System

Abstract

A system is provided.SOLUTION: A system comprising: means for uploading a document from a user terminal to a server; means for converting the uploaded document into a text format and performing preprocessing; means for reading scoring and examination criteria and making a generative artificial intelligence understand the criteria; means for analyzing the preprocessed text according to the criteria and performing automatic scoring; means for generating feedback based on the result of the automatic scoring and the basis thereof; means for generating a summary of the submitted document using the generative artificial intelligence; and means for providing the generated feedback and summary to a user and a scorer.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] Until now, document screening and written test grading has been performed by humans, but there have been problems with inconsistent scores due to subjectivity and fatigue. In addition, grading a large number of documents is time-consuming and costly, and the lack of fairness and transparency reduces the reliability of the evaluation. It is necessary to solve these issues. [Means for solving the problem]

[0005] The present invention provides a system including a means for uploading documents from a user terminal to a server, a means for converting the uploaded documents into text format and preprocessing them, a means for reading scoring and examination criteria and having a generating AI understand the criteria, a means for the generating AI to analyze the preprocessed text according to the criteria and perform automatic scoring, a means for generating feedback based on the results of the automatic scoring and the basis for the scoring, a means for generating a summary of the submitted documents using the generating AI, and a means for providing the generated feedback and summary to the user and the grader. The present invention also provides a system for analyzing scores and the basis for scoring using the generating AI and performing context-based scoring, and a system for summarizing the contents of submitted documents and providing them to the grader.

[0006] A "document" is a physical or electronic document containing textual information submitted by a user.

[0007] "Upload" is an operation for transmitting data from a user terminal to a server and storing the data.

[0008] A "server" is a computer system on a network that stores, processes, and serves data.

[0009] "Text format" is a data format for character strings that are treated as digital information.

[0010] "Preprocessing" is the process of preprocessing text data to remove noise and standardize formats.

[0011] "Scoring and evaluation criteria" are specific guidelines for scoring and judging specific evaluation targets.

[0012] "Generative AI" is an AI system that analyzes text based on scoring and review criteria and performs scoring and summarization.

[0013] "Automatic scoring" is the process by which a generative artificial intelligence analyzes text data and assigns scores according to established criteria.

[0014] "Feedback" is a response message that includes the results of the scoring and evaluation and points for improvement based on the scoring basis.

[0015] A "summary" is a short sentence that succinctly summarizes the content of a text and extracts important information.

[0016] A "user terminal" is an electronic device, such as a computer or mobile device, that a user uses to access a server. [Brief explanation of the drawings]

[0017] [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

[0018] 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.

[0019] First, the terms used in the following description will be explained.

[0020] 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).

[0021] 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.

[0022] 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.

[0023] 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.

[0024] 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."

[0025] [First embodiment]

[0026] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0027] 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.

[0028] 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).

[0029] 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.

[0030] 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.

[0031] 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.

[0032] 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.

[0033] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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."

[0038] This invention relates to a system that automates the grading of document screening and written exams. In this system, documents are uploaded from the user's terminal, and the server converts the documents into text format and performs preprocessing. The server then loads the grading and examination criteria, and has the AI ​​generative system understand the criteria, which then performs automatic grading. The system also includes a function to generate feedback based on the scores and grading rationale, and to generate summaries of the submitted documents.

[0039] User document upload

[0040] A user uploads documents to the server using a terminal, such as an answer sheet for a particular exam or a business proposal. The user can upload the documents through a web interface, and the server receives the uploaded documents.

[0041] Document text conversion and preprocessing

[0042] The server converts the uploaded documents into a text format, then preprocesses the text data, for example by removing unnecessary information and normalizing the text, to produce clean data suitable for analysis.

[0043] Loading scoring and judging criteria and training the AI

[0044] The server loads the scoring and judging criteria. These criteria are predefined rubrics or guidelines that the server uses to train the generative AI. The AI ​​understands these criteria and uses them as the basis for its analysis.

[0045] Run automatic grading

[0046] The server then passes the preprocessed text to a generative AI, which then automatically analyzes and scores the content, evaluating the content and logical consistency of the answers and calculating a score according to a rubric.

[0047] Generate feedback

[0048] Based on the results of the automatic scoring and the rationale behind it, the server generates feedback for the user. This feedback includes not only the score but also specific reasons for the scoring and suggestions for improvement, so the user can clearly understand what was evaluated and what areas need improvement.

[0049] Generate a summary of the submitted documents

[0050] Additionally, the server uses a generator to generate summaries of the submitted documents, which are provided by the server to graders and other interested parties for quick understanding.

[0051] Specific examples

[0052] For example, consider the case where a student uploads an answer sheet for a written exam. When the user scans and uploads the answer sheet from their device, the server converts it into text format, removes noise, and performs preprocessing. The server then trains an AI based on pre-defined scoring criteria, which then automatically scores the answer sheet. As a result, specific feedback is generated and provided to the student along with a score. An answer summary is also generated for the instructor, allowing them to quickly compare and consider the answers of multiple students.

[0053] This system ensures fair and efficient scoring, increases the credibility of the assessment, and allows assessors to focus on more important tasks.

[0054] The processing flow will be explained below.

[0055] Step 1:

[0056] The user uploads the document from the device.

[0057] Specifically, this is done by selecting document files such as answer sheets and proposals through the web interface and clicking the send button to the server, which receives and stores the uploaded files.

[0058] Step 2:

[0059] The server converts the uploaded document into text format.

[0060] Specifically, the saved file is read and images and PDF documents are converted to text format using OCR (Optical Character Recognition) technology, thereby obtaining the text data.

[0061] Step 3:

[0062] The server preprocesses the text data.

[0063] Specifically, unnecessary characters and spaces are removed from the acquired text data and normalization is performed. For example, this includes processing to combine multiple spaces into one and standardize line breaks. This processing generates clean text data that is easy to analyze.

[0064] Step 4:

[0065] The server loads the scoring and judging criteria.

[0066] Specifically, predefined rubrics and grading guidelines are stored as JSON or XML format files, which the server reads and treats as data, making the evaluation criteria clear.

[0067] Step 5:

[0068] The server passes the reference data to the generative artificial intelligence and trains the AI.

[0069] Specifically, the reference data is fed into a generative AI model, and machine learning techniques are used to train the model. At this stage, the AI ​​learns how to make evaluations.

[0070] Step 6:

[0071] The server passes the preprocessed text to a generative artificial intelligence for analysis.

[0072] Specifically, clean text data is fed into an AI model, which analyzes the content, understands the context, and calculates a score for each part based on a scoring criteria.

[0073] Step 7:

[0074] The server generates the results and rationale for the automatic scoring.

[0075] Specifically, the system organizes the scores calculated by the AI ​​and the analytical results that form the basis for the scores, and generates feedback data based on this. For example, it clearly shows which areas received high marks and which areas need improvement.

[0076] Step 8:

[0077] The server provides feedback to the user.

[0078] Specifically, the generated feedback data is returned to the user's device, where the user can check the score and specific evaluation comments through a web interface.

[0079] Step 9:

[0080] The server uses artificial intelligence to generate a summary of the submitted documents.

[0081] Specifically, the clean text data is fed into a summary generation module, which extracts key points and summarizes them into short sentences, generating a summary of the submitted document.

[0082] Step 10:

[0083] The server provides the summary results to the grader.

[0084] Specifically, the generated summary data is provided to the grader's device, allowing the grader to quickly evaluate and compare multiple documents.

[0085] The above processing steps enable fair and efficient scoring and provide appropriate feedback and summaries to users and scorers.

[0086] Example 1

[0087] 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."

[0088] Traditionally, document screening and written exam scoring have been done manually, often resulting in a lack of fairness and efficiency. Furthermore, bias among evaluators and variations in evaluation criteria can be problematic. Furthermore, summarizing documents to quickly grasp their contents has also been done manually, requiring time and effort. To address these issues, the present invention aims to provide a fair and efficient system that automates the entire process, from document uploading to automatic scoring, feedback generation, and summary generation.

[0089] 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.

[0090] In this invention, the server includes means for uploading documents from a user terminal to the server, means for converting the uploaded documents into text format using optical character recognition software and performing text cleaning, means for reading scoring and review criteria and having a generative artificial intelligence learn the criteria, means for the generative artificial intelligence to analyze the cleaned text data and perform automatic scoring, means for generating feedback based on the results of the automatic scoring and the basis for that scoring, means for generating summaries of the submitted documents using the generative artificial intelligence, and means for providing the generated feedback and summaries to the user and evaluator, thereby enabling fair and efficient scoring and rapid summary creation.

[0091] A "user terminal" is a device such as a computer or smartphone that is operated by a user.

[0092] A "server" is a computer system that provides information in response to requests from clients over a network.

[0093] "Document" refers to a file containing information in text, images, and other formats.

[0094] "Optical character recognition software" is a program for extracting text information from images.

[0095] "Text format" is a format in which text data is stored and displayed.

[0096] "Text cleaning" is the process of removing unnecessary information from text data and formatting the data.

[0097] "Scoring and judging criteria" are rules and guidelines that define specific evaluation items and how they will be evaluated.

[0098] "Generative AI" is an AI technology that analyzes and generates data.

[0099] "Automatic scoring" is the process by which generative artificial intelligence calculates an evaluation score based on data.

[0100] "Feedback" refers to providing information about the evaluation results and their rationale.

[0101] A "summary" is a concise summary of the main points or content of a text.

[0102] "Evaluator" is the person or system responsible for evaluating a particular document or data.

[0103] This invention relates to a system that automates the grading of document screening and written tests. This system is primarily implemented using a user terminal and a server. Users upload documents using their terminal, and the server converts the documents into text format and performs text cleaning. The server then loads the grading and review criteria, trains a generative AI to learn the criteria, and the AI ​​performs automatic grading. Furthermore, the system generates feedback based on the scores and grading rationale, and also generates summaries of the submitted documents.

[0104] A specific embodiment will be described.

[0105] Users access a web interface using their devices and upload documents, often in PDF or image format. Once uploaded, the documents are sent to a server, which uses OCR software (e.g., commonly used optical character recognition software) to convert the uploaded documents into text format. The server then performs a text cleaning process, removing unnecessary information and normalizing the text.

[0106] The server then loads predefined scoring and judging criteria, defined as rubrics or guidelines and stored in a file format. The server then uses this to train a generative artificial intelligence (e.g., a commonly used generative AI model) to understand the criteria. This process teaches the AI ​​the fundamentals of analysis.

[0107] The cleaned text data is input into a generative AI model, which then performs automatic scoring. The generative AI analyzes the text data and calculates a score according to criteria. For example, it evaluates the accuracy and logic of the answer and determines the score comprehensively. The server generates feedback based on the results of the automatic scoring and the reasons for it. This feedback includes not only the score, but also the specific reasons for the scoring and areas for improvement, allowing the user to clearly understand what was evaluated and what areas need improvement.

[0108] It also has a function to generate summaries for submitted documents. The server uses a generative AI model to generate a concise summary of the main points and content of the text data, which can be used by graders and other stakeholders to quickly understand the content.

[0109] For example, imagine a student uploading an answer sheet for a written exam. After the user scans and uploads the answer sheet, the server uses OCR software to convert it into text format and cleans the text. The server then loads pre-prepared grading criteria and trains the AI ​​using a generative AI model. The AI ​​then automatically grades the answer sheet and generates specific feedback for the student. In addition, teachers are provided with a summary of the answer sheet, allowing them to quickly compare and contrast the answers of multiple students.

[0110] This system enables fair and efficient scoring and rapid summary creation, significantly reducing the time and effort required for evaluation and allowing evaluators to focus on more important tasks.

[0111] And here's an example prompt for using a generative AI model: "Upload student answer sheets, grade them, and generate feedback."

[0112] By entering this prompt, the system will automatically execute a series of processes and provide high-quality evaluation results and feedback.

[0113] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0114] Step 1:

[0115] The user accesses the web interface using a terminal and uploads a document. Specifically, the user clicks the "Upload Document" button on the screen to open a file browser. The user selects the document file (e.g., a PDF file) they want to upload and presses the "Open" button. The selected file is sent to the server.

[0116] Input: A document file selected by the user

[0117] Output: The document file is sent to the server.

[0118] Step 2:

[0119] The server receives the submitted document file. It converts the uploaded document file into text format using OCR software (e.g., optical character recognition software). The server then performs text cleaning, removing unnecessary information and normalizing the text. Specifically, it removes whitespace, normalizes special characters, and standardizes line breaks.

[0120] Input: Document file sent by user

[0121] Output: Cleaned text data

[0122] Step 3:

[0123] The server loads predefined scoring and judging criteria. These criteria are defined as rubrics or guidelines and saved in a file format. The server then uses these to train a generative AI model to understand the criteria. During this process, the AI ​​learns the evaluation criteria and scoring methods.

[0124] Input: Rubric or guideline file

[0125] Output: A generative AI model that understands the criteria

[0126] Step 4:

[0127] The server inputs the cleaned text data into the generative AI model, which then analyzes the text data and calculates a score based on the scoring criteria. Specifically, the model evaluates the content, logic, grammar, etc. of the text data to determine an overall score.

[0128] Input: Cleaned text data

[0129] Output:Score result

[0130] Step 5:

[0131] The server generates feedback based on the score calculated by the generative AI model and the rationale for it. This feedback includes not only the score but also specific reasons for the score and areas for improvement. This allows the user to clearly understand which aspects were evaluated and which areas need improvement.

[0132] Input: Score results and scoring basis

[0133] Output: Feedback text

[0134] Step 6:

[0135] The server then feeds the cleaned text data back into the generative AI model to generate summaries of the submitted documents, which can be used by graders and other stakeholders to quickly understand the content.

[0136] Input: Cleaned text data

[0137] Output: Summary text

[0138] Step 7:

[0139] The server provides the generated feedback and summary to the user and the evaluator by posting the feedback to the user's account and displaying the summary on the grader's dashboard, allowing both parties to quickly check the evaluation results and summary information.

[0140] Input: Feedback text, Summary text

[0141] Output: Provided to users and evaluators

[0142] (Application example 1)

[0143] 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."

[0144] Currently, quality control and inspection work in factories is heavily dependent on human resources, resulting in issues with consistency and efficiency of evaluation. In particular, it is difficult to quickly process large amounts of inspection data and provide accurate feedback. For this reason, a new system is needed to automate and streamline quality control.

[0145] 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.

[0146] In this invention, the server includes a means for uploading documents from a user terminal to the server, a means for converting the uploaded documents into text format and preprocessing them, a means for reading grading and examination criteria and having a generating AI understand the criteria, a means for the generating AI to analyze the preprocessed text according to the criteria and perform automatic grading, a means for generating feedback based on the results of the automatic grading and the basis for that, a means for generating summaries of the submitted documents using the generating AI, a means for providing the generated feedback and summaries to the user and grader, and a machine for collecting and uploading inspection data, which enables the automation and efficiency of quality control and inspection work.

[0147] "Uploading a document" is the act of transferring a document file from a user terminal to a server.

[0148] "User terminal" refers to a device such as a computer, smartphone, or tablet connected to the Internet.

[0149] A "server" is a central processing unit that receives, processes, stores, and provides data.

[0150] "Convert to text format" is the process of converting uploaded documents into text data.

[0151] "Preprocessing" is the process of preprocessing data to make it suitable for analysis.

[0152] "Scoring and evaluation criteria" are guidelines that include evaluation items and rules.

[0153] "Generative artificial intelligence" refers to AI systems that use natural language processing and machine learning to perform specific tasks.

[0154] "Understanding the standards" means having the AI ​​learn the scoring and judging criteria and then perform analysis according to those rules.

[0155] "Automatic scoring" is the process by which AI analyzes text and calculates a score based on pre-set criteria.

[0156] "Feedback" is response information that includes the score results and the basis for the evaluation.

[0157] A summary is a short sentence that succinctly summarizes the content of a longer piece of text.

[0158] "Inspection data" refers to information collected in factories and other locations that is used to evaluate the quality of products and processes.

[0159] A "machine" is a hardware-based device used to collect inspection data.

[0160] The present invention relates to a system for automating quality control and inspection work in factories. This system is installed on a factory robot, uploads product inspection data to a server, and a generative artificial intelligence (AI) analyzes and evaluates the data. Specific embodiments of the present invention will be described.

[0161] System Program

[0162] The system consists of the following components:

[0163] Document upload: Uploads inspection data from the user's device to the server.

[0164] Data conversion and pre-processing: The server converts the uploaded data into text format and removes unnecessary information to produce clean data suitable for analysis.

[0165] Loading of scoring and judging criteria and training of AI: The server loads the quality criteria and trains the generative AI to understand the criteria.

[0166] Automatic scoring: The server passes the preprocessed data to the AI, which analyzes it and performs automatic scoring.

[0167] Feedback generation: Based on the scoring results and their rationale, feedback is generated to be returned to the user.

[0168] Summary generation: The server generates a summary of the test data and provides it to the assessor.

[0169] Hardware and Software

[0170] Hardware: Factory robots, servers, user terminals

[0171] Software: Python, requests module, TextBlob, AI model (generative AI model)

[0172] Data processing and calculation

[0173] Data processing: The test data uploaded by the user's device is converted into text format on the server, and then pre-processed to remove noise and unnecessary information.

[0174] Data calculation: The server trains the AI ​​based on the scoring criteria, and the AI ​​analyzes and evaluates the test data. Based on the evaluation results, feedback and summaries are generated.

[0175] Specific examples

[0176] For example, a quality control officer at a factory collects product inspection data and uploads it from the user's device to a server. The uploaded data is converted into text format, preprocessed, and stored on the server as clean data. The server then trains an AI based on preset quality standards, which then analyzes the inspection data and automatically evaluates it. The evaluation results, along with a score, are generated as specific feedback and provided to the person in charge. A summary is also generated, allowing for quick quality evaluation.

[0177] Prompt Sentence Examples

[0178] "Factory robots collect inspection data and upload it to a server. The server converts the data into text format, preprocesses it, and then uses AI to automatically evaluate it. Create an application that generates a score and feedback on areas for improvement, as well as a summary of the inspection data."

[0179] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0180] Step 1:

[0181] The user's device collects the test data and uploads it to the server. Specifically, the camera or sensor on the user's device acquires the test data and saves it in a file format. The file is then sent to the server via the Internet. In this case, the input is the test data file, and the output is the test data saved on the server.

[0182] Step 2:

[0183] The server converts the uploaded inspection data into text format and performs preprocessing. Specifically, it first converts images and other data formats into text using optical character recognition (OCR) technology, then normalizes the text, checks spelling, and removes unnecessary information. The input is the inspection data file, and the output is clean text data.

[0184] Step 3:

[0185] The server loads the scoring and review criteria and has the generative AI understand the criteria. Specifically, it loads predefined quality criteria and provides them to the AI ​​model for training. The input is the quality criteria data, and the output is the trained AI model.

[0186] Step 4:

[0187] The server passes the preprocessed text data to the AI ​​model, which performs automatic scoring. Specifically, the clean text data is input into the evaluation algorithm, which generates an evaluation score and commentary. The input is the preprocessed text data, and the output is the evaluation result and its rationale.

[0188] Step 5:

[0189] The server generates feedback based on the results of the automatic scoring and the rationale behind it. Specifically, it generates points for improvement and specific advice from the evaluation results and compiles them into a feedback document. The input is the evaluation results and their rationale, and the output is the feedback document.

[0190] Step 6:

[0191] The server generates a summary of the submitted document. Specifically, it uses natural language processing techniques to extract important parts of the text and summarize them concisely. The input is preprocessed text data, and the output is a summary.

[0192] Step 7:

[0193] The server provides the generated feedback and summary to the user and evaluator. Specifically, it sends the generated feedback document and summary to the user and evaluator via email or a dedicated web interface. The input is the feedback document and summary, and the output is the information sent to the user and evaluator.

[0194] 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.

[0195] This invention combines a system that automates document screening and written test scoring with an emotion engine that recognizes user emotions. Feedback can be generated based on the user's emotion data, improving the quality of evaluation. The basic flow of this system is that the user uploads documents from their device, and the server converts them into text format and preprocesses them. The server then passes the reference data to a generative AI, which then automatically scores the documents. The system also includes a function to generate feedback based on the scores and scoring rationale, and to generate summaries of the submitted documents. The server also uses the emotion engine to analyze the user's emotion data and reflect this in the feedback.

[0196] User document upload

[0197] A user uploads documents to the server using a terminal, such as an answer sheet for a particular exam or a business proposal. The user can upload documents through a web interface, and the server receives and stores the uploaded documents.

[0198] Document text conversion and preprocessing

[0199] The server converts the uploaded documents into a text format, then preprocesses the text data, for example by removing unnecessary information and normalizing the text, to produce clean data suitable for analysis.

[0200] Loading scoring and judging criteria and training the AI

[0201] The server loads the scoring and judging criteria. These criteria are predefined rubrics or guidelines that the server uses to train the generative AI. The AI ​​understands these criteria and uses them as the basis for its analysis.

[0202] Run automatic grading

[0203] The server then passes the preprocessed text to a generative AI, which then automatically analyzes and scores the content, evaluating the content and logical consistency of the answers and calculating a score according to a rubric.

[0204] Generate feedback

[0205] Based on the results of the automatic scoring and the rationale behind it, the server generates feedback for the user. This feedback includes not only the score but also specific reasons for the scoring and suggestions for improvement, so the user can clearly understand what was evaluated and what areas need improvement.

[0206] Generate a summary of the submitted documents

[0207] The server uses artificial intelligence to generate summaries of submitted documents, which are provided to graders and other stakeholders for quick understanding.

[0208] Acquiring and analyzing emotion data

[0209] The server uses an emotion engine to recognize the user's emotions. For example, it collects emotion data from the user's facial expressions and voice and analyzes that data. This allows it to understand the user's emotional state.

[0210] Emotion-based feedback adjustment

[0211] The content and tone of the feedback can be adjusted based on the emotional data recognized by the emotion engine. For example, if the user is nervous, the feedback can be given in a calmer tone, taking into consideration the user's emotions.

[0212] Specific examples

[0213] For example, consider the case where a student uploads an answer sheet for a written exam. When the user scans and uploads the answer sheet from their device, the server converts it into text format, removes noise, and performs preprocessing. The server then trains an AI based on pre-defined scoring criteria, which then automatically scores the answer sheet. The AI ​​generates a score and specific scoring rationale, which the server returns to the user as feedback. Furthermore, an emotion engine recognizes the user's emotional data and adjusts the tone of the feedback based on that data. For example, if the user is feeling anxious, the feedback will be adjusted to include more encouraging words. An answer summary is also generated for the instructor, allowing them to quickly compare and contrast the answers of multiple students.

[0214] This system achieves fair and efficient scoring and provides appropriate feedback and summaries to users and graders. Furthermore, by using an emotion engine, it is possible to provide feedback that takes into account users' emotions, thereby improving satisfaction with the evaluation.

[0215] The processing flow will be explained below.

[0216] Step 1:

[0217] The user uploads the document from the device.

[0218] Specifically, users select document files such as answer sheets or proposals through the web interface and click the submit button. The server receives the uploaded files and saves them in a designated folder.

[0219] Step 2:

[0220] The server converts the uploaded document into text format.

[0221] Specifically, the saved file is read and images and PDF documents are converted into text format using OCR (Optical Character Recognition) technology, and the document contents are obtained as text data.

[0222] Step 3:

[0223] The server preprocesses the text data.

[0224] Specifically, unnecessary characters and spaces are removed from the acquired text data and normalized. For example, multiple spaces are combined into one and line breaks are standardized. This puts the text data into a clean format that is easy to analyze.

[0225] Step 4:

[0226] The server loads the scoring and judging criteria.

[0227] Specifically, predefined rubrics and scoring guidelines are stored as JSON or XML files, which the server reads to obtain baseline data that is used for subsequent analysis.

[0228] Step 5:

[0229] The server passes the reference data to the generative artificial intelligence and trains the AI.

[0230] Specifically, the imported reference data is input into the AI ​​model, and the model is trained using machine learning algorithms. At this stage, the AI ​​learns how to evaluate the data and builds the foundation for analysis.

[0231] Step 6:

[0232] The server passes the preprocessed text to a generative artificial intelligence for analysis.

[0233] Specifically, clean text data is fed into an AI model, which analyzes the content. The AI ​​then understands the context and calculates a score for each part according to a set of criteria, such as logical consistency and accuracy of information.

[0234] Step 7:

[0235] The server generates the results and rationale for the automatic scoring.

[0236] Specifically, the scoring results are organized based on the scores calculated by the AI ​​and the rationale for each. Along with the scores, specific comments are generated to explain why the scores were given.

[0237] Step 8:

[0238] The server provides feedback to the user.

[0239] Specifically, the generated feedback data is returned to the user's device, where the user can check the score and specific evaluation comments through a web interface. The feedback includes areas for improvement and areas that are highly rated.

[0240] Step 9:

[0241] The server uses artificial intelligence to generate a summary of the submitted documents.

[0242] Specifically, the clean text data is fed into a summary generation module, which extracts key points and summarizes them into short sentences, generating a summary of the submitted document.

[0243] Step 10:

[0244] The server provides the summary results to the grader.

[0245] Specifically, the generated summary data is provided to the grader's device, allowing the grader to quickly evaluate and compare multiple documents.

[0246] Step 11:

[0247] The server uses an emotion engine to recognize the user's emotion.

[0248] Specifically, the system analyzes the user's facial expressions and voice to obtain emotional data. For example, it analyzes the video and audio uploaded by the user via a webcam to identify the user's emotional state (joy, anger, sadness, surprise, etc.).

[0249] Step 12:

[0250] The server adjusts the content and tone of the feedback based on the emotional data.

[0251] Specifically, the content of the feedback is adjusted based on the emotional data recognized by the emotion engine. For example, if a user is feeling anxious, feedback containing many encouraging words is provided. In this way, appropriate feedback is generated that takes into account the user's emotions.

[0252] Specific examples

[0253] For example, consider the case where a student uploads an answer sheet for a written exam. The user scans the answer sheet from their device and uploads it to the server, which converts the document into text format and preprocesses it by removing unnecessary information. The server then trains an AI system using pre-defined scoring criteria, which then automatically scores the answer sheet. The server generates feedback based on the scoring results and specific reasons and provides it to the user. Furthermore, an emotion engine recognizes emotions from the user's facial expressions and voice, and adjusts the tone of the feedback based on that data. For example, if the user is feeling anxious, the feedback could include more encouraging words. An answer summary is generated for the instructor, allowing them to quickly compare and consider multiple students' answers.

[0254] This system not only achieves fair and efficient scoring, but also provides feedback that takes users' emotions into consideration, improving satisfaction with the evaluation and allowing evaluators to focus on higher-level tasks.

[0255] Example 2

[0256] 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."

[0257] Conventional scoring systems rely on manual scoring, which inevitably leads to human error and bias. Furthermore, they lack sufficient summaries to quickly grasp the content of submitted documents, placing a heavy workload on the scorer. Furthermore, they are unable to provide feedback that takes into account the user's emotions, which hinders the improvement of satisfaction.

[0258] The specific processing by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes a means for uploading documents from an information terminal to the information processing device, a means for converting the uploaded documents into electronic text format and performing preprocessing, a means for reading grading and examination criteria and having a machine intelligence understand the criteria, a means for the machine intelligence to analyze the preprocessed electronic text according to the criteria and perform automatic grading, a means for generating feedback based on the results of the automatic grading and the basis thereof, a means for generating a summary of the submitted document using the machine intelligence, a means for providing the generated feedback and summary to the user and the grader, a means for acquiring and analyzing the user's emotions, and a means for adjusting the content and tone of the feedback based on the user's emotions. This enables fair and effective automatic grading and the provision of feedback that takes the user's emotions into consideration.

[0259] "Uploading a document" refers to the act of electronically transmitting a document from an information terminal to an information processing device.

[0260] "Information terminal" refers to electronic devices such as computers, smartphones, and tablets, which are devices operated by users.

[0261] An "information processing device" is a server that receives, stores, and processes transmitted documents.

[0262] An "electronic text format" is a format in which the contents of a document are expressed as digital data.

[0263] "Preprocessing" is the process of processing data to make documents converted into electronic text format easier to analyze.

[0264] "Grading and Review Criteria" refers to rubrics or guidelines for evaluating the content of a document based on specific criteria.

[0265] "Machine intelligence" is an artificial intelligence technology that learns and analyzes large amounts of data to perform specific tasks.

[0266] "Automatic scoring" is the process by which machine intelligence analyzes documents according to scoring and examination criteria and calculates a score.

[0267] "Feedback" is information that returns the document evaluation results and the basis for them to the user.

[0268] "Summary" refers to a concise summary of the main points or contents of the submitted documents.

[0269] "Emotion data" is information that represents the emotional state of a user, obtained from facial expressions, voice, context, and the like.

[0270] "Adjusting the content and tone of feedback" means changing the sentences and wording of the feedback depending on the user's emotional state.

[0271] The present invention is a system for implementing a series of processes including uploading documents and automatic marking. Detailed embodiments of the system are described below.

[0272] Uploading documents

[0273] Users use their devices to scan or electronically create documents and upload them to the server through a web interface. These documents can include answer sheets for written exams or business proposals.

[0274] Server: Receives uploaded documents and stores the data. The server checks the data for consistency and stores it in the appropriate folder.

[0275] Document text conversion and preprocessing

[0276] Server: The server uses OCR (Optical Character Recognition) software "Tesseract OCR" to convert the received documents into text format. The converted electronic text undergoes pre-processing to make it suitable for analysis. Pre-processing includes:

[0277] Removal of unnecessary information (e.g., page numbers and margins)

[0278] Text normalization (e.g., making all characters lowercase)

[0279] Loading scoring and judging criteria and training the AI

[0280] Server: Loads the scoring and judging criteria and trains a generative AI model (e.g., GPT-4) according to predefined rubrics and guidelines. This training process allows the AI ​​to understand the criteria and perform accurate analysis.

[0281] Run automatic grading

[0282] Server: Passes the preprocessed electronic text to the generative AI model. The AI ​​analyzes the text data and automatically scores it based on specific criteria. For example, it evaluates the logic of the text and the coherence of the content, and generates a score and its rationale.

[0283] Generate feedback

[0284] Server: Generates feedback to be returned to the user based on the automated scoring results and the rationale behind them. This feedback includes not only the score but also specific scoring rationale and areas for improvement. For example, it provides detailed comments such as "You received a high score because the logical structure of your answer was clear."

[0285] Generate a summary of the submitted documents

[0286] Server: Requests the AI ​​model to generate a summary of the submitted documents. The summary is intended to allow graders and other stakeholders to quickly understand the content, and concisely presents the main points and conclusions.

[0287] Acquiring and analyzing emotion data

[0288] Server: The server uses an emotion engine (such as IBM Watson Tone Analyzer or Microsoft Azure Face API) to obtain the user's emotional state. This is done by analyzing facial expressions and voice, and it is possible to understand the user's emotional state.

[0289] Emotion-based feedback adjustment

[0290] Server: Adjust the content and tone of the feedback based on the emotional data. For example, if the user is nervous, provide feedback in a gentler tone. A specific example would be something like, "Thank you for your hard work. Overall, you did a good job, but if you improve this next time, you'll get even better results."

[0291] Examples of concrete examples and prompts

[0292] Example: Consider a student uploading an answer sheet for a final exam. The user (student) scans and uploads the answer sheet from their device, and the server converts it into text format and performs preprocessing. Next, the server trains a generative AI model based on pre-defined scoring criteria, and the AI ​​automatically scores the answer sheet. The AI ​​generates a score and specific scoring reasons, which the server returns to the user as feedback. Furthermore, an emotion engine recognizes the user's emotional data and adjusts the tone of the feedback based on that data. For example, if the user is feeling anxious, the feedback is adjusted to include more encouraging words. An answer summary is also generated for the teacher, allowing them to quickly compare and consider the answers of multiple students.

[0293] Example prompt sentence:

[0294] "Upload students' answer sheets, grade them based on the grading criteria below, and generate feedback. Please be gentle with the tone of your feedback for students who seem nervous."

[0295] In this way, the present invention achieves fair and effective automatic scoring and provides feedback that takes into account the user's emotions. By using an emotion engine, flexible responses can be made according to the user's emotional state, improving satisfaction with the evaluation.

[0296] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0297] Step 1: Upload your documents

[0298] User:

[0299] A user scans or electronically creates a document using a terminal and uploads it to the server through a web interface. The input is a PDF or image file uploaded by the user, and the output is the document data stored on the server. Specifically, the user clicks a specific button, selects a file from a file selection dialog, and submits it.

[0300] Step 2: Text conversion and preprocessing of documents

[0301] server:

[0302] When the server receives the uploaded document, it converts the document into electronic text using OCR software such as "Tesseract OCR." The input is the saved document data, and the output is the converted text data. Specifically, the server inputs the document into the OCR software and temporarily stores the resulting text data.

[0303] server:

[0304] Next, the converted text data is preprocessed. The input is the converted text data, and the output is the preprocessed, clean text data. Specific preprocessing steps include removing unnecessary information (e.g., page numbers and margins) and normalizing the text (e.g., changing all characters to lowercase).

[0305] Step 3: Importing the scoring and judging criteria and training the AI

[0306] server:

[0307] The server loads the scoring and judging criteria and trains the generative AI model. The input is a predefined rubric or guideline, and the output is a trained AI model that understands the criteria. Specifically, the server provides the rubric data to the AI ​​training algorithm, which then learns the criteria.

[0308] Step 4: Run Auto-Scoring

[0309] server:

[0310] The server passes the preprocessed text data to a generative AI model, which analyzes the content and automatically scores the answers. The input is the preprocessed text data and the trained AI model, and the output is the score and the basis for the scoring. Specifically, the AI ​​analyzes the text data and calculates a score based on the content of each answer.

[0311] Step 5: Generate feedback

[0312] server:

[0313] The server generates feedback to be returned to the user based on the results of the automatic scoring and the rationale behind it. The input is the score and the rationale for the scoring, and the output is a feedback message. Specifically, it generates detailed comments such as "Your answer had a clear logical structure, so you got a high score," and provides them to the user.

[0314] Step 6: Generate a summary of the submission

[0315] server:

[0316] The server requests an AI model to generate a summary of the submitted document. The input is preprocessed text data, and the output is summary data. Specifically, the AI ​​analyzes the entire text, extracts key points and conclusions, and generates a summary.

[0317] Step 7: Acquire and analyze emotion data

[0318] server:

[0319] The server uses an emotion engine to acquire and analyze the user's emotional data. The input is the user's facial expressions and voice data, and the output is the analysis result of the user's emotional state. Specifically, data collected from the user's webcam and microphone is input into the analysis engine to determine the user's emotional state.

[0320] Step 8: Adjust your feedback based on emotion

[0321] server:

[0322] The server adjusts the content and tone of the feedback based on the emotional data. The input is the analysis result of the emotional state and the feedback message, and the output is the adjusted feedback message. Specifically, if the user is nervous, the server provides feedback in a gentle tone and includes encouraging words.

[0323] The above steps clearly explain how the system handles uploading documents, automatically scoring them, and providing feedback. It also takes into account users' emotional data to provide more personalized feedback.

[0324] (Application example 2)

[0325] 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."

[0326] Conventional scoring systems for document screening and written tests have limitations in the fairness and efficiency of scoring, making it difficult to provide feedback that takes emotions into account. Furthermore, automated analysis of customer surveys cannot take emotional data into account, making it difficult to accurately grasp customer satisfaction and areas for improvement. The present invention aims to solve these problems by providing a system that achieves fair and efficient scoring and enables the provision of emotional feedback.

[0327] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0328] In this invention, the server includes means for uploading documents from a user terminal to the server, means for converting the uploaded documents into text format and preprocessing them, means for reading grading and review criteria and having a generating AI understand the criteria, means for the generating AI to analyze the preprocessed text according to the criteria and perform automatic grading, means for generating feedback based on the results of the automatic grading and the basis for the results, means for generating summaries of the submitted documents using the generating AI, means for providing the generated feedback and summaries to the user and grader, means for using an emotion analysis engine to acquire user emotion data, and means for adjusting the tone and content of the feedback based on the emotion data, thereby enabling fair and efficient automatic grading and providing feedback that takes user emotions into consideration.

[0329] "Uploading a document" is the act of transferring document data from a user terminal to a server.

[0330] "Server" means a central computer for storing, processing, and analyzing document data.

[0331] "Convert to text format" refers to converting uploaded document image data, PDFs, etc. into text data.

[0332] "Preprocessing" refers to preprocessing such as normalizing text data and removing noise.

[0333] "Grading and review criteria" refers to pre-established rubrics and guidelines for evaluating the content of a document.

[0334] "Generative AI" is an AI model that is trained to perform a specific task.

[0335] "Automatic scoring" is the process of using AI to evaluate the content of a document and calculate a score.

[0336] "Feedback" is response information to the user, including the results of the assessment, a detailed explanation of the assessment, and points for improvement.

[0337] An "abstract" is a concise summary of the main contents of the submitted document.

[0338] The "emotion analysis engine" is a system for collecting and analyzing emotional data from a user's facial expressions and voice.

[0339] "Adjusting the tone and content of feedback" refers to optimizing the expression of feedback based on emotional data and providing an appropriate response according to the user's emotional state.

[0340] The following specific procedures and systems are used to implement the present invention: A server, a user terminal, and appropriate software and hardware are used in combination.

[0341] First, a user uploads document data to the server using their own device (such as a smartphone or tablet). The uploaded document is then converted into text format by the server. This conversion process uses optical character recognition (OCR) technology. The server then preprocesses the text data, removing unnecessary information and normalizing the text, among other processes. This generates clean data suitable for analysis.

[0342] The server trains a generative artificial intelligence (AI model) based on pre-set scoring and evaluation criteria. This trained AI model analyzes the pre-processed text according to the criteria and automatically scores the answers. The AI ​​model evaluates the content and logical consistency of the answers and calculates a score according to the rubric.

[0343] Based on the results of the automated scoring and the rationale behind it, the server generates feedback. This feedback includes not only the score but also specific scoring rationale and areas for improvement. The server also uses generative artificial intelligence to generate a summary of the submitted document, allowing graders and other stakeholders to quickly understand the content.

[0344] Furthermore, the server uses an emotion analysis engine to obtain the user's emotional data. Emotional data is collected and analyzed from the user's facial expressions and voice. This allows the server to understand the user's emotional state. Based on this emotional data, the server adjusts the tone and content of the feedback. For example, if the user is feeling anxious, the server will provide feedback in a calm tone.

[0345] A concrete example is a system that analyzes customer surveys in stores. Customers fill out and upload the survey using tablets provided in the store or their own smartphones. The server converts the survey data into text format, performs preprocessing, and then analyzes the data using an AI model to identify customer satisfaction levels and areas for improvement. Furthermore, a sentiment analysis engine is used to analyze the customer's facial expressions and voice, allowing the tone and content of the feedback to be adjusted to provide more personalized service.

[0346] As an example of a prompt sentence, in response to the feedback "The food was delicious, but the service was slow," feedback such as "Thank you for rating the food delicious. We will try to improve the slowness of the service" can be generated.

[0347] This enables fair and efficient automatic scoring and provides feedback that takes into account the user's feelings.

[0348] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0349] Step 1:

[0350] Users use their own devices (smartphones or tablets) to enter and upload document data.

[0351] Input: Images or PDF files of survey data and answer sheets

[0352] Output: Document data sent to the server

[0353] The server receives the uploaded document data, and the data sent from the user terminal is stored in a designated storage area on the server.

[0354] Step 2:

[0355] The server converts the received document data into text format.

[0356] Input: Document data (image or PDF)

[0357] Output: Text data

[0358] Specifically, the server uses optical character recognition (OCR) software (e.g., Tesseract OCR) to convert images and PDF files into text data, resulting in text data in a format suitable for analysis.

[0359] Step 3:

[0360] The server preprocesses the text data.

[0361] Input: Text data

[0362] Output: Cleaned text data

[0363] Pre-processing includes removing unnecessary information, normalizing text, checking spelling, etc. The server performs data cleansing using PANDAS and regular expression libraries.

[0364] Step 4:

[0365] The server loads the scoring and judging criteria and trains the AI ​​model.

[0366] Input: Scoring and judging criteria data

[0367] Output: A trained AI model

[0368] The server loads pre-prepared rubrics and guidelines and trains a generative AI model (e.g., BERT or GPT). Training also includes setting scores for each evaluation item and generating evaluation comments.

[0369] Step 5:

[0370] The server performs automatic scoring based on the preprocessed text.

[0371] Input: cleaned text data, trained AI model

[0372] Output: Score and grading rationale

[0373] Specifically, the server uses a trained generative AI model to analyze the content of the text data and generate a score and its rationale according to the rubric.

[0374] Step 6:

[0375] The server generates feedback based on the results of the automatic scoring and the reasons for it.

[0376] Input: Score and scoring rationale

[0377] Output: Feedback data

[0378] The feedback includes not only the score but also specific reasons for the score and areas for improvement. The server uses a generative AI model to generate sentences and create feedback to provide to the user.

[0379] Step 7:

[0380] The server generates a summary of the submission.

[0381] Input: Cleaned text data

[0382] Output: Summary data

[0383] To generate a summary, the server uses a generative AI model (e.g., a news article summarization model) to extract key points and create a concise summary.

[0384] Step 8:

[0385] The server acquires and analyzes the user's emotion data.

[0386] Input: User facial expression images, voice data

[0387] Output: Emotion data

[0388] The server uses an emotion analysis engine (e.g., OpenCV or TensorFlow) to analyze the user's facial expressions and voice obtained from smart glasses or a tablet to understand their emotional state.

[0389] Step 9:

[0390] The server adjusts the tone and content of the feedback based on the emotional data.

[0391] Input: Emotion data, feedback data

[0392] Output: Regulated feedback data

[0393] Adjust the content and tone of the feedback depending on the user's emotional state. For example, provide calmer, more encouraging feedback to a user who is feeling anxious.

[0394] Step 10:

[0395] The server provides the generated feedback and summaries to the user and grader.

[0396] Input: Adjusted feedback data, summary data

[0397] Output: Feedback and summaries provided to users and graders

[0398] The server generates feedback and summaries, which are then sent to the user's device and provided to the grader, enabling fair and efficient evaluation and emotionally sensitive feedback.

[0399] 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.

[0400] 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.

[0401] 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.

[0402] [Second embodiment]

[0403] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0404] 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.

[0405] 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).

[0406] 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.

[0407] 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.

[0408] 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).

[0409] 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.

[0410] 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.

[0411] 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.

[0412] 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.

[0413] 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.

[0414] 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."

[0415] This invention relates to a system that automates the grading of document screening and written exams. In this system, documents are uploaded from the user's terminal, and the server converts the documents into text format and performs preprocessing. The server then loads the grading and examination criteria, and has the AI ​​generative system understand the criteria, which then performs automatic grading. The system also includes a function to generate feedback based on the scores and grading rationale, and to generate summaries of the submitted documents.

[0416] User document upload

[0417] A user uploads documents to the server using a terminal, such as an answer sheet for a particular exam or a business proposal. The user can upload the documents through a web interface, and the server receives the uploaded documents.

[0418] Document text conversion and preprocessing

[0419] The server converts the uploaded documents into a text format, then preprocesses the text data, for example by removing unnecessary information and normalizing the text, to produce clean data suitable for analysis.

[0420] Loading scoring and judging criteria and training the AI

[0421] The server loads the scoring and judging criteria. These criteria are predefined rubrics or guidelines that the server uses to train the generative AI. The AI ​​understands these criteria and uses them as the basis for its analysis.

[0422] Run automatic grading

[0423] The server then passes the preprocessed text to a generative AI, which then automatically analyzes and scores the content, evaluating the content and logical consistency of the answers and calculating a score according to a rubric.

[0424] Generate feedback

[0425] Based on the results of the automatic scoring and the rationale behind it, the server generates feedback for the user. This feedback includes not only the score but also specific reasons for the scoring and suggestions for improvement, so the user can clearly understand what was evaluated and what areas need improvement.

[0426] Generate a summary of the submitted documents

[0427] Additionally, the server uses a generator to generate summaries of the submitted documents, which are provided by the server to graders and other interested parties for quick understanding.

[0428] Specific examples

[0429] For example, consider the case where a student uploads an answer sheet for a written exam. When the user scans and uploads the answer sheet from their device, the server converts it into text format, removes noise, and performs preprocessing. The server then trains an AI based on pre-defined scoring criteria, which then automatically scores the answer sheet. As a result, specific feedback is generated and provided to the student along with a score. An answer summary is also generated for the instructor, allowing them to quickly compare and consider the answers of multiple students.

[0430] This system ensures fair and efficient scoring, increases the credibility of the assessment, and allows assessors to focus on more important tasks.

[0431] The processing flow will be explained below.

[0432] Step 1:

[0433] The user uploads the document from the device.

[0434] Specifically, this is done by selecting document files such as answer sheets and proposals through the web interface and clicking the send button to the server, which receives and stores the uploaded files.

[0435] Step 2:

[0436] The server converts the uploaded document into text format.

[0437] Specifically, the saved file is read and images and PDF documents are converted to text format using OCR (Optical Character Recognition) technology, thereby obtaining the text data.

[0438] Step 3:

[0439] The server preprocesses the text data.

[0440] Specifically, unnecessary characters and spaces are removed from the acquired text data and normalization is performed. For example, this includes processing to combine multiple spaces into one and standardize line breaks. This processing generates clean text data that is easy to analyze.

[0441] Step 4:

[0442] The server loads the scoring and judging criteria.

[0443] Specifically, predefined rubrics and grading guidelines are stored as JSON or XML format files, which the server reads and treats as data, making the evaluation criteria clear.

[0444] Step 5:

[0445] The server passes the reference data to the generative artificial intelligence and trains the AI.

[0446] Specifically, the reference data is fed into a generative AI model, and machine learning techniques are used to train the model. At this stage, the AI ​​learns how to make evaluations.

[0447] Step 6:

[0448] The server passes the preprocessed text to a generative artificial intelligence for analysis.

[0449] Specifically, clean text data is fed into an AI model, which analyzes the content, understands the context, and calculates a score for each part based on a scoring criteria.

[0450] Step 7:

[0451] The server generates the results and rationale for the automatic scoring.

[0452] Specifically, the system organizes the scores calculated by the AI ​​and the analytical results that form the basis for the scores, and generates feedback data based on this. For example, it clearly shows which areas received high marks and which areas need improvement.

[0453] Step 8:

[0454] The server provides feedback to the user.

[0455] Specifically, the generated feedback data is returned to the user's device, where the user can check the score and specific evaluation comments through a web interface.

[0456] Step 9:

[0457] The server uses artificial intelligence to generate a summary of the submitted documents.

[0458] Specifically, the clean text data is fed into a summary generation module, which extracts key points and summarizes them into short sentences, generating a summary of the submitted document.

[0459] Step 10:

[0460] The server provides the summary results to the grader.

[0461] Specifically, the generated summary data is provided to the grader's device, allowing the grader to quickly evaluate and compare multiple documents.

[0462] The above processing steps enable fair and efficient scoring and provide appropriate feedback and summaries to users and scorers.

[0463] Example 1

[0464] 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."

[0465] Traditionally, document screening and written exam scoring have been done manually, often resulting in a lack of fairness and efficiency. Furthermore, bias among evaluators and variations in evaluation criteria can be problematic. Furthermore, summarizing documents to quickly grasp their contents has also been done manually, requiring time and effort. To address these issues, the present invention aims to provide a fair and efficient system that automates the entire process, from document uploading to automatic scoring, feedback generation, and summary generation.

[0466] 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.

[0467] In this invention, the server includes means for uploading documents from a user terminal to the server, means for converting the uploaded documents into text format using optical character recognition software and performing text cleaning, means for reading scoring and review criteria and having a generative artificial intelligence learn the criteria, means for the generative artificial intelligence to analyze the cleaned text data and perform automatic scoring, means for generating feedback based on the results of the automatic scoring and the basis for that scoring, means for generating summaries of the submitted documents using the generative artificial intelligence, and means for providing the generated feedback and summaries to the user and evaluator, thereby enabling fair and efficient scoring and rapid summary creation.

[0468] A "user terminal" is a device such as a computer or smartphone that is operated by a user.

[0469] A "server" is a computer system that provides information in response to requests from clients over a network.

[0470] "Document" refers to a file containing information in text, images, and other formats.

[0471] "Optical character recognition software" is a program for extracting text information from images.

[0472] "Text format" is a format in which text data is stored and displayed.

[0473] "Text cleaning" is the process of removing unnecessary information from text data and formatting the data.

[0474] "Scoring and judging criteria" are rules and guidelines that define specific evaluation items and how they will be evaluated.

[0475] "Generative AI" is an AI technology that analyzes and generates data.

[0476] "Automatic scoring" is the process by which generative artificial intelligence calculates an evaluation score based on data.

[0477] "Feedback" refers to providing information about the evaluation results and their rationale.

[0478] A "summary" is a concise summary of the main points or content of a text.

[0479] "Evaluator" is the person or system responsible for evaluating a particular document or data.

[0480] This invention relates to a system that automates the grading of document screening and written tests. This system is primarily implemented using a user terminal and a server. Users upload documents using their terminal, and the server converts the documents into text format and performs text cleaning. The server then loads the grading and review criteria, trains a generative AI to learn the criteria, and the AI ​​performs automatic grading. Furthermore, the system generates feedback based on the scores and grading rationale, and also generates summaries of the submitted documents.

[0481] A specific embodiment will be described.

[0482] Users access a web interface using their devices and upload documents, often in PDF or image format. Once uploaded, the documents are sent to a server, which uses OCR software (e.g., commonly used optical character recognition software) to convert the uploaded documents into text format. The server then performs a text cleaning process, removing unnecessary information and normalizing the text.

[0483] The server then loads predefined scoring and judging criteria, defined as rubrics or guidelines and stored in a file format. The server then uses this to train a generative artificial intelligence (e.g., a commonly used generative AI model) to understand the criteria. This process teaches the AI ​​the fundamentals of analysis.

[0484] The cleaned text data is input into a generative AI model, which then performs automatic scoring. The generative AI analyzes the text data and calculates a score according to criteria. For example, it evaluates the accuracy and logic of the answer and determines the score comprehensively. The server generates feedback based on the results of the automatic scoring and the reasons for it. This feedback includes not only the score, but also the specific reasons for the scoring and areas for improvement, allowing the user to clearly understand what was evaluated and what areas need improvement.

[0485] It also has a function to generate summaries for submitted documents. The server uses a generative AI model to generate a concise summary of the main points and content of the text data, which can be used by graders and other stakeholders to quickly understand the content.

[0486] For example, imagine a student uploading an answer sheet for a written exam. After the user scans and uploads the answer sheet, the server uses OCR software to convert it into text format and cleans the text. The server then loads pre-prepared grading criteria and trains the AI ​​using a generative AI model. The AI ​​then automatically grades the answer sheet and generates specific feedback for the student. In addition, teachers are provided with a summary of the answer sheet, allowing them to quickly compare and contrast the answers of multiple students.

[0487] This system enables fair and efficient scoring and rapid summary creation, significantly reducing the time and effort required for evaluation and allowing evaluators to focus on more important tasks.

[0488] And here's an example prompt for using a generative AI model: "Upload student answer sheets, grade them, and generate feedback."

[0489] By entering this prompt, the system will automatically execute a series of processes and provide high-quality evaluation results and feedback.

[0490] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0491] Step 1:

[0492] The user accesses the web interface using a terminal and uploads a document. Specifically, the user clicks the "Upload Document" button on the screen to open a file browser. The user selects the document file (e.g., a PDF file) they want to upload and presses the "Open" button. The selected file is sent to the server.

[0493] Input: A document file selected by the user

[0494] Output: The document file is sent to the server.

[0495] Step 2:

[0496] The server receives the submitted document file. It converts the uploaded document file into text format using OCR software (e.g., optical character recognition software). The server then performs text cleaning, removing unnecessary information and normalizing the text. Specifically, it removes whitespace, normalizes special characters, and standardizes line breaks.

[0497] Input: Document file sent by user

[0498] Output: Cleaned text data

[0499] Step 3:

[0500] The server loads predefined scoring and judging criteria. These criteria are defined as rubrics or guidelines and saved in a file format. The server then uses these to train a generative AI model to understand the criteria. During this process, the AI ​​learns the evaluation criteria and scoring methods.

[0501] Input: Rubric or guideline file

[0502] Output: A generative AI model that understands the criteria

[0503] Step 4:

[0504] The server inputs the cleaned text data into the generative AI model, which then analyzes the text data and calculates a score based on the scoring criteria. Specifically, the model evaluates the content, logic, grammar, etc. of the text data to determine an overall score.

[0505] Input: Cleaned text data

[0506] Output:Score result

[0507] Step 5:

[0508] The server generates feedback based on the score calculated by the generative AI model and the rationale for it. This feedback includes not only the score but also specific reasons for the score and areas for improvement. This allows the user to clearly understand which aspects were evaluated and which areas need improvement.

[0509] Input: Score results and scoring basis

[0510] Output: Feedback text

[0511] Step 6:

[0512] The server then feeds the cleaned text data back into the generative AI model to generate summaries of the submitted documents, which can be used by graders and other stakeholders to quickly understand the content.

[0513] Input: Cleaned text data

[0514] Output: Summary text

[0515] Step 7:

[0516] The server provides the generated feedback and summary to the user and the evaluator by posting the feedback to the user's account and displaying the summary on the grader's dashboard, allowing both parties to quickly check the evaluation results and summary information.

[0517] Input: Feedback text, Summary text

[0518] Output: Provided to users and evaluators

[0519] (Application example 1)

[0520] 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."

[0521] Currently, quality control and inspection work in factories is heavily dependent on human resources, resulting in issues with consistency and efficiency of evaluation. In particular, it is difficult to quickly process large amounts of inspection data and provide accurate feedback. For this reason, a new system is needed to automate and streamline quality control.

[0522] 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.

[0523] In this invention, the server includes a means for uploading documents from a user terminal to the server, a means for converting the uploaded documents into text format and preprocessing them, a means for reading grading and examination criteria and having a generating AI understand the criteria, a means for the generating AI to analyze the preprocessed text according to the criteria and perform automatic grading, a means for generating feedback based on the results of the automatic grading and the basis for that, a means for generating summaries of the submitted documents using the generating AI, a means for providing the generated feedback and summaries to the user and grader, and a machine for collecting and uploading inspection data, which enables the automation and efficiency of quality control and inspection work.

[0524] "Uploading a document" is the act of transferring a document file from a user terminal to a server.

[0525] "User terminal" refers to a device such as a computer, smartphone, or tablet connected to the Internet.

[0526] A "server" is a central processing unit that receives, processes, stores, and provides data.

[0527] "Convert to text format" is the process of converting uploaded documents into text data.

[0528] "Preprocessing" is the process of preprocessing data to make it suitable for analysis.

[0529] "Scoring and evaluation criteria" are guidelines that include evaluation items and rules.

[0530] "Generative artificial intelligence" refers to AI systems that use natural language processing and machine learning to perform specific tasks.

[0531] "Understanding the standards" means having the AI ​​learn the scoring and judging criteria and then perform analysis according to those rules.

[0532] "Automatic scoring" is the process by which AI analyzes text and calculates a score based on pre-set criteria.

[0533] "Feedback" is response information that includes the score results and the basis for the evaluation.

[0534] A summary is a short sentence that succinctly summarizes the content of a longer piece of text.

[0535] "Inspection data" refers to information collected in factories and other locations that is used to evaluate the quality of products and processes.

[0536] A "machine" is a hardware-based device used to collect inspection data.

[0537] The present invention relates to a system for automating quality control and inspection work in factories. This system is installed on a factory robot, uploads product inspection data to a server, and a generative artificial intelligence (AI) analyzes and evaluates the data. Specific embodiments of the present invention will be described.

[0538] System Program

[0539] The system consists of the following components:

[0540] Document upload: Uploads inspection data from the user's device to the server.

[0541] Data conversion and pre-processing: The server converts the uploaded data into text format and removes unnecessary information to produce clean data suitable for analysis.

[0542] Loading of scoring and judging criteria and training of AI: The server loads the quality criteria and trains the generative AI to understand the criteria.

[0543] Automatic scoring: The server passes the preprocessed data to the AI, which analyzes it and performs automatic scoring.

[0544] Feedback generation: Based on the scoring results and their rationale, feedback is generated to be returned to the user.

[0545] Summary generation: The server generates a summary of the test data and provides it to the assessor.

[0546] Hardware and Software

[0547] Hardware: Factory robots, servers, user terminals

[0548] Software: Python, requests module, TextBlob, AI model (generative AI model)

[0549] Data processing and calculation

[0550] Data processing: The test data uploaded by the user's device is converted into text format on the server, and then pre-processed to remove noise and unnecessary information.

[0551] Data calculation: The server trains the AI ​​based on the scoring criteria, and the AI ​​analyzes and evaluates the test data. Based on the evaluation results, feedback and summaries are generated.

[0552] Specific examples

[0553] For example, a quality control officer at a factory collects product inspection data and uploads it from the user's device to a server. The uploaded data is converted into text format, preprocessed, and stored on the server as clean data. The server then trains an AI based on preset quality standards, which then analyzes the inspection data and automatically evaluates it. The evaluation results, along with a score, are generated as specific feedback and provided to the person in charge. A summary is also generated, allowing for quick quality evaluation.

[0554] Prompt Sentence Examples

[0555] "Factory robots collect inspection data and upload it to a server. The server converts the data into text format, preprocesses it, and then uses AI to automatically evaluate it. Create an application that generates a score and feedback on areas for improvement, as well as a summary of the inspection data."

[0556] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0557] Step 1:

[0558] The user's device collects the test data and uploads it to the server. Specifically, the camera or sensor on the user's device acquires the test data and saves it in a file format. The file is then sent to the server via the Internet. In this case, the input is the test data file, and the output is the test data saved on the server.

[0559] Step 2:

[0560] The server converts the uploaded inspection data into text format and performs preprocessing. Specifically, it first converts images and other data formats into text using optical character recognition (OCR) technology, then normalizes the text, checks spelling, and removes unnecessary information. The input is the inspection data file, and the output is clean text data.

[0561] Step 3:

[0562] The server loads the scoring and review criteria and has the generative AI understand the criteria. Specifically, it loads predefined quality criteria and provides them to the AI ​​model for training. The input is the quality criteria data, and the output is the trained AI model.

[0563] Step 4:

[0564] The server passes the preprocessed text data to the AI ​​model, which performs automatic scoring. Specifically, the clean text data is input into the evaluation algorithm, which generates an evaluation score and commentary. The input is the preprocessed text data, and the output is the evaluation result and its rationale.

[0565] Step 5:

[0566] The server generates feedback based on the results of the automatic scoring and the rationale behind it. Specifically, it generates points for improvement and specific advice from the evaluation results and compiles them into a feedback document. The input is the evaluation results and their rationale, and the output is the feedback document.

[0567] Step 6:

[0568] The server generates a summary of the submitted document. Specifically, it uses natural language processing techniques to extract important parts of the text and summarize them concisely. The input is preprocessed text data, and the output is a summary.

[0569] Step 7:

[0570] The server provides the generated feedback and summary to the user and evaluator. Specifically, it sends the generated feedback document and summary to the user and evaluator via email or a dedicated web interface. The input is the feedback document and summary, and the output is the information sent to the user and evaluator.

[0571] 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.

[0572] This invention combines a system that automates document screening and written test scoring with an emotion engine that recognizes user emotions. Feedback can be generated based on the user's emotion data, improving the quality of evaluation. The basic flow of this system is that the user uploads documents from their device, and the server converts them into text format and preprocesses them. The server then passes the reference data to a generative AI, which then automatically scores the documents. The system also includes a function to generate feedback based on the scores and scoring rationale, and to generate summaries of the submitted documents. The server also uses the emotion engine to analyze the user's emotion data and reflect this in the feedback.

[0573] User document upload

[0574] A user uploads documents to the server using a terminal, such as an answer sheet for a particular exam or a business proposal. The user can upload documents through a web interface, and the server receives and stores the uploaded documents.

[0575] Document text conversion and preprocessing

[0576] The server converts the uploaded documents into a text format, then preprocesses the text data, for example by removing unnecessary information and normalizing the text, to produce clean data suitable for analysis.

[0577] Loading scoring and judging criteria and training the AI

[0578] The server loads the scoring and judging criteria. These criteria are predefined rubrics or guidelines that the server uses to train the generative AI. The AI ​​understands these criteria and uses them as the basis for its analysis.

[0579] Run automatic grading

[0580] The server then passes the preprocessed text to a generative AI, which then automatically analyzes and scores the content, evaluating the content and logical consistency of the answers and calculating a score according to a rubric.

[0581] Generate feedback

[0582] Based on the results of the automatic scoring and the rationale behind it, the server generates feedback for the user. This feedback includes not only the score but also specific reasons for the scoring and suggestions for improvement, so the user can clearly understand what was evaluated and what areas need improvement.

[0583] Generate a summary of the submitted documents

[0584] The server uses artificial intelligence to generate summaries of submitted documents, which are provided to graders and other stakeholders for quick understanding.

[0585] Acquiring and analyzing emotion data

[0586] The server uses an emotion engine to recognize the user's emotions. For example, it collects emotion data from the user's facial expressions and voice and analyzes that data. This allows it to understand the user's emotional state.

[0587] Emotion-based feedback adjustment

[0588] The content and tone of the feedback can be adjusted based on the emotional data recognized by the emotion engine. For example, if the user is nervous, the feedback can be given in a calmer tone, taking into consideration the user's emotions.

[0589] Specific examples

[0590] For example, consider the case where a student uploads an answer sheet for a written exam. When the user scans and uploads the answer sheet from their device, the server converts it into text format, removes noise, and performs preprocessing. The server then trains an AI based on pre-defined scoring criteria, which then automatically scores the answer sheet. The AI ​​generates a score and specific scoring rationale, which the server returns to the user as feedback. Furthermore, an emotion engine recognizes the user's emotional data and adjusts the tone of the feedback based on that data. For example, if the user is feeling anxious, the feedback will be adjusted to include more encouraging words. An answer summary is also generated for the instructor, allowing them to quickly compare and contrast the answers of multiple students.

[0591] This system achieves fair and efficient scoring and provides appropriate feedback and summaries to users and graders. Furthermore, by using an emotion engine, it is possible to provide feedback that takes into account users' emotions, thereby improving satisfaction with the evaluation.

[0592] The processing flow will be explained below.

[0593] Step 1:

[0594] The user uploads the document from the device.

[0595] Specifically, users select document files such as answer sheets or proposals through the web interface and click the submit button. The server receives the uploaded files and saves them in a designated folder.

[0596] Step 2:

[0597] The server converts the uploaded document into text format.

[0598] Specifically, the saved file is read and images and PDF documents are converted into text format using OCR (Optical Character Recognition) technology, and the document contents are obtained as text data.

[0599] Step 3:

[0600] The server preprocesses the text data.

[0601] Specifically, unnecessary characters and spaces are removed from the acquired text data and normalized. For example, multiple spaces are combined into one and line breaks are standardized. This puts the text data into a clean format that is easy to analyze.

[0602] Step 4:

[0603] The server loads the scoring and judging criteria.

[0604] Specifically, predefined rubrics and scoring guidelines are stored as JSON or XML files, which the server reads to obtain baseline data that is used for subsequent analysis.

[0605] Step 5:

[0606] The server passes the reference data to the generative artificial intelligence and trains the AI.

[0607] Specifically, the imported reference data is input into the AI ​​model, and the model is trained using machine learning algorithms. At this stage, the AI ​​learns how to evaluate the data and builds the foundation for analysis.

[0608] Step 6:

[0609] The server passes the preprocessed text to a generative artificial intelligence for analysis.

[0610] Specifically, clean text data is fed into an AI model, which analyzes the content. The AI ​​then understands the context and calculates a score for each part according to a set of criteria, such as logical consistency and accuracy of information.

[0611] Step 7:

[0612] The server generates the results and rationale for the automatic scoring.

[0613] Specifically, the scoring results are organized based on the scores calculated by the AI ​​and the rationale for each. Along with the scores, specific comments are generated to explain why the scores were given.

[0614] Step 8:

[0615] The server provides feedback to the user.

[0616] Specifically, the generated feedback data is returned to the user's device, where the user can check the score and specific evaluation comments through a web interface. The feedback includes areas for improvement and areas that are highly rated.

[0617] Step 9:

[0618] The server uses artificial intelligence to generate a summary of the submitted documents.

[0619] Specifically, the clean text data is fed into a summary generation module, which extracts key points and summarizes them into short sentences, generating a summary of the submitted document.

[0620] Step 10:

[0621] The server provides the summary results to the grader.

[0622] Specifically, the generated summary data is provided to the grader's device, allowing the grader to quickly evaluate and compare multiple documents.

[0623] Step 11:

[0624] The server uses an emotion engine to recognize the user's emotion.

[0625] Specifically, the system analyzes the user's facial expressions and voice to obtain emotional data. For example, it analyzes the video and audio uploaded by the user via a webcam to identify the user's emotional state (joy, anger, sadness, surprise, etc.).

[0626] Step 12:

[0627] The server adjusts the content and tone of the feedback based on the emotional data.

[0628] Specifically, the content of the feedback is adjusted based on the emotional data recognized by the emotion engine. For example, if a user is feeling anxious, feedback containing many encouraging words is provided. In this way, appropriate feedback is generated that takes into account the user's emotions.

[0629] Specific examples

[0630] For example, consider the case where a student uploads an answer sheet for a written exam. The user scans the answer sheet from their device and uploads it to the server, which converts the document into text format and preprocesses it by removing unnecessary information. The server then trains an AI system using pre-defined scoring criteria, which then automatically scores the answer sheet. The server generates feedback based on the scoring results and specific reasons and provides it to the user. Furthermore, an emotion engine recognizes emotions from the user's facial expressions and voice, and adjusts the tone of the feedback based on that data. For example, if the user is feeling anxious, the feedback could include more encouraging words. An answer summary is generated for the instructor, allowing them to quickly compare and consider multiple students' answers.

[0631] This system not only achieves fair and efficient scoring, but also provides feedback that takes users' emotions into consideration, improving satisfaction with the evaluation and allowing evaluators to focus on higher-level tasks.

[0632] Example 2

[0633] 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."

[0634] Conventional scoring systems rely on manual scoring, which inevitably leads to human error and bias. Furthermore, they lack sufficient summaries to quickly grasp the content of submitted documents, placing a heavy workload on the scorer. Furthermore, they are unable to provide feedback that takes into account the user's emotions, which hinders the improvement of satisfaction.

[0635] The specific processing by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes a means for uploading documents from an information terminal to the information processing device, a means for converting the uploaded documents into electronic text format and performing preprocessing, a means for reading grading and examination criteria and having a machine intelligence understand the criteria, a means for the machine intelligence to analyze the preprocessed electronic text according to the criteria and perform automatic grading, a means for generating feedback based on the results of the automatic grading and the basis thereof, a means for generating a summary of the submitted document using the machine intelligence, a means for providing the generated feedback and summary to the user and the grader, a means for acquiring and analyzing the user's emotions, and a means for adjusting the content and tone of the feedback based on the user's emotions. This enables fair and effective automatic grading and the provision of feedback that takes the user's emotions into consideration.

[0636] "Uploading a document" refers to the act of electronically transmitting a document from an information terminal to an information processing device.

[0637] "Information terminal" refers to electronic devices such as computers, smartphones, and tablets, which are devices operated by users.

[0638] An "information processing device" is a server that receives, stores, and processes transmitted documents.

[0639] An "electronic text format" is a format in which the contents of a document are expressed as digital data.

[0640] "Preprocessing" is the process of processing data to make documents converted into electronic text format easier to analyze.

[0641] "Grading and Review Criteria" refers to rubrics or guidelines for evaluating the content of a document based on specific criteria.

[0642] "Machine intelligence" is an artificial intelligence technology that learns and analyzes large amounts of data to perform specific tasks.

[0643] "Automatic scoring" is the process by which machine intelligence analyzes documents according to scoring and examination criteria and calculates a score.

[0644] "Feedback" is information that returns the document evaluation results and the basis for them to the user.

[0645] "Summary" refers to a concise summary of the main points or contents of the submitted documents.

[0646] "Emotion data" is information that represents the emotional state of a user, obtained from facial expressions, voice, context, and the like.

[0647] "Adjusting the content and tone of feedback" means changing the sentences and wording of the feedback depending on the user's emotional state.

[0648] The present invention is a system for implementing a series of processes including uploading documents and automatic marking. Detailed embodiments of the system are described below.

[0649] Uploading documents

[0650] Users use their devices to scan or electronically create documents and upload them to the server through a web interface. These documents can include answer sheets for written exams or business proposals.

[0651] Server: Receives uploaded documents and stores the data. The server checks the data for consistency and stores it in the appropriate folder.

[0652] Document text conversion and preprocessing

[0653] Server: The server uses OCR (Optical Character Recognition) software "Tesseract OCR" to convert the received documents into text format. The converted electronic text undergoes pre-processing to make it suitable for analysis. Pre-processing includes:

[0654] Removal of unnecessary information (e.g., page numbers and margins)

[0655] Text normalization (e.g., making all characters lowercase)

[0656] Loading scoring and judging criteria and training the AI

[0657] Server: Loads the scoring and judging criteria and trains a generative AI model (e.g., GPT-4) according to predefined rubrics and guidelines. This training process allows the AI ​​to understand the criteria and perform accurate analysis.

[0658] Run automatic grading

[0659] Server: Passes the preprocessed electronic text to the generative AI model. The AI ​​analyzes the text data and automatically scores it based on specific criteria. For example, it evaluates the logic of the text and the coherence of the content, and generates a score and its rationale.

[0660] Generate feedback

[0661] Server: Generates feedback to be returned to the user based on the automated scoring results and the rationale behind them. This feedback includes not only the score but also specific scoring rationale and areas for improvement. For example, it provides detailed comments such as "You received a high score because the logical structure of your answer was clear."

[0662] Generate a summary of the submitted documents

[0663] Server: Requests the AI ​​model to generate a summary of the submitted documents. The summary is intended to allow graders and other stakeholders to quickly understand the content, and concisely presents the main points and conclusions.

[0664] Acquiring and analyzing emotion data

[0665] Server: The server uses an emotion engine (such as IBM Watson Tone Analyzer or Microsoft Azure Face API) to obtain the user's emotional state. This is done by analyzing facial expressions and voice, and it is possible to understand the user's emotional state.

[0666] Emotion-based feedback adjustment

[0667] Server: Adjust the content and tone of the feedback based on the emotional data. For example, if the user is nervous, provide feedback in a gentler tone. A specific example would be something like, "Thank you for your hard work. Overall, you did a good job, but if you improve this next time, you'll get even better results."

[0668] Examples of concrete examples and prompts

[0669] Example: Consider a student uploading an answer sheet for a final exam. The user (student) scans and uploads the answer sheet from their device, and the server converts it into text format and performs preprocessing. Next, the server trains a generative AI model based on pre-defined scoring criteria, and the AI ​​automatically scores the answer sheet. The AI ​​generates a score and specific scoring reasons, which the server returns to the user as feedback. Furthermore, an emotion engine recognizes the user's emotional data and adjusts the tone of the feedback based on that data. For example, if the user is feeling anxious, the feedback is adjusted to include more encouraging words. An answer summary is also generated for the teacher, allowing them to quickly compare and consider the answers of multiple students.

[0670] Example prompt sentence:

[0671] "Upload students' answer sheets, grade them based on the grading criteria below, and generate feedback. Please be gentle with the tone of your feedback for students who seem nervous."

[0672] In this way, the present invention achieves fair and effective automatic scoring and provides feedback that takes into account the user's emotions. By using an emotion engine, flexible responses can be made according to the user's emotional state, improving satisfaction with the evaluation.

[0673] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0674] Step 1: Upload your documents

[0675] User:

[0676] A user scans or electronically creates a document using a terminal and uploads it to the server through a web interface. The input is a PDF or image file uploaded by the user, and the output is the document data stored on the server. Specifically, the user clicks a specific button, selects a file from a file selection dialog, and submits it.

[0677] Step 2: Text conversion and preprocessing of documents

[0678] server:

[0679] When the server receives the uploaded document, it converts the document into electronic text using OCR software such as "Tesseract OCR." The input is the saved document data, and the output is the converted text data. Specifically, the server inputs the document into the OCR software and temporarily stores the resulting text data.

[0680] server:

[0681] Next, the converted text data is preprocessed. The input is the converted text data, and the output is the preprocessed, clean text data. Specific preprocessing steps include removing unnecessary information (e.g., page numbers and margins) and normalizing the text (e.g., changing all characters to lowercase).

[0682] Step 3: Importing the scoring and judging criteria and training the AI

[0683] server:

[0684] The server loads the scoring and judging criteria and trains the generative AI model. The input is a predefined rubric or guideline, and the output is a trained AI model that understands the criteria. Specifically, the server provides the rubric data to the AI ​​training algorithm, which then learns the criteria.

[0685] Step 4: Run Auto-Scoring

[0686] server:

[0687] The server passes the preprocessed text data to a generative AI model, which analyzes the content and automatically scores the answers. The input is the preprocessed text data and the trained AI model, and the output is the score and the basis for the scoring. Specifically, the AI ​​analyzes the text data and calculates a score based on the content of each answer.

[0688] Step 5: Generate feedback

[0689] server:

[0690] The server generates feedback to be returned to the user based on the results of the automatic scoring and the rationale behind it. The input is the score and the rationale for the scoring, and the output is a feedback message. Specifically, it generates detailed comments such as "Your answer had a clear logical structure, so you got a high score," and provides them to the user.

[0691] Step 6: Generate a summary of the submission

[0692] server:

[0693] The server requests an AI model to generate a summary of the submitted document. The input is preprocessed text data, and the output is summary data. Specifically, the AI ​​analyzes the entire text, extracts key points and conclusions, and generates a summary.

[0694] Step 7: Acquire and analyze emotion data

[0695] server:

[0696] The server uses an emotion engine to acquire and analyze the user's emotional data. The input is the user's facial expressions and voice data, and the output is the analysis result of the user's emotional state. Specifically, data collected from the user's webcam and microphone is input into the analysis engine to determine the user's emotional state.

[0697] Step 8: Adjust your feedback based on emotion

[0698] server:

[0699] The server adjusts the content and tone of the feedback based on the emotional data. The input is the analysis result of the emotional state and the feedback message, and the output is the adjusted feedback message. Specifically, if the user is nervous, the server provides feedback in a gentle tone and includes encouraging words.

[0700] The above steps clearly explain how the system handles uploading documents, automatically scoring them, and providing feedback. It also takes into account users' emotional data to provide more personalized feedback.

[0701] (Application example 2)

[0702] 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."

[0703] Conventional scoring systems for document screening and written tests have limitations in the fairness and efficiency of scoring, making it difficult to provide feedback that takes emotions into account. Furthermore, automated analysis of customer surveys cannot take emotional data into account, making it difficult to accurately grasp customer satisfaction and areas for improvement. The present invention aims to solve these problems by providing a system that achieves fair and efficient scoring and enables the provision of emotional feedback.

[0704] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0705] In this invention, the server includes means for uploading documents from a user terminal to the server, means for converting the uploaded documents into text format and preprocessing them, means for reading grading and review criteria and having a generating AI understand the criteria, means for the generating AI to analyze the preprocessed text according to the criteria and perform automatic grading, means for generating feedback based on the results of the automatic grading and the basis for the results, means for generating summaries of the submitted documents using the generating AI, means for providing the generated feedback and summaries to the user and grader, means for using an emotion analysis engine to acquire user emotion data, and means for adjusting the tone and content of the feedback based on the emotion data, thereby enabling fair and efficient automatic grading and providing feedback that takes user emotions into consideration.

[0706] "Uploading a document" is the act of transferring document data from a user terminal to a server.

[0707] "Server" means a central computer for storing, processing, and analyzing document data.

[0708] "Convert to text format" refers to converting uploaded document image data, PDFs, etc. into text data.

[0709] "Preprocessing" refers to preprocessing such as normalizing text data and removing noise.

[0710] "Grading and review criteria" refers to pre-established rubrics and guidelines for evaluating the content of a document.

[0711] "Generative AI" is an AI model that is trained to perform a specific task.

[0712] "Automatic scoring" is the process of using AI to evaluate the content of a document and calculate a score.

[0713] "Feedback" is response information to the user, including the results of the assessment, a detailed explanation of the assessment, and points for improvement.

[0714] An "abstract" is a concise summary of the main contents of the submitted document.

[0715] The "emotion analysis engine" is a system for collecting and analyzing emotional data from a user's facial expressions and voice.

[0716] "Adjusting the tone and content of feedback" refers to optimizing the expression of feedback based on emotional data and providing an appropriate response according to the user's emotional state.

[0717] The following specific procedures and systems are used to implement the present invention: A server, a user terminal, and appropriate software and hardware are used in combination.

[0718] First, a user uploads document data to the server using their own device (such as a smartphone or tablet). The uploaded document is then converted into text format by the server. This conversion process uses optical character recognition (OCR) technology. The server then preprocesses the text data, removing unnecessary information and normalizing the text, among other processes. This generates clean data suitable for analysis.

[0719] The server trains a generative artificial intelligence (AI model) based on pre-set scoring and evaluation criteria. This trained AI model analyzes the pre-processed text according to the criteria and automatically scores the answers. The AI ​​model evaluates the content and logical consistency of the answers and calculates a score according to the rubric.

[0720] Based on the results of the automated scoring and the rationale behind it, the server generates feedback. This feedback includes not only the score but also specific scoring rationale and areas for improvement. The server also uses generative artificial intelligence to generate a summary of the submitted document, allowing graders and other stakeholders to quickly understand the content.

[0721] Furthermore, the server uses an emotion analysis engine to obtain the user's emotional data. Emotional data is collected and analyzed from the user's facial expressions and voice. This allows the server to understand the user's emotional state. Based on this emotional data, the server adjusts the tone and content of the feedback. For example, if the user is feeling anxious, the server will provide feedback in a calm tone.

[0722] A concrete example is a system that analyzes customer surveys in stores. Customers fill out and upload the survey using tablets provided in the store or their own smartphones. The server converts the survey data into text format, performs preprocessing, and then analyzes the data using an AI model to identify customer satisfaction levels and areas for improvement. Furthermore, a sentiment analysis engine is used to analyze the customer's facial expressions and voice, allowing the tone and content of the feedback to be adjusted to provide more personalized service.

[0723] As an example of a prompt sentence, in response to the feedback "The food was delicious, but the service was slow," feedback such as "Thank you for rating the food delicious. We will try to improve the slowness of the service" can be generated.

[0724] This enables fair and efficient automatic scoring and provides feedback that takes into account the user's feelings.

[0725] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0726] Step 1:

[0727] Users use their own devices (smartphones or tablets) to enter and upload document data.

[0728] Input: Images or PDF files of survey data and answer sheets

[0729] Output: Document data sent to the server

[0730] The server receives the uploaded document data, and the data sent from the user terminal is stored in a designated storage area on the server.

[0731] Step 2:

[0732] The server converts the received document data into text format.

[0733] Input: Document data (image or PDF)

[0734] Output: Text data

[0735] Specifically, the server uses optical character recognition (OCR) software (e.g., Tesseract OCR) to convert images and PDF files into text data, resulting in text data in a format suitable for analysis.

[0736] Step 3:

[0737] The server preprocesses the text data.

[0738] Input: Text data

[0739] Output: Cleaned text data

[0740] Pre-processing includes removing unnecessary information, normalizing text, checking spelling, etc. The server performs data cleansing using PANDAS and regular expression libraries.

[0741] Step 4:

[0742] The server loads the scoring and judging criteria and trains the AI ​​model.

[0743] Input: Scoring and judging criteria data

[0744] Output: A trained AI model

[0745] The server loads pre-prepared rubrics and guidelines and trains a generative AI model (e.g., BERT or GPT). Training also includes setting scores for each evaluation item and generating evaluation comments.

[0746] Step 5:

[0747] The server performs automatic scoring based on the preprocessed text.

[0748] Input: cleaned text data, trained AI model

[0749] Output: Score and grading rationale

[0750] Specifically, the server uses a trained generative AI model to analyze the content of the text data and generate a score and its rationale according to the rubric.

[0751] Step 6:

[0752] The server generates feedback based on the results of the automatic scoring and the reasons for it.

[0753] Input: Score and scoring rationale

[0754] Output: Feedback data

[0755] The feedback includes not only the score but also specific reasons for the score and areas for improvement. The server uses a generative AI model to generate sentences and create feedback to provide to the user.

[0756] Step 7:

[0757] The server generates a summary of the submission.

[0758] Input: Cleaned text data

[0759] Output: Summary data

[0760] To generate a summary, the server uses a generative AI model (e.g., a news article summarization model) to extract key points and create a concise summary.

[0761] Step 8:

[0762] The server acquires and analyzes the user's emotion data.

[0763] Input: User facial expression images, voice data

[0764] Output: Emotion data

[0765] The server uses an emotion analysis engine (e.g., OpenCV or TensorFlow) to analyze the user's facial expressions and voice obtained from smart glasses or a tablet to understand their emotional state.

[0766] Step 9:

[0767] The server adjusts the tone and content of the feedback based on the emotional data.

[0768] Input: Emotion data, feedback data

[0769] Output: Regulated feedback data

[0770] Adjust the content and tone of the feedback depending on the user's emotional state. For example, provide calmer, more encouraging feedback to a user who is feeling anxious.

[0771] Step 10:

[0772] The server provides the generated feedback and summaries to the user and grader.

[0773] Input: Adjusted feedback data, summary data

[0774] Output: Feedback and summaries provided to users and graders

[0775] The server generates feedback and summaries, which are then sent to the user's device and provided to the grader, enabling fair and efficient evaluation and emotionally sensitive feedback.

[0776] 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.

[0777] 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.

[0778] 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.

[0779] [Third embodiment]

[0780] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0781] 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.

[0782] 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).

[0783] 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.

[0784] 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.

[0785] 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).

[0786] 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. 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.

[0787] 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.

[0788] 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.

[0789] 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.

[0790] 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.

[0791] 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."

[0792] This invention relates to a system that automates the grading of document screening and written exams. In this system, documents are uploaded from the user's terminal, and the server converts the documents into text format and performs preprocessing. The server then loads the grading and examination criteria, trains a generating AI to understand the criteria, and the AI ​​performs automatic grading. The system also includes a function to generate feedback based on the scores and grading rationale, and to generate summaries of the submitted documents.

[0793] User document upload

[0794] A user uploads documents to the server using a terminal, such as an answer sheet for a particular exam or a business proposal. The user can upload the documents through a web interface, and the server receives the uploaded documents.

[0795] Document text conversion and preprocessing

[0796] The server converts the uploaded documents into a text format, then preprocesses the text data, for example by removing unnecessary information and normalizing the text, to produce clean data suitable for analysis.

[0797] Loading scoring and judging criteria and training the AI

[0798] The server loads the scoring and judging criteria. These criteria are predefined rubrics or guidelines that the server uses to train the generative AI. The AI ​​understands these criteria and uses them as the basis for its analysis.

[0799] Run automatic grading

[0800] The server then passes the preprocessed text to a generative AI, which then automatically analyzes and scores the content, evaluating the content and logical consistency of the answers and calculating a score according to a rubric.

[0801] Generate feedback

[0802] Based on the results of the automatic scoring and the rationale behind it, the server generates feedback for the user. This feedback includes not only the score but also specific reasons for the scoring and suggestions for improvement, so the user can clearly understand what was evaluated and what areas need improvement.

[0803] Generate a summary of the submitted documents

[0804] Additionally, the server uses a generator to generate summaries of the submitted documents, which are provided by the server to graders and other interested parties for quick understanding.

[0805] Specific examples

[0806] For example, consider the case where a student uploads an answer sheet for a written exam. When the user scans and uploads the answer sheet from their device, the server converts it into text format, removes noise, and performs preprocessing. The server then trains an AI based on pre-defined scoring criteria, which then automatically scores the answer sheet. As a result, specific feedback is generated and provided to the student along with a score. An answer summary is also generated for the instructor, allowing them to quickly compare and consider the answers of multiple students.

[0807] This system ensures fair and efficient scoring, increases the credibility of the assessment, and allows assessors to focus on more important tasks.

[0808] The processing flow will be explained below.

[0809] Step 1:

[0810] The user uploads the document from the device.

[0811] Specifically, the user selects document files such as answer sheets and proposals through the web interface and clicks the send button to the server, which receives and stores the uploaded files.

[0812] Step 2:

[0813] The server converts the uploaded document into a text format.

[0814] Specifically, the saved file is read and images and PDF documents are converted to text format using OCR (Optical Character Recognition) technology, thereby obtaining the text data.

[0815] Step 3:

[0816] The server preprocesses the text data.

[0817] Specifically, unnecessary characters and spaces are removed from the acquired text data and normalization is performed. For example, this includes combining multiple spaces into one and standardizing line breaks. This process generates clean text data that is easy to analyze.

[0818] Step 4:

[0819] The server loads the scoring and judging criteria.

[0820] Specifically, predefined rubrics and grading guidelines are stored as JSON or XML format files, which the server reads and treats as data, making the evaluation criteria clear.

[0821] Step 5:

[0822] The server passes the reference data to the generative artificial intelligence and trains the AI.

[0823] Specifically, the reference data is fed into a generative AI model, and machine learning techniques are used to train the model. At this stage, the AI ​​learns how to make evaluations.

[0824] Step 6:

[0825] The server passes the preprocessed text to a generative artificial intelligence for analysis.

[0826] Specifically, clean text data is fed into an AI model, which analyzes the content, understands the context, and calculates a score for each part based on a scoring criteria.

[0827] Step 7:

[0828] The server generates the results and rationale for the automatic scoring.

[0829] Specifically, the system organizes the scores calculated by the AI ​​and the analytical results that form the basis for the scores, and generates feedback data based on this. For example, it clearly shows which areas received high marks and which areas need improvement.

[0830] Step 8:

[0831] The server provides feedback to the user.

[0832] Specifically, the generated feedback data is returned to the user's device, where the user can check the score and specific evaluation comments through a web interface.

[0833] Step 9:

[0834] The server generates a summary of the submitted documents using artificial intelligence.

[0835] Specifically, the clean text data is fed into a summary generation module, which extracts key points and summarizes them into short sentences, generating a summary of the submitted document.

[0836] Step 10:

[0837] The server provides the summary results to the grader.

[0838] Specifically, the generated summary data is provided to the grader's terminal, allowing the grader to quickly evaluate and compare multiple documents.

[0839] The above processing steps enable fair and efficient scoring and provide appropriate feedback and summaries to users and scorers.

[0840] Example 1

[0841] 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."

[0842] Traditionally, document screening and written exam scoring have been done manually, often resulting in a lack of fairness and efficiency. Furthermore, bias among evaluators and variations in evaluation criteria can be problematic. Furthermore, summarizing documents to quickly grasp their contents has also been done manually, requiring time and effort. To address these issues, the present invention aims to provide a fair and efficient system that automates the entire process, from document uploading to automatic scoring, feedback generation, and summary generation.

[0843] 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.

[0844] In this invention, the server includes means for uploading documents from a user terminal to the server, means for converting the uploaded documents into text format using optical character recognition software and performing text cleaning, means for reading scoring and review criteria and having a generative artificial intelligence learn the criteria, means for the generative artificial intelligence to analyze the cleaned text data and perform automatic scoring, means for generating feedback based on the results of the automatic scoring and the basis for that scoring, means for generating summaries of the submitted documents using the generative artificial intelligence, and means for providing the generated feedback and summaries to the user and evaluator, thereby enabling fair and efficient scoring and rapid summary creation.

[0845] A "user terminal" is a device such as a computer or smartphone that is operated by a user.

[0846] A "server" is a computer system that provides information in response to requests from clients over a network.

[0847] "Document" refers to a file containing information in text, images, and other formats.

[0848] "Optical character recognition software" is a program for extracting text information from images.

[0849] "Text format" is a format in which text data is stored and displayed.

[0850] "Text cleaning" is the process of removing unnecessary information from text data and formatting the data.

[0851] "Scoring and judging criteria" are rules and guidelines that define specific evaluation items and how they will be evaluated.

[0852] "Generative AI" is an AI technology that analyzes and generates data.

[0853] "Automatic scoring" is the process by which generative artificial intelligence calculates an evaluation score based on data.

[0854] "Feedback" refers to providing information about the evaluation results and their rationale.

[0855] A "summary" is a concise summary of the main points or content of a text.

[0856] "Evaluator" is the person or system responsible for evaluating a particular document or data.

[0857] This invention relates to a system that automates the grading of document screening and written tests. This system is primarily implemented using a user terminal and a server. Users upload documents using their terminal, and the server converts the documents into text format and performs text cleaning. The server then loads the grading and review criteria, trains a generative AI to learn the criteria, and the AI ​​performs automatic grading. Furthermore, the system generates feedback based on the scores and grading rationale, and also generates summaries of the submitted documents.

[0858] A specific embodiment will be described.

[0859] Users access a web interface using their devices and upload documents, often in PDF or image format. Once uploaded, the documents are sent to a server, which uses OCR software (e.g., commonly used optical character recognition software) to convert the uploaded documents into text format. The server then performs a text cleaning process, removing unnecessary information and normalizing the text.

[0860] The server then loads predefined scoring and judging criteria, defined as rubrics or guidelines and stored in a file format. The server then uses this to train a generative artificial intelligence (e.g., a commonly used generative AI model) to understand the criteria. This process teaches the AI ​​the fundamentals of analysis.

[0861] The cleaned text data is input into a generative AI model, which then performs automatic scoring. The generative AI analyzes the text data and calculates a score according to criteria. For example, it evaluates the accuracy and logic of the answer and determines the score comprehensively. The server generates feedback based on the results of the automatic scoring and the reasons for it. This feedback includes not only the score, but also the specific reasons for the scoring and areas for improvement, allowing the user to clearly understand what was evaluated and what areas need improvement.

[0862] It also has a function to generate summaries for submitted documents. The server uses a generative AI model to generate a concise summary of the main points and content of the text data, which can be used by graders and other stakeholders to quickly understand the content.

[0863] For example, imagine a student uploading an answer sheet for a written exam. After the user scans and uploads the answer sheet, the server uses OCR software to convert it into text format and cleans the text. The server then loads pre-prepared grading criteria and trains the AI ​​using a generative AI model. The AI ​​then automatically grades the answer sheet and generates specific feedback for the student. In addition, teachers are provided with a summary of the answer sheet, allowing them to quickly compare and contrast the answers of multiple students.

[0864] This system enables fair and efficient scoring and rapid summary creation, significantly reducing the time and effort required for evaluation and allowing evaluators to focus on more important tasks.

[0865] And here's an example prompt for using a generative AI model: "Upload student answer sheets, grade them, and generate feedback."

[0866] By entering this prompt, the system will automatically execute a series of processes and provide high-quality evaluation results and feedback.

[0867] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0868] Step 1:

[0869] The user accesses the web interface using a terminal and uploads a document. Specifically, the user clicks the "Upload Document" button on the screen to open a file browser. The user selects the document file (e.g., a PDF file) they want to upload and presses the "Open" button. The selected file is sent to the server.

[0870] Input: A document file selected by the user

[0871] Output: The document file is sent to the server.

[0872] Step 2:

[0873] The server receives the submitted document file. It converts the uploaded document file into text format using OCR software (e.g., optical character recognition software). The server then performs text cleaning, removing unnecessary information and normalizing the text. Specifically, it removes whitespace, normalizes special characters, and standardizes line breaks.

[0874] Input: Document file sent by user

[0875] Output: Cleaned text data

[0876] Step 3:

[0877] The server loads predefined scoring and judging criteria. These criteria are defined as rubrics or guidelines and saved in a file format. The server then uses these to train a generative AI model to understand the criteria. During this process, the AI ​​learns the evaluation criteria and scoring methods.

[0878] Input: Rubric or guideline file

[0879] Output: A generative AI model that understands the criteria

[0880] Step 4:

[0881] The server inputs the cleaned text data into the generative AI model, which then analyzes the text data and calculates a score based on the scoring criteria. Specifically, the model evaluates the content, logic, grammar, etc. of the text data to determine an overall score.

[0882] Input: Cleaned text data

[0883] Output:Score result

[0884] Step 5:

[0885] The server generates feedback based on the score calculated by the generative AI model and the rationale for it. This feedback includes not only the score but also specific reasons for the score and areas for improvement. This allows the user to clearly understand which aspects were evaluated and which areas need improvement.

[0886] Input: Score results and scoring basis

[0887] Output: Feedback text

[0888] Step 6:

[0889] The server then feeds the cleaned text data back into the generative AI model to generate summaries of the submitted documents, which can be used by graders and other stakeholders to quickly understand the content.

[0890] Input: Cleaned text data

[0891] Output: Summary text

[0892] Step 7:

[0893] The server provides the generated feedback and summary to the user and the evaluator by posting the feedback to the user's account and displaying the summary on the grader's dashboard, allowing both parties to quickly check the evaluation results and summary information.

[0894] Input: Feedback text, Summary text

[0895] Output: Provided to users and evaluators

[0896] (Application example 1)

[0897] 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."

[0898] Currently, quality control and inspection work in factories is heavily dependent on human resources, resulting in issues with consistency and efficiency of evaluation. In particular, it is difficult to quickly process large amounts of inspection data and provide accurate feedback. For this reason, a new system is needed to automate and streamline quality control.

[0899] 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.

[0900] In this invention, the server includes a means for uploading documents from a user terminal to the server, a means for converting the uploaded documents into text format and preprocessing them, a means for reading grading and examination criteria and having a generating AI understand the criteria, a means for the generating AI to analyze the preprocessed text according to the criteria and perform automatic grading, a means for generating feedback based on the results of the automatic grading and the basis for that, a means for generating summaries of the submitted documents using the generating AI, a means for providing the generated feedback and summaries to the user and grader, and a machine for collecting and uploading inspection data, which enables the automation and efficiency of quality control and inspection work.

[0901] "Uploading a document" is the act of transferring a document file from a user terminal to a server.

[0902] "User terminal" refers to a device such as a computer, smartphone, or tablet connected to the Internet.

[0903] A "server" is a central processing unit that receives, processes, stores, and provides data.

[0904] "Convert to text format" is the process of converting uploaded documents into text data.

[0905] "Preprocessing" is the process of preprocessing data to make it suitable for analysis.

[0906] "Scoring and evaluation criteria" are guidelines that include evaluation items and rules.

[0907] "Generative artificial intelligence" refers to AI systems that use natural language processing and machine learning to perform specific tasks.

[0908] "Understanding the standards" means having the AI ​​learn the scoring and judging criteria and then perform analysis according to those rules.

[0909] "Automatic scoring" is the process by which AI analyzes text and calculates a score based on pre-set criteria.

[0910] "Feedback" is response information that includes the score results and the basis for the evaluation.

[0911] A summary is a short sentence that succinctly summarizes the content of a longer piece of text.

[0912] "Inspection data" refers to information collected in factories and other locations that is used to evaluate the quality of products and processes.

[0913] A "machine" is a hardware-based device used to collect inspection data.

[0914] The present invention relates to a system for automating quality control and inspection work in factories. This system is installed on a factory robot, uploads product inspection data to a server, and a generative artificial intelligence (AI) analyzes and evaluates the data. Specific embodiments of the present invention will be described.

[0915] System Program

[0916] The system consists of the following components:

[0917] Document upload: Uploads inspection data from the user's device to the server.

[0918] Data conversion and pre-processing: The server converts the uploaded data into text format and removes unnecessary information to produce clean data suitable for analysis.

[0919] Loading of scoring and judging criteria and training of AI: The server loads the quality criteria and trains the generative AI to understand the criteria.

[0920] Automatic scoring: The server passes the preprocessed data to the AI, which analyzes it and performs automatic scoring.

[0921] Feedback generation: Based on the scoring results and their rationale, feedback is generated to be returned to the user.

[0922] Summary generation: The server generates a summary of the test data and provides it to the assessor.

[0923] Hardware and Software

[0924] Hardware: Factory robots, servers, user terminals

[0925] Software: Python, requests module, TextBlob, AI model (generative AI model)

[0926] Data processing and calculation

[0927] Data processing: The test data uploaded by the user's device is converted into text format on the server, and then pre-processed to remove noise and unnecessary information.

[0928] Data calculation: The server trains the AI ​​based on the scoring criteria, and the AI ​​analyzes and evaluates the test data. Based on the evaluation results, feedback and summaries are generated.

[0929] Specific examples

[0930] For example, a quality control officer at a factory collects product inspection data and uploads it from the user's device to a server. The uploaded data is converted into text format, preprocessed, and stored on the server as clean data. The server then trains an AI based on preset quality standards, which then analyzes the inspection data and automatically evaluates it. The evaluation results, along with a score, are generated as specific feedback and provided to the person in charge. A summary is also generated, allowing for quick quality evaluation.

[0931] Prompt Sentence Examples

[0932] "Factory robots collect inspection data and upload it to a server. The server converts the data into text format, preprocesses it, and then uses AI to automatically evaluate it. Create an application that generates a score and feedback on areas for improvement, as well as a summary of the inspection data."

[0933] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0934] Step 1:

[0935] The user's device collects the test data and uploads it to the server. Specifically, the camera or sensor on the user's device acquires the test data and saves it in a file format. The file is then sent to the server via the Internet. In this case, the input is the test data file, and the output is the test data saved on the server.

[0936] Step 2:

[0937] The server converts the uploaded inspection data into text format and performs preprocessing. Specifically, it first converts images and other data formats into text using optical character recognition (OCR) technology, then normalizes the text, checks spelling, and removes unnecessary information. The input is the inspection data file, and the output is clean text data.

[0938] Step 3:

[0939] The server loads the scoring and review criteria and has the generative AI understand the criteria. Specifically, it loads predefined quality criteria and provides them to the AI ​​model for training. The input is the quality criteria data, and the output is the trained AI model.

[0940] Step 4:

[0941] The server passes the preprocessed text data to the AI ​​model, which performs automatic scoring. Specifically, the clean text data is input into the evaluation algorithm, which generates an evaluation score and commentary. The input is the preprocessed text data, and the output is the evaluation result and its rationale.

[0942] Step 5:

[0943] The server generates feedback based on the results of the automatic scoring and the rationale behind it. Specifically, it generates points for improvement and specific advice from the evaluation results and compiles them into a feedback document. The input is the evaluation results and their rationale, and the output is the feedback document.

[0944] Step 6:

[0945] The server generates a summary of the submitted document. Specifically, it uses natural language processing techniques to extract important parts of the text and summarize them concisely. The input is preprocessed text data, and the output is a summary.

[0946] Step 7:

[0947] The server provides the generated feedback and summary to the user and evaluator. Specifically, it sends the generated feedback document and summary to the user and evaluator via email or a dedicated web interface. The input is the feedback document and summary, and the output is the information sent to the user and evaluator.

[0948] 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.

[0949] This invention combines a system that automates document screening and written test scoring with an emotion engine that recognizes user emotions. Feedback can be generated based on the user's emotion data, improving the quality of evaluation. The basic flow of this system is that the user uploads documents from their device, and the server converts them into text format and preprocesses them. The server then passes the reference data to a generative AI, which then automatically scores the documents. The system also includes a function to generate feedback based on the scores and scoring rationale, and to generate summaries of the submitted documents. The server also uses the emotion engine to analyze the user's emotion data and reflect this in the feedback.

[0950] User document upload

[0951] A user uploads documents to the server using a terminal, such as an answer sheet for a particular exam or a business proposal. The user can upload documents through a web interface, and the server receives and stores the uploaded documents.

[0952] Document text conversion and preprocessing

[0953] The server converts the uploaded documents into a text format, then preprocesses the text data, for example by removing unnecessary information and normalizing the text, to produce clean data suitable for analysis.

[0954] Loading scoring and judging criteria and training the AI

[0955] The server loads the scoring and judging criteria. These criteria are predefined rubrics or guidelines that the server uses to train the generative AI. The AI ​​understands these criteria and uses them as the basis for its analysis.

[0956] Run automatic grading

[0957] The server then passes the preprocessed text to a generative AI, which then automatically analyzes and scores the content, evaluating the content and logical consistency of the answers and calculating a score according to a rubric.

[0958] Generate feedback

[0959] Based on the results of the automatic scoring and the rationale behind it, the server generates feedback for the user. This feedback includes not only the score but also specific reasons for the scoring and suggestions for improvement, so the user can clearly understand what was evaluated and what areas need improvement.

[0960] Generate a summary of the submitted documents

[0961] The server uses artificial intelligence to generate summaries of submitted documents, which are provided to graders and other stakeholders for quick understanding.

[0962] Acquiring and analyzing emotion data

[0963] The server uses an emotion engine to recognize the user's emotions. For example, it collects emotion data from the user's facial expressions and voice and analyzes that data. This allows it to understand the user's emotional state.

[0964] Emotion-based feedback adjustment

[0965] The content and tone of the feedback can be adjusted based on the emotional data recognized by the emotion engine. For example, if the user is nervous, the feedback can be given in a calmer tone, taking into consideration the user's emotions.

[0966] Specific examples

[0967] For example, consider the case where a student uploads an answer sheet for a written exam. When the user scans and uploads the answer sheet from their device, the server converts it into text format, removes noise, and performs preprocessing. The server then trains an AI based on pre-defined scoring criteria, which then automatically scores the answer sheet. The AI ​​generates a score and specific scoring rationale, which the server returns to the user as feedback. Furthermore, an emotion engine recognizes the user's emotional data and adjusts the tone of the feedback based on that data. For example, if the user is feeling anxious, the feedback will be adjusted to include more encouraging words. An answer summary is also generated for the instructor, allowing them to quickly compare and contrast the answers of multiple students.

[0968] This system achieves fair and efficient scoring and provides appropriate feedback and summaries to users and graders. Furthermore, by using an emotion engine, it is possible to provide feedback that takes into account users' emotions, thereby improving satisfaction with the evaluation.

[0969] The processing flow will be explained below.

[0970] Step 1:

[0971] The user uploads the document from the device.

[0972] Specifically, users select document files such as answer sheets or proposals through the web interface and click the submit button. The server receives the uploaded files and saves them in a designated folder.

[0973] Step 2:

[0974] The server converts the uploaded document into text format.

[0975] Specifically, the saved file is read and images and PDF documents are converted into text format using OCR (Optical Character Recognition) technology, and the document contents are obtained as text data.

[0976] Step 3:

[0977] The server preprocesses the text data.

[0978] Specifically, unnecessary characters and spaces are removed from the acquired text data and normalized. For example, multiple spaces are combined into one and line breaks are standardized. This puts the text data into a clean format that is easy to analyze.

[0979] Step 4:

[0980] The server loads the scoring and judging criteria.

[0981] Specifically, predefined rubrics and scoring guidelines are stored as JSON or XML files, which the server reads to obtain baseline data that is used for subsequent analysis.

[0982] Step 5:

[0983] The server passes the reference data to the generative artificial intelligence and trains the AI.

[0984] Specifically, the imported reference data is input into the AI ​​model, and the model is trained using machine learning algorithms. At this stage, the AI ​​learns how to evaluate the data and builds the foundation for analysis.

[0985] Step 6:

[0986] The server passes the preprocessed text to a generative artificial intelligence for analysis.

[0987] Specifically, clean text data is fed into an AI model, which analyzes the content. The AI ​​then understands the context and calculates a score for each part according to a set of criteria, such as logical consistency and accuracy of information.

[0988] Step 7:

[0989] The server generates the results and rationale for the automatic scoring.

[0990] Specifically, the scoring results are organized based on the scores calculated by the AI ​​and the rationale for each. Along with the scores, specific comments are generated to explain why the scores were given.

[0991] Step 8:

[0992] The server provides feedback to the user.

[0993] Specifically, the generated feedback data is returned to the user's device, where the user can check the score and specific evaluation comments through a web interface. The feedback includes areas for improvement and areas that are highly rated.

[0994] Step 9:

[0995] The server uses artificial intelligence to generate a summary of the submitted documents.

[0996] Specifically, the clean text data is fed into a summary generation module, which extracts key points and summarizes them into short sentences, generating a summary of the submitted document.

[0997] Step 10:

[0998] The server provides the summary results to the grader.

[0999] Specifically, the generated summary data is provided to the grader's device, allowing the grader to quickly evaluate and compare multiple documents.

[1000] Step 11:

[1001] The server uses an emotion engine to recognize the user's emotion.

[1002] Specifically, the system analyzes the user's facial expressions and voice to obtain emotional data. For example, it analyzes the video and audio uploaded by the user via a webcam to identify the user's emotional state (joy, anger, sadness, surprise, etc.).

[1003] Step 12:

[1004] The server adjusts the content and tone of the feedback based on the emotional data.

[1005] Specifically, the content of the feedback is adjusted based on the emotional data recognized by the emotion engine. For example, if a user is feeling anxious, feedback containing many encouraging words is provided. In this way, appropriate feedback is generated that takes into account the user's emotions.

[1006] Specific examples

[1007] For example, consider the case where a student uploads an answer sheet for a written exam. The user scans the answer sheet from their device and uploads it to the server, which converts the document into text format and preprocesses it by removing unnecessary information. The server then trains an AI system using pre-defined scoring criteria, which then automatically scores the answer sheet. The server generates feedback based on the scoring results and specific reasons and provides it to the user. Furthermore, an emotion engine recognizes emotions from the user's facial expressions and voice, and adjusts the tone of the feedback based on that data. For example, if the user is feeling anxious, the feedback could include more encouraging words. An answer summary is generated for the instructor, allowing them to quickly compare and consider multiple students' answers.

[1008] This system not only achieves fair and efficient scoring, but also provides feedback that takes users' emotions into consideration, improving satisfaction with the evaluation and allowing evaluators to focus on higher-level tasks.

[1009] Example 2

[1010] 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."

[1011] Conventional scoring systems rely on manual scoring, which inevitably leads to human error and bias. Furthermore, they lack sufficient summaries to quickly grasp the content of submitted documents, placing a heavy workload on the scorer. Furthermore, they are unable to provide feedback that takes into account the user's emotions, which hinders the improvement of satisfaction.

[1012] The specific processing by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes a means for uploading documents from an information terminal to the information processing device, a means for converting the uploaded documents into electronic text format and performing preprocessing, a means for reading grading and examination criteria and having a machine intelligence understand the criteria, a means for the machine intelligence to analyze the preprocessed electronic text according to the criteria and perform automatic grading, a means for generating feedback based on the results of the automatic grading and the basis thereof, a means for generating a summary of the submitted document using the machine intelligence, a means for providing the generated feedback and summary to the user and the grader, a means for acquiring and analyzing the user's emotions, and a means for adjusting the content and tone of the feedback based on the user's emotions. This enables fair and effective automatic grading and the provision of feedback that takes the user's emotions into consideration.

[1013] "Uploading a document" refers to the act of electronically transmitting a document from an information terminal to an information processing device.

[1014] "Information terminal" refers to electronic devices such as computers, smartphones, and tablets, which are devices operated by users.

[1015] An "information processing device" is a server that receives, stores, and processes transmitted documents.

[1016] An "electronic text format" is a format in which the contents of a document are expressed as digital data.

[1017] "Preprocessing" is the process of processing data to make documents converted into electronic text format easier to analyze.

[1018] "Grading and Review Criteria" refers to rubrics or guidelines for evaluating the content of a document based on specific criteria.

[1019] "Machine intelligence" is an artificial intelligence technology that learns and analyzes large amounts of data to perform specific tasks.

[1020] "Automatic scoring" is the process by which machine intelligence analyzes documents according to scoring and examination criteria and calculates a score.

[1021] "Feedback" is information that returns the document evaluation results and the basis for them to the user.

[1022] "Summary" refers to a concise summary of the main points or contents of the submitted documents.

[1023] "Emotion data" is information that represents the emotional state of a user, obtained from facial expressions, voice, context, and the like.

[1024] "Adjusting the content and tone of feedback" means changing the sentences and wording of the feedback depending on the user's emotional state.

[1025] The present invention is a system for implementing a series of processes including uploading documents and automatic marking. Detailed embodiments of the system are described below.

[1026] Uploading documents

[1027] Users use their devices to scan or electronically create documents and upload them to the server through a web interface. These documents can include answer sheets for written exams or business proposals.

[1028] Server: Receives uploaded documents and stores the data. The server checks the data for consistency and stores it in the appropriate folder.

[1029] Document text conversion and preprocessing

[1030] Server: The server uses OCR (Optical Character Recognition) software "Tesseract OCR" to convert the received documents into text format. The converted electronic text undergoes pre-processing to make it suitable for analysis. Pre-processing includes:

[1031] Removal of unnecessary information (e.g., page numbers and margins)

[1032] Text normalization (e.g., making all characters lowercase)

[1033] Loading scoring and judging criteria and training the AI

[1034] Server: Loads the scoring and judging criteria and trains a generative AI model (e.g., GPT-4) according to predefined rubrics and guidelines. This training process allows the AI ​​to understand the criteria and perform accurate analysis.

[1035] Run automatic grading

[1036] Server: Passes the preprocessed electronic text to the generative AI model. The AI ​​analyzes the text data and automatically scores it based on specific criteria. For example, it evaluates the logic of the text and the coherence of the content, and generates a score and its rationale.

[1037] Generate feedback

[1038] Server: Generates feedback to be returned to the user based on the automated scoring results and the rationale behind them. This feedback includes not only the score but also specific scoring rationale and areas for improvement. For example, it provides detailed comments such as "You received a high score because the logical structure of your answer was clear."

[1039] Generate a summary of the submitted documents

[1040] Server: Requests the AI ​​model to generate a summary of the submitted documents. The summary is intended to allow graders and other stakeholders to quickly understand the content, and concisely presents the main points and conclusions.

[1041] Acquiring and analyzing emotion data

[1042] Server: The server uses an emotion engine (such as IBM Watson Tone Analyzer or Microsoft Azure Face API) to obtain the user's emotional state. This is done by analyzing facial expressions and voice, and it is possible to understand the user's emotional state.

[1043] Emotion-based feedback adjustment

[1044] Server: Adjust the content and tone of the feedback based on the emotional data. For example, if the user is nervous, provide feedback in a gentler tone. A specific example would be something like, "Thank you for your hard work. Overall, you did a good job, but if you improve this next time, you'll get even better results."

[1045] Examples of concrete examples and prompts

[1046] Example: Consider a student uploading an answer sheet for a final exam. The user (student) scans and uploads the answer sheet from their device, and the server converts it into text format and performs preprocessing. Next, the server trains a generative AI model based on pre-defined scoring criteria, and the AI ​​automatically scores the answer sheet. The AI ​​generates a score and specific scoring reasons, which the server returns to the user as feedback. Furthermore, an emotion engine recognizes the user's emotional data and adjusts the tone of the feedback based on that data. For example, if the user is feeling anxious, the feedback is adjusted to include more encouraging words. An answer summary is also generated for the teacher, allowing them to quickly compare and consider the answers of multiple students.

[1047] Example prompt sentence:

[1048] "Upload students' answer sheets, grade them based on the grading criteria below, and generate feedback. Please be gentle with the tone of your feedback for students who seem nervous."

[1049] In this way, the present invention achieves fair and effective automatic scoring and provides feedback that takes into account the user's emotions. By using an emotion engine, flexible responses can be made according to the user's emotional state, improving satisfaction with the evaluation.

[1050] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1051] Step 1: Upload your documents

[1052] User:

[1053] A user scans or electronically creates a document using a terminal and uploads it to the server through a web interface. The input is a PDF or image file uploaded by the user, and the output is the document data stored on the server. Specifically, the user clicks a specific button, selects a file from a file selection dialog, and submits it.

[1054] Step 2: Text conversion and preprocessing of documents

[1055] server:

[1056] When the server receives the uploaded document, it converts the document into electronic text using OCR software such as "Tesseract OCR." The input is the saved document data, and the output is the converted text data. Specifically, the server inputs the document into the OCR software and temporarily stores the resulting text data.

[1057] server:

[1058] Next, the converted text data is preprocessed. The input is the converted text data, and the output is the preprocessed, clean text data. Specific preprocessing steps include removing unnecessary information (e.g., page numbers and margins) and normalizing the text (e.g., changing all characters to lowercase).

[1059] Step 3: Importing the scoring and judging criteria and training the AI

[1060] server:

[1061] The server loads the scoring and judging criteria and trains the generative AI model. The input is a predefined rubric or guideline, and the output is a trained AI model that understands the criteria. Specifically, the server provides the rubric data to the AI ​​training algorithm, which then learns the criteria.

[1062] Step 4: Run Auto-Scoring

[1063] server:

[1064] The server passes the preprocessed text data to a generative AI model, which analyzes the content and automatically scores the answers. The input is the preprocessed text data and the trained AI model, and the output is the score and the basis for the scoring. Specifically, the AI ​​analyzes the text data and calculates a score based on the content of each answer.

[1065] Step 5: Generate feedback

[1066] server:

[1067] The server generates feedback to be returned to the user based on the results of the automatic scoring and the rationale behind it. The input is the score and the rationale for the scoring, and the output is a feedback message. Specifically, it generates detailed comments such as "Your answer had a clear logical structure, so you got a high score," and provides them to the user.

[1068] Step 6: Generate a summary of the submission

[1069] server:

[1070] The server requests an AI model to generate a summary of the submitted document. The input is preprocessed text data, and the output is summary data. Specifically, the AI ​​analyzes the entire text, extracts key points and conclusions, and generates a summary.

[1071] Step 7: Acquire and analyze emotion data

[1072] server:

[1073] The server uses an emotion engine to acquire and analyze the user's emotional data. The input is the user's facial expressions and voice data, and the output is the analysis result of the user's emotional state. Specifically, data collected from the user's webcam and microphone is input into the analysis engine to determine the user's emotional state.

[1074] Step 8: Adjust your feedback based on emotion

[1075] server:

[1076] The server adjusts the content and tone of the feedback based on the emotional data. The input is the analysis result of the emotional state and the feedback message, and the output is the adjusted feedback message. Specifically, if the user is nervous, the server provides feedback in a gentle tone and includes encouraging words.

[1077] The above steps clearly explain how the system handles uploading documents, automatically scoring them, and providing feedback. It also takes into account users' emotional data to provide more personalized feedback.

[1078] (Application example 2)

[1079] 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."

[1080] Conventional scoring systems for document screening and written tests have limitations in the fairness and efficiency of scoring, making it difficult to provide feedback that takes emotions into account. Furthermore, automated analysis of customer surveys cannot take emotional data into account, making it difficult to accurately grasp customer satisfaction and areas for improvement. The present invention aims to solve these problems by providing a system that achieves fair and efficient scoring and enables the provision of emotional feedback.

[1081] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1082] In this invention, the server includes means for uploading documents from a user terminal to the server, means for converting the uploaded documents into text format and preprocessing them, means for reading grading and review criteria and having a generating AI understand the criteria, means for the generating AI to analyze the preprocessed text according to the criteria and perform automatic grading, means for generating feedback based on the results of the automatic grading and the basis for the results, means for generating summaries of the submitted documents using the generating AI, means for providing the generated feedback and summaries to the user and grader, means for using an emotion analysis engine to acquire user emotion data, and means for adjusting the tone and content of the feedback based on the emotion data, thereby enabling fair and efficient automatic grading and providing feedback that takes user emotions into consideration.

[1083] "Uploading a document" is the act of transferring document data from a user terminal to a server.

[1084] "Server" means a central computer for storing, processing, and analyzing document data.

[1085] "Convert to text format" refers to converting uploaded document image data, PDFs, etc. into text data.

[1086] "Preprocessing" refers to preprocessing such as normalizing text data and removing noise.

[1087] "Grading and review criteria" refers to pre-established rubrics and guidelines for evaluating the content of a document.

[1088] "Generative AI" is an AI model that is trained to perform a specific task.

[1089] "Automatic scoring" is the process of using AI to evaluate the content of a document and calculate a score.

[1090] "Feedback" is response information to the user, including the results of the assessment, a detailed explanation of the assessment, and points for improvement.

[1091] An "abstract" is a concise summary of the main contents of the submitted document.

[1092] The "emotion analysis engine" is a system for collecting and analyzing emotional data from a user's facial expressions and voice.

[1093] "Adjusting the tone and content of feedback" refers to optimizing the expression of feedback based on emotional data and providing an appropriate response according to the user's emotional state.

[1094] The following specific procedures and systems are used to implement the present invention: A server, a user terminal, and appropriate software and hardware are used in combination.

[1095] First, a user uploads document data to the server using their own device (such as a smartphone or tablet). The uploaded document is then converted into text format by the server. This conversion process uses optical character recognition (OCR) technology. The server then preprocesses the text data, removing unnecessary information and normalizing the text, among other processes. This generates clean data suitable for analysis.

[1096] The server trains a generative artificial intelligence (AI model) based on pre-set scoring and evaluation criteria. This trained AI model analyzes the pre-processed text according to the criteria and automatically scores the answers. The AI ​​model evaluates the content and logical consistency of the answers and calculates a score according to the rubric.

[1097] Based on the results of the automated scoring and the rationale behind it, the server generates feedback. This feedback includes not only the score but also specific scoring rationale and areas for improvement. The server also uses generative artificial intelligence to generate a summary of the submitted document, allowing graders and other stakeholders to quickly understand the content.

[1098] Furthermore, the server uses an emotion analysis engine to obtain the user's emotional data. Emotional data is collected and analyzed from the user's facial expressions and voice. This allows the server to understand the user's emotional state. Based on this emotional data, the server adjusts the tone and content of the feedback. For example, if the user is feeling anxious, the server will provide feedback in a calm tone.

[1099] A concrete example is a system that analyzes customer surveys in stores. Customers fill out and upload the survey using tablets provided in the store or their own smartphones. The server converts the survey data into text format, performs preprocessing, and then analyzes the data using an AI model to identify customer satisfaction levels and areas for improvement. Furthermore, a sentiment analysis engine is used to analyze the customer's facial expressions and voice, allowing the tone and content of the feedback to be adjusted to provide more personalized service.

[1100] As an example of a prompt sentence, in response to the feedback "The food was delicious, but the service was slow," feedback such as "Thank you for rating the food delicious. We will try to improve the slowness of the service" can be generated.

[1101] This enables fair and efficient automatic scoring and provides feedback that takes into account the user's feelings.

[1102] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1103] Step 1:

[1104] Users use their own devices (smartphones or tablets) to enter and upload document data.

[1105] Input: Images or PDF files of survey data and answer sheets

[1106] Output: Document data sent to the server

[1107] The server receives the uploaded document data, and the data sent from the user terminal is stored in a designated storage area on the server.

[1108] Step 2:

[1109] The server converts the received document data into text format.

[1110] Input: Document data (image or PDF)

[1111] Output: Text data

[1112] Specifically, the server uses optical character recognition (OCR) software (e.g., Tesseract OCR) to convert images and PDF files into text data, resulting in text data in a format suitable for analysis.

[1113] Step 3:

[1114] The server preprocesses the text data.

[1115] Input: Text data

[1116] Output: Cleaned text data

[1117] Pre-processing includes removing unnecessary information, normalizing text, checking spelling, etc. The server performs data cleansing using PANDAS and regular expression libraries.

[1118] Step 4:

[1119] The server loads the scoring and judging criteria and trains the AI ​​model.

[1120] Input: Scoring and judging criteria data

[1121] Output: A trained AI model

[1122] The server loads pre-prepared rubrics and guidelines and trains a generative AI model (e.g., BERT or GPT). Training also includes setting scores for each evaluation item and generating evaluation comments.

[1123] Step 5:

[1124] The server performs automatic scoring based on the preprocessed text.

[1125] Input: cleaned text data, trained AI model

[1126] Output: Score and grading rationale

[1127] Specifically, the server uses a trained generative AI model to analyze the content of the text data and generate a score and its rationale according to the rubric.

[1128] Step 6:

[1129] The server generates feedback based on the results of the automatic scoring and the reasons for it.

[1130] Input: Score and scoring rationale

[1131] Output: Feedback data

[1132] The feedback includes not only the score but also specific reasons for the score and areas for improvement. The server uses a generative AI model to generate sentences and create feedback to provide to the user.

[1133] Step 7:

[1134] The server generates a summary of the submission.

[1135] Input: Cleaned text data

[1136] Output: Summary data

[1137] To generate a summary, the server uses a generative AI model (e.g., a news article summarization model) to extract key points and create a concise summary.

[1138] Step 8:

[1139] The server acquires and analyzes the user's emotion data.

[1140] Input: User facial expression images, voice data

[1141] Output: Emotion data

[1142] The server uses an emotion analysis engine (e.g., OpenCV or TensorFlow) to analyze the user's facial expressions and voice obtained from smart glasses or a tablet to understand their emotional state.

[1143] Step 9:

[1144] The server adjusts the tone and content of the feedback based on the emotional data.

[1145] Input: Emotion data, feedback data

[1146] Output: Regulated feedback data

[1147] Adjust the content and tone of the feedback depending on the user's emotional state. For example, provide calmer, more encouraging feedback to a user who is feeling anxious.

[1148] Step 10:

[1149] The server provides the generated feedback and summaries to the user and grader.

[1150] Input: Adjusted feedback data, summary data

[1151] Output: Feedback and summaries provided to users and graders

[1152] The server generates feedback and summaries, which are then sent to the user's device and provided to the grader, enabling fair and efficient evaluation and emotionally sensitive feedback.

[1153] 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.

[1154] 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.

[1155] 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.

[1156] [Fourth embodiment]

[1157] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1158] 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.

[1159] 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).

[1160] 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.

[1161] 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.

[1162] 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).

[1163] 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. 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.

[1164] 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.

[1165] 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.

[1166] 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.

[1167] 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.

[1168] 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.

[1169] 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."

[1170] This invention relates to a system that automates the grading of document screening and written exams. In this system, documents are uploaded from the user's terminal, and the server converts the documents into text format and performs preprocessing. The server then loads the grading and examination criteria, trains a generating AI to understand the criteria, and the AI ​​performs automatic grading. The system also includes a function to generate feedback based on the scores and grading rationale, and to generate summaries of the submitted documents.

[1171] User document upload

[1172] A user uploads documents to the server using a terminal, such as an answer sheet for a particular exam or a business proposal. The user can upload the documents through a web interface, and the server receives the uploaded documents.

[1173] Document text conversion and preprocessing

[1174] The server converts the uploaded documents into a text format, then preprocesses the text data, for example by removing unnecessary information and normalizing the text, to produce clean data suitable for analysis.

[1175] Loading scoring and judging criteria and training the AI

[1176] The server loads the scoring and judging criteria. These criteria are predefined rubrics or guidelines that the server uses to train the generative AI. The AI ​​understands these criteria and uses them as the basis for its analysis.

[1177] Run automatic grading

[1178] The server then passes the preprocessed text to a generative AI, which then automatically analyzes and scores the content, evaluating the content and logical consistency of the answers and calculating a score according to a rubric.

[1179] Generate feedback

[1180] Based on the results of the automatic scoring and the rationale behind it, the server generates feedback for the user. This feedback includes not only the score but also specific reasons for the scoring and suggestions for improvement, so the user can clearly understand what was evaluated and what areas need improvement.

[1181] Generate a summary of the submitted documents

[1182] Additionally, the server uses a generator to generate summaries of the submitted documents, which are provided by the server to graders and other interested parties for quick understanding.

[1183] Specific examples

[1184] For example, consider the case where a student uploads an answer sheet for a written exam. When the user scans and uploads the answer sheet from their device, the server converts it into text format, removes noise, and performs preprocessing. The server then trains an AI based on pre-defined scoring criteria, which then automatically scores the answer sheet. As a result, specific feedback is generated and provided to the student along with a score. An answer summary is also generated for the instructor, allowing them to quickly compare and consider the answers of multiple students.

[1185] This system ensures fair and efficient scoring, increases the credibility of the assessment, and allows assessors to focus on more important tasks.

[1186] The processing flow will be explained below.

[1187] Step 1:

[1188] The user uploads the document from the device.

[1189] Specifically, the user selects document files such as answer sheets and proposals through the web interface and clicks the send button to the server, which receives and stores the uploaded files.

[1190] Step 2:

[1191] The server converts the uploaded document into a text format.

[1192] Specifically, the saved file is read and images and PDF documents are converted to text format using OCR (Optical Character Recognition) technology, thereby obtaining the text data.

[1193] Step 3:

[1194] The server preprocesses the text data.

[1195] Specifically, unnecessary characters and spaces are removed from the acquired text data and normalization is performed. For example, this includes combining multiple spaces into one and standardizing line breaks. This process generates clean text data that is easy to analyze.

[1196] Step 4:

[1197] The server loads the scoring and judging criteria.

[1198] Specifically, predefined rubrics and grading guidelines are stored as JSON or XML format files, which the server reads and treats as data, making the evaluation criteria clear.

[1199] Step 5:

[1200] The server passes the reference data to the generative artificial intelligence and trains the AI.

[1201] Specifically, the reference data is fed into a generative AI model, and machine learning techniques are used to train the model. At this stage, the AI ​​learns how to make evaluations.

[1202] Step 6:

[1203] The server passes the preprocessed text to a generative artificial intelligence for analysis.

[1204] Specifically, clean text data is fed into an AI model, which analyzes the content, understands the context, and calculates a score for each part based on a scoring criteria.

[1205] Step 7:

[1206] The server generates the results and rationale for the automatic scoring.

[1207] Specifically, the system organizes the scores calculated by the AI ​​and the analytical results that form the basis for the scores, and generates feedback data based on this. For example, it clearly shows which areas received high marks and which areas need improvement.

[1208] Step 8:

[1209] The server provides feedback to the user.

[1210] Specifically, the generated feedback data is returned to the user's device, where the user can check the score and specific evaluation comments through a web interface.

[1211] Step 9:

[1212] The server generates a summary of the submitted documents using artificial intelligence.

[1213] Specifically, the clean text data is fed into a summary generation module, which extracts key points and summarizes them into short sentences, generating a summary of the submitted document.

[1214] Step 10:

[1215] The server provides the summary results to the grader.

[1216] Specifically, the generated summary data is provided to the grader's terminal, allowing the grader to quickly evaluate and compare multiple documents.

[1217] The above processing steps enable fair and efficient scoring and provide appropriate feedback and summaries to users and scorers.

[1218] Example 1

[1219] 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."

[1220] Traditionally, document screening and written exam scoring have been done manually, often resulting in a lack of fairness and efficiency. Furthermore, bias among evaluators and variations in evaluation criteria can be problematic. Furthermore, summarizing documents to quickly grasp their contents has also been done manually, requiring time and effort. To address these issues, the present invention aims to provide a fair and efficient system that automates the entire process, from document uploading to automatic scoring, feedback generation, and summary generation.

[1221] 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.

[1222] In this invention, the server includes means for uploading documents from a user terminal to the server, means for converting the uploaded documents into text format using optical character recognition software and performing text cleaning, means for reading scoring and review criteria and having a generative artificial intelligence learn the criteria, means for the generative artificial intelligence to analyze the cleaned text data and perform automatic scoring, means for generating feedback based on the results of the automatic scoring and the basis for that scoring, means for generating summaries of the submitted documents using the generative artificial intelligence, and means for providing the generated feedback and summaries to the user and evaluator, thereby enabling fair and efficient scoring and rapid summary creation.

[1223] A "user terminal" is a device such as a computer or smartphone that is operated by a user.

[1224] A "server" is a computer system that provides information in response to requests from clients over a network.

[1225] "Document" refers to a file containing information in text, images, and other formats.

[1226] "Optical character recognition software" is a program for extracting text information from images.

[1227] "Text format" is a format in which text data is stored and displayed.

[1228] "Text cleaning" is the process of removing unnecessary information from text data and formatting the data.

[1229] "Scoring and judging criteria" are rules and guidelines that define specific evaluation items and how they will be evaluated.

[1230] "Generative AI" is an AI technology that analyzes and generates data.

[1231] "Automatic scoring" is the process by which generative artificial intelligence calculates an evaluation score based on data.

[1232] "Feedback" refers to providing information about the evaluation results and their rationale.

[1233] A "summary" is a concise summary of the main points or content of a text.

[1234] "Evaluator" is the person or system responsible for evaluating a particular document or data.

[1235] This invention relates to a system that automates the grading of document screening and written tests. This system is primarily implemented using a user terminal and a server. Users upload documents using their terminal, and the server converts the documents into text format and performs text cleaning. The server then loads the grading and review criteria, trains a generative AI to learn the criteria, and the AI ​​performs automatic grading. Furthermore, the system generates feedback based on the scores and grading rationale, and also generates summaries of the submitted documents.

[1236] A specific embodiment will be described.

[1237] Users access a web interface using their devices and upload documents, often in PDF or image format. Once uploaded, the documents are sent to a server, which uses OCR software (e.g., commonly used optical character recognition software) to convert the uploaded documents into text format. The server then performs a text cleaning process, removing unnecessary information and normalizing the text.

[1238] The server then loads predefined scoring and judging criteria, defined as rubrics or guidelines and stored in a file format. The server then uses this to train a generative artificial intelligence (e.g., a commonly used generative AI model) to understand the criteria. This process teaches the AI ​​the fundamentals of analysis.

[1239] The cleaned text data is input into a generative AI model, which then performs automatic scoring. The generative AI analyzes the text data and calculates a score according to criteria. For example, it evaluates the accuracy and logic of the answer and determines the score comprehensively. The server generates feedback based on the results of the automatic scoring and the reasons for it. This feedback includes not only the score, but also the specific reasons for the scoring and areas for improvement, allowing the user to clearly understand what was evaluated and what areas need improvement.

[1240] It also has a function to generate summaries for submitted documents. The server uses a generative AI model to generate a concise summary of the main points and content of the text data, which can be used by graders and other stakeholders to quickly understand the content.

[1241] For example, imagine a student uploading an answer sheet for a written exam. After the user scans and uploads the answer sheet, the server uses OCR software to convert it into text format and cleans the text. The server then loads pre-prepared grading criteria and trains the AI ​​using a generative AI model. The AI ​​then automatically grades the answer sheet and generates specific feedback for the student. In addition, teachers are provided with a summary of the answer sheet, allowing them to quickly compare and contrast the answers of multiple students.

[1242] This system enables fair and efficient scoring and rapid summary creation, significantly reducing the time and effort required for evaluation and allowing evaluators to focus on more important tasks.

[1243] And here's an example prompt for using a generative AI model: "Upload student answer sheets, grade them, and generate feedback."

[1244] By entering this prompt, the system will automatically execute a series of processes and provide high-quality evaluation results and feedback.

[1245] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1246] Step 1:

[1247] The user accesses the web interface using a terminal and uploads a document. Specifically, the user clicks the "Upload Document" button on the screen to open a file browser. The user selects the document file (e.g., a PDF file) they want to upload and presses the "Open" button. The selected file is sent to the server.

[1248] Input: A document file selected by the user

[1249] Output: The document file is sent to the server.

[1250] Step 2:

[1251] The server receives the submitted document file. It converts the uploaded document file into text format using OCR software (e.g., optical character recognition software). The server then performs text cleaning, removing unnecessary information and normalizing the text. Specifically, it removes whitespace, normalizes special characters, and standardizes line breaks.

[1252] Input: Document file sent by user

[1253] Output: Cleaned text data

[1254] Step 3:

[1255] The server loads predefined scoring and judging criteria. These criteria are defined as rubrics or guidelines and saved in a file format. The server then uses these to train a generative AI model to understand the criteria. During this process, the AI ​​learns the evaluation criteria and scoring methods.

[1256] Input: Rubric or guideline file

[1257] Output: A generative AI model that understands the criteria

[1258] Step 4:

[1259] The server inputs the cleaned text data into the generative AI model, which then analyzes the text data and calculates a score based on the scoring criteria. Specifically, the model evaluates the content, logic, grammar, etc. of the text data to determine an overall score.

[1260] Input: Cleaned text data

[1261] Output:Score result

[1262] Step 5:

[1263] The server generates feedback based on the score calculated by the generative AI model and the rationale for it. This feedback includes not only the score but also specific reasons for the score and areas for improvement. This allows the user to clearly understand which aspects were evaluated and which areas need improvement.

[1264] Input: Score results and scoring basis

[1265] Output: Feedback text

[1266] Step 6:

[1267] The server then feeds the cleaned text data back into the generative AI model to generate summaries of the submitted documents, which can be used by graders and other stakeholders to quickly understand the content.

[1268] Input: Cleaned text data

[1269] Output: Summary text

[1270] Step 7:

[1271] The server provides the generated feedback and summary to the user and the evaluator by posting the feedback to the user's account and displaying the summary on the grader's dashboard, allowing both parties to quickly check the evaluation results and summary information.

[1272] Input: Feedback text, Summary text

[1273] Output: Provided to users and evaluators

[1274] (Application example 1)

[1275] 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."

[1276] Currently, quality control and inspection work in factories is heavily dependent on human resources, resulting in issues with consistency and efficiency of evaluation. In particular, it is difficult to quickly process large amounts of inspection data and provide accurate feedback. For this reason, a new system is needed to automate and streamline quality control.

[1277] 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.

[1278] In this invention, the server includes a means for uploading documents from a user terminal to the server, a means for converting the uploaded documents into text format and preprocessing them, a means for reading grading and examination criteria and having a generating AI understand the criteria, a means for the generating AI to analyze the preprocessed text according to the criteria and perform automatic grading, a means for generating feedback based on the results of the automatic grading and the basis for that, a means for generating summaries of the submitted documents using the generating AI, a means for providing the generated feedback and summaries to the user and grader, and a machine for collecting and uploading inspection data, which enables the automation and efficiency of quality control and inspection work.

[1279] "Uploading a document" is the act of transferring a document file from a user terminal to a server.

[1280] "User terminal" refers to a device such as a computer, smartphone, or tablet connected to the Internet.

[1281] A "server" is a central processing unit that receives, processes, stores, and provides data.

[1282] "Convert to text format" is the process of converting uploaded documents into text data.

[1283] "Preprocessing" is the process of preprocessing data to make it suitable for analysis.

[1284] "Scoring and evaluation criteria" are guidelines that include evaluation items and rules.

[1285] "Generative artificial intelligence" refers to AI systems that use natural language processing and machine learning to perform specific tasks.

[1286] "Understanding the standards" means having the AI ​​learn the scoring and judging criteria and then perform analysis according to those rules.

[1287] "Automatic scoring" is the process by which AI analyzes text and calculates a score based on pre-set criteria.

[1288] "Feedback" is response information that includes the score results and the basis for the evaluation.

[1289] A summary is a short sentence that succinctly summarizes the content of a longer piece of text.

[1290] "Inspection data" refers to information collected in factories and other locations that is used to evaluate the quality of products and processes.

[1291] A "machine" is a hardware-based device used to collect inspection data.

[1292] The present invention relates to a system for automating quality control and inspection work in factories. This system is installed on a factory robot, uploads product inspection data to a server, and a generative artificial intelligence (AI) analyzes and evaluates the data. Specific embodiments of the present invention will be described.

[1293] System Program

[1294] The system consists of the following components:

[1295] Document upload: Uploads inspection data from the user's device to the server.

[1296] Data conversion and pre-processing: The server converts the uploaded data into text format and removes unnecessary information to produce clean data suitable for analysis.

[1297] Loading of scoring and judging criteria and training of AI: The server loads the quality criteria and trains the generative AI to understand the criteria.

[1298] Automatic scoring: The server passes the preprocessed data to the AI, which analyzes it and performs automatic scoring.

[1299] Feedback generation: Based on the scoring results and their rationale, feedback is generated to be returned to the user.

[1300] Summary generation: The server generates a summary of the test data and provides it to the assessor.

[1301] Hardware and Software

[1302] Hardware: Factory robots, servers, user terminals

[1303] Software: Python, requests module, TextBlob, AI model (generative AI model)

[1304] Data processing and calculation

[1305] Data processing: The test data uploaded by the user's device is converted into text format on the server, and then pre-processed to remove noise and unnecessary information.

[1306] Data calculation: The server trains the AI ​​based on the scoring criteria, and the AI ​​analyzes and evaluates the test data. Based on the evaluation results, feedback and summaries are generated.

[1307] Specific examples

[1308] For example, a quality control officer at a factory collects product inspection data and uploads it from the user's device to a server. The uploaded data is converted into text format, preprocessed, and stored on the server as clean data. The server then trains an AI based on preset quality standards, which then analyzes the inspection data and automatically evaluates it. The evaluation results, along with a score, are generated as specific feedback and provided to the person in charge. A summary is also generated, allowing for quick quality evaluation.

[1309] Prompt Sentence Examples

[1310] "Factory robots collect inspection data and upload it to a server. The server converts the data into text format, preprocesses it, and then uses AI to automatically evaluate it. Create an application that generates a score and feedback on areas for improvement, as well as a summary of the inspection data."

[1311] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1312] Step 1:

[1313] The user's device collects the test data and uploads it to the server. Specifically, the camera or sensor on the user's device acquires the test data and saves it in a file format. The file is then sent to the server via the Internet. In this case, the input is the test data file, and the output is the test data saved on the server.

[1314] Step 2:

[1315] The server converts the uploaded inspection data into text format and performs preprocessing. Specifically, it first converts images and other data formats into text using optical character recognition (OCR) technology, then normalizes the text, checks spelling, and removes unnecessary information. The input is the inspection data file, and the output is clean text data.

[1316] Step 3:

[1317] The server loads the scoring and review criteria and has the generative AI understand the criteria. Specifically, it loads predefined quality criteria and provides the content to the AI ​​model for training. The input is the quality criteria data, and the output is the trained AI model.

[1318] Step 4:

[1319] The server passes the preprocessed text data to the AI ​​model, which performs automatic scoring. Specifically, the clean text data is input into the evaluation algorithm, which generates an evaluation score and commentary. The input is the preprocessed text data, and the output is the evaluation result and its rationale.

[1320] Step 5:

[1321] The server generates feedback based on the results of the automatic scoring and the rationale behind it. Specifically, it generates points for improvement and specific advice from the evaluation results and compiles them into a feedback document. The input is the evaluation results and their rationale, and the output is the feedback document.

[1322] Step 6:

[1323] The server generates a summary of the submitted document. Specifically, it uses natural language processing techniques to extract important parts of the text and summarize them concisely. The input is preprocessed text data, and the output is a summary.

[1324] Step 7:

[1325] The server provides the generated feedback and summary to the user and evaluator. Specifically, it sends the generated feedback document and summary to the user and evaluator via email or a dedicated web interface. The input is the feedback document and summary, and the output is the information sent to the user and evaluator.

[1326] 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.

[1327] This invention combines a system that automates document screening and written test scoring with an emotion engine that recognizes user emotions. Feedback can be generated based on the user's emotion data, improving the quality of evaluation. The basic flow of this system is that the user uploads documents from their device, and the server converts them into text format and preprocesses them. The server then passes the reference data to a generative AI, which then automatically scores the documents. The system also includes a function to generate feedback based on the scores and scoring rationale, and to generate summaries of the submitted documents. The server also uses the emotion engine to analyze the user's emotion data and reflect this in the feedback.

[1328] User document upload

[1329] A user uploads documents to the server using a terminal, such as an answer sheet for a particular exam or a business proposal. The user can upload documents through a web interface, and the server receives and stores the uploaded documents.

[1330] Document text conversion and preprocessing

[1331] The server converts the uploaded documents into a text format, then preprocesses the text data, for example by removing unnecessary information and normalizing the text, to produce clean data suitable for analysis.

[1332] Loading scoring and judging criteria and training the AI

[1333] The server loads the scoring and judging criteria. These criteria are predefined rubrics or guidelines that the server uses to train the generative AI. The AI ​​understands these criteria and uses them as the basis for its analysis.

[1334] Run automatic grading

[1335] The server then passes the preprocessed text to a generative AI, which then automatically analyzes and scores the content, evaluating the content and logical consistency of the answers and calculating a score according to a rubric.

[1336] Generate feedback

[1337] Based on the results of the automatic scoring and the rationale behind it, the server generates feedback for the user. This feedback includes not only the score but also specific reasons for the scoring and suggestions for improvement, so the user can clearly understand what was evaluated and what areas need improvement.

[1338] Generate a summary of the submitted documents

[1339] The server uses artificial intelligence to generate summaries of submitted documents, which are provided to graders and other stakeholders for quick understanding.

[1340] Acquiring and analyzing emotion data

[1341] The server uses an emotion engine to recognize the user's emotions. For example, it collects emotion data from the user's facial expressions and voice and analyzes that data. This allows it to understand the user's emotional state.

[1342] Emotion-based feedback adjustment

[1343] The content and tone of the feedback can be adjusted based on the emotional data recognized by the emotion engine. For example, if the user is nervous, the feedback can be given in a calmer tone, taking into consideration the user's emotions.

[1344] Specific examples

[1345] For example, consider the case where a student uploads an answer sheet for a written exam. When the user scans and uploads the answer sheet from their device, the server converts it into text format, removes noise, and performs preprocessing. The server then trains an AI based on pre-defined scoring criteria, which then automatically scores the answer sheet. The AI ​​generates a score and specific scoring rationale, which the server returns to the user as feedback. Furthermore, an emotion engine recognizes the user's emotional data and adjusts the tone of the feedback based on that data. For example, if the user is feeling anxious, the feedback will be adjusted to include more encouraging words. An answer summary is also generated for the instructor, allowing them to quickly compare and contrast the answers of multiple students.

[1346] This system achieves fair and efficient scoring and provides appropriate feedback and summaries to users and graders. Furthermore, by using an emotion engine, it is possible to provide feedback that takes into account users' emotions, thereby improving satisfaction with the evaluation.

[1347] The processing flow will be explained below.

[1348] Step 1:

[1349] The user uploads the document from the device.

[1350] Specifically, users select document files such as answer sheets or proposals through the web interface and click the submit button. The server receives the uploaded files and saves them in a designated folder.

[1351] Step 2:

[1352] The server converts the uploaded document into a text format.

[1353] Specifically, the saved file is read and images and PDF documents are converted into text format using OCR (Optical Character Recognition) technology, and the document contents are obtained as text data.

[1354] Step 3:

[1355] The server preprocesses the text data.

[1356] Specifically, unnecessary characters and spaces are removed from the acquired text data and normalized. For example, multiple spaces are combined into one and line breaks are standardized. This puts the text data into a clean format that is easy to analyze.

[1357] Step 4:

[1358] The server loads the scoring and judging criteria.

[1359] Specifically, predefined rubrics and scoring guidelines are stored as JSON or XML files, which the server reads to obtain baseline data that is used for subsequent analysis.

[1360] Step 5:

[1361] The server passes the reference data to the generative artificial intelligence and trains the AI.

[1362] Specifically, the imported reference data is input into the AI ​​model, and the model is trained using machine learning algorithms. At this stage, the AI ​​learns how to evaluate the data and builds the foundation for analysis.

[1363] Step 6:

[1364] The server passes the preprocessed text to a generative artificial intelligence for analysis.

[1365] Specifically, clean text data is fed into an AI model, which analyzes the content. The AI ​​then understands the context and calculates a score for each part according to a set of criteria, such as logical consistency and accuracy of information.

[1366] Step 7:

[1367] The server generates the results and rationale for the automatic scoring.

[1368] Specifically, the scoring results are organized based on the scores calculated by the AI ​​and the rationale for each. Along with the scores, specific comments are generated to explain why the scores were given.

[1369] Step 8:

[1370] The server provides feedback to the user.

[1371] Specifically, the generated feedback data is returned to the user's device, where the user can check the score and specific evaluation comments through a web interface. The feedback includes areas for improvement and areas that are highly rated.

[1372] Step 9:

[1373] The server generates a summary of the submitted documents using artificial intelligence.

[1374] Specifically, the clean text data is fed into a summary generation module, which extracts key points and summarizes them into short sentences, generating a summary of the submitted document.

[1375] Step 10:

[1376] The server provides the summary results to the grader.

[1377] Specifically, the generated summary data is provided to the grader's terminal, allowing the grader to quickly evaluate and compare multiple documents.

[1378] Step 11:

[1379] The server uses an emotion engine to recognize the user's emotion.

[1380] Specifically, the system analyzes the user's facial expressions and voice to obtain emotional data. For example, it analyzes the video and audio uploaded by the user via a webcam to identify the user's emotional state (joy, anger, sadness, surprise, etc.).

[1381] Step 12:

[1382] The server adjusts the content and tone of the feedback based on the emotional data.

[1383] Specifically, the content of the feedback is adjusted based on the emotional data recognized by the emotion engine. For example, if a user is feeling anxious, feedback containing many encouraging words is provided. In this way, appropriate feedback is generated that takes into account the user's emotions.

[1384] Specific examples

[1385] For example, consider the case where a student uploads an answer sheet for a written exam. The user scans the answer sheet from their device and uploads it to the server, which converts the document into text format and preprocesses it by removing unnecessary information. The server then trains an AI system using pre-defined scoring criteria, which then automatically scores the answer sheet. The server generates feedback based on the scoring results and specific reasons and provides it to the user. Furthermore, an emotion engine recognizes emotions from the user's facial expressions and voice, and adjusts the tone of the feedback based on that data. For example, if the user is feeling anxious, the feedback could include more encouraging words. An answer summary is generated for the instructor, allowing them to quickly compare and consider multiple students' answers.

[1386] This system not only achieves fair and efficient scoring, but also provides feedback that takes users' emotions into consideration, improving satisfaction with the evaluation and allowing evaluators to focus on higher-level tasks.

[1387] Example 2

[1388] 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."

[1389] Conventional scoring systems rely on manual scoring, which inevitably leads to human error and bias. Furthermore, they lack sufficient summaries to quickly grasp the content of submitted documents, placing a heavy workload on the scorer. Furthermore, they are unable to provide feedback that takes into account the user's emotions, which hinders the improvement of satisfaction.

[1390] The specific processing by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes a means for uploading documents from an information terminal to the information processing device, a means for converting the uploaded documents into electronic text format and performing preprocessing, a means for reading grading and examination criteria and having a machine intelligence understand the criteria, a means for the machine intelligence to analyze the preprocessed electronic text according to the criteria and perform automatic grading, a means for generating feedback based on the results of the automatic grading and the basis thereof, a means for generating summaries of the submitted documents using the machine intelligence, a means for providing the generated feedback and summaries to the user and the grader, a means for acquiring and analyzing the user's emotions, and a means for adjusting the content and tone of the feedback based on the user's emotions. This enables fair and effective automatic grading and the provision of feedback that takes the user's emotions into consideration.

[1391] "Uploading a document" refers to the act of electronically transmitting a document from an information terminal to an information processing device.

[1392] "Information terminal" refers to electronic devices such as computers, smartphones, and tablets, which are devices operated by users.

[1393] An "information processing device" is a server that receives, stores, and processes transmitted documents.

[1394] An "electronic text format" is a format in which the contents of a document are expressed as digital data.

[1395] "Preprocessing" is the process of processing data to make documents converted into electronic text format easier to analyze.

[1396] "Grading and Review Criteria" refers to rubrics or guidelines for evaluating the content of a document based on specific criteria.

[1397] "Machine intelligence" is an artificial intelligence technology that learns and analyzes large amounts of data to perform specific tasks.

[1398] "Automatic scoring" is the process by which machine intelligence analyzes documents according to scoring and examination criteria and calculates a score.

[1399] "Feedback" is information that returns the document evaluation results and the basis for them to the user.

[1400] "Summary" refers to a concise summary of the main points or contents of the submitted documents.

[1401] "Emotion data" is information that represents the emotional state of a user, obtained from facial expressions, voice, context, and the like.

[1402] "Adjusting the content and tone of feedback" means changing the sentences and wording of the feedback depending on the user's emotional state.

[1403] The present invention is a system for implementing a series of processes including uploading documents and automatic marking. Detailed embodiments of the system are described below.

[1404] Uploading documents

[1405] Users use their devices to scan or electronically create documents and upload them to the server through a web interface. These documents can include answer sheets for written exams or business proposals.

[1406] Server: Receives uploaded documents and stores the data. The server checks the data for consistency and stores it in the appropriate folder.

[1407] Document text conversion and preprocessing

[1408] Server: The server uses OCR (Optical Character Recognition) software "Tesseract OCR" to convert the received documents into text format. The converted electronic text undergoes pre-processing to make it suitable for analysis. Pre-processing includes:

[1409] Removal of unnecessary information (e.g., page numbers and margins)

[1410] Text normalization (e.g., making all characters lowercase)

[1411] Loading scoring and judging criteria and training the AI

[1412] Server: Loads the scoring and judging criteria and trains a generative AI model (e.g., GPT-4) according to predefined rubrics and guidelines. This training process allows the AI ​​to understand the criteria and perform accurate analysis.

[1413] Run automatic grading

[1414] Server: Passes the preprocessed electronic text to the generative AI model. The AI ​​analyzes the text data and automatically scores it based on specific criteria. For example, it evaluates the logic of the text and the coherence of the content, and generates a score and its rationale.

[1415] Generate feedback

[1416] Server: Generates feedback to be returned to the user based on the automated scoring results and the rationale behind them. This feedback includes not only the score but also specific scoring rationale and areas for improvement. For example, it provides detailed comments such as "You received a high score because the logical structure of your answer was clear."

[1417] Generate a summary of the submitted documents

[1418] Server: Requests the AI ​​model to generate a summary of the submitted documents. The summary is intended to allow graders and other stakeholders to quickly understand the content, and concisely presents the main points and conclusions.

[1419] Acquiring and analyzing emotion data

[1420] Server: The server uses an emotion engine (such as IBM Watson Tone Analyzer or Microsoft Azure Face API) to obtain the user's emotional state. This is done by analyzing facial expressions and voice, and it is possible to understand the user's emotional state.

[1421] Emotion-based feedback adjustment

[1422] Server: Adjust the content and tone of the feedback based on the emotional data. For example, if the user is nervous, provide feedback in a gentler tone. A specific example would be something like, "Thank you for your hard work. Overall, you did a good job, but if you improve this next time, you'll get even better results."

[1423] Examples of concrete examples and prompts

[1424] Example: Consider a student uploading an answer sheet for a final exam. The user (student) scans and uploads the answer sheet from their device, and the server converts it into text format and performs preprocessing. Next, the server trains a generative AI model based on pre-defined scoring criteria, and the AI ​​automatically scores the answer sheet. The AI ​​generates a score and specific scoring reasons, which the server returns to the user as feedback. Furthermore, an emotion engine recognizes the user's emotional data and adjusts the tone of the feedback based on that data. For example, if the user is feeling anxious, the feedback is adjusted to include more encouraging words. An answer summary is also generated for the teacher, allowing them to quickly compare and consider the answers of multiple students.

[1425] Example prompt sentence:

[1426] "Upload students' answer sheets, grade them based on the grading criteria below, and generate feedback. Please be gentle with the tone of your feedback for students who seem nervous."

[1427] In this way, the present invention achieves fair and effective automatic scoring and provides feedback that takes into account the user's emotions. By using an emotion engine, flexible responses can be made according to the user's emotional state, improving satisfaction with the evaluation.

[1428] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1429] Step 1: Upload your documents

[1430] User:

[1431] A user scans or electronically creates a document using a terminal and uploads it to the server through a web interface. The input is a PDF or image file uploaded by the user, and the output is the document data stored on the server. Specifically, the user clicks a specific button, selects a file from a file selection dialog, and submits it.

[1432] Step 2: Text conversion and preprocessing of documents

[1433] server:

[1434] When the server receives the uploaded document, it converts the document into electronic text using OCR software such as "Tesseract OCR." The input is the saved document data, and the output is the converted text data. Specifically, the server inputs the document into the OCR software and temporarily stores the resulting text data.

[1435] server:

[1436] Next, the converted text data is preprocessed. The input is the converted text data, and the output is the preprocessed, clean text data. Specific preprocessing steps include removing unnecessary information (e.g., page numbers and margins) and normalizing the text (e.g., changing all characters to lowercase).

[1437] Step 3: Importing the scoring and judging criteria and training the AI

[1438] server:

[1439] The server loads the scoring and judging criteria and trains the generative AI model. The input is a predefined rubric or guideline, and the output is a trained AI model that understands the criteria. Specifically, the server provides the rubric data to the AI ​​training algorithm, which then learns the criteria.

[1440] Step 4: Run Auto-Scoring

[1441] server:

[1442] The server passes the preprocessed text data to a generative AI model, which analyzes the content and automatically scores the answers. The input is the preprocessed text data and the trained AI model, and the output is the score and the basis for the scoring. Specifically, the AI ​​analyzes the text data and calculates a score based on the content of each answer.

[1443] Step 5: Generate feedback

[1444] server:

[1445] The server generates feedback to be returned to the user based on the results of the automatic scoring and the rationale behind it. The input is the score and the rationale for the scoring, and the output is a feedback message. Specifically, it generates detailed comments such as "Your answer had a clear logical structure, so you got a high score," and provides them to the user.

[1446] Step 6: Generate a summary of the submission

[1447] server:

[1448] The server requests an AI model to generate a summary of the submitted document. The input is preprocessed text data, and the output is summary data. Specifically, the AI ​​analyzes the entire text, extracts key points and conclusions, and generates a summary.

[1449] Step 7: Acquire and analyze emotion data

[1450] server:

[1451] The server uses an emotion engine to acquire and analyze the user's emotional data. The input is the user's facial expressions and voice data, and the output is the analysis result of the user's emotional state. Specifically, data collected from the user's webcam and microphone is input into the analysis engine to determine the user's emotional state.

[1452] Step 8: Adjust your feedback based on emotion

[1453] server:

[1454] The server adjusts the content and tone of the feedback based on the emotional data. The input is the analysis result of the emotional state and the feedback message, and the output is the adjusted feedback message. Specifically, if the user is nervous, the server provides feedback in a gentle tone and includes encouraging words.

[1455] The above steps clearly explain how the system handles uploading documents, automatically scoring them, and providing feedback. It also takes into account users' emotional data to provide more personalized feedback.

[1456] (Application example 2)

[1457] 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."

[1458] Conventional scoring systems for document screening and written tests have limitations in the fairness and efficiency of scoring, making it difficult to provide feedback that takes emotions into account. Furthermore, automated analysis of customer surveys cannot take emotional data into account, making it difficult to accurately grasp customer satisfaction and areas for improvement. The present invention aims to solve these problems by providing a system that achieves fair and efficient scoring and enables the provision of emotional feedback.

[1459] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1460] In this invention, the server includes means for uploading documents from a user terminal to the server, means for converting the uploaded documents into text format and preprocessing them, means for reading grading and review criteria and having a generating AI understand the criteria, means for the generating AI to analyze the preprocessed text according to the criteria and perform automatic grading, means for generating feedback based on the results of the automatic grading and the basis for the results, means for generating summaries of the submitted documents using the generating AI, means for providing the generated feedback and summaries to the user and grader, means for using an emotion analysis engine to acquire user emotion data, and means for adjusting the tone and content of the feedback based on the emotion data, thereby enabling fair and efficient automatic grading and providing feedback that takes user emotions into consideration.

[1461] "Uploading a document" is the act of transferring document data from a user terminal to a server.

[1462] "Server" means a central computer for storing, processing, and analyzing document data.

[1463] "Convert to text format" refers to converting uploaded document image data, PDFs, etc. into text data.

[1464] "Preprocessing" refers to preprocessing such as normalizing text data and removing noise.

[1465] "Grading and review criteria" refers to pre-established rubrics and guidelines for evaluating the content of a document.

[1466] "Generative AI" is an AI model that is trained to perform a specific task.

[1467] "Automatic scoring" is the process of using AI to evaluate the content of a document and calculate a score.

[1468] "Feedback" is response information to the user, including the results of the assessment, detailed explanations of the evaluation, and areas for improvement.

[1469] An "abstract" is a concise summary of the main contents of the submitted document.

[1470] The "emotion analysis engine" is a system for collecting and analyzing emotional data from a user's facial expressions and voice.

[1471] "Adjusting the tone and content of feedback" refers to optimizing the expression of feedback based on emotional data and providing an appropriate response according to the user's emotional state.

[1472] The following specific procedures and systems are used to implement the present invention: A server, a user terminal, and appropriate software and hardware are used in combination.

[1473] First, a user uploads document data to the server using their own device (such as a smartphone or tablet). The uploaded document is then converted into text format by the server. This conversion process uses optical character recognition (OCR) technology. The server then preprocesses the text data, removing unnecessary information and normalizing the text, among other processes. This generates clean data suitable for analysis.

[1474] The server trains a generative artificial intelligence (AI model) based on pre-set scoring and evaluation criteria. This trained AI model analyzes the pre-processed text according to the criteria and automatically scores the answers. The AI ​​model evaluates the content and logical consistency of the answers and calculates a score according to the rubric.

[1475] Based on the results of the automated scoring and the rationale behind it, the server generates feedback. This feedback includes not only the score but also specific scoring rationale and areas for improvement. The server also uses generative artificial intelligence to generate a summary of the submitted document, allowing graders and other stakeholders to quickly understand the content.

[1476] Furthermore, the server uses an emotion analysis engine to obtain the user's emotional data. Emotional data is collected and analyzed from the user's facial expressions and voice. This allows the server to understand the user's emotional state. Based on this emotional data, the server adjusts the tone and content of the feedback. For example, if the user is feeling anxious, the server will provide feedback in a calm tone.

[1477] A concrete example is a system that analyzes customer surveys in stores. Customers fill out and upload the survey using tablets provided in the store or their own smartphones. The server converts the survey data into text format, performs preprocessing, and then analyzes the data using an AI model to identify customer satisfaction levels and areas for improvement. Furthermore, a sentiment analysis engine is used to analyze the customer's facial expressions and voice, allowing the tone and content of the feedback to be adjusted to provide more personalized service.

[1478] As an example of a prompt sentence, in response to the feedback "The food was delicious, but the service was slow," feedback such as "Thank you for rating the food delicious. We will try to improve the slowness of the service" can be generated.

[1479] This enables fair and efficient automatic scoring and provides feedback that takes into account the user's feelings.

[1480] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1481] Step 1:

[1482] Users use their own devices (smartphones or tablets) to enter and upload document data.

[1483] Input: Images or PDF files of survey data and answer sheets

[1484] Output: Document data sent to the server

[1485] The server receives the uploaded document data, and the data sent from the user terminal is stored in a designated storage area on the server.

[1486] Step 2:

[1487] The server converts the received document data into text format.

[1488] Input: Document data (image or PDF)

[1489] Output: Text data

[1490] Specifically, the server uses optical character recognition (OCR) software (e.g., Tesseract OCR) to convert images and PDF files into text data, resulting in text data in a format suitable for analysis.

[1491] Step 3:

[1492] The server preprocesses the text data.

[1493] Input: Text data

[1494] Output: Cleaned text data

[1495] Pre-processing includes removing unnecessary information, normalizing text, checking spelling, etc. The server performs data cleansing using PANDAS and regular expression libraries.

[1496] Step 4:

[1497] The server loads the scoring and judging criteria and trains the AI ​​model.

[1498] Input: Scoring and judging criteria data

[1499] Output: A trained AI model

[1500] The server loads pre-prepared rubrics and guidelines and trains a generative AI model (e.g., BERT or GPT). Training also includes setting scores for each evaluation item and generating evaluation comments.

[1501] Step 5:

[1502] The server performs automatic scoring based on the preprocessed text.

[1503] Input: cleaned text data, trained AI model

[1504] Output: Score and scoring rationale

[1505] Specifically, the server uses a trained generative AI model to analyze the content of the text data and generate a score and its rationale according to the rubric.

[1506] Step 6:

[1507] The server generates feedback based on the results of the automatic scoring and the reasons for it.

[1508] Input: Score and grading rationale

[1509] Output: Feedback data

[1510] The feedback includes not only the score but also specific reasons for the score and areas for improvement. The server uses a generative AI model to generate sentences and create feedback to provide to the user.

[1511] Step 7:

[1512] The server generates a summary of the submission.

[1513] Input: Cleaned text data

[1514] Output: Summary data

[1515] To generate a summary, the server uses a generative AI model (e.g., a news article summarization model) to extract key points and create a concise summary.

[1516] Step 8:

[1517] The server acquires and analyzes the user's emotion data.

[1518] Input: User facial expression images, voice data

[1519] Output: Emotion data

[1520] The server uses an emotion analysis engine (e.g., OpenCV or TensorFlow) to analyze the user's facial expressions and voice obtained from smart glasses or a tablet to understand their emotional state.

[1521] Step 9:

[1522] The server adjusts the tone and content of the feedback based on the emotional data.

[1523] Input: Emotion data, feedback data

[1524] Output: Regulated feedback data

[1525] Adjust the content and tone of the feedback depending on the user's emotional state. For example, provide calmer, more encouraging feedback to a user who is feeling anxious.

[1526] Step 10:

[1527] The server provides the generated feedback and summaries to the user and grader.

[1528] Input: Adjusted feedback data, summary data

[1529] Output: Feedback and summaries provided to users and graders

[1530] The server generates feedback and summaries, which are then sent to the user's device and provided to the grader, enabling fair and efficient evaluation and emotionally sensitive feedback.

[1531] 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.

[1532] 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.

[1533] 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.

[1534] 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.

[1535] 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.

[1536] 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.

[1537] 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).

[1538] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1539] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1540] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1541] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1542] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1543] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1544] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1545] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1546] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1547] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1548] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1549] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1550] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1551] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1552] The following is further disclosed regarding the above embodiment.

[1553] (Claim 1)

[1554] A means for uploading documents from a user terminal to a server;

[1555] means for converting uploaded documents into text format and pre-processing them;

[1556] A means to read the scoring and judging criteria and have the generating AI understand the criteria,

[1557] A means for generating artificial intelligence to analyze and automatically score the preprocessed text according to a standard;

[1558] A means for generating feedback based on the results of automatic scoring and the rationale for doing so;

[1559] means for generating a summary of the submitted documents using artificial intelligence;

[1560] The system includes a means for providing generated feedback and summaries to users and graders.

[1561] (Claim 2)

[1562] The system of claim 1, wherein the generative artificial intelligence analyzes the scores and the scoring basis to perform context-based scoring.

[1563] (Claim 3)

[1564] The system of claim 1, wherein the system provides a summary of the contents of the submitted documents to the grader.

[1565] "Example 1"

[1566] (Claim 1)

[1567] means for uploading documents from a user terminal to a server;

[1568] means for converting the uploaded documents into text format using optical character recognition software and for text cleaning;

[1569] A means for reading the scoring and judging criteria and having the generative AI learn the criteria;

[1570] A generative artificial intelligence analyzes the cleaned text data and automatically scores it.

[1571] A means for generating feedback based on the results of the automatic scoring and the rationale for the same;

[1572] means for generating a summary of the submitted document using generative artificial intelligence;

[1573] The system includes a means for providing generated feedback and summaries to users and evaluators.

[1574] (Claim 2)

[1575] The system of claim 1 employs generative artificial intelligence to analyze scores and scoring rationale and perform context-based scoring.

[1576] (Claim 3)

[1577] 2. The system of claim 1, wherein the system provides a summary of the contents of submitted documents to the evaluator.

[1578] "Application Example 1"

[1579] (Claim 1)

[1580] A means for uploading documents from a user terminal to a server;

[1581] means for converting uploaded documents into text format and pre-processing them;

[1582] A means to read the scoring and judging criteria and have the generating AI understand the criteria,

[1583] A means for generating artificial intelligence to analyze and automatically score the preprocessed text according to a standard;

[1584] A means for generating feedback based on the results of automatic scoring and the rationale for doing so;

[1585] means for generating a summary of the submitted documents using artificial intelligence;

[1586] means for providing the generated feedback and summaries to the user and grader;

[1587] and a machine for collecting and uploading inspection data.

[1588] (Claim 2)

[1589] The system of claim 1, wherein the generative artificial intelligence analyzes the scores and the scoring basis to perform context-based scoring.

[1590] (Claim 3)

[1591] 10. The system of claim 1, wherein a summary of the submitted documents and collected data is generated and provided to the assessor.

[1592] "Example 2: Combining Emotion Engines"

[1593] (Claim 1)

[1594] A means for uploading documents from an information terminal to an information processing device;

[1595] means for converting and pre-processing the uploaded documents into electronic text format;

[1596] A means to read the scoring and judging criteria and have machine intelligence understand the criteria,

[1597] A means for automatically scoring the preprocessed electronic text by machine intelligence according to a standard;

[1598] A means for generating feedback based on the results of automatic scoring and the rationale for doing so;

[1599] means for generating a summary of the submission using machine intelligence;

[1600] means for providing the generated feedback and summaries to the user and grader;

[1601] A means for acquiring and analyzing user emotions;

[1602] A system that includes a means for adjusting the content and tone of feedback based on the user's emotions.

[1603] (Claim 2)

[1604] 10. The system of claim 1, wherein the system uses machine intelligence to analyze the scores and the scoring rationale and perform context-based scoring.

[1605] (Claim 3)

[1606] The system of claim 1, wherein the system provides a summary of the contents of the submitted documents to the grader.

[1607] "Application example 2 when combining emotion engines"

[1608] (Claim 1)

[1609] A means for uploading documents from a user terminal to a server;

[1610] means for converting uploaded documents into text format and pre-processing them;

[1611] A means to read the scoring and judging criteria and have the generating AI understand the criteria,

[1612] A means for generating artificial intelligence to analyze and automatically score the preprocessed text according to a standard;

[1613] A means for generating feedback based on the results of automatic scoring and the rationale for doing so;

[1614] means for generating a summary of the submitted documents using artificial intelligence;

[1615] means for providing the generated feedback and summaries to the user and grader;

[1616] a means using an emotion analysis engine to acquire emotion data of a user;

[1617] The system includes a means for adjusting the tone and content of feedback based on emotional data.

[1618] (Claim 2)

[1619] The system of claim 1, wherein the generative artificial intelligence analyzes the scores and the scoring basis to perform context-based scoring.

[1620] (Claim 3)

[1621] The system of claim 1, wherein the system provides a summary of the contents of the submitted documents to the grader. [Explanation of symbols]

[1622] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for uploading documents from a user terminal to a server; means for converting uploaded documents into text format and pre-processing them; A means to read the scoring and judging criteria and have the generating AI understand the criteria, A means for generating artificial intelligence to analyze and automatically score the preprocessed text according to a standard; A means for generating feedback based on the results of automatic scoring and the rationale for doing so; means for generating a summary of the submitted documents using artificial intelligence; The system includes a means for providing generated feedback and summaries to users and graders.

2. The system of claim 1, wherein the generative artificial intelligence analyzes the scores and the scoring rationale to perform context-based scoring.

3. The system according to claim 1, wherein the content of the submitted document is summarized and provided to the grader.

Images