[0066] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their terminal to the server.

[0067] Receiving and analyzing response data (server):

[0068] The server receives the submitted answer data and categorizes it into multiple choice and essay questions.

[0069] Automatic grading (server):

[0070] The server uses an AI model to match multiple-choice questions with correct answer data and evaluate essay questions.

[0071] Distribution (server):

[0072] The grading results are automatically tallied and sent to the teacher's device.

[0073] Example of creating club activity guidelines

[0074] Request Submission (User):

[0075] An inexperienced soccer club advisor requests a "one-year coaching plan for a junior high school soccer club."

[0076] Receiving and processing the request (server):

[0077] The server receives this request and analyzes the contents.

[0078] Course guide generation (server):

[0079] The server uses a trained AI model to generate a one-year coaching plan for a typical junior high school soccer club.

[0080] Distribution (server):

[0081] The generated teaching guidelines are converted into PDF format and sent to the teacher's device.

[0082] The above system can reduce the workload of teachers and provide efficient educational support.

[0083] The processing flow will be explained below.

[0084] Program processing flow

[0085] Quiz creation process

[0086] Step 1:

[0087] A user (teacher) sends a request to create a quiz from a terminal, which includes detailed parameters such as grade, subject, topic, and number of questions.

[0088] Step 2:

[0089] The server receives the request from the user and analyzes its contents, which includes extracting the request parameters.

[0090] Step 3:

[0091] The server inputs the extracted parameters into a trained generative artificial intelligence model.

[0092] Step 4:

[0093] The generative AI model automatically generates quizzes based on input parameters, with questions tailored to the specified grade, subject, and topic.

[0094] Step 5:

[0095] The server converts the generated quiz into PDF or HTML format.

[0096] Step 6:

[0097] The server delivers the converted quiz to the user's device, where the user can review the quiz and distribute it to students by printing or online.

[0098] Auto-grading process

[0099] Step 1:

[0100] The user (teacher) sends the students' answers from their terminal to the server. The answers include answers to multiple-choice questions and answers to written questions.

[0101] Step 2:

[0102] The server receives the answer data sent by the user and classifies the answer into multiple choice and written answer.

[0103] Step 3:

[0104] The server compares the answers to the multiple-choice questions with the correct answer data and performs automatic grading.

[0105] Step 4:

[0106] The server inputs the answers to the essay questions into a trained generative AI model for evaluation, and the model automatically calculates a score based on this evaluation.

[0107] Step 5:

[0108] The server aggregates the multiple choice and essay grading results and generates a final score.

[0109] Step 6:

[0110] The server delivers the results to the user's device, where the user can check the results and provide feedback to the student.

[0111] Creation of club activity guidelines

[0112] Step 1:

[0113] The user (teacher) sends a request for club activity guidance from a terminal. The request includes detailed parameters such as the type of club activity, teaching goals, and duration.

[0114] Step 2:

[0115] The server receives the request from the user and analyzes its contents, which includes extracting the request parameters.

[0116] Step 3:

[0117] The server inputs the extracted parameters into a trained generative artificial intelligence model.

[0118] Step 4:

[0119] The generative AI model automatically generates standard club activity teaching guidelines based on input parameters. The generated teaching guidelines are appropriate for the type of club activity and teaching goals.

[0120] Step 5:

[0121] The server converts the generated teaching guidelines into PDF or HTML format.

[0122] Step 6:

[0123] The server delivers the converted teaching guidelines to the user's terminal, where the user can check and apply the teaching guidelines.

[0124] This will enable teachers to significantly improve their work efficiency and reduce the burden on the educational field.

[0125] Example 1

[0126] 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."

[0127] In the traditional education system, teachers have a heavy workload, especially when it comes to creating and grading quizzes and planning extracurricular activities. Furthermore, existing tools for streamlining these tasks are limited and lack sufficient flexibility and precision. This makes it difficult for teachers to devote sufficient time to their primary educational activities.

[0128] 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.

[0129] In this invention, the server includes: means for collecting and storing past assessment materials, entrance examination information, educational guidelines, and extracurricular activity instruction materials from educational institutions; means for preprocessing the collected materials and converting them into a unified format; means for training an AI model using the preprocessed materials; means for receiving requests from users to create exams, grade evaluations, and extracurricular activity guidelines; means for generating exams, grade evaluations, and extracurricular activity guidelines based on the requests using the trained AI model; means for delivering the generated exams, grade evaluations, or extracurricular activity guidelines to the user's terminal; means for analyzing the request content and generating prompts appropriate for the AI ​​model; and means for feeding the AI ​​model based on the analyzed request to generate optimal output. This reduces the workload of teachers and enables efficient support of educational activities.

[0130] An "educational institution" is a formal institution that provides teaching and learning, such as a school, university, college, or other institution.

[0131] "Assessment materials" are materials used to assess student learning outcomes, such as previously administered tests, exams, or quizzes.

[0132] "Entrance Examination Information" means data and information relating to examinations for admission to a particular school or educational institution.

[0133] "Educational guidelines" are guidelines or instructional documents for achieving a particular educational curriculum or learning objectives.

[0134] "Extracurricular activity instruction materials" are manuals and plans for instructing extracurricular activities, such as club activities and extracurricular club activities.

[0135] "Preprocessing" is the process of preparing collected data in a format that is easy to analyze and use, and performing noise removal and normalization.

[0136] A "generative artificial intelligence model" is an AI system that builds knowledge to perform education-related tasks from training data.

[0137] "Training" is the process by which an artificial intelligence model learns from specified data and improves its ability to solve specified problems.

[0138] A "request" is an instruction or command from a user to a system requesting a specific process or operation.

[0139] "Exam" refers to a test or quiz administered to assess a student's understanding or knowledge.

[0140] "Grading" is the process of assessing a student's achievements on a test or quiz and expressing that performance in terms of a number or grade.

[0141] The "Extracurricular Activities Guidelines" are detailed procedures and guidelines for guiding the planning and conduct of club and extracurricular activities.

[0142] A "prompt sentence" is an input sentence that is used to obtain a specific answer or product from the AI ​​model being generated.

[0143] "Feeding" is the process of inputting learning data and request content into an artificial intelligence model to obtain optimal output.

[0144] This invention is a system that utilizes generative artificial intelligence (AI) to reduce the workload of elementary and junior high school teachers. Specifically, the system automatically creates quizzes, automates grading, and generates club activity teaching guidelines by learning from past test questions, entrance exam information, educational guidelines, and club activity teaching materials.

[0145] System configuration

[0146] Data collection methods

[0147] The server automatically collects past assessment materials, entrance examination information, educational guidelines, and extracurricular activity guidance materials from educational institutions using APIs and FTP servers. The collected data is stored in a database.

[0148] Data preprocessing measures

[0149] The server preprocesses the collected data using text analysis tools (e.g., SpaCy, NLTK), removing noise from the text data, normalizing it, and converting it into a unified format so that it can be used as training data.

[0150] Generative AI Model

[0151] The server uses the preprocessed data to train an AI model (e.g., GPT-3, BERT), and uses machine learning libraries (e.g., TensorFlow, PyTorch) to build a model specialized for a given education-related task (quiz creation, grading, extracurricular activity curriculum generation).

[0152] User request receiving means

[0153] Teachers using the devices can use a dedicated application to request the creation and grading of quizzes and the generation of club activity guidelines. These requests are sent to the server in real time.

[0154] Test Generation Methods

[0155] The server uses a trained AI model to automatically generate quizzes based on user requests. It analyzes the request content and generates prompts appropriate for the AI ​​model. The generated quiz is converted into PDF or HTML format and sent to the teacher's device.

[0156] Scoring method

[0157] The server receives the student's response data sent by the user and automatically grades the multiple-choice and essay questions using an AI model. Multiple-choice questions are compared with the correct answer data, and essay questions are evaluated using NLP technology. The grading results are compiled and sent as feedback to the teacher's device.

[0158] Curriculum guideline generation means

[0159] The server uses a trained AI model to generate extracurricular activity guidelines based on the user's request. It analyzes the request and generates a prompt appropriate for the AI ​​model. The generated guidelines are converted into PDF format and sent to the teacher's device.

[0160] Data Delivery Methods

[0161] The server converts the generated quizzes, marking results, or extracurricular activity guidelines into PDF files or HTML format and delivers them to the teacher's terminal.

[0162] Specific example explanation

[0163] Example of creating a quiz

[0164] Request submission (user)

[0165] The teacher sends a request from the terminal saying, "Create a 10-question quiz on linear equations in second-year junior high school mathematics."

[0166] Receiving and processing requests (server)

[0167] The server receives this request in real time and analyzes the content.

[0168] Quiz generation (server)

[0169] The server uses a trained AI model to generate quizzes based on the specified parameters.

[0170] Distribution (server)

[0171] The generated quiz is converted into PDF format and sent to the teacher's device.

[0172] Specific examples of automatic scoring

[0173] Submit Answer (User)

[0174] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their terminal to the server.

[0175] Receiving and analyzing response data (server)

[0176] The server receives the submitted response data in real time and categorizes it into multiple choice and essay questions.

[0177] Automatic scoring (server)

[0178] The server uses an AI model to match multiple-choice questions with correct answer data and evaluate essay questions.

[0179] Distribution (server)

[0180] The grading results are automatically tallied and sent to the teacher's device.

[0181] Example of creating club activity guidelines

[0182] Request submission (user)

[0183] An inexperienced soccer club advisor requests a "one-year coaching plan for a junior high school soccer club."

[0184] Receiving and processing requests (server)

[0185] The server receives this request in real time and analyzes the content.

[0186] Teaching guideline generation (server)

[0187] The server uses a trained AI model to generate a one-year coaching plan for a typical junior high school soccer club.

[0188] Distribution (server)

[0189] The generated teaching guidelines are converted into PDF format and sent to the teacher's device.

[0190] This system will reduce the workload of teachers and provide efficient, high-quality educational support.

[0191] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0192] Step 1: Collect data

[0193] The server collects past evaluation materials, entrance examination information, educational guidelines, and extracurricular activity guidance materials from educational institutions and stores them in a database. Specifically, the server obtains data through APIs and FTP servers. The collected materials are stored in the database. Various educational materials are input, and educational materials in a unified data format are generated as output.

[0194] Step 2: Preprocessing the data

[0195] The server preprocesses the collected data. This involves using text analysis tools (e.g., SpaCy, NLTK) to remove noise and normalize the data. The input is the collected educational materials, and the output is preprocessed, cleaned data. Specifically, special characters are removed and the data is formatted.

[0196] Step 3: Training the AI ​​model

[0197] The server uses the preprocessed data to train a generative AI model (e.g., GPT-3, BERT). Machine learning libraries used include TensorFlow and PyTorch. The input is the preprocessed data, and the output is a trained AI model. Specifically, the data is fed into the AI ​​model, and iterative learning occurs to improve the model's accuracy.

[0198] Step 4: Receiving the request

[0199] Teachers using the devices make requests from their devices to create and grade quizzes and generate curriculum guidelines for club activities. These requests are sent to the server. The input is the teacher's request, and the output is the request content saved on the server. Specifically, the teacher inputs the request into a dedicated application, which is then sent to the server.

[0200] Step 5: Parsing the request and generating a prompt

[0201] The server analyzes the request content and generates a prompt sentence appropriate for the generative AI model. The input is the request received from the teacher, and the output is the generated prompt sentence. Specifically, the server analyzes the request content and generates a prompt sentence in an appropriate format.

[0202] Step 6: Generate the quiz

[0203] The server uses a trained AI model to automatically generate quizzes based on prompts. The input is the prompt, and the output is the generated quiz. Specifically, the generated prompt is fed into the AI ​​model to generate quiz questions.

[0204] Step 7: Convert and distribute the quiz

[0205] The generated quiz is converted into PDF or HTML format and sent to the teacher's device. The input is the generated quiz, and the output is a file converted into PDF or HTML format. Specifically, the test questions are converted into a document format and distributed to the device.

[0206] Step 8: Receiving and parsing responses

[0207] The teacher sends the students' answer data (multiple choice and written questions) from their terminal to the server. The server receives this data and classifies it into multiple choice and written questions. The input is the students' answer data, and the output is the classified answer data. Specifically, the data is classified.

[0208] Step 9: Automated scoring

[0209] The server uses an AI model to automatically score the received student answers. Multiple-choice questions are compared with correct answer data, and essay questions are evaluated using NLP technology. The input is classified answer data, and the output is the scoring results. Specifically, multiple-choice questions are automatically matched, and essay questions are evaluated using semantic analysis.

[0210] Step 10: Calculation and distribution of scores

[0211] The grading results are automatically compiled and sent to the teacher's device as feedback. The input is the grading results, and the output is the compiled feedback. Specifically, the grading results are compiled, and feedback is generated and distributed.

[0212] Step 11: Generate and distribute curriculum guidelines

[0213] The server uses a trained AI model to generate standard teaching guidelines based on the user's request. The generated teaching guidelines are converted into PDF format and sent to the teacher's device. The input is the prompt text and the AI ​​model, and the output is the teaching guidelines in PDF format. Specifically, the server analyzes the request content, generates the teaching guidelines, converts them into PDF, and distributes them.

[0214] (Application example 1)

[0215] 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."

[0216] In-factory worker education and training takes a great deal of time and effort, and if there is a shortage of skilled instructors, it can be difficult to improve worker skills and ensure safety. In addition, it can be difficult to provide consistent teaching materials and manage training progress, which can lead to inconsistent quality of education.

[0217] 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.

[0218] In this invention, the server includes a means for collecting and storing past training data, operation information, safety guidelines, and work manuals, a means for preprocessing the collected data and converting it into a unified format, and a means for training a generative artificial intelligence model using the preprocessed data, thereby enabling training materials, evaluation results, and troubleshooting tests to be automatically generated and distributed to terminals based on user requests.

[0219] "Training data" refers to past data collected for the purpose of educating and training workers.

[0220] "Operation information" refers to information about specific operating procedures and methods for machines and equipment.

[0221] "Safety guidelines" are guidelines and rules for ensuring safety during work and operations.

[0222] A "work manual" is a document or instruction manual that shows specific tasks and procedures.

[0223] "Preprocessing" refers to the process of organizing and cleaning raw data to convert it into a usable format.

[0224] A "uniform format" is a standard format for converting different types of data into a consistent structure.

[0225] "Training means" refers to the data and methods used to train an AI model.

[0226] A "request" means that a user requests something from the system.

[0227] "Training materials" are materials created for workers to learn and improve their skills.

[0228] "Evaluation data" is data based on the activities and responses performed by the worker.

[0229] "Troubleshooting tests" are tests to identify and resolve problems or faults.

[0230] "Automatic generation" means that the system autonomously creates data and information without human intervention.

[0231] A "terminal" is a device used to access the system.

[0232] This invention provides a system for optimizing worker education and training in a factory and improving work efficiency. This system automatically generates and evaluates training materials for worker education and provides troubleshooting tests through the following processing steps.

[0233] System Overview

[0234] The system uses smartphones, smart glasses, and head-mounted displays (HMDs) as hardware, and Flask (a Python microweb framework) and GPT-2 (a generative AI model) as software.

[0235] Program processing

[0236] Data collection and storage

[0237] The server collects and stores past training data, operating information, safety guidelines, and work manuals from databases within the factory and external sources.

[0238] Data preprocessing and conversion to a unified format

[0239] The server performs preprocessing such as noise removal and normalization on the collected data, and converts it into an appropriate unified format, making it possible to process various types of data in a consistent format.

[0240] Training generative artificial intelligence models

[0241] Using the pre-processed data, the server trains generative artificial intelligence (AI) models that are specialized to perform education-related tasks within the factory (training material creation, evaluation, troubleshooting test generation).

[0242] Receiving a user request

[0243] Users use their devices (smartphones, smart glasses, HMDs) to request the creation and evaluation of training materials and the generation of troubleshooting tests. These requests are sent to the server.

[0244] Automatic generation of training materials

[0245] The server automatically generates training materials using a trained AI model based on the received user requests, converts the generated materials into PDF or HTML format, and delivers them to the user's device.

[0246] Specific examples

[0247] For example, to create training materials for new employees on how to operate a lathe, the user might enter the following prompt:

[0248] Develop training materials for new employees on lathe operation, including basic safety guidelines and detailed operating procedures.

[0249] The server receives the prompt and generates training materials using a trained generative AI model, allowing users to instantly access the training materials and efficiently train new employees in the factory.

[0250] This system is a powerful tool for quickly and efficiently educating and training workers, contributing to improved productivity and safety within factories.

[0251] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0252] Step 1:

[0253] The server collects and stores past training data, operation information, safety guidelines, and work manuals from databases within the factory and external sources. This allows the various data required for training to be collected and stored within the system. The input is various materials such as training data and operation information, and the output is converted into a unified format and stored in the database.

[0254] Step 2:

[0255] The server performs preprocessing on the collected data, such as noise removal and normalization, and converts it into an appropriate unified format. This allows various data to be processed in a consistent format. The input is the data saved in the previous step, and the output is the preprocessed data in a consistent format.

[0256] Step 3:

[0257] Using the preprocessed data, the server trains a generative artificial intelligence (AI) model that is specialized to perform education-related tasks within the factory (training material creation, evaluation, troubleshooting test generation). The input is the preprocessed data, and the output is the trained generative AI model.

[0258] Step 4:

[0259] Users use devices (smartphones, smart glasses, HMDs) to request the creation and evaluation of training materials and the generation of troubleshooting tests. These requests are sent to the server. The input is the user's request, and the output is the receipt and analysis of the request content.

[0260] Step 5:

[0261] The server automatically generates training materials using a trained AI model based on the received user request. The generated materials are converted into PDF or HTML format and delivered to the user's device. The input is the user request and the AI ​​model, and the output is the generated training materials.

[0262] For example, when creating training materials for new employees on how to operate a lathe, the user would enter the following prompt:

[0263] Develop training materials for new employees on lathe operation, including basic safety guidelines and detailed operating procedures.

[0264] The server receives this prompt and generates training materials using a trained generative AI model, and the output is sent to the device as training materials in PDF or HTML format.

[0265] 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.

[0266] System Overview

[0267] This invention utilizes generative artificial intelligence (AI) and an emotion engine to reduce the workload of elementary and junior high school teachers. Specifically, by learning from past test questions, entrance exam information, and teaching guidelines, the system automatically creates quizzes, automates grading, and automatically generates teaching guidelines for club activities. This system also incorporates an emotion engine that recognizes the user's emotions, improving the quality of educational resources and feedback.

[0268] System configuration

[0269] 1. Data collection methods:

[0270] The server collects data provided by teachers and educational institutions, such as past test questions, entrance exam information, teaching guidelines, and club activity instruction manuals. This data is stored in the system's database.

[0271] 2. Data preprocessing methods:

[0272] The server cleans the collected data to standardize it into an appropriate format, removing noise from the text data and normalizing it, making it usable as education-related data.

[0273] 3. Generative AI Models:

[0274] The server uses the preprocessed data to train generative artificial intelligence (AI) models that are specialized to perform education-related tasks (quiz creation, grading, curriculum generation).

[0275] 4. User request receiving means:

[0276] Teachers using the devices can request the creation and grading of quizzes and the generation of club activity guidelines. The requests are sent to the server and analyzed.

[0277] 5. Emotion Engine:

[0278] The server has an emotion engine that recognizes the emotion of the user when he / she sends a request, and provides a function to optimize the request content based on the user's emotional state.

[0279] 6. Test generation methods:

[0280] Based on user requests, the server automatically generates quizzes using an AI model. The generated quizzes are structured according to the specified subject area and number of questions, but are appropriately adjusted to reflect the results of the emotion engine.

[0281] 7. Scoring Procedure:

[0282] The server receives the student's response data sent by the user and automatically scores multiple-choice and essay questions using an AI model. The results are automatically compiled and feedback is provided using an emotion engine.

[0283] 8. Curriculum syllabus generation means:

[0284] For extracurricular activities, the server uses the trained AI model to generate standard teaching guidelines based on the user's requests, which are then optimized according to the user's emotional state.

[0285] 9. Delivery Method:

[0286] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's terminal.

[0287] Program processing and specific examples

[0288] Example of creating a quiz

[0289] Request Submission (User):

[0290] A teacher sends a request from a terminal saying, "Create a quiz with 10 questions on linear equations in second-year junior high school mathematics." If the emotion engine determines that the emotion is "stress," the server simplifies the request or suggests a template.

[0291] Receiving and processing the request (server):

[0292] The server receives this request, analyzes the content, and evaluates the emotional state using an emotion engine.

[0293] Quiz generation (server):

[0294] The server uses a trained AI model to generate quizzes based on the specified parameters, and optimizes the test content based on the results of the emotion engine.

[0295] Distribution (server):

[0296] The generated quiz is converted into PDF format and sent to the teacher's device.

[0297] Specific examples of automatic scoring

[0298] Submit Answer (User):

[0299] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their device to the server. If the emotion engine determines that the emotion is "anxiety," detailed feedback is provided.

[0300] Receiving and analyzing response data (server):

[0301] The server receives the submitted answer data and categorizes it into multiple choice and essay questions.

[0302] Automatic grading (server):

[0303] The server uses AI models to match multiple-choice questions with correct answers and evaluate essay questions, using an emotion engine to tailor feedback.

[0304] Distribution (server):

[0305] The grading results are automatically tallied and sent to the teacher's device.

[0306] Example of creating club activity guidelines

[0307] Request Submission (User):

[0308] An inexperienced soccer club advisor requests a "one-year coaching plan for a junior high school soccer club." If the emotion engine determines that the emotion is "lack of confidence," the coaching guidelines are supplemented with more detailed information.

[0309] Receiving and processing the request (server):

[0310] The server receives this request, analyzes the content, and evaluates the emotional state using an emotion engine.

[0311] Course guide generation (server):

[0312] The server uses the trained AI model to generate a one-year coaching plan for a typical junior high school soccer club, and optimizes the coaching content by reflecting the results of the emotion engine.

[0313] Distribution (server):

[0314] The generated teaching guidelines are converted into PDF format and sent to the teacher's device.

[0315] This will significantly reduce the workload of teachers and allow them to receive high-quality educational support that responds to their emotions.

[0316] The processing flow will be explained below.

[0317] Specific flow of program processing

[0318] Quiz creation process

[0319] Step 1:

[0320] The user (teacher) sends a request to create a quiz from their device. The request includes detailed parameters such as grade, subject, topic, and number of questions. At the same time as sending the request, the device's camera and microphone are used to collect the user's facial expressions and voice to capture their emotions.

[0321] Step 2:

[0322] The server receives requests from users and analyzes their content. During the analysis, the emotion engine evaluates the user's emotional state. For example, if the user is feeling stressed, the parameters are adjusted to reflect that.

[0323] Step 3:

[0324] The server inputs the extracted parameters and the results of the emotion engine into a trained generative artificial intelligence model.

[0325] Step 4:

[0326] The generative AI model automatically generates quizzes based on input parameters and emotional assessments, adjusting the difficulty and format of the questions depending on the emotional state.

[0327] Step 5:

[0328] The server converts the generated quiz into PDF or HTML format.

[0329] Step 6:

[0330] The server delivers the converted quiz to the user's device, where the user can review the quiz and distribute it to students by printing or online.

[0331] Auto-grading process

[0332] Step 1:

[0333] The user (teacher) sends the students' answers from their terminal to the server. The answers include answers to multiple-choice questions and answers to written questions.

[0334] Step 2:

[0335] The server receives the answer data sent by the user and classifies them into multiple choice and written answers. At the same time, the emotion engine evaluates the teacher's stress and anxiety.

[0336] Step 3:

[0337] The server automatically scores multiple-choice questions by comparing answers with correct answer data, and the emotion engine's evaluation adjusts the detailed grading process to a minimum for teachers who are particularly stressed.

[0338] Step 4:

[0339] The server inputs the answers to the essay questions into a trained generative AI model for evaluation. The model automatically calculates a score based on the evaluation. The detail and tone of the feedback are adjusted based on the evaluation by the emotion engine.

[0340] Step 5:

[0341] The server aggregates the multiple choice and essay grading results and generates a final score.

[0342] Step 6:

[0343] The server delivers the aggregated results to the user's device. This may include individual feedback based on the emotion engine. The user checks the results and provides feedback to the student.

[0344] Creation of club activity guidelines

[0345] Step 1:

[0346] The user (teacher) sends a request for club activity guidance from their device. The request includes detailed parameters such as the type of club activity, teaching goals, and duration. At the same time as sending the request, the device's camera and microphone are used to collect the user's emotions from their facial expressions and voice.

[0347] Step 2:

[0348] The server receives the user's request and analyzes its contents. This analysis includes extracting request parameters and evaluating the user's emotional state using an emotion engine. For example, if the user feels a lack of confidence, detailed instructional guidance reflecting this is provided.

[0349] Step 3:

[0350] The server inputs the extracted parameters and emotion evaluations into a trained generative artificial intelligence model.

[0351] Step 4:

[0352] The generative AI model automatically generates standard club activity instruction guidelines based on input parameters and emotional evaluations, adjusting instruction content and supplementary information according to the emotional state.

[0353] Step 5:

[0354] The server converts the generated teaching guidelines into PDF or HTML format.

[0355] Step 6:

[0356] The server delivers the converted teaching guidelines to the user's terminal, where the user can check and apply the teaching guidelines.

[0357] Specific examples

[0358] Example of creating a quiz:

[0359] If a teacher sends a request to "create a 10-question quiz on linear equations in second-year junior high school mathematics" and the emotion is recognized as "stress," the server will suggest simplified problem sets and recommended problem templates, and also adjust the difficulty of the problems appropriately.

[0360] Examples of automatic grading:

[0361] When a teacher submits a student's answer and the emotion is recognized as "anxiety," the server provides detailed feedback and attaches special explanations to some of the more difficult answers.

[0362] Examples of club activity guidelines:

[0363] If a coaching teacher who requests a soccer club coaching plan is identified as lacking confidence, the server will provide a standardized teaching guideline including detailed coaching content, the purpose of each practice, points to note, etc. It will also provide encouraging messages and additional reference materials according to the coaching situation.

[0364] This will significantly reduce the workload of teachers and allow them to receive high-quality educational support that responds to their emotions.

[0365] Example 2

[0366] 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."

[0367] Teachers and educational institutions have a wide range of tasks to perform, such as creating tests, grading, and creating club activity guidelines, which require a great deal of time and effort. Furthermore, the heavy workload can lead to insufficient feedback on the quality of education and instruction. To solve these problems, an efficient educational support system is needed.

[0368] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0369] In this invention, the server includes means for collecting and storing past document data and information, means for preprocessing the collected data and converting it into a unified format, means for training a generation algorithm using the preprocessed data, means for receiving a data generation request from a user, means for generating data based on the request using the trained generation algorithm, means for recognizing the user's emotions and optimizing the request content, and means for delivering the generated data to the user's terminal. This significantly reduces the workload of teachers and enables high-quality educational support that responds to emotions.

[0370] "Past document data and information" refers to documents and related information such as test questions, entrance examination information, curriculum guidelines, and club activity instruction manuals that have been created and used by educational institutions and teachers in the past.

[0371] "Means of collection" refers to the function of obtaining data provided by educational institutions and teachers via the Internet or a dedicated interface and storing it in the system.

[0372] "Means of storage" refers to the databases and storage within the system used to properly manage and store collected data.

[0373] "Preprocessing means" refers to processes such as data cleaning, noise removal, and normalization that are performed to convert the collected data into an appropriate format.

[0374] The "means of converting to a unified format" refers to an operation for converting data stored in different formats into a consistent format so that it can be used for subsequent processing.

[0375] A "generative algorithm" is a machine learning model or AI used to automatically generate new test questions, curriculum guidelines, etc. using preprocessed data as input.

[0376] "Training" refers to the machine learning training process that uses collected data to improve the performance of the generative algorithm.

[0377] "Means for receiving data generation requests from users" refers to a mechanism for receiving requests from users via a request form, API, etc., and using them as the starting point for processing.

[0378] "Means for recognizing emotions and optimizing request content" is a function that analyzes the emotional state of the user from their input and behavior, and adjusts the system's response and the content of the output based on the results.

[0379] "Means for delivering generated data to the user's terminal" refers to a mechanism for transmitting generated quizzes, marking results, teaching guidelines, etc. to the user's terminal in an appropriate format.

[0380] This invention utilizes a generative AI model and an emotion engine to reduce the workload of teachers in educational institutions. Specifically, by learning from past test questions, entrance exam information, and teaching guidelines, the system automatically creates quizzes, automates grading, and automatically generates teaching guidelines for club activities. By combining this with an emotion engine that recognizes the user's emotions, the system improves the quality of educational resources and feedback.

[0381] System configuration and operation

[0382] 1. Data Collection Methods

[0383] The server collects data provided by teachers and educational institutions, such as past test questions, entrance exam information, teaching guidelines, and club activity instruction manuals. This data is stored in the system's database. Specifically, the data is collected using an FTP server and API and stored in the system's database.

[0384] 2. Data preprocessing methods

[0385] The server cleans the collected data to standardize it into an appropriate format. For example, it removes noise from the text data and normalizes it. It uses a text analysis tool (e.g., a Python text processing library) to remove meaningless strings and noise and standardize the data format.

[0386] 3. Generative AI Model

[0387] The server uses the preprocessed data to train a generative AI model, which is specialized for automating education-related tasks. Specifically, the training data is used as input for the AI ​​model, and the model parameters are optimized. Training is performed using a machine learning framework (e.g., TensorFlow or PyTorch).

[0388] 4. User Request Reception Method

[0389] Teachers use their terminals to request the creation and grading of quizzes and the generation of club activity guidelines. These requests are sent to the server and analyzed. The user interface is provided through a web browser, and the request contents are sent as HTTP requests.

[0390] 5. Emotion Engine

[0391] The server has an emotion engine that recognizes emotions when a user sends a request. Specifically, it analyzes emotions from the user's input text and voice. The emotion engine uses natural language processing (NLP) algorithms (e.g., BERT) to assign emotion tags (e.g., stress, anxiety, lack of confidence).

[0392] 6. Test Generation Methods

[0393] Based on user requests, the server automatically generates quizzes using a generative AI model. The results of the emotion engine are reflected and the test content is optimized. As a specific example, a prompt such as "junior high school second-year mathematics, linear equations, 10-question quiz" is entered to generate test questions. The difficulty and question format are adjusted according to the evaluation results of the emotion engine.

[0394] Example prompt sentence:

[0395] "Second year junior high school mathematics, linear equations, 10-question quiz"

[0396] 7. Scoring Method

[0397] The server receives the student's response data sent by the user and grades it using a generative AI model. It also provides feedback using an emotion engine. For example, it inputs the received response data into the AI ​​model, compares multiple-choice questions with correct answer data, and evaluates essay questions. It then aggregates the grades and adjusts the feedback based on the emotion tags generated by the emotion engine.

[0398] 8. Curriculum guideline generation means

[0399] The server uses a generative AI model to automatically generate club activity teaching guidelines based on the user's requests. Optimization is performed by reflecting the results of the emotion engine. As a specific example, a "one-year teaching plan for a junior high school soccer club" is input as a prompt, and a standard teaching guideline is generated. Specific activity content and points to note are added based on the evaluation results of the emotion engine.

[0400] Example prompt sentence:

[0401] "One-year coaching plan for junior high school soccer club"

[0402] 9. Distribution Method

[0403] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's device. Specifically, the server converts the generated content into PDF format using a PDF generation tool (e.g., ReportLab), and then sends it to the user's device using the HTTP protocol.

[0404] This system will significantly reduce the workload of teachers and provide them with high-quality educational support that responds to their emotions.

[0405] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0406] Step 1:

[0407] The server collects and stores past document data and information provided by teachers and educational institutions. Specifically, it retrieves data using an FTP server or API and stores it in the system's database. Input data includes past test questions, entrance exam information, curriculum guidelines, and club activity instruction manuals, and the output is structured data stored in the database.

[0408] Step 2:

[0409] The server preprocesses the collected data and converts it into a unified format. It uses text analysis tools (e.g., Python text processing libraries) to remove meaningless strings and noise and standardize the data format. The input is raw data from the database, and the output is formatted data that has been cleaned and normalized. Specific operations include removing unnecessary strings, standardizing character codes, and normalizing the text.

[0410] Step 3:

[0411] The server uses the preprocessed data to train a generative AI model. Using a machine learning framework (e.g., TensorFlow or PyTorch), it provides the training data as input to the AI ​​model and optimizes the model's parameters. The input is the preprocessed data, and the output is a trained generative AI model. Specific operations include splitting the dataset, building a model, running training, and evaluating and refining the model.

[0412] Step 4:

[0413] A user uses a terminal to request the creation and grading of quizzes and the generation of club activity curriculum guidelines. The user interface is provided through a web browser, and the request contents are sent to the server as an HTTP request. The input is the user's request data (the scope of the quiz, the number of questions, etc.), and the output is the HTTP request sent to the server. Specific actions include filling out the request form and clicking the submit button.

[0414] Step 5:

[0415] The server uses an emotion engine to recognize emotions when a user submits a request. It analyzes emotions from the user's input text or voice and generates emotion tags. It uses an NLP algorithm (e.g., BERT) to assign emotion tags (e.g., stress, anxiety, lack of confidence). The input is the user's request data, and the output is the emotion-tagged request data. Specific operations include performing text analysis, speech recognition, and emotion analysis.

[0416] Step 6:

[0417] The server uses a trained generative AI model to automatically generate quizzes. For example, a prompt such as "Junior high school second-year mathematics, linear equations, 10-question quiz" is input and test questions are generated. The difficulty and question format are adjusted according to the emotion tag. The input is the emotion-tagged request data and the prompt, and the output is an automatically generated quiz. Specific operations include inputting the prompt into the generative AI model and adjusting the format of the generated test.

[0418] Step 7:

[0419] The server receives the students' response data sent by the user and automatically grades them using the generative AI model. The input is the response data (multiple-choice questions and essay questions), and the output is the grading results and feedback. Specific operations include receiving the response data, checking the correct answers for the multiple-choice questions, evaluating the essay questions, and generating feedback.

[0420] Step 8:

[0421] The server uses a generative AI model to automatically generate club activity guidelines based on the user's request. For example, a "one-year teaching plan for a junior high school soccer club" is input as a prompt, and a standard teaching guideline is generated. Specific activity content and important points are added according to the emotion tag. The input is the emotion-tagged request data and the prompt text, and the output is the generated teaching guideline. Specific operations include inputting the prompt text into the generative AI model and adjusting the content of the generated teaching guideline.

[0422] Step 9:

[0423] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's device. The content is converted using a PDF generation tool (e.g., ReportLab) and then sent to the user's device using the HTTP protocol. The input is the generated content, and the output is the data delivered to the user's device. Specific operations include PDF generation and HTTP communication execution.

[0424] (Application example 2)

[0425] 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."

[0426] Currently, elementary and junior high school teachers are burdened with a wide range of tasks, including creating and grading quizzes and creating curriculum guidelines for club activities. There is also a lack of support that takes into account teachers' emotions and stress levels, raising concerns that this could lead to a decline in the quality of education. Furthermore, there are currently few tools available to support teachers in real time in the classroom. These issues need to be addressed.

[0427] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting and storing past test questions, entrance examination information, and teaching guidelines, means for preprocessing the collected data and converting it into a unified format, means for training a generative AI model using the preprocessed data, means for receiving a request from a user and recognizing the user's emotions, means for using the generative AI model to generate a quiz, test score, or teaching guide based on the request and optimizing it based on the recognized emotions, means for delivering the generated test, score results, or teaching guide to the user's terminal, and means for providing real-time support to the user in a virtual space. This significantly reduces the workload of teachers and enables them to receive high-quality educational support that responds to their emotions.

[0428] "Past test questions" refer to test questions previously used by an educational institution.

[0429] "Entrance exam information" refers to all data related to school and university entrance exams, including exam dates, exam topics, and past passing scores.

[0430] "Course of instruction" refers to a document that describes the content and procedures for lessons and activities that an educational institution will provide to students.

[0431] "Data preprocessing" refers to processing collected raw data to convert it into a format suitable for analysis and model generation.

[0432] A "uniform format" refers to a format for expressing data in multiple different formats according to consistent rules.

[0433] A "generative artificial intelligence model" refers to an AI system that has been trained using machine learning or deep learning techniques.

[0434] "Means for recognizing user emotions" refers to technology that detects emotions from the user's voice, facial expressions, and writing, and provides appropriate feedback based on that.

[0435] "Virtual space" refers to a virtual environment that is different from the real world and is generated using computer graphics and virtual reality technology.

[0436] "Means for providing support in real time" refers to technologies and systems that provide appropriate assistance immediately according to the user's situation, rather than content prepared in advance.

[0437] "Generated Test" refers to a test question newly created by a generative artificial intelligence model based on a user request.

[0438] "Scoring results" refers to the data compiled after evaluating the test scores given by students.

[0439] "Course generation" refers to the creation of new lesson plans using generative artificial intelligence models for specific educational tasks or activities.

[0440] System Overview

[0441] This system provides real-time support in a virtual space, reducing teachers' workload and providing high-quality educational support. The system collects past test questions, entrance exam information, and teaching guidelines, and uses a generative artificial intelligence model to automatically create quizzes, grade them, and generate teaching guidelines. Furthermore, it utilizes an emotion engine to provide optimized educational resources and feedback based on the user's emotions.

[0442] System configuration

[0443] 1. Data collection methods:

[0444] The server collects and stores past test questions, entrance exam information, and curriculum guidelines from educational institutions. This data is stored in various formats, including CSV and PDF.

[0445] 2. Data preprocessing methods:

[0446] The server preprocesses the collected raw data, for example using the Python Pandas library to cleanse the data and convert it into a unified format.

[0447] 3. Training the generative AI model:

[0448] The server uses the preprocessed data to train a generative artificial intelligence model using TensorFlow, which specializes in creating and grading quizzes and generating lesson plans.

[0449] 4. Receiving user requests and emotion recognition:

[0450] A teacher using a device wears a head-mounted display and sends a request to "Create a short math quiz for second-year junior high school students" in the virtual space. The server uses an emotion engine to recognize emotions from the user's voice and facial expressions. If the emotion is determined to be "stress," the request is simplified.

[0451] 5. Test generation methods and sentiment optimization:

[0452] The server uses a trained generative AI model to generate quizzes based on requests, such as creating 10 questions about linear equations in second-grade math, and adjusts the difficulty of the questions based on the results of the emotion engine.

[0453] 6. Automated scoring and feedback:

[0454] The teacher sends the student's answers (10 multiple-choice questions and 5 written questions) to the server. The server receives this data and automatically scores it using a generative AI model. If the emotion engine determines that the student is "anxious," it provides detailed feedback.

[0455] 7. Means of Delivery of Product:

[0456] The generated quizzes, marking results, and teaching guidelines are converted into PDF format and sent to the device.

[0457] Examples of concrete examples and prompts

[0458] Specific examples of quizzes

[0459] Request Submission (User):

[0460] Send the following prompt from the device: "Please create 10 short quizzes in mathematics for second-year junior high school students. Please simplify the content as teachers are feeling stressed. For example, please include a lot of basic questions on linear equations."

[0461] Specific examples of scoring

[0462] Submit Answer (User):

[0463] A teacher sends the following prompt: "You are grading a 8th grade math quiz. There are 10 multiple choice questions and 5 essay questions. The teacher is concerned, so please provide detailed explanations. For example, if the answers are incorrect, please explain in detail the correct way to solve the problem and the reason why."

[0464] Main technologies used

[0465] Hardware:

[0466] Head-mounted displays (e.g., Oculus Rift, HTC Vive)

[0467] High-Performance Server

[0468] software:

[0469] Data preprocessing: Python + Pandas

[0470] AI model generation: TensorFlow

[0471] Emotion engine: OpenCV + TensorFlow

[0472] This will significantly reduce the workload of teachers and allow them to receive high-quality educational support that responds to their emotions.

[0473] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0474] Step 1: Data collection

[0475] The server collects past test questions, entrance exam information, and curriculum guidelines from educational institutions and stores them in a database. This data is provided in various formats, including CSV and PDF.

[0476] Input: Test questions, entrance exam information, and curriculum guidelines provided by educational institutions

[0477] Output: Saved test questions, entrance exam information, and curriculum data

[0478] Step 2: Data Preprocessing

[0479] The server preprocesses the collected raw data, for example using Python's Pandas library to cleanse the data and convert it into a unified format.

[0480] Input: Stored raw data (test questions, entrance exam information, curriculum guidelines)

[0481] Output: Cleansed and unified data

[0482] Step 3: Training the AI ​​model

[0483] The server uses the preprocessed data to train a generative artificial intelligence model using TensorFlow, which specializes in creating quizzes, grading, and generating lesson plans.

[0484] Input: Data converted into a unified format

[0485] Output: A trained generative AI model

[0486] Step 4: Receiving user requests and recognizing emotions

[0487] A teacher using a device wears a head-mounted display and sends a request to the virtual space, such as "Create a short math quiz for second-year junior high school students." The server receives this request and uses an emotion engine to recognize emotions from the user's voice and facial expressions.

[0488] Input: Request submitted by teacher (e.g., "Create a quiz for 8th grade math")

[0489] Output: Optimized request when emotion is recognized as "stress"

[0490] Step 5: Quiz generation and sentiment optimization

[0491] The server uses a trained generative AI model to generate quizzes based on user requests, adjusting the difficulty of the quiz based on the results of the emotion engine.

[0492] Input: Optimized request (e.g., "8th grade math quiz with basic linear equations")

[0493] Output: The generated quiz (e.g. 10 questions on linear equations)

[0494] Step 6: Sending and receiving response data

[0495] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their device to the server. The server receives this data and proceeds to the next step.

[0496] Input: Student response data (multiple choice, written)

[0497] Output: Answer data saved on the server

[0498] Step 7: Automated scoring and feedback

[0499] The server automatically grades the saved answer data using a generative AI model, and uses an emotion engine to recognize the teacher's emotions and provide detailed feedback.

[0500] Input: Student response data, Teacher sentiment

[0501] Output: Marking results and detailed feedback

[0502] Step 8: Delivering results

[0503] The server converts the generated quizzes, marking results, and teaching guidelines into PDF format and sends them to the teacher's device.

[0504] Input: Generated quizzes, grades, and teaching guidelines

[0505] Output: PDF file delivered to the teacher's device

[0506] This details the inputs delivered at each processing step, the data manipulations or operations performed, and the final results achieved.

[0507] 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.

[0508] 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.

[0509] 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.

[0510] [Second embodiment]

[0511] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0512] 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.

[0513] 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).

[0514] 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.

[0515] 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.

[0516] 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).

[0517] 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.

[0518] 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.

[0519] 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.

[0520] 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.

[0521] 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.

[0522] 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."

[0523] System Overview

[0524] This invention is a system that uses generative artificial intelligence (AI) to reduce the workload of elementary and junior high school teachers. Specifically, by learning from past test questions, entrance exam information, and teaching guidelines, it automatically creates quizzes, automates grading, and generates teaching guidelines for club activities.

[0525] System configuration

[0526] 1. Data collection methods:

[0527] The server collects data provided by teachers and educational institutions, such as past test questions, entrance exam information, teaching guidelines, and club activity instruction manuals. This data is stored in the system's database.

[0528] 2. Data preprocessing methods:

[0529] The server cleans the collected data to standardize it into an appropriate format, removing noise from the text data and normalizing it, making it usable as education-related data.

[0530] 3. Generative AI Models:

[0531] The server uses the preprocessed data to train generative artificial intelligence (AI) models that are specialized to perform education-related tasks (quiz creation, grading, curriculum generation).

[0532] 4. User request receiving means:

[0533] Teachers using the devices can request the creation and grading of quizzes and the generation of club activity guidelines. The requests are sent to the server and analyzed.

[0534] 5. Test generation methods:

[0535] Based on user requests, the server automatically generates quizzes using AI models, which are structured according to the user's specified subject matter and number of questions.

[0536] 6. Scoring Procedure:

[0537] The server receives the student's response data sent by the user and automatically scores multiple-choice and essay questions using an AI model. The results are automatically compiled and feedback is provided.

[0538] 7. Curriculum syllabus generation means:

[0539] For club activity teaching guidelines, the server uses a trained AI model to generate standard teaching guidelines based on user requests, and these guidelines are provided in a format that is easy to use even for inexperienced teachers.

[0540] 8. Delivery Method:

[0541] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's terminal.

[0542] Program processing and specific examples

[0543] Example of creating a quiz

[0544] Request Submission (User):

[0545] A teacher sends a request from a terminal saying, "Create a 10-question quiz on linear equations in second-year junior high school mathematics."

[0546] Receiving and processing the request (server):

[0547] The server receives this request and analyzes the contents.

[0548] Quiz generation (server):

[0549] The server uses a trained AI model to generate quizzes based on the specified parameters.

[0550] Distribution (server):

[0551] The generated quiz is converted into PDF format and sent to the teacher's device.

[0552] Specific examples of automatic scoring

[0553] Submit Answer (User):

[0554] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their terminal to the server.

[0555] Receiving and analyzing response data (server):

[0556] The server receives the submitted answer data and categorizes it into multiple choice and essay questions.

[0557] Automatic grading (server):

[0558] The server uses an AI model to match multiple-choice questions with correct answer data and evaluate essay questions.

[0559] Distribution (server):

[0560] The grading results are automatically tallied and sent to the teacher's device.

[0561] Example of creating club activity guidelines

[0562] Request Submission (User):

[0563] An inexperienced soccer club advisor requests a "one-year coaching plan for a junior high school soccer club."

[0564] Receiving and processing the request (server):

[0565] The server receives this request and analyzes the contents.

[0566] Course guide generation (server):

[0567] The server uses a trained AI model to generate a one-year coaching plan for a typical junior high school soccer club.

[0568] Distribution (server):

[0569] The generated teaching guidelines are converted into PDF format and sent to the teacher's device.

[0570] The above system can reduce the workload of teachers and provide efficient educational support.

[0571] The processing flow will be explained below.

[0572] Program processing flow

[0573] Quiz creation process

[0574] Step 1:

[0575] A user (teacher) sends a request to create a quiz from a terminal, which includes detailed parameters such as grade, subject, topic, and number of questions.

[0576] Step 2:

[0577] The server receives the request from the user and analyzes its contents, which includes extracting the request parameters.

[0578] Step 3:

[0579] The server inputs the extracted parameters into a trained generative artificial intelligence model.

[0580] Step 4:

[0581] The generative AI model automatically generates quizzes based on input parameters, with questions tailored to the specified grade, subject, and topic.

[0582] Step 5:

[0583] The server converts the generated quiz into PDF or HTML format.

[0584] Step 6:

[0585] The server delivers the converted quiz to the user's device, where the user can review the quiz and distribute it to students by printing or online.

[0586] Auto-grading process

[0587] Step 1:

[0588] The user (teacher) sends the students' answers from their terminal to the server. The answers include answers to multiple-choice questions and answers to written questions.

[0589] Step 2:

[0590] The server receives the answer data sent by the user and classifies the answer into multiple choice and written answer.

[0591] Step 3:

[0592] The server compares the answers to the multiple-choice questions with the correct answer data and performs automatic grading.

[0593] Step 4:

[0594] The server inputs the answers to the essay questions into a trained generative AI model for evaluation, and the model automatically calculates a score based on this evaluation.

[0595] Step 5:

[0596] The server aggregates the multiple choice and essay grading results and generates a final score.

[0597] Step 6:

[0598] The server delivers the results to the user's device, where the user can check the results and provide feedback to the student.

[0599] Creation of club activity guidelines

[0600] Step 1:

[0601] The user (teacher) sends a request for club activity guidance from a terminal. The request includes detailed parameters such as the type of club activity, teaching goals, and duration.

[0602] Step 2:

[0603] The server receives the request from the user and analyzes its contents, which includes extracting the request parameters.

[0604] Step 3:

[0605] The server inputs the extracted parameters into a trained generative artificial intelligence model.

[0606] Step 4:

[0607] The generative AI model automatically generates standard club activity teaching guidelines based on input parameters. The generated teaching guidelines are appropriate for the type of club activity and teaching goals.

[0608] Step 5:

[0609] The server converts the generated teaching guidelines into PDF or HTML format.

[0610] Step 6:

[0611] The server delivers the converted teaching guidelines to the user's terminal, where the user can check and apply the teaching guidelines.

[0612] This will enable teachers to significantly improve their work efficiency and reduce the burden on the educational field.

[0613] Example 1

[0614] 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."

[0615] In the traditional education system, teachers have a heavy workload, especially when it comes to creating and grading quizzes and planning extracurricular activities. Furthermore, existing tools for streamlining these tasks are limited and lack sufficient flexibility and precision. This makes it difficult for teachers to devote sufficient time to their primary educational activities.

[0616] 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.

[0617] In this invention, the server includes: means for collecting and storing past assessment materials, entrance examination information, educational guidelines, and extracurricular activity instruction materials from educational institutions; means for preprocessing the collected materials and converting them into a unified format; means for training an AI model using the preprocessed materials; means for receiving requests from users to create exams, grade evaluations, and extracurricular activity guidelines; means for generating exams, grade evaluations, and extracurricular activity guidelines based on the requests using the trained AI model; means for delivering the generated exams, grade evaluations, or extracurricular activity guidelines to the user's terminal; means for analyzing the request content and generating prompts appropriate for the AI ​​model; and means for feeding the AI ​​model based on the analyzed request to generate optimal output. This reduces the workload of teachers and enables efficient support of educational activities.

[0618] An "educational institution" is a formal institution that provides teaching and learning, such as a school, university, college, or other institution.

[0619] "Assessment materials" are materials used to assess student learning outcomes, such as previously administered tests, exams, or quizzes.

[0620] "Entrance Examination Information" means data and information relating to examinations for admission to a particular school or educational institution.

[0621] "Educational guidelines" are guidelines or instructional documents for achieving a particular educational curriculum or learning objectives.

[0622] "Extracurricular activity instruction materials" are manuals and plans for instructing extracurricular activities, such as club activities and extracurricular club activities.

[0623] "Preprocessing" is the process of preparing collected data in a format that is easy to analyze and use, and performing noise removal and normalization.

[0624] A "generative artificial intelligence model" is an AI system that builds knowledge to perform education-related tasks from training data.

[0625] "Training" is the process by which an artificial intelligence model learns from specified data and improves its ability to solve specified problems.

[0626] A "request" is an instruction or command from a user to a system requesting a specific process or operation.

[0627] "Exam" refers to a test or quiz administered to assess a student's understanding or knowledge.

[0628] "Grading" is the process of assessing a student's achievements on a test or quiz and expressing that performance in terms of a number or grade.

[0629] The "Extracurricular Activities Guidelines" are detailed procedures and guidelines for guiding the planning and conduct of club and extracurricular activities.

[0630] A "prompt sentence" is an input sentence that is used to obtain a specific answer or product from the AI ​​model being generated.

[0631] "Feeding" is the process of inputting learning data and request content into an artificial intelligence model to obtain optimal output.

[0632] This invention is a system that utilizes generative artificial intelligence (AI) to reduce the workload of elementary and junior high school teachers. Specifically, the system automatically creates quizzes, automates grading, and generates club activity teaching guidelines by learning from past test questions, entrance exam information, educational guidelines, and club activity teaching materials.

[0633] System configuration

[0634] Data collection methods

[0635] The server automatically collects past assessment materials, entrance examination information, educational guidelines, and extracurricular activity guidance materials from educational institutions using APIs and FTP servers. The collected data is stored in a database.

[0636] Data preprocessing measures

[0637] The server preprocesses the collected data using text analysis tools (e.g., SpaCy, NLTK), removing noise from the text data, normalizing it, and converting it into a unified format so that it can be used as training data.

[0638] Generative AI Model

[0639] The server uses the preprocessed data to train an AI model (e.g., GPT-3, BERT), and uses machine learning libraries (e.g., TensorFlow, PyTorch) to build a model specialized for a given education-related task (quiz creation, grading, extracurricular activity curriculum generation).

[0640] User request receiving means

[0641] Teachers using the devices can use a dedicated application to request the creation and grading of quizzes and the generation of club activity guidelines. These requests are sent to the server in real time.

[0642] Test Generation Methods

[0643] The server uses a trained AI model to automatically generate quizzes based on user requests. It analyzes the request content and generates prompts appropriate for the AI ​​model. The generated quiz is converted into PDF or HTML format and sent to the teacher's device.

[0644] Scoring method

[0645] The server receives the student's response data sent by the user and automatically grades the multiple-choice and essay questions using an AI model. Multiple-choice questions are compared with the correct answer data, and essay questions are evaluated using NLP technology. The grading results are compiled and sent as feedback to the teacher's device.

[0646] Curriculum guideline generation means

[0647] The server uses a trained AI model to generate extracurricular activity guidelines based on the user's request. It analyzes the request and generates a prompt appropriate for the AI ​​model. The generated guidelines are converted into PDF format and sent to the teacher's device.

[0648] Data Delivery Methods

[0649] The server converts the generated quizzes, marking results, or extracurricular activity guidelines into PDF files or HTML format and delivers them to the teacher's terminal.

[0650] Specific example explanation

[0651] Example of creating a quiz

[0652] Request submission (user)

[0653] The teacher sends a request from the terminal saying, "Create a 10-question quiz on linear equations in second-year junior high school mathematics."

[0654] Receiving and processing requests (server)

[0655] The server receives this request in real time and analyzes the content.

[0656] Quiz generation (server)

[0657] The server uses a trained AI model to generate quizzes based on the specified parameters.

[0658] Distribution (server)

[0659] The generated quiz is converted into PDF format and sent to the teacher's device.

[0660] Specific examples of automatic scoring

[0661] Submit Answer (User)

[0662] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their terminal to the server.

[0663] Receiving and analyzing response data (server)

[0664] The server receives the submitted response data in real time and categorizes it into multiple choice and essay questions.

[0665] Automatic scoring (server)

[0666] The server uses an AI model to match multiple-choice questions with correct answer data and evaluate essay questions.

[0667] Distribution (server)

[0668] The grading results are automatically tallied and sent to the teacher's device.

[0669] Example of creating club activity guidelines

[0670] Request submission (user)

[0671] An inexperienced soccer club advisor requests a "one-year coaching plan for a junior high school soccer club."

[0672] Receiving and processing requests (server)

[0673] The server receives this request in real time and analyzes the content.

[0674] Teaching guideline generation (server)

[0675] The server uses a trained AI model to generate a one-year coaching plan for a typical junior high school soccer club.

[0676] Distribution (server)

[0677] The generated teaching guidelines are converted into PDF format and sent to the teacher's device.

[0678] This system will reduce the workload of teachers and provide efficient, high-quality educational support.

[0679] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0680] Step 1: Collect data

[0681] The server collects past evaluation materials, entrance examination information, educational guidelines, and extracurricular activity guidance materials from educational institutions and stores them in a database. Specifically, the server obtains data through APIs and FTP servers. The collected materials are stored in the database. Various educational materials are input, and educational materials in a unified data format are generated as output.

[0682] Step 2: Preprocessing the data

[0683] The server preprocesses the collected data. This involves using text analysis tools (e.g., SpaCy, NLTK) to remove noise and normalize the data. The input is the collected educational materials, and the output is preprocessed, cleaned data. Specifically, special characters are removed and the data is formatted.

[0684] Step 3: Training the AI ​​model

[0685] The server uses the preprocessed data to train a generative AI model (e.g., GPT-3, BERT). Machine learning libraries used include TensorFlow and PyTorch. The input is the preprocessed data, and the output is a trained AI model. Specifically, the data is fed into the AI ​​model, and iterative learning occurs to improve the model's accuracy.

[0686] Step 4: Receiving the request

[0687] Teachers using the devices make requests from their devices to create and grade quizzes and generate curriculum guidelines for club activities. These requests are sent to the server. The input is the teacher's request, and the output is the request content saved on the server. Specifically, the teacher inputs the request into a dedicated application, which is then sent to the server.

[0688] Step 5: Parsing the request and generating a prompt

[0689] The server analyzes the request content and generates a prompt sentence appropriate for the generative AI model. The input is the request received from the teacher, and the output is the generated prompt sentence. Specifically, the server analyzes the request content and generates a prompt sentence in an appropriate format.

[0690] Step 6: Generate the quiz

[0691] The server uses a trained AI model to automatically generate quizzes based on prompts. The input is the prompt, and the output is the generated quiz. Specifically, the generated prompt is fed into the AI ​​model to generate quiz questions.

[0692] Step 7: Convert and distribute the quiz

[0693] The generated quiz is converted into PDF or HTML format and sent to the teacher's device. The input is the generated quiz, and the output is a file converted into PDF or HTML format. Specifically, the test questions are converted into a document format and distributed to the device.

[0694] Step 8: Receiving and parsing responses

[0695] The teacher sends the students' answer data (multiple choice and written questions) from their terminal to the server. The server receives this data and classifies it into multiple choice and written questions. The input is the students' answer data, and the output is the classified answer data. Specifically, the data is classified.

[0696] Step 9: Automated scoring

[0697] The server uses an AI model to automatically score the received student answers. Multiple-choice questions are compared with correct answer data, and essay questions are evaluated using NLP technology. The input is classified answer data, and the output is the scoring results. Specifically, multiple-choice questions are automatically matched, and essay questions are evaluated using semantic analysis.

[0698] Step 10: Calculation and distribution of scores

[0699] The grading results are automatically compiled and sent to the teacher's device as feedback. The input is the grading results, and the output is the compiled feedback. Specifically, the grading results are compiled, and feedback is generated and distributed.

[0700] Step 11: Generate and distribute curriculum guidelines

[0701] The server uses a trained AI model to generate standard teaching guidelines based on the user's request. The generated teaching guidelines are converted into PDF format and sent to the teacher's device. The input is the prompt text and the AI ​​model, and the output is the teaching guidelines in PDF format. Specifically, the server analyzes the request content, generates the teaching guidelines, converts them into PDF, and distributes them.

[0702] (Application example 1)

[0703] 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."

[0704] In-factory worker education and training takes a great deal of time and effort, and if there is a shortage of skilled instructors, it can be difficult to improve worker skills and ensure safety. In addition, it can be difficult to provide consistent teaching materials and manage training progress, which can lead to inconsistent quality of education.

[0705] 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.

[0706] In this invention, the server includes a means for collecting and storing past training data, operation information, safety guidelines, and work manuals, a means for preprocessing the collected data and converting it into a unified format, and a means for training a generative artificial intelligence model using the preprocessed data, thereby enabling training materials, evaluation results, and troubleshooting tests to be automatically generated and distributed to terminals based on user requests.

[0707] "Training data" refers to past data collected for the purpose of educating and training workers.

[0708] "Operation information" refers to information about specific operating procedures and methods for machines and equipment.

[0709] "Safety guidelines" are guidelines and rules for ensuring safety during work and operations.

[0710] A "work manual" is a document or instruction manual that shows specific tasks and procedures.

[0711] "Preprocessing" refers to the process of organizing and cleaning raw data to convert it into a usable format.

[0712] A "uniform format" is a standard format for converting different types of data into a consistent structure.

[0713] "Training means" refers to the data and methods used to train an AI model.

[0714] A "request" means that a user requests something from the system.

[0715] "Training materials" are materials created for workers to learn and improve their skills.

[0716] "Evaluation data" is data based on the activities and responses performed by the worker.

[0717] "Troubleshooting tests" are tests to identify and resolve problems or faults.

[0718] "Automatic generation" means that the system autonomously creates data and information without human intervention.

[0719] A "terminal" is a device used to access the system.

[0720] This invention provides a system for optimizing worker education and training in a factory and improving work efficiency. This system automatically generates and evaluates training materials for worker education and provides troubleshooting tests through the following processing steps.

[0721] System Overview

[0722] The system uses smartphones, smart glasses, and head-mounted displays (HMDs) as hardware, and Flask (a Python microweb framework) and GPT-2 (a generative AI model) as software.

[0723] Program processing

[0724] Data collection and storage

[0725] The server collects and stores past training data, operating information, safety guidelines, and work manuals from databases within the factory and external sources.

[0726] Data preprocessing and conversion to a unified format

[0727] The server performs preprocessing such as noise removal and normalization on the collected data, and converts it into an appropriate unified format, making it possible to process various types of data in a consistent format.

[0728] Training generative artificial intelligence models

[0729] Using the pre-processed data, the server trains generative artificial intelligence (AI) models that are specialized to perform education-related tasks within the factory (training material creation, evaluation, troubleshooting test generation).

[0730] Receiving a user request

[0731] Users use their devices (smartphones, smart glasses, HMDs) to request the creation and evaluation of training materials and the generation of troubleshooting tests. These requests are sent to the server.

[0732] Automatic generation of training materials

[0733] The server automatically generates training materials using a trained AI model based on the received user requests, converts the generated materials into PDF or HTML format, and delivers them to the user's device.

[0734] Specific examples

[0735] For example, to create training materials for new employees on how to operate a lathe, the user might enter the following prompt:

[0736] Develop training materials for new employees on lathe operation, including basic safety guidelines and detailed operating procedures.

[0737] The server receives the prompt and generates training materials using a trained generative AI model, allowing users to instantly access the training materials and efficiently train new employees in the factory.

[0738] This system is a powerful tool for quickly and efficiently educating and training workers, contributing to improved productivity and safety within factories.

[0739] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0740] Step 1:

[0741] The server collects and stores past training data, operation information, safety guidelines, and work manuals from databases within the factory and external sources. This allows the various data required for training to be collected and stored within the system. The input is various materials such as training data and operation information, and the output is converted into a unified format and stored in the database.

[0742] Step 2:

[0743] The server performs preprocessing on the collected data, such as noise removal and normalization, and converts it into an appropriate unified format. This allows various data to be processed in a consistent format. The input is the data saved in the previous step, and the output is the preprocessed data in a consistent format.

[0744] Step 3:

[0745] Using the preprocessed data, the server trains a generative artificial intelligence (AI) model that is specialized to perform education-related tasks within the factory (training material creation, evaluation, troubleshooting test generation). The input is the preprocessed data, and the output is the trained generative AI model.

[0746] Step 4:

[0747] Users use devices (smartphones, smart glasses, HMDs) to request the creation and evaluation of training materials and the generation of troubleshooting tests. These requests are sent to the server. The input is the user's request, and the output is the receipt and analysis of the request content.

[0748] Step 5:

[0749] The server automatically generates training materials using a trained AI model based on the received user request. The generated materials are converted into PDF or HTML format and delivered to the user's device. The input is the user request and the AI ​​model, and the output is the generated training materials.

[0750] For example, when creating training materials for new employees on how to operate a lathe, the user would enter the following prompt:

[0751] Develop training materials for new employees on lathe operation, including basic safety guidelines and detailed operating procedures.

[0752] The server receives this prompt and generates training materials using a trained generative AI model, and the output is sent to the device as training materials in PDF or HTML format.

[0753] 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.

[0754] System Overview

[0755] This invention utilizes generative artificial intelligence (AI) and an emotion engine to reduce the workload of elementary and junior high school teachers. Specifically, by learning from past test questions, entrance exam information, and teaching guidelines, the system automatically creates quizzes, automates grading, and automatically generates teaching guidelines for club activities. This system also incorporates an emotion engine that recognizes the user's emotions, improving the quality of educational resources and feedback.

[0756] System configuration

[0757] 1. Data collection methods:

[0758] The server collects data provided by teachers and educational institutions, such as past test questions, entrance exam information, teaching guidelines, and club activity instruction manuals. This data is stored in the system's database.

[0759] 2. Data preprocessing methods:

[0760] The server cleans the collected data to standardize it into an appropriate format, removing noise from the text data and normalizing it, making it usable as education-related data.

[0761] 3. Generative AI Models:

[0762] The server uses the preprocessed data to train generative artificial intelligence (AI) models that are specialized to perform education-related tasks (quiz creation, grading, curriculum generation).

[0763] 4. User request receiving means:

[0764] Teachers using the devices can request the creation and grading of quizzes and the generation of club activity guidelines. The requests are sent to the server and analyzed.

[0765] 5. Emotion Engine:

[0766] The server has an emotion engine that recognizes the emotion of the user when he / she sends a request, and provides a function to optimize the request content based on the user's emotional state.

[0767] 6. Test generation methods:

[0768] Based on user requests, the server automatically generates quizzes using an AI model. The generated quizzes are structured according to the specified subject area and number of questions, but are appropriately adjusted to reflect the results of the emotion engine.

[0769] 7. Scoring Procedure:

[0770] The server receives the student's response data sent by the user and automatically scores multiple-choice and essay questions using an AI model. The results are automatically compiled and feedback is provided using an emotion engine.

[0771] 8. Curriculum syllabus generation means:

[0772] For extracurricular activities, the server uses the trained AI model to generate standard teaching guidelines based on the user's requests, which are then optimized according to the user's emotional state.

[0773] 9. Delivery Method:

[0774] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's terminal.

[0775] Program processing and specific examples

[0776] Example of creating a quiz

[0777] Request Submission (User):

[0778] A teacher sends a request from a terminal saying, "Create a quiz with 10 questions on linear equations in second-year junior high school mathematics." If the emotion engine determines that the emotion is "stress," the server simplifies the request or suggests a template.

[0779] Receiving and processing the request (server):

[0780] The server receives this request, analyzes the content, and evaluates the emotional state using an emotion engine.

[0781] Quiz generation (server):

[0782] The server uses a trained AI model to generate quizzes based on the specified parameters, and optimizes the test content based on the results of the emotion engine.

[0783] Distribution (server):

[0784] The generated quiz is converted into PDF format and sent to the teacher's device.

[0785] Specific examples of automatic scoring

[0786] Submit Answer (User):

[0787] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their device to the server. If the emotion engine determines that the emotion is "anxiety," detailed feedback is provided.

[0788] Receiving and analyzing response data (server):

[0789] The server receives the submitted answer data and categorizes it into multiple choice and essay questions.

[0790] Automatic grading (server):

[0791] The server uses AI models to match multiple-choice questions with correct answers and evaluate essay questions, using an emotion engine to tailor feedback.

[0792] Distribution (server):

[0793] The grading results are automatically tallied and sent to the teacher's device.

[0794] Example of creating club activity guidelines

[0795] Request Submission (User):

[0796] An inexperienced soccer club advisor requests a "one-year coaching plan for a junior high school soccer club." If the emotion engine determines that the emotion is "lack of confidence," the coaching guidelines are supplemented with more detailed information.

[0797] Receiving and processing the request (server):

[0798] The server receives this request, analyzes the content, and evaluates the emotional state using an emotion engine.

[0799] Course guide generation (server):

[0800] The server uses the trained AI model to generate a one-year coaching plan for a typical junior high school soccer club, and optimizes the coaching content by reflecting the results of the emotion engine.

[0801] Distribution (server):

[0802] The generated teaching guidelines are converted into PDF format and sent to the teacher's device.

[0803] This will significantly reduce the workload of teachers and allow them to receive high-quality educational support that responds to their emotions.

[0804] The processing flow will be explained below.

[0805] Specific flow of program processing

[0806] Quiz creation process

[0807] Step 1:

[0808] The user (teacher) sends a request to create a quiz from their device. The request includes detailed parameters such as grade, subject, topic, and number of questions. At the same time as sending the request, the device's camera and microphone are used to collect the user's facial expressions and voice to capture their emotions.

[0809] Step 2:

[0810] The server receives requests from users and analyzes their content. During the analysis, the emotion engine evaluates the user's emotional state. For example, if the user is feeling stressed, the parameters are adjusted to reflect that.

[0811] Step 3:

[0812] The server inputs the extracted parameters and the results of the emotion engine into a trained generative artificial intelligence model.

[0813] Step 4:

[0814] The generative AI model automatically generates quizzes based on input parameters and emotional assessments, adjusting the difficulty and format of the questions depending on the emotional state.

[0815] Step 5:

[0816] The server converts the generated quiz into PDF or HTML format.

[0817] Step 6:

[0818] The server delivers the converted quiz to the user's device, where the user can review the quiz and distribute it to students by printing or online.

[0819] Auto-grading process

[0820] Step 1:

[0821] The user (teacher) sends the students' answers from their terminal to the server. The answers include answers to multiple-choice questions and answers to written questions.

[0822] Step 2:

[0823] The server receives the answer data sent by the user and classifies them into multiple choice and written answers. At the same time, the emotion engine evaluates the teacher's stress and anxiety.

[0824] Step 3:

[0825] The server automatically scores multiple-choice questions by comparing answers with correct answer data, and the emotion engine's evaluation adjusts the detailed grading process to a minimum for teachers who are particularly stressed.

[0826] Step 4:

[0827] The server inputs the answers to the essay questions into a trained generative AI model for evaluation. The model automatically calculates a score based on the evaluation. The detail and tone of the feedback are adjusted based on the evaluation by the emotion engine.

[0828] Step 5:

[0829] The server aggregates the multiple choice and essay grading results and generates a final score.

[0830] Step 6:

[0831] The server delivers the aggregated results to the user's device. This may include individual feedback based on the emotion engine. The user checks the results and provides feedback to the student.

[0832] Creation of club activity guidelines

[0833] Step 1:

[0834] The user (teacher) sends a request for club activity guidance from their device. The request includes detailed parameters such as the type of club activity, teaching goals, and duration. At the same time as sending the request, the device's camera and microphone are used to collect the user's emotions from their facial expressions and voice.

[0835] Step 2:

[0836] The server receives the user's request and analyzes its contents. This analysis includes extracting request parameters and evaluating the user's emotional state using an emotion engine. For example, if the user feels a lack of confidence, detailed instructional guidance reflecting this is provided.

[0837] Step 3:

[0838] The server inputs the extracted parameters and emotion evaluations into a trained generative artificial intelligence model.

[0839] Step 4:

[0840] The generative AI model automatically generates standard club activity instruction guidelines based on input parameters and emotional evaluations, adjusting instruction content and supplementary information according to the emotional state.

[0841] Step 5:

[0842] The server converts the generated teaching guidelines into PDF or HTML format.

[0843] Step 6:

[0844] The server delivers the converted teaching guidelines to the user's terminal, where the user can check and apply the teaching guidelines.

[0845] Specific examples

[0846] Example of creating a quiz:

[0847] If a teacher sends a request to "create a 10-question quiz on linear equations in second-year junior high school mathematics" and the emotion is recognized as "stress," the server will suggest simplified problem sets and recommended problem templates, and also adjust the difficulty of the problems appropriately.

[0848] Examples of automatic grading:

[0849] When a teacher submits a student's answer and the emotion is recognized as "anxiety," the server provides detailed feedback and attaches special explanations to some of the more difficult answers.

[0850] Examples of club activity guidelines:

[0851] If a coaching teacher who requests a soccer club coaching plan is identified as lacking confidence, the server will provide a standardized teaching guideline including detailed coaching content, the purpose of each practice, points to note, etc. It will also provide encouraging messages and additional reference materials according to the coaching situation.

[0852] This will significantly reduce the workload of teachers and allow them to receive high-quality educational support that responds to their emotions.

[0853] Example 2

[0854] 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."

[0855] Teachers and educational institutions have a wide range of tasks to perform, such as creating tests, grading, and creating club activity guidelines, which require a great deal of time and effort. Furthermore, the heavy workload can lead to insufficient feedback on the quality of education and instruction. To solve these problems, an efficient educational support system is needed.

[0856] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0857] In this invention, the server includes means for collecting and storing past document data and information, means for preprocessing the collected data and converting it into a unified format, means for training a generation algorithm using the preprocessed data, means for receiving a data generation request from a user, means for generating data based on the request using the trained generation algorithm, means for recognizing the user's emotions and optimizing the request content, and means for delivering the generated data to the user's terminal. This significantly reduces the workload of teachers and enables high-quality educational support that responds to emotions.

[0858] "Past document data and information" refers to documents and related information such as test questions, entrance examination information, curriculum guidelines, and club activity instruction manuals that have been created and used by educational institutions and teachers in the past.

[0859] "Means of collection" refers to the function of obtaining data provided by educational institutions and teachers via the Internet or a dedicated interface and storing it in the system.

[0860] "Means of storage" refers to the databases and storage within the system used to properly manage and store collected data.

[0861] "Preprocessing means" refers to processes such as data cleaning, noise removal, and normalization that are performed to convert the collected data into an appropriate format.

[0862] The "means of converting to a unified format" refers to an operation for converting data stored in different formats into a consistent format so that it can be used for subsequent processing.

[0863] A "generative algorithm" is a machine learning model or AI used to automatically generate new test questions, curriculum guidelines, etc. using preprocessed data as input.

[0864] "Training" refers to the machine learning training process that uses collected data to improve the performance of the generative algorithm.

[0865] "Means for receiving data generation requests from users" refers to a mechanism for receiving requests from users via a request form, API, etc., and using them as the starting point for processing.

[0866] "Means for recognizing emotions and optimizing request content" is a function that analyzes the emotional state of the user from their input and behavior, and adjusts the system's response and the content of the output based on the results.

[0867] "Means for delivering generated data to the user's terminal" refers to a mechanism for transmitting generated quizzes, marking results, teaching guidelines, etc. to the user's terminal in an appropriate format.

[0868] This invention utilizes a generative AI model and an emotion engine to reduce the workload of teachers in educational institutions. Specifically, by learning from past test questions, entrance exam information, and teaching guidelines, the system automatically creates quizzes, automates grading, and automatically generates teaching guidelines for club activities. By combining this with an emotion engine that recognizes the user's emotions, the system improves the quality of educational resources and feedback.

[0869] System configuration and operation

[0870] 1. Data Collection Methods

[0871] The server collects data provided by teachers and educational institutions, such as past test questions, entrance exam information, teaching guidelines, and club activity instruction manuals. This data is stored in the system's database. Specifically, the data is collected using an FTP server and API and stored in the system's database.

[0872] 2. Data preprocessing methods

[0873] The server cleans the collected data to standardize it into an appropriate format. For example, it removes noise from the text data and normalizes it. It uses a text analysis tool (e.g., a Python text processing library) to remove meaningless strings and noise and standardize the data format.

[0874] 3. Generative AI Model

[0875] The server uses the preprocessed data to train a generative AI model, which is specialized for automating education-related tasks. Specifically, the training data is used as input for the AI ​​model, and the model parameters are optimized. Training is performed using a machine learning framework (e.g., TensorFlow or PyTorch).

[0876] 4. User Request Reception Method

[0877] Teachers use their terminals to request the creation and grading of quizzes and the generation of club activity guidelines. These requests are sent to the server and analyzed. The user interface is provided through a web browser, and the request contents are sent as HTTP requests.

[0878] 5. Emotion Engine

[0879] The server has an emotion engine that recognizes emotions when a user sends a request. Specifically, it analyzes emotions from the user's input text and voice. The emotion engine uses natural language processing (NLP) algorithms (e.g., BERT) to assign emotion tags (e.g., stress, anxiety, lack of confidence).

[0880] 6. Test Generation Methods

[0881] Based on user requests, the server automatically generates quizzes using a generative AI model. The results of the emotion engine are reflected and the test content is optimized. As a specific example, a prompt such as "junior high school second-year mathematics, linear equations, 10-question quiz" is entered to generate test questions. The difficulty and question format are adjusted according to the evaluation results of the emotion engine.

[0882] Example prompt sentence:

[0883] "Second year junior high school mathematics, linear equations, 10-question quiz"

[0884] 7. Scoring Method

[0885] The server receives the student's response data sent by the user and grades it using a generative AI model. It also provides feedback using an emotion engine. For example, it inputs the received response data into the AI ​​model, compares multiple-choice questions with correct answer data, and evaluates essay questions. It then aggregates the grades and adjusts the feedback based on the emotion tags generated by the emotion engine.

[0886] 8. Curriculum guideline generation means

[0887] The server uses a generative AI model to automatically generate club activity teaching guidelines based on the user's requests. Optimization is performed by reflecting the results of the emotion engine. As a specific example, a "one-year teaching plan for a junior high school soccer club" is input as a prompt, and a standard teaching guideline is generated. Specific activity content and points to note are added based on the evaluation results of the emotion engine.

[0888] Example prompt sentence:

[0889] "One-year coaching plan for junior high school soccer club"

[0890] 9. Distribution Method

[0891] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's device. Specifically, the server converts the generated content into PDF format using a PDF generation tool (e.g., ReportLab), and then sends it to the user's device using the HTTP protocol.

[0892] This system will significantly reduce the workload of teachers and provide them with high-quality educational support that responds to their emotions.

[0893] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0894] Step 1:

[0895] The server collects and stores past document data and information provided by teachers and educational institutions. Specifically, it retrieves data using an FTP server or API and stores it in the system's database. Input data includes past test questions, entrance exam information, curriculum guidelines, and club activity instruction manuals, and the output is structured data stored in the database.

[0896] Step 2:

[0897] The server preprocesses the collected data and converts it into a unified format. It uses text analysis tools (e.g., Python text processing libraries) to remove meaningless strings and noise and standardize the data format. The input is raw data from the database, and the output is formatted data that has been cleaned and normalized. Specific operations include removing unnecessary strings, standardizing character codes, and normalizing the text.

[0898] Step 3:

[0899] The server uses the preprocessed data to train a generative AI model. Using a machine learning framework (e.g., TensorFlow or PyTorch), it provides the training data as input to the AI ​​model and optimizes the model's parameters. The input is the preprocessed data, and the output is a trained generative AI model. Specific operations include splitting the dataset, building a model, running training, and evaluating and refining the model.

[0900] Step 4:

[0901] A user uses a terminal to request the creation and grading of quizzes and the generation of club activity curriculum guidelines. The user interface is provided through a web browser, and the request contents are sent to the server as an HTTP request. The input is the user's request data (the scope of the quiz, the number of questions, etc.), and the output is the HTTP request sent to the server. Specific actions include filling out the request form and clicking the submit button.

[0902] Step 5:

[0903] The server uses an emotion engine to recognize emotions when a user submits a request. It analyzes emotions from the user's input text or voice and generates emotion tags. It uses an NLP algorithm (e.g., BERT) to assign emotion tags (e.g., stress, anxiety, lack of confidence). The input is the user's request data, and the output is the emotion-tagged request data. Specific operations include performing text analysis, speech recognition, and emotion analysis.

[0904] Step 6:

[0905] The server uses a trained generative AI model to automatically generate quizzes. For example, a prompt such as "Junior high school second-year mathematics, linear equations, 10-question quiz" is input and test questions are generated. The difficulty and question format are adjusted according to the emotion tag. The input is the emotion-tagged request data and the prompt, and the output is an automatically generated quiz. Specific operations include inputting the prompt into the generative AI model and adjusting the format of the generated test.

[0906] Step 7:

[0907] The server receives the students' response data sent by the user and automatically grades them using the generative AI model. The input is the response data (multiple-choice questions and essay questions), and the output is the grading results and feedback. Specific operations include receiving the response data, checking the correct answers for the multiple-choice questions, evaluating the essay questions, and generating feedback.

[0908] Step 8:

[0909] The server uses a generative AI model to automatically generate club activity guidelines based on the user's request. For example, a "one-year teaching plan for a junior high school soccer club" is input as a prompt, and a standard teaching guideline is generated. Specific activity content and important points are added according to the emotion tag. The input is the emotion-tagged request data and the prompt text, and the output is the generated teaching guideline. Specific operations include inputting the prompt text into the generative AI model and adjusting the content of the generated teaching guideline.

[0910] Step 9:

[0911] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's device. The content is converted using a PDF generation tool (e.g., ReportLab) and then sent to the user's device using the HTTP protocol. The input is the generated content, and the output is the data delivered to the user's device. Specific operations include PDF generation and HTTP communication execution.

[0912] (Application example 2)

[0913] 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."

[0914] Currently, elementary and junior high school teachers are burdened with a wide range of tasks, including creating and grading quizzes and creating curriculum guidelines for club activities. There is also a lack of support that takes into account teachers' emotions and stress levels, raising concerns that this could lead to a decline in the quality of education. Furthermore, there are currently few tools available to support teachers in real time in the classroom. These issues need to be addressed.

[0915] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting and storing past test questions, entrance examination information, and teaching guidelines, means for preprocessing the collected data and converting it into a unified format, means for training a generative AI model using the preprocessed data, means for receiving a request from a user and recognizing the user's emotions, means for using the generative AI model to generate a quiz, test score, or teaching guide based on the request and optimizing it based on the recognized emotions, means for delivering the generated test, score results, or teaching guide to the user's terminal, and means for providing real-time support to the user in a virtual space. This significantly reduces the workload of teachers and enables them to receive high-quality educational support that responds to their emotions.

[0916] "Past test questions" refer to test questions previously used by an educational institution.

[0917] "Entrance exam information" refers to all data related to school and university entrance exams, including exam dates, exam topics, and past passing scores.

[0918] "Course of instruction" refers to a document that describes the content and procedures for lessons and activities that an educational institution will provide to students.

[0919] "Data preprocessing" refers to processing collected raw data to convert it into a format suitable for analysis and model generation.

[0920] A "uniform format" refers to a format for expressing data in multiple different formats according to consistent rules.

[0921] A "generative artificial intelligence model" refers to an AI system that has been trained using machine learning or deep learning techniques.

[0922] "Means for recognizing user emotions" refers to technology that detects emotions from the user's voice, facial expressions, and writing, and provides appropriate feedback based on that.

[0923] "Virtual space" refers to a virtual environment that is different from the real world and is generated using computer graphics and virtual reality technology.

[0924] "Means for providing support in real time" refers to technologies and systems that provide appropriate assistance immediately according to the user's situation, rather than content prepared in advance.

[0925] "Generated Test" refers to a test question newly created by a generative artificial intelligence model based on a user request.

[0926] "Scoring results" refers to the data compiled after evaluating the test scores given by students.

[0927] "Course generation" refers to the creation of new lesson plans using generative artificial intelligence models for specific educational tasks or activities.

[0928] System Overview

[0929] This system provides real-time support in a virtual space, reducing teachers' workload and providing high-quality educational support. The system collects past test questions, entrance exam information, and teaching guidelines, and uses a generative artificial intelligence model to automatically create quizzes, grade them, and generate teaching guidelines. Furthermore, it utilizes an emotion engine to provide optimized educational resources and feedback based on the user's emotions.

[0930] System configuration

[0931] 1. Data collection methods:

[0932] The server collects and stores past test questions, entrance exam information, and curriculum guidelines from educational institutions. This data is stored in various formats, including CSV and PDF.

[0933] 2. Data preprocessing methods:

[0934] The server preprocesses the collected raw data, for example using the Python Pandas library to cleanse the data and convert it into a unified format.

[0935] 3. Training the generative AI model:

[0936] The server uses the preprocessed data to train a generative artificial intelligence model using TensorFlow, which specializes in creating and grading quizzes and generating lesson plans.

[0937] 4. Receiving user requests and emotion recognition:

[0938] A teacher using a device wears a head-mounted display and sends a request to "Create a short math quiz for second-year junior high school students" in the virtual space. The server uses an emotion engine to recognize emotions from the user's voice and facial expressions. If the emotion is determined to be "stress," the request is simplified.

[0939] 5. Test generation methods and sentiment optimization:

[0940] The server uses a trained generative AI model to generate quizzes based on requests, such as creating 10 questions about linear equations in second-grade math, and adjusts the difficulty of the questions based on the results of the emotion engine.

[0941] 6. Automated scoring and feedback:

[0942] The teacher sends the student's answers (10 multiple-choice questions and 5 written questions) to the server. The server receives this data and automatically scores it using a generative AI model. If the emotion engine determines that the student is "anxious," it provides detailed feedback.

[0943] 7. Means of Delivery of Product:

[0944] The generated quizzes, marking results, and teaching guidelines are converted into PDF format and sent to the device.

[0945] Examples of concrete examples and prompts

[0946] Specific examples of quizzes

[0947] Request Submission (User):

[0948] Send the following prompt from the device: "Please create 10 short quizzes in mathematics for second-year junior high school students. Please simplify the content as teachers are feeling stressed. For example, please include a lot of basic questions on linear equations."

[0949] Specific examples of scoring

[0950] Submit Answer (User):

[0951] A teacher sends the following prompt: "You are grading a 8th grade math quiz. There are 10 multiple choice questions and 5 essay questions. The teacher is concerned, so please provide detailed explanations. For example, if the answers are incorrect, please explain in detail the correct way to solve the problem and the reason why."

[0952] Main technologies used

[0953] Hardware:

[0954] Head-mounted displays (e.g., Oculus Rift, HTC Vive)

[0955] High-Performance Server

[0956] software:

[0957] Data preprocessing: Python + Pandas

[0958] AI model generation: TensorFlow

[0959] Emotion engine: OpenCV + TensorFlow

[0960] This will significantly reduce the workload of teachers and allow them to receive high-quality educational support that responds to their emotions.

[0961] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0962] Step 1: Data collection

[0963] The server collects past test questions, entrance exam information, and curriculum guidelines from educational institutions and stores them in a database. This data is provided in various formats, including CSV and PDF.

[0964] Input: Test questions, entrance exam information, and curriculum guidelines provided by educational institutions

[0965] Output: Saved test questions, entrance exam information, and curriculum data

[0966] Step 2: Data Preprocessing

[0967] The server preprocesses the collected raw data, for example using Python's Pandas library to cleanse the data and convert it into a unified format.

[0968] Input: Stored raw data (test questions, entrance exam information, curriculum guidelines)

[0969] Output: Cleansed and unified data

[0970] Step 3: Training the AI ​​model

[0971] The server uses the preprocessed data to train a generative artificial intelligence model using TensorFlow, which specializes in creating quizzes, grading, and generating lesson plans.

[0972] Input: Data converted into a unified format

[0973] Output: A trained generative AI model

[0974] Step 4: Receiving user requests and recognizing emotions

[0975] A teacher using a device wears a head-mounted display and sends a request to the virtual space, such as "Create a short math quiz for second-year junior high school students." The server receives this request and uses an emotion engine to recognize emotions from the user's voice and facial expressions.

[0976] Input: Request submitted by teacher (e.g., "Create a quiz for 8th grade math")

[0977] Output: Optimized request when emotion is recognized as "stress"

[0978] Step 5: Quiz generation and sentiment optimization

[0979] The server uses a trained generative AI model to generate quizzes based on user requests, adjusting the difficulty of the quiz based on the results of the emotion engine.

[0980] Input: Optimized request (e.g., "8th grade math quiz with basic linear equations")

[0981] Output: The generated quiz (e.g. 10 questions on linear equations)

[0982] Step 6: Sending and receiving response data

[0983] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their device to the server. The server receives this data and proceeds to the next step.

[0984] Input: Student response data (multiple choice, written)

[0985] Output: Answer data saved on the server

[0986] Step 7: Automated scoring and feedback

[0987] The server automatically grades the saved answer data using a generative AI model, and uses an emotion engine to recognize the teacher's emotions and provide detailed feedback.

[0988] Input: Student response data, Teacher sentiment

[0989] Output: Marking results and detailed feedback

[0990] Step 8: Delivering results

[0991] The server converts the generated quizzes, marking results, and teaching guidelines into PDF format and sends them to the teacher's device.

[0992] Input: Generated quizzes, grades, and teaching guidelines

[0993] Output: PDF file delivered to the teacher's device

[0994] This details the inputs delivered at each processing step, the data manipulations or operations performed, and the final results achieved.

[0995] 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.

[0996] 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.

[0997] 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.

[0998] [Third embodiment]

[0999] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1000] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1001] 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).

[1002] 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.

[1003] 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.

[1004] 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).

[1005] 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.

[1006] 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.

[1007] 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.

[1008] 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.

[1009] 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.

[1010] 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."

[1011] System Overview

[1012] This invention is a system that uses generative artificial intelligence (AI) to reduce the workload of elementary and junior high school teachers. Specifically, by learning from past test questions, entrance exam information, and teaching guidelines, it automatically creates quizzes, automates grading, and generates teaching guidelines for club activities.

[1013] System configuration

[1014] 1. Data collection methods:

[1015] The server collects data provided by teachers and educational institutions, such as past test questions, entrance exam information, teaching guidelines, and club activity instruction manuals. This data is stored in the system's database.

[1016] 2. Data preprocessing methods:

[1017] The server cleans the collected data to standardize it into an appropriate format, removing noise from the text data and normalizing it, making it usable as education-related data.

[1018] 3. Generative AI Models:

[1019] The server uses the preprocessed data to train generative artificial intelligence (AI) models that are specialized to perform education-related tasks (quiz creation, grading, curriculum generation).

[1020] 4. User request receiving means:

[1021] Teachers using the devices can request the creation and grading of quizzes and the generation of club activity guidelines. The requests are sent to the server and analyzed.

[1022] 5. Test generation methods:

[1023] Based on user requests, the server automatically generates quizzes using AI models, which are structured according to the user's specified subject matter and number of questions.

[1024] 6. Scoring Procedure:

[1025] The server receives the student's response data sent by the user and automatically scores multiple-choice and essay questions using an AI model. The results are automatically compiled and feedback is provided.

[1026] 7. Curriculum syllabus generation means:

[1027] For club activity teaching guidelines, the server uses a trained AI model to generate standard teaching guidelines based on user requests, and these guidelines are provided in a format that is easy to use even for inexperienced teachers.

[1028] 8. Delivery Method:

[1029] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's terminal.

[1030] Program processing and specific examples

[1031] Example of creating a quiz

[1032] Request Submission (User):

[1033] A teacher sends a request from a terminal saying, "Create a 10-question quiz on linear equations in second-year junior high school mathematics."

[1034] Receiving and processing the request (server):

[1035] The server receives this request and analyzes the contents.

[1036] Quiz generation (server):

[1037] The server uses a trained AI model to generate quizzes based on the specified parameters.

[1038] Distribution (server):

[1039] The generated quiz is converted into PDF format and sent to the teacher's device.

[1040] Specific examples of automatic scoring

[1041] Submit Answer (User):

[1042] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their terminal to the server.

[1043] Receiving and analyzing response data (server):

[1044] The server receives the submitted answer data and categorizes it into multiple choice and essay questions.

[1045] Automatic grading (server):

[1046] The server uses an AI model to match multiple-choice questions with correct answer data and evaluate essay questions.

[1047] Distribution (server):

[1048] The grading results are automatically tallied and sent to the teacher's device.

[1049] Example of creating club activity guidelines

[1050] Request Submission (User):

[1051] An inexperienced soccer club advisor requests a "one-year coaching plan for a junior high school soccer club."

[1052] Receiving and processing the request (server):

[1053] The server receives this request and analyzes the contents.

[1054] Course guide generation (server):

[1055] The server uses a trained AI model to generate a one-year coaching plan for a typical junior high school soccer club.

[1056] Distribution (server):

[1057] The generated teaching guidelines are converted into PDF format and sent to the teacher's device.

[1058] The above system can reduce the workload of teachers and provide efficient educational support.

[1059] The processing flow will be explained below.

[1060] Program processing flow

[1061] Quiz creation process

[1062] Step 1:

[1063] A user (teacher) sends a request to create a quiz from a terminal, which includes detailed parameters such as grade, subject, topic, and number of questions.

[1064] Step 2:

[1065] The server receives the request from the user and analyzes its contents, which includes extracting the request parameters.

[1066] Step 3:

[1067] The server inputs the extracted parameters into a trained generative artificial intelligence model.

[1068] Step 4:

[1069] The generative AI model automatically generates quizzes based on input parameters, with questions tailored to the specified grade, subject, and topic.

[1070] Step 5:

[1071] The server converts the generated quiz into PDF or HTML format.

[1072] Step 6:

[1073] The server delivers the converted quiz to the user's device, where the user can review the quiz and distribute it to students by printing or online.

[1074] Auto-grading process

[1075] Step 1:

[1076] The user (teacher) sends the students' answers from their terminal to the server. The answers include answers to multiple-choice questions and answers to written questions.

[1077] Step 2:

[1078] The server receives the answer data sent by the user and classifies the answer into multiple choice and written answer.

[1079] Step 3:

[1080] The server compares the answers to the multiple-choice questions with the correct answer data and performs automatic grading.

[1081] Step 4:

[1082] The server inputs the answers to the essay questions into a trained generative AI model for evaluation, and the model automatically calculates a score based on this evaluation.

[1083] Step 5:

[1084] The server aggregates the multiple choice and essay grading results and generates a final score.

[1085] Step 6:

[1086] The server delivers the results to the user's device, where the user can check the results and provide feedback to the student.

[1087] Creation of club activity guidelines

[1088] Step 1:

[1089] The user (teacher) sends a request for club activity guidance from a terminal. The request includes detailed parameters such as the type of club activity, teaching goals, and duration.

[1090] Step 2:

[1091] The server receives the request from the user and analyzes its contents, which includes extracting the request parameters.

[1092] Step 3:

[1093] The server inputs the extracted parameters into a trained generative artificial intelligence model.

[1094] Step 4:

[1095] The generative AI model automatically generates standard club activity teaching guidelines based on input parameters. The generated teaching guidelines are appropriate for the type of club activity and teaching goals.

[1096] Step 5:

[1097] The server converts the generated teaching guidelines into PDF or HTML format.

[1098] Step 6:

[1099] The server delivers the converted teaching guidelines to the user's terminal, where the user can check and apply the teaching guidelines.

[1100] This will enable teachers to significantly improve their work efficiency and reduce the burden on the educational field.

[1101] Example 1

[1102] 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."

[1103] In the traditional education system, teachers have a heavy workload, especially when it comes to creating and grading quizzes and planning extracurricular activities. Furthermore, existing tools for streamlining these tasks are limited and lack sufficient flexibility and precision. This makes it difficult for teachers to devote sufficient time to their primary educational activities.

[1104] 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.

[1105] In this invention, the server includes: means for collecting and storing past assessment materials, entrance examination information, educational guidelines, and extracurricular activity instruction materials from educational institutions; means for preprocessing the collected materials and converting them into a unified format; means for training an AI model using the preprocessed materials; means for receiving requests from users to create exams, grade evaluations, and extracurricular activity guidelines; means for generating exams, grade evaluations, and extracurricular activity guidelines based on the requests using the trained AI model; means for delivering the generated exams, grade evaluations, or extracurricular activity guidelines to the user's terminal; means for analyzing the request content and generating prompts appropriate for the AI ​​model; and means for feeding the AI ​​model based on the analyzed request to generate optimal output. This reduces the workload of teachers and enables efficient support of educational activities.

[1106] An "educational institution" is a formal institution that provides teaching and learning, such as a school, university, college, or other institution.

[1107] "Assessment materials" are materials used to assess student learning outcomes, such as previously administered tests, exams, or quizzes.

[1108] "Entrance Examination Information" means data and information relating to examinations for admission to a particular school or educational institution.

[1109] "Educational guidelines" are guidelines or instructional documents for achieving a particular educational curriculum or learning objectives.

[1110] "Extracurricular activity instruction materials" are manuals and plans for instructing extracurricular activities, such as club activities and extracurricular club activities.

[1111] "Preprocessing" is the process of preparing collected data in a format that is easy to analyze and use, and performing noise removal and normalization.

[1112] A "generative artificial intelligence model" is an AI system that builds knowledge to perform education-related tasks from training data.

[1113] "Training" is the process by which an artificial intelligence model learns from specified data and improves its ability to solve specified problems.

[1114] A "request" is an instruction or command from a user to a system requesting a specific process or operation.

[1115] "Exam" refers to a test or quiz administered to assess a student's understanding or knowledge.

[1116] "Grading" is the process of assessing a student's achievements on a test or quiz and expressing that performance in terms of a number or grade.

[1117] The "Extracurricular Activities Guidelines" are detailed procedures and guidelines for guiding the planning and conduct of club and extracurricular activities.

[1118] A "prompt sentence" is an input sentence that is used to obtain a specific answer or product from the AI ​​model being generated.

[1119] "Feeding" is the process of inputting learning data and request content into an artificial intelligence model to obtain optimal output.

[1120] This invention is a system that utilizes generative artificial intelligence (AI) to reduce the workload of elementary and junior high school teachers. Specifically, the system automatically creates quizzes, automates grading, and generates club activity teaching guidelines by learning from past test questions, entrance exam information, educational guidelines, and club activity teaching materials.

[1121] System configuration

[1122] Data collection methods

[1123] The server automatically collects past assessment materials, entrance examination information, educational guidelines, and extracurricular activity guidance materials from educational institutions using APIs and FTP servers. The collected data is stored in a database.

[1124] Data preprocessing measures

[1125] The server preprocesses the collected data using text analysis tools (e.g., SpaCy, NLTK), removing noise from the text data, normalizing it, and converting it into a unified format so that it can be used as training data.

[1126] Generative AI Model

[1127] The server uses the preprocessed data to train an AI model (e.g., GPT-3, BERT), and uses machine learning libraries (e.g., TensorFlow, PyTorch) to build a model specialized for a given education-related task (quiz creation, grading, extracurricular activity curriculum generation).

[1128] User request receiving means

[1129] Teachers using the devices can use a dedicated application to request the creation and grading of quizzes and the generation of club activity guidelines. These requests are sent to the server in real time.

[1130] Test Generation Methods

[1131] The server uses a trained AI model to automatically generate quizzes based on user requests. It analyzes the request content and generates prompts appropriate for the AI ​​model. The generated quiz is converted into PDF or HTML format and sent to the teacher's device.

[1132] Scoring method

[1133] The server receives the student's response data sent by the user and automatically grades the multiple-choice and essay questions using an AI model. Multiple-choice questions are compared with the correct answer data, and essay questions are evaluated using NLP technology. The grading results are compiled and sent as feedback to the teacher's device.

[1134] Curriculum guideline generation means

[1135] The server uses a trained AI model to generate extracurricular activity guidelines based on the user's request. It analyzes the request and generates a prompt appropriate for the AI ​​model. The generated guidelines are converted into PDF format and sent to the teacher's device.

[1136] Data Delivery Methods

[1137] The server converts the generated quizzes, marking results, or extracurricular activity guidelines into PDF files or HTML format and delivers them to the teacher's terminal.

[1138] Specific example explanation

[1139] Example of creating a quiz

[1140] Request submission (user)

[1141] The teacher sends a request from the terminal saying, "Create a 10-question quiz on linear equations in second-year junior high school mathematics."

[1142] Receiving and processing requests (server)

[1143] The server receives this request in real time and analyzes the content.

[1144] Quiz generation (server)

[1145] The server uses a trained AI model to generate quizzes based on the specified parameters.

[1146] Distribution (server)

[1147] The generated quiz is converted into PDF format and sent to the teacher's device.

[1148] Specific examples of automatic scoring

[1149] Submit Answer (User)

[1150] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their terminal to the server.

[1151] Receiving and analyzing response data (server)

[1152] The server receives the submitted response data in real time and categorizes it into multiple choice and essay questions.

[1153] Automatic scoring (server)

[1154] The server uses an AI model to match multiple-choice questions with correct answer data and evaluate essay questions.

[1155] Distribution (server)

[1156] The grading results are automatically tallied and sent to the teacher's device.

[1157] Example of creating club activity guidelines

[1158] Request submission (user)

[1159] An inexperienced soccer club advisor requests a "one-year coaching plan for a junior high school soccer club."

[1160] Receiving and processing requests (server)

[1161] The server receives this request in real time and analyzes the content.

[1162] Teaching guideline generation (server)

[1163] The server uses a trained AI model to generate a one-year coaching plan for a typical junior high school soccer club.

[1164] Distribution (server)

[1165] The generated teaching guidelines are converted into PDF format and sent to the teacher's device.

[1166] This system will reduce the workload of teachers and provide efficient, high-quality educational support.

[1167] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1168] Step 1: Collect data

[1169] The server collects past evaluation materials, entrance examination information, educational guidelines, and extracurricular activity guidance materials from educational institutions and stores them in a database. Specifically, the server obtains data through APIs and FTP servers. The collected materials are stored in the database. Various educational materials are input, and educational materials in a unified data format are generated as output.

[1170] Step 2: Preprocessing the data

[1171] The server preprocesses the collected data. This involves using text analysis tools (e.g., SpaCy, NLTK) to remove noise and normalize the data. The input is the collected educational materials, and the output is preprocessed, cleaned data. Specifically, special characters are removed and the data is formatted.

[1172] Step 3: Training the AI ​​model

[1173] The server uses the preprocessed data to train a generative AI model (e.g., GPT-3, BERT). Machine learning libraries used include TensorFlow and PyTorch. The input is the preprocessed data, and the output is a trained AI model. Specifically, the data is fed into the AI ​​model, and iterative learning occurs to improve the model's accuracy.

[1174] Step 4: Receiving the request

[1175] Teachers using the devices make requests from their devices to create and grade quizzes and generate curriculum guidelines for club activities. These requests are sent to the server. The input is the teacher's request, and the output is the request content saved on the server. Specifically, the teacher inputs the request into a dedicated application, which is then sent to the server.

[1176] Step 5: Parsing the request and generating a prompt

[1177] The server analyzes the request content and generates a prompt sentence appropriate for the generative AI model. The input is the request received from the teacher, and the output is the generated prompt sentence. Specifically, the server analyzes the request content and generates a prompt sentence in an appropriate format.

[1178] Step 6: Generate the quiz

[1179] The server uses a trained AI model to automatically generate quizzes based on prompts. The input is the prompt, and the output is the generated quiz. Specifically, the generated prompt is fed into the AI ​​model to generate quiz questions.

[1180] Step 7: Convert and distribute the quiz

[1181] The generated quiz is converted into PDF or HTML format and sent to the teacher's device. The input is the generated quiz, and the output is a file converted into PDF or HTML format. Specifically, the test questions are converted into a document format and distributed to the device.

[1182] Step 8: Receiving and parsing responses

[1183] The teacher sends the students' answer data (multiple choice and written questions) from their terminal to the server. The server receives this data and classifies it into multiple choice and written questions. The input is the students' answer data, and the output is the classified answer data. Specifically, the data is classified.

[1184] Step 9: Automated scoring

[1185] The server uses an AI model to automatically score the received student answers. Multiple-choice questions are compared with correct answer data, and essay questions are evaluated using NLP technology. The input is classified answer data, and the output is the scoring results. Specifically, multiple-choice questions are automatically matched, and essay questions are evaluated using semantic analysis.

[1186] Step 10: Calculation and distribution of scores

[1187] The grading results are automatically compiled and sent to the teacher's device as feedback. The input is the grading results, and the output is the compiled feedback. Specifically, the grading results are compiled, and feedback is generated and distributed.

[1188] Step 11: Generate and distribute curriculum guidelines

[1189] The server uses a trained AI model to generate standard teaching guidelines based on the user's request. The generated teaching guidelines are converted into PDF format and sent to the teacher's device. The input is the prompt text and the AI ​​model, and the output is the teaching guidelines in PDF format. Specifically, the server analyzes the request content, generates the teaching guidelines, converts them into PDF, and distributes them.

[1190] (Application example 1)

[1191] 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."

[1192] In-factory worker education and training takes a great deal of time and effort, and if there is a shortage of skilled instructors, it can be difficult to improve worker skills and ensure safety. In addition, it can be difficult to provide consistent teaching materials and manage training progress, which can lead to inconsistent quality of education.

[1193] 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.

[1194] In this invention, the server includes a means for collecting and storing past training data, operation information, safety guidelines, and work manuals, a means for preprocessing the collected data and converting it into a unified format, and a means for training a generative artificial intelligence model using the preprocessed data, thereby enabling training materials, evaluation results, and troubleshooting tests to be automatically generated and distributed to terminals based on user requests.

[1195] "Training data" refers to past data collected for the purpose of educating and training workers.

[1196] "Operation information" refers to information about specific operating procedures and methods for machines and equipment.

[1197] "Safety guidelines" are guidelines and rules for ensuring safety during work and operations.

[1198] A "work manual" is a document or instruction manual that shows specific tasks and procedures.

[1199] "Preprocessing" refers to the process of organizing and cleaning raw data to convert it into a usable format.

[1200] A "uniform format" is a standard format for converting different types of data into a consistent structure.

[1201] "Training means" refers to the data and methods used to train an AI model.

[1202] A "request" means that a user requests something from the system.

[1203] "Training materials" are materials created for workers to learn and improve their skills.

[1204] "Evaluation data" is data based on the activities and responses performed by the worker.

[1205] "Troubleshooting tests" are tests to identify and resolve problems or faults.

[1206] "Automatic generation" means that the system autonomously creates data and information without human intervention.

[1207] A "terminal" is a device used to access the system.

[1208] This invention provides a system for optimizing worker education and training in a factory and improving work efficiency. This system automatically generates and evaluates training materials for worker education and provides troubleshooting tests through the following processing steps.

[1209] System Overview

[1210] The system uses smartphones, smart glasses, and head-mounted displays (HMDs) as hardware, and Flask (a Python microweb framework) and GPT-2 (a generative AI model) as software.

[1211] Program processing

[1212] Data collection and storage

[1213] The server collects and stores past training data, operating information, safety guidelines, and work manuals from databases within the factory and external sources.

[1214] Data preprocessing and conversion to a unified format

[1215] The server performs preprocessing such as noise removal and normalization on the collected data, and converts it into an appropriate unified format, making it possible to process various types of data in a consistent format.

[1216] Training generative artificial intelligence models

[1217] Using the pre-processed data, the server trains generative artificial intelligence (AI) models that are specialized to perform education-related tasks within the factory (training material creation, evaluation, troubleshooting test generation).

[1218] Receiving a user request

[1219] Users use their devices (smartphones, smart glasses, HMDs) to request the creation and evaluation of training materials and the generation of troubleshooting tests. These requests are sent to the server.

[1220] Automatic generation of training materials

[1221] The server automatically generates training materials using a trained AI model based on the received user requests, converts the generated materials into PDF or HTML format, and delivers them to the user's device.

[1222] Specific examples

[1223] For example, to create training materials for new employees on how to operate a lathe, the user might enter the following prompt:

[1224] Develop training materials for new employees on lathe operation, including basic safety guidelines and detailed operating procedures.

[1225] The server receives the prompt and generates training materials using a trained generative AI model, allowing users to instantly access the training materials and efficiently train new employees in the factory.

[1226] This system is a powerful tool for quickly and efficiently educating and training workers, contributing to improved productivity and safety within factories.

[1227] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1228] Step 1:

[1229] The server collects and stores past training data, operation information, safety guidelines, and work manuals from databases within the factory and external sources. This allows the various data required for training to be collected and stored within the system. The input is various materials such as training data and operation information, and the output is converted into a unified format and stored in the database.

[1230] Step 2:

[1231] The server performs preprocessing on the collected data, such as noise removal and normalization, and converts it into an appropriate unified format. This allows various data to be processed in a consistent format. The input is the data saved in the previous step, and the output is the preprocessed data in a consistent format.

[1232] Step 3:

[1233] Using the preprocessed data, the server trains a generative artificial intelligence (AI) model that is specialized to perform education-related tasks within the factory (training material creation, evaluation, troubleshooting test generation). The input is the preprocessed data, and the output is the trained generative AI model.

[1234] Step 4:

[1235] Users use devices (smartphones, smart glasses, HMDs) to request the creation and evaluation of training materials and the generation of troubleshooting tests. These requests are sent to the server. The input is the user's request, and the output is the receipt and analysis of the request content.

[1236] Step 5:

[1237] The server automatically generates training materials using a trained AI model based on the received user request. The generated materials are converted into PDF or HTML format and delivered to the user's device. The input is the user request and the AI ​​model, and the output is the generated training materials.

[1238] For example, when creating training materials for new employees on how to operate a lathe, the user would enter the following prompt:

[1239] Develop training materials for new employees on lathe operation, including basic safety guidelines and detailed operating procedures.

[1240] The server receives this prompt and generates training materials using a trained generative AI model, and the output is sent to the device as training materials in PDF or HTML format.

[1241] 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.

[1242] System Overview

[1243] This invention utilizes generative artificial intelligence (AI) and an emotion engine to reduce the workload of elementary and junior high school teachers. Specifically, by learning from past test questions, entrance exam information, and teaching guidelines, the system automatically creates quizzes, automates grading, and automatically generates teaching guidelines for club activities. This system also incorporates an emotion engine that recognizes the user's emotions, improving the quality of educational resources and feedback.

[1244] System configuration

[1245] 1. Data collection methods:

[1246] The server collects data provided by teachers and educational institutions, such as past test questions, entrance exam information, teaching guidelines, and club activity instruction manuals. This data is stored in the system's database.

[1247] 2. Data preprocessing methods:

[1248] The server cleans the collected data to standardize it into an appropriate format, removing noise from the text data and normalizing it, making it usable as education-related data.

[1249] 3. Generative AI Models:

[1250] The server uses the preprocessed data to train generative artificial intelligence (AI) models that are specialized to perform education-related tasks (quiz creation, grading, curriculum generation).

[1251] 4. User request receiving means:

[1252] Teachers using the devices can request the creation and grading of quizzes and the generation of club activity guidelines. The requests are sent to the server and analyzed.

[1253] 5. Emotion Engine:

[1254] The server has an emotion engine that recognizes the emotion of the user when he / she sends a request, and provides a function to optimize the request content based on the user's emotional state.

[1255] 6. Test generation methods:

[1256] Based on user requests, the server automatically generates quizzes using an AI model. The generated quizzes are structured according to the specified subject area and number of questions, but are appropriately adjusted to reflect the results of the emotion engine.

[1257] 7. Scoring Procedure:

[1258] The server receives the student's response data sent by the user and automatically scores multiple-choice and essay questions using an AI model. The results are automatically compiled and feedback is provided using an emotion engine.

[1259] 8. Curriculum syllabus generation means:

[1260] For extracurricular activities, the server uses the trained AI model to generate standard teaching guidelines based on the user's requests, which are then optimized according to the user's emotional state.

[1261] 9. Delivery Method:

[1262] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's terminal.

[1263] Program processing and specific examples

[1264] Example of creating a quiz

[1265] Request Submission (User):

[1266] A teacher sends a request from a terminal saying, "Create a quiz with 10 questions on linear equations in second-year junior high school mathematics." If the emotion engine determines that the emotion is "stress," the server simplifies the request or suggests a template.

[1267] Receiving and processing the request (server):

[1268] The server receives this request, analyzes the content, and evaluates the emotional state using an emotion engine.

[1269] Quiz generation (server):

[1270] The server uses a trained AI model to generate quizzes based on the specified parameters, and optimizes the test content based on the results of the emotion engine.

[1271] Distribution (server):

[1272] The generated quiz is converted into PDF format and sent to the teacher's device.

[1273] Specific examples of automatic scoring

[1274] Submit Answer (User):

[1275] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their device to the server. If the emotion engine determines that the emotion is "anxiety," detailed feedback is provided.

[1276] Receiving and analyzing response data (server):

[1277] The server receives the submitted answer data and categorizes it into multiple choice and essay questions.

[1278] Automatic grading (server):

[1279] The server uses AI models to match multiple-choice questions with correct answers and evaluate essay questions, using an emotion engine to tailor feedback.

[1280] Distribution (server):

[1281] The grading results are automatically tallied and sent to the teacher's device.

[1282] Example of creating club activity guidelines

[1283] Request Submission (User):

[1284] An inexperienced soccer club advisor requests a "one-year coaching plan for a junior high school soccer club." If the emotion engine determines that the emotion is "lack of confidence," the coaching guidelines are supplemented with more detailed information.

[1285] Receiving and processing the request (server):

[1286] The server receives this request, analyzes the content, and evaluates the emotional state using an emotion engine.

[1287] Course guide generation (server):

[1288] The server uses the trained AI model to generate a one-year coaching plan for a typical junior high school soccer club, and optimizes the coaching content by reflecting the results of the emotion engine.

[1289] Distribution (server):

[1290] The generated teaching guidelines are converted into PDF format and sent to the teacher's device.

[1291] This will significantly reduce the workload of teachers and allow them to receive high-quality educational support that responds to their emotions.

[1292] The processing flow will be explained below.

[1293] Specific flow of program processing

[1294] Quiz creation process

[1295] Step 1:

[1296] The user (teacher) sends a request to create a quiz from their device. The request includes detailed parameters such as grade, subject, topic, and number of questions. At the same time as sending the request, the device's camera and microphone are used to collect the user's facial expressions and voice to capture their emotions.

[1297] Step 2:

[1298] The server receives requests from users and analyzes their content. During the analysis, the emotion engine evaluates the user's emotional state. For example, if the user is feeling stressed, the parameters are adjusted to reflect that.

[1299] Step 3:

[1300] The server inputs the extracted parameters and the results of the emotion engine into a trained generative artificial intelligence model.

[1301] Step 4:

[1302] The generative AI model automatically generates quizzes based on input parameters and emotional assessments, adjusting the difficulty and format of the questions depending on the emotional state.

[1303] Step 5:

[1304] The server converts the generated quiz into PDF or HTML format.

[1305] Step 6:

[1306] The server delivers the converted quiz to the user's device, where the user can review the quiz and distribute it to students by printing or online.

[1307] Auto-grading process

[1308] Step 1:

[1309] The user (teacher) sends the students' answers from their terminal to the server. The answers include answers to multiple-choice questions and answers to written questions.

[1310] Step 2:

[1311] The server receives the answer data sent by the user and classifies them into multiple choice and written answers. At the same time, the emotion engine evaluates the teacher's stress and anxiety.

[1312] Step 3:

[1313] The server automatically scores multiple-choice questions by comparing answers with correct answer data, and the emotion engine's evaluation adjusts the detailed grading process to a minimum for teachers who are particularly stressed.

[1314] Step 4:

[1315] The server inputs the answers to the essay questions into a trained generative AI model for evaluation. The model automatically calculates a score based on the evaluation. The detail and tone of the feedback are adjusted based on the evaluation by the emotion engine.

[1316] Step 5:

[1317] The server aggregates the multiple choice and essay grading results and generates a final score.

[1318] Step 6:

[1319] The server delivers the aggregated results to the user's device. This may include individual feedback based on the emotion engine. The user checks the results and provides feedback to the student.

[1320] Creation of club activity guidelines

[1321] Step 1:

[1322] The user (teacher) sends a request for club activity guidance from their device. The request includes detailed parameters such as the type of club activity, teaching goals, and duration. At the same time as sending the request, the device's camera and microphone are used to collect the user's emotions from their facial expressions and voice.

[1323] Step 2:

[1324] The server receives the user's request and analyzes its contents. This analysis includes extracting request parameters and evaluating the user's emotional state using an emotion engine. For example, if the user feels a lack of confidence, detailed instructional guidance reflecting this is provided.

[1325] Step 3:

[1326] The server inputs the extracted parameters and emotion evaluations into a trained generative artificial intelligence model.

[1327] Step 4:

[1328] The generative AI model automatically generates standard club activity instruction guidelines based on input parameters and emotional evaluations, adjusting instruction content and supplementary information according to the emotional state.

[1329] Step 5:

[1330] The server converts the generated teaching guidelines into PDF or HTML format.

[1331] Step 6:

[1332] The server delivers the converted teaching guidelines to the user's terminal, where the user can check and apply the teaching guidelines.

[1333] Specific examples

[1334] Example of creating a quiz:

[1335] If a teacher sends a request to "create a 10-question quiz on linear equations in second-year junior high school mathematics" and the emotion is recognized as "stress," the server will suggest simplified problem sets and recommended problem templates, and also adjust the difficulty of the problems appropriately.

[1336] Examples of automatic grading:

[1337] When a teacher submits a student's answer and the emotion is recognized as "anxiety," the server provides detailed feedback and attaches special explanations to some of the more difficult answers.

[1338] Examples of club activity guidelines:

[1339] If a coaching teacher who requests a soccer club coaching plan is identified as lacking confidence, the server will provide a standardized teaching guideline including detailed coaching content, the purpose of each practice, points to note, etc. It will also provide encouraging messages and additional reference materials according to the coaching situation.

[1340] This will significantly reduce the workload of teachers and allow them to receive high-quality educational support that responds to their emotions.

[1341] Example 2

[1342] 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."

[1343] Teachers and educational institutions have a wide range of tasks to perform, such as creating tests, grading, and creating club activity guidelines, which require a great deal of time and effort. Furthermore, the heavy workload can lead to insufficient feedback on the quality of education and instruction. To solve these problems, an efficient educational support system is needed.

[1344] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1345] In this invention, the server includes means for collecting and storing past document data and information, means for preprocessing the collected data and converting it into a unified format, means for training a generation algorithm using the preprocessed data, means for receiving a data generation request from a user, means for generating data based on the request using the trained generation algorithm, means for recognizing the user's emotions and optimizing the request content, and means for delivering the generated data to the user's terminal. This significantly reduces the workload of teachers and enables high-quality educational support that responds to emotions.

[1346] "Past document data and information" refers to documents and related information such as test questions, entrance examination information, curriculum guidelines, and club activity instruction manuals that have been created and used by educational institutions and teachers in the past.

[1347] "Means of collection" refers to the function of obtaining data provided by educational institutions and teachers via the Internet or a dedicated interface and storing it in the system.

[1348] "Means of storage" refers to the databases and storage within the system used to properly manage and store collected data.

[1349] "Preprocessing means" refers to processes such as data cleaning, noise removal, and normalization that are performed to convert the collected data into an appropriate format.

[1350] The "means of converting to a unified format" refers to an operation for converting data stored in different formats into a consistent format so that it can be used for subsequent processing.

[1351] A "generative algorithm" is a machine learning model or AI used to automatically generate new test questions, curriculum guidelines, etc. using preprocessed data as input.

[1352] "Training" refers to the machine learning training process that uses collected data to improve the performance of the generative algorithm.

[1353] "Means for receiving data generation requests from users" refers to a mechanism for receiving requests from users via a request form, API, etc., and using them as the starting point for processing.

[1354] "Means for recognizing emotions and optimizing request content" is a function that analyzes the emotional state of the user from their input and behavior, and adjusts the system's response and the content of the output based on the results.

[1355] "Means for delivering generated data to the user's terminal" refers to a mechanism for transmitting generated quizzes, marking results, teaching guidelines, etc. to the user's terminal in an appropriate format.

[1356] This invention utilizes a generative AI model and an emotion engine to reduce the workload of teachers in educational institutions. Specifically, by learning from past test questions, entrance exam information, and teaching guidelines, the system automatically creates quizzes, automates grading, and automatically generates teaching guidelines for club activities. By combining this with an emotion engine that recognizes the user's emotions, the system improves the quality of educational resources and feedback.

[1357] System configuration and operation

[1358] 1. Data Collection Methods

[1359] The server collects data provided by teachers and educational institutions, such as past test questions, entrance exam information, teaching guidelines, and club activity instruction manuals. This data is stored in the system's database. Specifically, the data is collected using an FTP server and API and stored in the system's database.

[1360] 2. Data preprocessing methods

[1361] The server cleans the collected data to standardize it into an appropriate format. For example, it removes noise from the text data and normalizes it. It uses a text analysis tool (e.g., a Python text processing library) to remove meaningless strings and noise and standardize the data format.

[1362] 3. Generative AI Model

[1363] The server uses the preprocessed data to train a generative AI model, which is specialized for automating education-related tasks. Specifically, the training data is used as input for the AI ​​model, and the model parameters are optimized. Training is performed using a machine learning framework (e.g., TensorFlow or PyTorch).

[1364] 4. User Request Reception Method

[1365] Teachers use their terminals to request the creation and grading of quizzes and the generation of club activity guidelines. These requests are sent to the server and analyzed. The user interface is provided through a web browser, and the request contents are sent as HTTP requests.

[1366] 5. Emotion Engine

[1367] The server has an emotion engine that recognizes emotions when a user sends a request. Specifically, it analyzes emotions from the user's input text and voice. The emotion engine uses natural language processing (NLP) algorithms (e.g., BERT) to assign emotion tags (e.g., stress, anxiety, lack of confidence).

[1368] 6. Test Generation Methods

[1369] Based on user requests, the server automatically generates quizzes using a generative AI model. The results of the emotion engine are reflected and the test content is optimized. As a specific example, a prompt such as "junior high school second-year mathematics, linear equations, 10-question quiz" is entered to generate test questions. The difficulty and question format are adjusted according to the evaluation results of the emotion engine.

[1370] Example prompt sentence:

[1371] "Second year junior high school mathematics, linear equations, 10-question quiz"

[1372] 7. Scoring Method

[1373] The server receives the student's response data sent by the user and grades it using a generative AI model. It also provides feedback using an emotion engine. For example, it inputs the received response data into the AI ​​model, compares multiple-choice questions with correct answer data, and evaluates essay questions. It then aggregates the grades and adjusts the feedback based on the emotion tags generated by the emotion engine.

[1374] 8. Curriculum guideline generation means

[1375] The server uses a generative AI model to automatically generate club activity teaching guidelines based on the user's requests. Optimization is performed by reflecting the results of the emotion engine. As a specific example, a "one-year teaching plan for a junior high school soccer club" is input as a prompt, and a standard teaching guideline is generated. Specific activity content and points to note are added based on the evaluation results of the emotion engine.

[1376] Example prompt sentence:

[1377] "One-year coaching plan for junior high school soccer club"

[1378] 9. Distribution Method

[1379] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's device. Specifically, the server converts the generated content into PDF format using a PDF generation tool (e.g., ReportLab), and then sends it to the user's device using the HTTP protocol.

[1380] This system will significantly reduce the workload of teachers and provide them with high-quality educational support that responds to their emotions.

[1381] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1382] Step 1:

[1383] The server collects and stores past document data and information provided by teachers and educational institutions. Specifically, it retrieves data using an FTP server or API and stores it in the system's database. Input data includes past test questions, entrance exam information, curriculum guidelines, and club activity instruction manuals, and the output is structured data stored in the database.

[1384] Step 2:

[1385] The server preprocesses the collected data and converts it into a unified format. It uses text analysis tools (e.g., Python text processing libraries) to remove meaningless strings and noise and standardize the data format. The input is raw data from the database, and the output is formatted data that has been cleaned and normalized. Specific operations include removing unnecessary strings, standardizing character codes, and normalizing the text.

[1386] Step 3:

[1387] The server uses the preprocessed data to train a generative AI model. Using a machine learning framework (e.g., TensorFlow or PyTorch), it provides the training data as input to the AI ​​model and optimizes the model's parameters. The input is the preprocessed data, and the output is a trained generative AI model. Specific operations include splitting the dataset, building a model, running training, and evaluating and refining the model.

[1388] Step 4:

[1389] A user uses a terminal to request the creation and grading of quizzes and the generation of club activity curriculum guidelines. The user interface is provided through a web browser, and the request contents are sent to the server as an HTTP request. The input is the user's request data (the scope of the quiz, the number of questions, etc.), and the output is the HTTP request sent to the server. Specific actions include filling out the request form and clicking the submit button.

[1390] Step 5:

[1391] The server uses an emotion engine to recognize emotions when a user submits a request. It analyzes emotions from the user's input text or voice and generates emotion tags. It uses an NLP algorithm (e.g., BERT) to assign emotion tags (e.g., stress, anxiety, lack of confidence). The input is the user's request data, and the output is the emotion-tagged request data. Specific operations include performing text analysis, speech recognition, and emotion analysis.

[1392] Step 6:

[1393] The server uses a trained generative AI model to automatically generate quizzes. For example, a prompt such as "Junior high school second-year mathematics, linear equations, 10-question quiz" is input and test questions are generated. The difficulty and question format are adjusted according to the emotion tag. The input is the emotion-tagged request data and the prompt, and the output is an automatically generated quiz. Specific operations include inputting the prompt into the generative AI model and adjusting the format of the generated test.

[1394] Step 7:

[1395] The server receives the students' response data sent by the user and automatically grades them using the generative AI model. The input is the response data (multiple-choice questions and essay questions), and the output is the grading results and feedback. Specific operations include receiving the response data, checking the correct answers for the multiple-choice questions, evaluating the essay questions, and generating feedback.

[1396] Step 8:

[1397] The server uses a generative AI model to automatically generate club activity guidelines based on the user's request. For example, a "one-year teaching plan for a junior high school soccer club" is input as a prompt, and a standard teaching guideline is generated. Specific activity content and important points are added according to the emotion tag. The input is the emotion-tagged request data and the prompt text, and the output is the generated teaching guideline. Specific operations include inputting the prompt text into the generative AI model and adjusting the content of the generated teaching guideline.

[1398] Step 9:

[1399] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's device. The content is converted using a PDF generation tool (e.g., ReportLab) and then sent to the user's device using the HTTP protocol. The input is the generated content, and the output is the data delivered to the user's device. Specific operations include PDF generation and HTTP communication execution.

[1400] (Application example 2)

[1401] 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."

[1402] Currently, elementary and junior high school teachers are burdened with a wide range of tasks, including creating and grading quizzes and creating curriculum guidelines for club activities. There is also a lack of support that takes into account teachers' emotions and stress levels, raising concerns that this could lead to a decline in the quality of education. Furthermore, there are currently few tools available to support teachers in real time in the classroom. These issues need to be addressed.

[1403] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting and storing past test questions, entrance examination information, and teaching guidelines, means for preprocessing the collected data and converting it into a unified format, means for training a generative AI model using the preprocessed data, means for receiving a request from a user and recognizing the user's emotions, means for using the generative AI model to generate a quiz, test score, or teaching guide based on the request and optimizing it based on the recognized emotions, means for delivering the generated test, score results, or teaching guide to the user's terminal, and means for providing real-time support to the user in a virtual space. This significantly reduces the workload of teachers and enables them to receive high-quality educational support that responds to their emotions.

[1404] "Past test questions" refer to test questions previously used by an educational institution.

[1405] "Entrance exam information" refers to all data related to school and university entrance exams, including exam dates, exam topics, and past passing scores.

[1406] "Course of instruction" refers to a document that describes the content and procedures for lessons and activities that an educational institution will provide to students.

[1407] "Data preprocessing" refers to processing collected raw data to convert it into a format suitable for analysis and model generation.

[1408] A "uniform format" refers to a format for expressing data in multiple different formats according to consistent rules.

[1409] A "generative artificial intelligence model" refers to an AI system that has been trained using machine learning or deep learning techniques.

[1410] "Means for recognizing user emotions" refers to technology that detects emotions from the user's voice, facial expressions, and writing, and provides appropriate feedback based on that.

[1411] "Virtual space" refers to a virtual environment that is different from the real world and is generated using computer graphics and virtual reality technology.

[1412] "Means for providing support in real time" refers to technologies and systems that provide appropriate assistance immediately according to the user's situation, rather than content prepared in advance.

[1413] "Generated Test" refers to a test question newly created by a generative artificial intelligence model based on a user request.

[1414] "Scoring results" refers to the data compiled after evaluating the test scores given by students.

[1415] "Course generation" refers to the creation of new lesson plans using generative artificial intelligence models for specific educational tasks or activities.

[1416] System Overview

[1417] This system provides real-time support in a virtual space, reducing teachers' workload and providing high-quality educational support. The system collects past test questions, entrance exam information, and teaching guidelines, and uses a generative artificial intelligence model to automatically create quizzes, grade them, and generate teaching guidelines. Furthermore, it utilizes an emotion engine to provide optimized educational resources and feedback based on the user's emotions.

[1418] System configuration

[1419] 1. Data collection methods:

[1420] The server collects and stores past test questions, entrance exam information, and curriculum guidelines from educational institutions. This data is stored in various formats, including CSV and PDF.

[1421] 2. Data preprocessing methods:

[1422] The server preprocesses the collected raw data, for example using the Python Pandas library to cleanse the data and convert it into a unified format.

[1423] 3. Training the generative AI model:

[1424] The server uses the preprocessed data to train a generative artificial intelligence model using TensorFlow, which specializes in creating and grading quizzes and generating lesson plans.

[1425] 4. Receiving user requests and emotion recognition:

[1426] A teacher using a device wears a head-mounted display and sends a request to "Create a short math quiz for second-year junior high school students" in the virtual space. The server uses an emotion engine to recognize emotions from the user's voice and facial expressions. If the emotion is determined to be "stress," the request is simplified.

[1427] 5. Test generation methods and sentiment optimization:

[1428] The server uses a trained generative AI model to generate quizzes based on requests, such as creating 10 questions about linear equations in second-grade math, and adjusts the difficulty of the questions based on the results of the emotion engine.

[1429] 6. Automated scoring and feedback:

[1430] The teacher sends the student's answers (10 multiple-choice questions and 5 written questions) to the server. The server receives this data and automatically scores it using a generative AI model. If the emotion engine determines that the student is "anxious," it provides detailed feedback.

[1431] 7. Means of Delivery of Product:

[1432] The generated quizzes, marking results, and teaching guidelines are converted into PDF format and sent to the device.

[1433] Examples of concrete examples and prompts

[1434] Specific examples of quizzes

[1435] Request Submission (User):

[1436] Send the following prompt from the device: "Please create 10 short quizzes in mathematics for second-year junior high school students. Please simplify the content as teachers are feeling stressed. For example, please include a lot of basic questions on linear equations."

[1437] Specific examples of scoring

[1438] Submit Answer (User):

[1439] A teacher sends the following prompt: "You are grading a 8th grade math quiz. There are 10 multiple choice questions and 5 essay questions. The teacher is concerned, so please provide detailed explanations. For example, if the answers are incorrect, please explain in detail the correct way to solve the problem and the reason why."

[1440] Main technologies used

[1441] Hardware:

[1442] Head-mounted displays (e.g., Oculus Rift, HTC Vive)

[1443] High-Performance Server

[1444] software:

[1445] Data preprocessing: Python + Pandas

[1446] AI model generation: TensorFlow

[1447] Emotion engine: OpenCV + TensorFlow

[1448] This will significantly reduce the workload of teachers and allow them to receive high-quality educational support that responds to their emotions.

[1449] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1450] Step 1: Data collection

[1451] The server collects past test questions, entrance exam information, and curriculum guidelines from educational institutions and stores them in a database. This data is provided in various formats, including CSV and PDF.

[1452] Input: Test questions, entrance exam information, and curriculum guidelines provided by educational institutions

[1453] Output: Saved test questions, entrance exam information, and curriculum data

[1454] Step 2: Data Preprocessing

[1455] The server preprocesses the collected raw data, for example using Python's Pandas library to cleanse the data and convert it into a unified format.

[1456] Input: Stored raw data (test questions, entrance exam information, curriculum guidelines)

[1457] Output: Cleansed and unified data

[1458] Step 3: Training the AI ​​model

[1459] The server uses the preprocessed data to train a generative artificial intelligence model using TensorFlow, which specializes in creating quizzes, grading, and generating lesson plans.

[1460] Input: Data converted into a unified format

[1461] Output: A trained generative AI model

[1462] Step 4: Receiving user requests and recognizing emotions

[1463] A teacher using a device wears a head-mounted display and sends a request to the virtual space, such as "Create a short math quiz for second-year junior high school students." The server receives this request and uses an emotion engine to recognize emotions from the user's voice and facial expressions.

[1464] Input: Request submitted by teacher (e.g., "Create a quiz for 8th grade math")

[1465] Output: Optimized request when emotion is recognized as "stress"

[1466] Step 5: Quiz generation and sentiment optimization

[1467] The server uses a trained generative AI model to generate quizzes based on user requests, adjusting the difficulty of the quiz based on the results of the emotion engine.

[1468] Input: Optimized request (e.g., "8th grade math quiz with basic linear equations")

[1469] Output: The generated quiz (e.g. 10 questions on linear equations)

[1470] Step 6: Sending and receiving response data

[1471] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their device to the server. The server receives this data and proceeds to the next step.

[1472] Input: Student response data (multiple choice, written)

[1473] Output: Answer data saved on the server

[1474] Step 7: Automated scoring and feedback

[1475] The server automatically grades the saved answer data using a generative AI model, and uses an emotion engine to recognize the teacher's emotions and provide detailed feedback.

[1476] Input: Student response data, Teacher sentiment

[1477] Output: Marking results and detailed feedback

[1478] Step 8: Delivering results

[1479] The server converts the generated quizzes, marking results, and teaching guidelines into PDF format and sends them to the teacher's device.

[1480] Input: Generated quizzes, grades, and teaching guidelines

[1481] Output: PDF file delivered to the teacher's device

[1482] This details the inputs delivered at each processing step, the data manipulations or operations performed, and the final results achieved.

[1483] 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.

[1484] 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.

[1485] 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.

[1486] [Fourth embodiment]

[1487] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1488] 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.

[1489] 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).

[1490] 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.

[1491] 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.

[1492] 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).

[1493] 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.

[1494] 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.

[1495] 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.

[1496] 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.

[1497] 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.

[1498] 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.

[1499] 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."

[1500] System Overview

[1501] This invention is a system that uses generative artificial intelligence (AI) to reduce the workload of elementary and junior high school teachers. Specifically, by learning from past test questions, entrance exam information, and teaching guidelines, it automatically creates quizzes, automates grading, and generates teaching guidelines for club activities.

[1502] System configuration

[1503] 1. Data collection methods:

[1504] The server collects data provided by teachers and educational institutions, such as past test questions, entrance exam information, teaching guidelines, and club activity instruction manuals. This data is stored in the system's database.

[1505] 2. Data preprocessing methods:

[1506] The server cleans the collected data to standardize it into an appropriate format, removing noise from the text data and normalizing it, making it usable as education-related data.

[1507] 3. Generative AI Models:

[1508] The server uses the preprocessed data to train generative artificial intelligence (AI) models that are specialized to perform education-related tasks (quiz creation, grading, curriculum generation).

[1509] 4. User request receiving means:

[1510] Teachers using the devices can request the creation and grading of quizzes and the generation of club activity guidelines. The requests are sent to the server and analyzed.

[1511] 5. Test generation methods:

[1512] Based on user requests, the server automatically generates quizzes using AI models, which are structured according to the user's specified subject matter and number of questions.

[1513] 6. Scoring Procedure:

[1514] The server receives the student's response data sent by the user and automatically scores multiple-choice and essay questions using an AI model. The results are automatically compiled and feedback is provided.

[1515] 7. Curriculum syllabus generation means:

[1516] For club activity teaching guidelines, the server uses a trained AI model to generate standard teaching guidelines based on user requests, and these guidelines are provided in a format that is easy to use even for inexperienced teachers.

[1517] 8. Delivery Method:

[1518] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's terminal.

[1519] Program processing and specific examples

[1520] Example of creating a quiz

[1521] Request Submission (User):

[1522] A teacher sends a request from a terminal saying, "Create a 10-question quiz on linear equations in second-year junior high school mathematics."

[1523] Receiving and processing the request (server):

[1524] The server receives this request and analyzes the contents.

[1525] Quiz generation (server):

[1526] The server uses a trained AI model to generate quizzes based on the specified parameters.

[1527] Distribution (server):

[1528] The generated quiz is converted into PDF format and sent to the teacher's device.

[1529] Specific examples of automatic scoring

[1530] Submit Answer (User):

[1531] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their terminal to the server.

[1532] Receiving and analyzing response data (server):

[1533] The server receives the submitted answer data and categorizes it into multiple choice and essay questions.

[1534] Automatic grading (server):

[1535] The server uses an AI model to match multiple-choice questions with correct answer data and evaluate essay questions.

[1536] Distribution (server):

[1537] The grading results are automatically tallied and sent to the teacher's device.

[1538] Example of creating club activity guidelines

[1539] Request Submission (User):

[1540] An inexperienced soccer club advisor requests a "one-year coaching plan for a junior high school soccer club."

[1541] Receiving and processing the request (server):

[1542] The server receives this request and analyzes the contents.

[1543] Course guide generation (server):

[1544] The server uses a trained AI model to generate a one-year coaching plan for a typical junior high school soccer club.

[1545] Distribution (server):

[1546] The generated teaching guidelines are converted into PDF format and sent to the teacher's device.

[1547] The above system can reduce the workload of teachers and provide efficient educational support.

[1548] The processing flow will be explained below.

[1549] Program processing flow

[1550] Quiz creation process

[1551] Step 1:

[1552] A user (teacher) sends a request to create a quiz from a terminal, which includes detailed parameters such as grade, subject, topic, and number of questions.

[1553] Step 2:

[1554] The server receives the request from the user and analyzes its contents, which includes extracting the request parameters.

[1555] Step 3:

[1556] The server inputs the extracted parameters into a trained generative artificial intelligence model.

[1557] Step 4:

[1558] The generative AI model automatically generates quizzes based on input parameters, with questions tailored to the specified grade, subject, and topic.

[1559] Step 5:

[1560] The server converts the generated quiz into PDF or HTML format.

[1561] Step 6:

[1562] The server delivers the converted quiz to the user's device, where the user can review the quiz and distribute it to students by printing or online.

[1563] Auto-grading process

[1564] Step 1:

[1565] The user (teacher) sends the students' answers from their terminal to the server. The answers include answers to multiple-choice questions and answers to written questions.

[1566] Step 2:

[1567] The server receives the answer data sent by the user and classifies the answer into multiple choice and written answer.

[1568] Step 3:

[1569] The server compares the answers to the multiple-choice questions with the correct answer data and performs automatic grading.

[1570] Step 4:

[1571] The server inputs the answers to the essay questions into a trained generative AI model for evaluation, and the model automatically calculates a score based on this evaluation.

[1572] Step 5:

[1573] The server aggregates the multiple choice and essay grading results and generates a final score.

[1574] Step 6:

[1575] The server delivers the results to the user's device, where the user can check the results and provide feedback to the student.

[1576] Creation of club activity guidelines

[1577] Step 1:

[1578] The user (teacher) sends a request for club activity guidance from a terminal. The request includes detailed parameters such as the type of club activity, teaching goals, and duration.

[1579] Step 2:

[1580] The server receives the request from the user and analyzes its contents, which includes extracting the request parameters.

[1581] Step 3:

[1582] The server inputs the extracted parameters into a trained generative artificial intelligence model.

[1583] Step 4:

[1584] The generative AI model automatically generates standard club activity teaching guidelines based on input parameters. The generated teaching guidelines are appropriate for the type of club activity and teaching goals.

[1585] Step 5:

[1586] The server converts the generated teaching guidelines into PDF or HTML format.

[1587] Step 6:

[1588] The server delivers the converted teaching guidelines to the user's terminal, where the user can check and apply the teaching guidelines.

[1589] This will enable teachers to significantly improve their work efficiency and reduce the burden on the educational field.

[1590] Example 1

[1591] 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."

[1592] In the traditional education system, teachers have a heavy workload, especially when it comes to creating and grading quizzes and planning extracurricular activities. Furthermore, existing tools for streamlining these tasks are limited and lack sufficient flexibility and precision. This makes it difficult for teachers to devote sufficient time to their primary educational activities.

[1593] 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.

[1594] In this invention, the server includes: means for collecting and storing past assessment materials, entrance examination information, educational guidelines, and extracurricular activity instruction materials from educational institutions; means for preprocessing the collected materials and converting them into a unified format; means for training an AI model using the preprocessed materials; means for receiving requests from users to create exams, grade evaluations, and extracurricular activity guidelines; means for generating exams, grade evaluations, and extracurricular activity guidelines based on the requests using the trained AI model; means for delivering the generated exams, grade evaluations, or extracurricular activity guidelines to the user's terminal; means for analyzing the request content and generating prompts appropriate for the AI ​​model; and means for feeding the AI ​​model based on the analyzed request to generate optimal output. This reduces the workload of teachers and enables efficient support of educational activities.

[1595] An "educational institution" is a formal institution that provides teaching and learning, such as a school, university, college, or other institution.

[1596] "Assessment materials" are materials used to assess student learning outcomes, such as previously administered tests, exams, or quizzes.

[1597] "Entrance Examination Information" means data and information relating to examinations for admission to a particular school or educational institution.

[1598] "Educational guidelines" are guidelines or instructional documents for achieving a particular educational curriculum or learning objectives.

[1599] "Extracurricular activity instruction materials" are manuals and plans for instructing extracurricular activities, such as club activities and extracurricular club activities.

[1600] "Preprocessing" is the process of preparing collected data in a format that is easy to analyze and use, and performing noise removal and normalization.

[1601] A "generative artificial intelligence model" is an AI system that builds knowledge to perform education-related tasks from training data.

[1602] "Training" is the process by which an artificial intelligence model learns from specified data and improves its ability to solve specified problems.

[1603] A "request" is an instruction or command from a user to a system requesting a specific process or operation.

[1604] "Exam" refers to a test or quiz administered to assess a student's understanding or knowledge.

[1605] "Grading" is the process of assessing a student's achievements on a test or quiz and expressing that performance in terms of a number or grade.

[1606] The "Extracurricular Activities Guidelines" are detailed procedures and guidelines for guiding the planning and conduct of club and extracurricular activities.

[1607] A "prompt sentence" is an input sentence that is used to obtain a specific answer or product from the AI ​​model being generated.

[1608] "Feeding" is the process of inputting learning data and request content into an artificial intelligence model to obtain optimal output.

[1609] This invention is a system that utilizes generative artificial intelligence (AI) to reduce the workload of elementary and junior high school teachers. Specifically, the system automatically creates quizzes, automates grading, and generates club activity teaching guidelines by learning from past test questions, entrance exam information, educational guidelines, and club activity teaching materials.

[1610] System configuration

[1611] Data collection methods

[1612] The server automatically collects past assessment materials, entrance examination information, educational guidelines, and extracurricular activity guidance materials from educational institutions using APIs and FTP servers. The collected data is stored in a database.

[1613] Data preprocessing measures

[1614] The server preprocesses the collected data using text analysis tools (e.g., SpaCy, NLTK), removing noise from the text data, normalizing it, and converting it into a unified format so that it can be used as training data.

[1615] Generative AI Model

[1616] The server uses the preprocessed data to train an AI model (e.g., GPT-3, BERT), and uses machine learning libraries (e.g., TensorFlow, PyTorch) to build a model specialized for a given education-related task (quiz creation, grading, extracurricular activity curriculum generation).

[1617] User request receiving means

[1618] Teachers using the devices can use a dedicated application to request the creation and grading of quizzes and the generation of club activity guidelines. These requests are sent to the server in real time.

[1619] Test Generation Methods

[1620] The server uses a trained AI model to automatically generate quizzes based on user requests. It analyzes the request content and generates prompts appropriate for the AI ​​model. The generated quiz is converted into PDF or HTML format and sent to the teacher's device.

[1621] Scoring method

[1622] The server receives the student's response data sent by the user and automatically grades the multiple-choice and essay questions using an AI model. Multiple-choice questions are compared with the correct answer data, and essay questions are evaluated using NLP technology. The grading results are compiled and sent as feedback to the teacher's device.

[1623] Curriculum guideline generation means

[1624] The server uses a trained AI model to generate extracurricular activity guidelines based on the user's request. It analyzes the request and generates a prompt appropriate for the AI ​​model. The generated guidelines are converted into PDF format and sent to the teacher's device.

[1625] Data Delivery Methods

[1626] The server converts the generated quizzes, marking results, or extracurricular activity guidelines into PDF files or HTML format and delivers them to the teacher's terminal.

[1627] Specific example explanation

[1628] Example of creating a quiz

[1629] Request submission (user)

[1630] The teacher sends a request from the terminal saying, "Create a 10-question quiz on linear equations in second-year junior high school mathematics."

[1631] Receiving and processing requests (server)

[1632] The server receives this request in real time and analyzes the content.

[1633] Quiz generation (server)

[1634] The server uses a trained AI model to generate quizzes based on the specified parameters.

[1635] Distribution (server)

[1636] The generated quiz is converted into PDF format and sent to the teacher's device.

[1637] Specific examples of automatic scoring

[1638] Submit Answer (User)

[1639] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their terminal to the server.

[1640] Receiving and analyzing response data (server)

[1641] The server receives the submitted response data in real time and categorizes it into multiple choice and essay questions.

[1642] Automatic scoring (server)

[1643] The server uses an AI model to match multiple-choice questions with correct answer data and evaluate essay questions.

[1644] Distribution (server)

[1645] The grading results are automatically tallied and sent to the teacher's device.

[1646] Example of creating club activity guidelines

[1647] Request submission (user)

[1648] An inexperienced soccer club advisor requests a "one-year coaching plan for a junior high school soccer club."

[1649] Receiving and processing requests (server)

[1650] The server receives this request in real time and analyzes the content.

[1651] Teaching guideline generation (server)

[1652] The server uses a trained AI model to generate a one-year coaching plan for a typical junior high school soccer club.

[1653] Distribution (server)

[1654] The generated teaching guidelines are converted into PDF format and sent to the teacher's device.

[1655] This system will reduce the workload of teachers and provide efficient, high-quality educational support.

[1656] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1657] Step 1: Collect data

[1658] The server collects past evaluation materials, entrance examination information, educational guidelines, and extracurricular activity guidance materials from educational institutions and stores them in a database. Specifically, the server obtains data through APIs and FTP servers. The collected materials are stored in the database. Various educational materials are input, and educational materials in a unified data format are generated as output.

[1659] Step 2: Preprocessing the data

[1660] The server preprocesses the collected data. This involves using text analysis tools (e.g., SpaCy, NLTK) to remove noise and normalize the data. The input is the collected educational materials, and the output is preprocessed, cleaned data. Specifically, special characters are removed and the data is formatted.

[1661] Step 3: Training the AI ​​model

[1662] The server uses the preprocessed data to train a generative AI model (e.g., GPT-3, BERT). Machine learning libraries used include TensorFlow and PyTorch. The input is the preprocessed data, and the output is a trained AI model. Specifically, the data is fed into the AI ​​model, and iterative learning occurs to improve the model's accuracy.

[1663] Step 4: Receiving the request

[1664] Teachers using the devices make requests from their devices to create and grade quizzes and generate curriculum guidelines for club activities. These requests are sent to the server. The input is the teacher's request, and the output is the request content saved on the server. Specifically, the teacher inputs the request into a dedicated application, which is then sent to the server.

[1665] Step 5: Parsing the request and generating a prompt

[1666] The server analyzes the request content and generates a prompt sentence appropriate for the generative AI model. The input is the request received from the teacher, and the output is the generated prompt sentence. Specifically, the server analyzes the request content and generates a prompt sentence in an appropriate format.

[1667] Step 6: Generate the quiz

[1668] The server uses a trained AI model to automatically generate quizzes based on prompts. The input is the prompt, and the output is the generated quiz. Specifically, the generated prompt is fed into the AI ​​model to generate quiz questions.

[1669] Step 7: Convert and distribute the quiz

[1670] The generated quiz is converted into PDF or HTML format and sent to the teacher's device. The input is the generated quiz, and the output is a file converted into PDF or HTML format. Specifically, the test questions are converted into a document format and distributed to the device.

[1671] Step 8: Receiving and parsing responses

[1672] The teacher sends the students' answer data (multiple choice and written questions) from their terminal to the server. The server receives this data and classifies it into multiple choice and written questions. The input is the students' answer data, and the output is the classified answer data. Specifically, the data is classified.

[1673] Step 9: Automated scoring

[1674] The server uses an AI model to automatically score the received student answers. Multiple-choice questions are compared with correct answer data, and essay questions are evaluated using NLP technology. The input is classified answer data, and the output is the scoring results. Specifically, multiple-choice questions are automatically matched, and essay questions are evaluated using semantic analysis.

[1675] Step 10: Calculation and distribution of scores

[1676] The grading results are automatically compiled and sent to the teacher's device as feedback. The input is the grading results, and the output is the compiled feedback. Specifically, the grading results are compiled, and feedback is generated and distributed.

[1677] Step 11: Generate and distribute curriculum guidelines

[1678] The server uses a trained AI model to generate standard teaching guidelines based on the user's request. The generated teaching guidelines are converted into PDF format and sent to the teacher's device. The input is the prompt text and the AI ​​model, and the output is the teaching guidelines in PDF format. Specifically, the server analyzes the request content, generates the teaching guidelines, converts them into PDF, and distributes them.

[1679] (Application example 1)

[1680] 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."

[1681] In-factory worker education and training takes a great deal of time and effort, and if there is a shortage of skilled instructors, it can be difficult to improve worker skills and ensure safety. In addition, it can be difficult to provide consistent teaching materials and manage training progress, which can lead to inconsistent quality of education.

[1682] 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.

[1683] In this invention, the server includes a means for collecting and storing past training data, operation information, safety guidelines, and work manuals, a means for preprocessing the collected data and converting it into a unified format, and a means for training a generative artificial intelligence model using the preprocessed data, thereby enabling training materials, evaluation results, and troubleshooting tests to be automatically generated and distributed to terminals based on user requests.

[1684] "Training data" refers to past data collected for the purpose of educating and training workers.

[1685] "Operation information" refers to information about specific operating procedures and methods for machines and equipment.

[1686] "Safety guidelines" are guidelines and rules for ensuring safety during work and operations.

[1687] A "work manual" is a document or instruction manual that shows specific tasks and procedures.

[1688] "Preprocessing" refers to the process of organizing and cleaning raw data to convert it into a usable format.

[1689] A "uniform format" is a standard format for converting different types of data into a consistent structure.

[1690] "Training means" refers to the data and methods used to train an AI model.

[1691] A "request" means that a user requests something from the system.

[1692] "Training materials" are materials created for workers to learn and improve their skills.

[1693] "Evaluation data" is data based on the activities and responses performed by the worker.

[1694] "Troubleshooting tests" are tests to identify and resolve problems or faults.

[1695] "Automatic generation" means that the system autonomously creates data and information without human intervention.

[1696] A "terminal" is a device used to access the system.

[1697] This invention provides a system for optimizing worker education and training in a factory and improving work efficiency. This system automatically generates and evaluates training materials for worker education and provides troubleshooting tests through the following processing steps.

[1698] System Overview

[1699] The system uses smartphones, smart glasses, and head-mounted displays (HMDs) as hardware, and Flask (a Python microweb framework) and GPT-2 (a generative AI model) as software.

[1700] Program processing

[1701] Data collection and storage

[1702] The server collects and stores past training data, operating information, safety guidelines, and work manuals from databases within the factory and external sources.

[1703] Data preprocessing and conversion to a unified format

[1704] The server performs preprocessing such as noise removal and normalization on the collected data, and converts it into an appropriate unified format, making it possible to process various types of data in a consistent format.

[1705] Training generative artificial intelligence models

[1706] Using the pre-processed data, the server trains generative artificial intelligence (AI) models that are specialized to perform education-related tasks within the factory (training material creation, evaluation, troubleshooting test generation).

[1707] Receiving a user request

[1708] Users use their devices (smartphones, smart glasses, HMDs) to request the creation and evaluation of training materials and the generation of troubleshooting tests. These requests are sent to the server.

[1709] Automatic generation of training materials

[1710] The server automatically generates training materials using a trained AI model based on the received user requests, converts the generated materials into PDF or HTML format, and delivers them to the user's device.

[1711] Specific examples

[1712] For example, to create training materials for new employees on how to operate a lathe, the user might enter the following prompt:

[1713] Develop training materials for new employees on lathe operation, including basic safety guidelines and detailed operating procedures.

[1714] The server receives the prompt and generates training materials using a trained generative AI model, allowing users to instantly access the training materials and efficiently train new employees in the factory.

[1715] This system is a powerful tool for quickly and efficiently educating and training workers, contributing to improved productivity and safety within factories.

[1716] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1717] Step 1:

[1718] The server collects and stores past training data, operation information, safety guidelines, and work manuals from databases within the factory and external sources. This allows the various data required for training to be collected and stored within the system. The input is various materials such as training data and operation information, and the output is converted into a unified format and stored in the database.

[1719] Step 2:

[1720] The server performs preprocessing on the collected data, such as noise removal and normalization, and converts it into an appropriate unified format. This allows various data to be processed in a consistent format. The input is the data saved in the previous step, and the output is the preprocessed data in a consistent format.

[1721] Step 3:

[1722] Using the preprocessed data, the server trains a generative artificial intelligence (AI) model that is specialized to perform education-related tasks within the factory (training material creation, evaluation, troubleshooting test generation). The input is the preprocessed data, and the output is the trained generative AI model.

[1723] Step 4:

[1724] Users use devices (smartphones, smart glasses, HMDs) to request the creation and evaluation of training materials and the generation of troubleshooting tests. These requests are sent to the server. The input is the user's request, and the output is the receipt and analysis of the request content.

[1725] Step 5:

[1726] The server automatically generates training materials using a trained AI model based on the received user request. The generated materials are converted into PDF or HTML format and delivered to the user's device. The input is the user request and the AI ​​model, and the output is the generated training materials.

[1727] For example, when creating training materials for new employees on how to operate a lathe, the user would enter the following prompt:

[1728] Develop training materials for new employees on lathe operation, including basic safety guidelines and detailed operating procedures.

[1729] The server receives this prompt and generates training materials using a trained generative AI model, and the output is sent to the device as training materials in PDF or HTML format.

[1730] 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.

[1731] System Overview

[1732] This invention utilizes generative artificial intelligence (AI) and an emotion engine to reduce the workload of elementary and junior high school teachers. Specifically, by learning from past test questions, entrance exam information, and teaching guidelines, the system automatically creates quizzes, automates grading, and automatically generates teaching guidelines for club activities. This system also incorporates an emotion engine that recognizes the user's emotions, improving the quality of educational resources and feedback.

[1733] System configuration

[1734] 1. Data collection methods:

[1735] The server collects data provided by teachers and educational institutions, such as past test questions, entrance exam information, teaching guidelines, and club activity instruction manuals. This data is stored in the system's database.

[1736] 2. Data preprocessing methods:

[1737] The server cleans the collected data to standardize it into an appropriate format, removing noise from the text data and normalizing it, making it usable as education-related data.

[1738] 3. Generative AI Models:

[1739] The server uses the preprocessed data to train generative artificial intelligence (AI) models that are specialized to perform education-related tasks (quiz creation, grading, curriculum generation).

[1740] 4. User request receiving means:

[1741] Teachers using the devices can request the creation and grading of quizzes and the generation of club activity guidelines. The requests are sent to the server and analyzed.

[1742] 5. Emotion Engine:

[1743] The server has an emotion engine that recognizes the emotion of the user when he / she sends a request, and provides a function to optimize the request content based on the user's emotional state.

[1744] 6. Test generation methods:

[1745] Based on user requests, the server automatically generates quizzes using an AI model. The generated quizzes are structured according to the specified subject area and number of questions, but are appropriately adjusted to reflect the results of the emotion engine.

[1746] 7. Scoring Procedure:

[1747] The server receives the student's response data sent by the user and automatically scores multiple-choice and essay questions using an AI model. The results are automatically compiled and feedback is provided using an emotion engine.

[1748] 8. Curriculum syllabus generation means:

[1749] For extracurricular activities, the server uses the trained AI model to generate standard teaching guidelines based on the user's requests, which are then optimized according to the user's emotional state.

[1750] 9. Delivery Method:

[1751] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's terminal.

[1752] Program processing and specific examples

[1753] Example of creating a quiz

[1754] Request Submission (User):

[1755] A teacher sends a request from a terminal saying, "Create a quiz with 10 questions on linear equations in second-year junior high school mathematics." If the emotion engine determines that the emotion is "stress," the server simplifies the request or suggests a template.

[1756] Receiving and processing the request (server):

[1757] The server receives this request, analyzes the content, and evaluates the emotional state using an emotion engine.

[1758] Quiz generation (server):

[1759] The server uses a trained AI model to generate quizzes based on the specified parameters, and optimizes the test content based on the results of the emotion engine.

[1760] Distribution (server):

[1761] The generated quiz is converted into PDF format and sent to the teacher's device.

[1762] Specific examples of automatic scoring

[1763] Submit Answer (User):

[1764] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their device to the server. If the emotion engine determines that the emotion is "anxiety," detailed feedback is provided.

[1765] Receiving and analyzing response data (server):

[1766] The server receives the submitted answer data and categorizes it into multiple choice and essay questions.

[1767] Automatic grading (server):

[1768] The server uses AI models to match multiple-choice questions with correct answers and evaluate essay questions, using an emotion engine to tailor feedback.

[1769] Distribution (server):

[1770] The grading results are automatically tallied and sent to the teacher's device.

[1771] Example of creating club activity guidelines

[1772] Request Submission (User):

[1773] An inexperienced soccer club advisor requests a "one-year coaching plan for a junior high school soccer club." If the emotion engine determines that the emotion is "lack of confidence," the coaching guidelines are supplemented with more detailed information.

[1774] Receiving and processing the request (server):

[1775] The server receives this request, analyzes the content, and evaluates the emotional state using an emotion engine.

[1776] Course guide generation (server):

[1777] The server uses the trained AI model to generate a one-year coaching plan for a typical junior high school soccer club, and optimizes the coaching content by reflecting the results of the emotion engine.

[1778] Distribution (server):

[1779] The generated teaching guidelines are converted into PDF format and sent to the teacher's device.

[1780] This will significantly reduce the workload of teachers and allow them to receive high-quality educational support that responds to their emotions.

[1781] The processing flow will be explained below.

[1782] Specific flow of program processing

[1783] Quiz creation process

[1784] Step 1:

[1785] The user (teacher) sends a request to create a quiz from their device. The request includes detailed parameters such as grade, subject, topic, and number of questions. At the same time as sending the request, the device's camera and microphone are used to collect the user's facial expressions and voice to capture their emotions.

[1786] Step 2:

[1787] The server receives requests from users and analyzes their content. During the analysis, the emotion engine evaluates the user's emotional state. For example, if the user is feeling stressed, the parameters are adjusted to reflect that.

[1788] Step 3:

[1789] The server inputs the extracted parameters and the results of the emotion engine into a trained generative artificial intelligence model.

[1790] Step 4:

[1791] The generative AI model automatically generates quizzes based on input parameters and emotional assessments, adjusting the difficulty and format of the questions depending on the emotional state.

[1792] Step 5:

[1793] The server converts the generated quiz into PDF or HTML format.

[1794] Step 6:

[1795] The server delivers the converted quiz to the user's device, where the user can review the quiz and distribute it to students by printing or online.

[1796] Auto-grading process

[1797] Step 1:

[1798] The user (teacher) sends the students' answers from their terminal to the server. The answers include answers to multiple-choice questions and answers to written questions.

[1799] Step 2:

[1800] The server receives the answer data sent by the user and classifies them into multiple choice and written answers. At the same time, the emotion engine evaluates the teacher's stress and anxiety.

[1801] Step 3:

[1802] The server automatically scores multiple-choice questions by comparing answers with correct answer data, and the emotion engine's evaluation adjusts the detailed grading process to a minimum for teachers who are particularly stressed.

[1803] Step 4:

[1804] The server inputs the answers to the essay questions into a trained generative AI model for evaluation. The model automatically calculates a score based on the evaluation. The detail and tone of the feedback are adjusted based on the evaluation by the emotion engine.

[1805] Step 5:

[1806] The server aggregates the multiple choice and essay grading results and generates a final score.

[1807] Step 6:

[1808] The server delivers the aggregated results to the user's device. This may include individual feedback based on the emotion engine. The user checks the results and provides feedback to the student.

[1809] Creation of club activity guidelines

[1810] Step 1:

[1811] The user (teacher) sends a request for club activity guidance from their device. The request includes detailed parameters such as the type of club activity, teaching goals, and duration. At the same time as sending the request, the device's camera and microphone are used to collect the user's emotions from their facial expressions and voice.

[1812] Step 2:

[1813] The server receives the user's request and analyzes its contents. This analysis includes extracting request parameters and evaluating the user's emotional state using an emotion engine. For example, if the user feels a lack of confidence, detailed instructional guidance reflecting this is provided.

[1814] Step 3:

[1815] The server inputs the extracted parameters and emotion evaluations into a trained generative artificial intelligence model.

[1816] Step 4:

[1817] The generative AI model automatically generates standard club activity instruction guidelines based on input parameters and emotional evaluations, adjusting instruction content and supplementary information according to the emotional state.

[1818] Step 5:

[1819] The server converts the generated teaching guidelines into PDF or HTML format.

[1820] Step 6:

[1821] The server delivers the converted teaching guidelines to the user's terminal, where the user can check and apply the teaching guidelines.

[1822] Specific examples

[1823] Example of creating a quiz:

[1824] If a teacher sends a request to "create a 10-question quiz on linear equations in second-year junior high school mathematics" and the emotion is recognized as "stress," the server will suggest simplified problem sets and recommended problem templates, and also adjust the difficulty of the problems appropriately.

[1825] Examples of automatic grading:

[1826] When a teacher submits a student's answer and the emotion is recognized as "anxiety," the server provides detailed feedback and attaches special explanations to some of the more difficult answers.

[1827] Examples of club activity guidelines:

[1828] If a coaching teacher who requests a soccer club coaching plan is identified as lacking confidence, the server will provide a standardized teaching guideline including detailed coaching content, the purpose of each practice, points to note, etc. It will also provide encouraging messages and additional reference materials according to the coaching situation.

[1829] This will significantly reduce the workload of teachers and allow them to receive high-quality educational support that responds to their emotions.

[1830] Example 2

[1831] 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."

[1832] Teachers and educational institutions have a wide range of tasks to perform, such as creating tests, grading, and creating club activity guidelines, which require a great deal of time and effort. Furthermore, the heavy workload can lead to insufficient feedback on the quality of education and instruction. To solve these problems, an efficient educational support system is needed.

[1833] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1834] In this invention, the server includes means for collecting and storing past document data and information, means for preprocessing the collected data and converting it into a unified format, means for training a generation algorithm using the preprocessed data, means for receiving a data generation request from a user, means for generating data based on the request using the trained generation algorithm, means for recognizing the user's emotions and optimizing the request content, and means for delivering the generated data to the user's terminal. This significantly reduces the workload of teachers and enables high-quality educational support that responds to emotions.

[1835] "Past document data and information" refers to documents and related information such as test questions, entrance examination information, curriculum guidelines, and club activity instruction manuals that have been created and used by educational institutions and teachers in the past.

[1836] "Means of collection" refers to the function of obtaining data provided by educational institutions and teachers via the Internet or a dedicated interface and storing it in the system.

[1837] "Means of storage" refers to the databases and storage within the system used to properly manage and store collected data.

[1838] "Preprocessing means" refers to processes such as data cleaning, noise removal, and normalization that are performed to convert the collected data into an appropriate format.

[1839] The "means of converting to a unified format" refers to an operation for converting data stored in different formats into a consistent format so that it can be used for subsequent processing.

[1840] A "generative algorithm" is a machine learning model or AI used to automatically generate new test questions, curriculum guidelines, etc. using preprocessed data as input.

[1841] "Training" refers to the machine learning training process that uses collected data to improve the performance of the generative algorithm.

[1842] "Means for receiving data generation requests from users" refers to a mechanism for receiving requests from users via a request form, API, etc., and using them as the starting point for processing.

[1843] "Means for recognizing emotions and optimizing request content" is a function that analyzes the emotional state of the user from their input and behavior, and adjusts the system's response and the content of the output based on the results.

[1844] "Means for delivering generated data to the user's terminal" refers to a mechanism for transmitting generated quizzes, marking results, teaching guidelines, etc. to the user's terminal in an appropriate format.

[1845] This invention utilizes a generative AI model and an emotion engine to reduce the workload of teachers in educational institutions. Specifically, by learning from past test questions, entrance exam information, and teaching guidelines, the system automatically creates quizzes, automates grading, and automatically generates teaching guidelines for club activities. By combining this with an emotion engine that recognizes the user's emotions, the system improves the quality of educational resources and feedback.

[1846] System configuration and operation

[1847] 1. Data Collection Methods

[1848] The server collects data provided by teachers and educational institutions, such as past test questions, entrance exam information, teaching guidelines, and club activity instruction manuals. This data is stored in the system's database. Specifically, the data is collected using an FTP server and API and stored in the system's database.

[1849] 2. Data preprocessing methods

[1850] The server cleans the collected data to standardize it into an appropriate format. For example, it removes noise from the text data and normalizes it. It uses a text analysis tool (e.g., a Python text processing library) to remove meaningless strings and noise and standardize the data format.

[1851] 3. Generative AI Model

[1852] The server uses the preprocessed data to train a generative AI model, which is specialized for automating education-related tasks. Specifically, the training data is used as input for the AI ​​model, and the model parameters are optimized. Training is performed using a machine learning framework (e.g., TensorFlow or PyTorch).

[1853] 4. User Request Reception Method

[1854] Teachers use their terminals to request the creation and grading of quizzes and the generation of club activity guidelines. These requests are sent to the server and analyzed. The user interface is provided through a web browser, and the request contents are sent as HTTP requests.

[1855] 5. Emotion Engine

[1856] The server has an emotion engine that recognizes emotions when a user sends a request. Specifically, it analyzes emotions from the user's input text and voice. The emotion engine uses natural language processing (NLP) algorithms (e.g., BERT) to assign emotion tags (e.g., stress, anxiety, lack of confidence).

[1857] 6. Test Generation Methods

[1858] Based on user requests, the server automatically generates quizzes using a generative AI model. The results of the emotion engine are reflected and the test content is optimized. As a specific example, a prompt such as "junior high school second-year mathematics, linear equations, 10-question quiz" is entered to generate test questions. The difficulty and question format are adjusted according to the evaluation results of the emotion engine.

[1859] Example prompt sentence:

[1860] "Second year junior high school mathematics, linear equations, 10-question quiz"

[1861] 7. Scoring Method

[1862] The server receives the student's response data sent by the user and grades it using a generative AI model. It also provides feedback using an emotion engine. For example, it inputs the received response data into the AI ​​model, compares multiple-choice questions with correct answer data, and evaluates essay questions. It then aggregates the grades and adjusts the feedback based on the emotion tags generated by the emotion engine.

[1863] 8. Curriculum guideline generation means

[1864] The server uses a generative AI model to automatically generate club activity teaching guidelines based on the user's requests. Optimization is performed by reflecting the results of the emotion engine. As a specific example, a "one-year teaching plan for a junior high school soccer club" is input as a prompt, and a standard teaching guideline is generated. Specific activity content and points to note are added based on the evaluation results of the emotion engine.

[1865] Example prompt sentence:

[1866] "One-year coaching plan for junior high school soccer club"

[1867] 9. Distribution Method

[1868] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's device. Specifically, the server converts the generated content into PDF format using a PDF generation tool (e.g., ReportLab), and then sends it to the user's device using the HTTP protocol.

[1869] This system will significantly reduce the workload of teachers and provide them with high-quality educational support that responds to their emotions.

[1870] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1871] Step 1:

[1872] The server collects and stores past document data and information provided by teachers and educational institutions. Specifically, it retrieves data using an FTP server or API and stores it in the system's database. Input data includes past test questions, entrance exam information, curriculum guidelines, and club activity instruction manuals, and the output is structured data stored in the database.

[1873] Step 2:

[1874] The server preprocesses the collected data and converts it into a unified format. It uses text analysis tools (e.g., Python text processing libraries) to remove meaningless strings and noise and standardize the data format. The input is raw data from the database, and the output is formatted data that has been cleaned and normalized. Specific operations include removing unnecessary strings, standardizing character codes, and normalizing the text.

[1875] Step 3:

[1876] The server uses the preprocessed data to train a generative AI model. Using a machine learning framework (e.g., TensorFlow or PyTorch), it provides the training data as input to the AI ​​model and optimizes the model's parameters. The input is the preprocessed data, and the output is a trained generative AI model. Specific operations include splitting the dataset, building a model, running training, and evaluating and refining the model.

[1877] Step 4:

[1878] A user uses a terminal to request the creation and grading of quizzes and the generation of club activity curriculum guidelines. The user interface is provided through a web browser, and the request contents are sent to the server as an HTTP request. The input is the user's request data (the scope of the quiz, the number of questions, etc.), and the output is the HTTP request sent to the server. Specific actions include filling out the request form and clicking the submit button.

[1879] Step 5:

[1880] The server uses an emotion engine to recognize emotions when a user submits a request. It analyzes emotions from the user's input text or voice and generates emotion tags. It uses an NLP algorithm (e.g., BERT) to assign emotion tags (e.g., stress, anxiety, lack of confidence). The input is the user's request data, and the output is the emotion-tagged request data. Specific operations include performing text analysis, speech recognition, and emotion analysis.

[1881] Step 6:

[1882] The server uses a trained generative AI model to automatically generate quizzes. For example, a prompt such as "Junior high school second-year mathematics, linear equations, 10-question quiz" is input and test questions are generated. The difficulty and question format are adjusted according to the emotion tag. The input is the emotion-tagged request data and the prompt, and the output is an automatically generated quiz. Specific operations include inputting the prompt into the generative AI model and adjusting the format of the generated test.

[1883] Step 7:

[1884] The server receives the students' response data sent by the user and automatically grades them using the generative AI model. The input is the response data (multiple-choice questions and essay questions), and the output is the grading results and feedback. Specific operations include receiving the response data, checking the correct answers for the multiple-choice questions, evaluating the essay questions, and generating feedback.

[1885] Step 8:

[1886] The server uses a generative AI model to automatically generate club activity guidelines based on the user's request. For example, a "one-year teaching plan for a junior high school soccer club" is input as a prompt, and a standard teaching guideline is generated. Specific activity content and important points are added according to the emotion tag. The input is the emotion-tagged request data and the prompt text, and the output is the generated teaching guideline. Specific operations include inputting the prompt text into the generative AI model and adjusting the content of the generated teaching guideline.

[1887] Step 9:

[1888] The server converts the generated quizzes, marking results, or club activity guidelines into PDF files or HTML format and delivers them to the user's device. The content is converted using a PDF generation tool (e.g., ReportLab) and then sent to the user's device using the HTTP protocol. The input is the generated content, and the output is the data delivered to the user's device. Specific operations include PDF generation and HTTP communication execution.

[1889] (Application example 2)

[1890] 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."

[1891] Currently, elementary and junior high school teachers are burdened with a wide range of tasks, including creating and grading quizzes and creating curriculum guidelines for club activities. There is also a lack of support that takes into account teachers' emotions and stress levels, raising concerns that this could lead to a decline in the quality of education. Furthermore, there are currently few tools available to support teachers in real time in the classroom. These issues need to be addressed.

[1892] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting and storing past test questions, entrance examination information, and teaching guidelines, means for preprocessing the collected data and converting it into a unified format, means for training a generative AI model using the preprocessed data, means for receiving a request from a user and recognizing the user's emotions, means for using the generative AI model to generate a quiz, test score, or teaching guide based on the request and optimizing it based on the recognized emotions, means for delivering the generated test, score results, or teaching guide to the user's terminal, and means for providing real-time support to the user in a virtual space. This significantly reduces the workload of teachers and enables them to receive high-quality educational support that responds to their emotions.

[1893] "Past test questions" refer to test questions previously used by an educational institution.

[1894] "Entrance exam information" refers to all data related to school and university entrance exams, including exam dates, exam topics, and past passing scores.

[1895] "Course of instruction" refers to a document that describes the content and procedures for lessons and activities that an educational institution will provide to students.

[1896] "Data preprocessing" refers to processing collected raw data to convert it into a format suitable for analysis and model generation.

[1897] A "uniform format" refers to a format for expressing data in multiple different formats according to consistent rules.

[1898] A "generative artificial intelligence model" refers to an AI system that has been trained using machine learning or deep learning techniques.

[1899] "Means for recognizing user emotions" refers to technology that detects emotions from the user's voice, facial expressions, and writing, and provides appropriate feedback based on that.

[1900] "Virtual space" refers to a virtual environment that is different from the real world and is generated using computer graphics and virtual reality technology.

[1901] "Means for providing support in real time" refers to technologies and systems that provide appropriate assistance immediately according to the user's situation, rather than content prepared in advance.

[1902] "Generated Test" refers to a test question newly created by a generative artificial intelligence model based on a user request.

[1903] "Scoring results" refers to the data compiled after evaluating the test scores given by students.

[1904] "Course generation" refers to the creation of new lesson plans using generative artificial intelligence models for specific educational tasks or activities.

[1905] System Overview

[1906] This system provides real-time support in a virtual space, reducing teachers' workload and providing high-quality educational support. The system collects past test questions, entrance exam information, and teaching guidelines, and uses a generative artificial intelligence model to automatically create quizzes, grade them, and generate teaching guidelines. Furthermore, it utilizes an emotion engine to provide optimized educational resources and feedback based on the user's emotions.

[1907] System configuration

[1908] 1. Data collection methods:

[1909] The server collects and stores past test questions, entrance exam information, and curriculum guidelines from educational institutions. This data is stored in various formats, including CSV and PDF.

[1910] 2. Data preprocessing methods:

[1911] The server preprocesses the collected raw data, for example using the Python Pandas library to cleanse the data and convert it into a unified format.

[1912] 3. Training the generative AI model:

[1913] The server uses the preprocessed data to train a generative artificial intelligence model using TensorFlow, which specializes in creating and grading quizzes and generating lesson plans.

[1914] 4. Receiving user requests and emotion recognition:

[1915] A teacher using a device wears a head-mounted display and sends a request to "Create a short math quiz for second-year junior high school students" in the virtual space. The server uses an emotion engine to recognize emotions from the user's voice and facial expressions. If the emotion is determined to be "stress," the request is simplified.

[1916] 5. Test generation methods and sentiment optimization:

[1917] The server uses a trained generative AI model to generate quizzes based on requests, such as creating 10 questions about linear equations in second-grade math, and adjusts the difficulty of the questions based on the results of the emotion engine.

[1918] 6. Automated scoring and feedback:

[1919] The teacher sends the student's answers (10 multiple-choice questions and 5 written questions) to the server. The server receives this data and automatically scores it using a generative AI model. If the emotion engine determines that the student is "anxious," it provides detailed feedback.

[1920] 7. Means of Delivery of Product:

[1921] The generated quizzes, marking results, and teaching guidelines are converted into PDF format and sent to the device.

[1922] Examples of concrete examples and prompts

[1923] Specific examples of quizzes

[1924] Request Submission (User):

[1925] Send the following prompt from the device: "Please create 10 short quizzes in mathematics for second-year junior high school students. Please simplify the content as teachers are feeling stressed. For example, please include a lot of basic questions on linear equations."

[1926] Specific examples of scoring

[1927] Submit Answer (User):

[1928] A teacher sends the following prompt: "You are grading a 8th grade math quiz. There are 10 multiple choice questions and 5 essay questions. The teacher is concerned, so please provide detailed explanations. For example, if the answers are incorrect, please explain in detail the correct way to solve the problem and the reason why."

[1929] Main technologies used

[1930] Hardware:

[1931] Head-mounted displays (e.g., Oculus Rift, HTC Vive)

[1932] High-Performance Server

[1933] software:

[1934] Data preprocessing: Python + Pandas

[1935] AI model generation: TensorFlow

[1936] Emotion engine: OpenCV + TensorFlow

[1937] This will significantly reduce the workload of teachers and allow them to receive high-quality educational support that responds to their emotions.

[1938] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1939] Step 1: Data collection

[1940] The server collects past test questions, entrance exam information, and curriculum guidelines from educational institutions and stores them in a database. This data is provided in various formats, including CSV and PDF.

[1941] Input: Test questions, entrance exam information, and curriculum guidelines provided by educational institutions

[1942] Output: Saved test questions, entrance exam information, and curriculum data

[1943] Step 2: Data Preprocessing

[1944] The server preprocesses the collected raw data, for example using Python's Pandas library to cleanse the data and convert it into a unified format.

[1945] Input: Stored raw data (test questions, entrance exam information, curriculum guidelines)

[1946] Output: Cleansed and unified data

[1947] Step 3: Training the AI ​​model

[1948] The server uses the preprocessed data to train a generative artificial intelligence model using TensorFlow, which specializes in creating quizzes, grading, and generating lesson plans.

[1949] Input: Data converted into a unified format

[1950] Output: A trained generative AI model

[1951] Step 4: Receiving user requests and recognizing emotions

[1952] A teacher using a device wears a head-mounted display and sends a request to the virtual space, such as "Create a short math quiz for second-year junior high school students." The server receives this request and uses an emotion engine to recognize emotions from the user's voice and facial expressions.

[1953] Input: Request submitted by teacher (e.g., "Create a quiz for 8th grade math")

[1954] Output: Optimized request when emotion is recognized as "stress"

[1955] Step 5: Quiz generation and sentiment optimization

[1956] The server uses a trained generative AI model to generate quizzes based on user requests, adjusting the difficulty of the quiz based on the results of the emotion engine.

[1957] Input: Optimized request (e.g., "8th grade math quiz with basic linear equations")

[1958] Output: The generated quiz (e.g. 10 questions on linear equations)

[1959] Step 6: Sending and receiving response data

[1960] The teacher sends the students' answers (10 multiple choice questions and 5 written questions) from their device to the server. The server receives this data and proceeds to the next step.

[1961] Input: Student response data (multiple choice, written)

[1962] Output: Answer data saved on the server

[1963] Step 7: Automated scoring and feedback

[1964] The server automatically grades the saved answer data using a generative AI model, and uses an emotion engine to recognize the teacher's emotions and provide detailed feedback.

[1965] Input: Student response data, Teacher sentiment

[1966] Output: Marking results and detailed feedback

[1967] Step 8: Delivering results

[1968] The server converts the generated quizzes, marking results, and teaching guidelines into PDF format and sends them to the teacher's device.

[1969] Input: Generated quizzes, grades, and teaching guidelines

[1970] Output: PDF file delivered to the teacher's device

[1971] This details the inputs delivered at each processing step, the data manipulations or operations performed, and the final results achieved.

[1972] 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.

[1973] 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.

[1974] 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.

[1975] 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.

[1976] 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.

[1977] 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.

[1978] 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).

[1979] 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 ma...

Claims

1. A means of collecting and storing past test questions, entrance exam information, curriculum guidelines, and club activity instruction manuals, a means of preprocessing the collected data and converting it into a unified format; means for training a generative artificial intelligence model using the preprocessed data; A means for receiving requests from users to create quizzes, grade tests, and generate club activity guidelines; A means for generating quizzes, test scores, and extracurricular activity guidelines based on requests using a trained generative artificial intelligence model; A system including means for delivering generated tests, scores, or teaching guidelines to a user's terminal.

2. 10. The system of claim 1, further comprising means for automatically generating quizzes based on requests.

3. 2. The system according to claim 1, further comprising means for receiving answer data sent by a user and automatically scoring the answer data using a generative artificial intelligence model.

4. 2. The system according to claim 1, further comprising means for automatically generating a club activity instruction guideline based on a user request.

5. The system according to claim 1, further comprising means for converting the generated quiz, marking results, or extracurricular activity instruction guidelines into a PDF file or HTML format and distributing the converted quiz, marking results, or extracurricular activity instruction guidelines.

Images