system

Abstract

We provide the system. [Solution] A terminal for asking interactive questions to employees, A server that receives employee response data, A means for analyzing the aforementioned response data and assigning scores to the evaluation items, means for generating a prompt based on the scored evaluation items, An image generation AI that generates an expression image using the aforementioned prompt and the employee's facial image, Means for saving and displaying the aforementioned facial expression images in chronological order, A system that includes this.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of the chatbot's 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

Summary of the Invention

Problems to be Solved by the Invention

[0004] The mental health of employees is an important factor in improving productivity and working environment in enterprises. However, there is a problem that it is difficult to effectively manage by the conventional method because there is a lack of means to timely grasp the change of employees' mental status and provide appropriate support. In the current questionnaire and interview, it is difficult to elicit the true feelings of employees, and analysis and support based on the results cannot be provided quickly. As a result, stress and anxiety of employees may accumulate, which may result in a decline in the performance of the enterprise and an increase in the turnover rate.

Means for Solving the Problems

[0005] To solve the above problems, this invention provides a terminal for asking interactive questions to employees, a server for receiving and analyzing employee response data, and a system for visualizing employees' mental state based on the analysis results. Specifically, it includes means for analyzing employee response data using natural language processing technology and scoring evaluation items, and means for generating prompts based on the scored evaluation items. It also includes an image generation AI that generates facial expression images using prompts and employee facial images, and means for saving and displaying these facial expression images in chronological order. This system enables timely understanding of employees' mental health status and promptly providing appropriate support. As a result, it can contribute to improving corporate productivity and the working environment.

[0006] A "terminal" refers to a device used to ask employees interactive questions and obtain their answers.

[0007] A "server" refers to a computer system that receives employee response data sent from terminals, analyzes it, and stores it.

[0008] "Response data" refers to the text data of responses entered by employees using their devices.

[0009] "Evaluation items" refer to indicators used to assess the mental health status of employees, which are set based on employee response data.

[0010] "Scoring" refers to the process of assigning numerical values ​​to evaluation items in order to quantitatively assess them.

[0011] A "prompt" refers to a set of instructions generated for input into an image generation AI, based on a scoring system for evaluation.

[0012] "Image generation AI" refers to an artificial intelligence system that uses prompts and employee facial images to generate specified facial expression images.

[0013] "Facial expression images" refer to facial images generated by image generation AI that indicate a specific mental health state of an employee.

[0014] "Saving in chronological order" refers to the process of saving generated facial expression images in chronological order.

[0015] "Display" refers to the process of visualizing and outputting saved facial expression images in a format that allows for review. [Brief explanation of the drawing]

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

Mode for Carrying Out the Invention

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

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

[0019] In the following embodiments, the numbered processor (hereinafter simply referred to as "processor") may be one arithmetic unit or a combination of a plurality of arithmetic units. Also, the processor may be one type of arithmetic unit or a combination of a plurality of types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

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

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

[0022] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0023] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0024] [First Embodiment]

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

[0026] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

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

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

[0029] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0030] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0031] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

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

[0033] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

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

[0035] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

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

[0037] This invention relates to a system for regularly monitoring the mental health status of employees. This system utilizes a dialogue generation AI to conduct conversations and score evaluation items, and then uses these scored evaluation items as prompts to input into an image generation AI, thereby visualizing changes in facial expressions over time.

[0038] System Overview

[0039] 1. Generating dialogue with employees

[0040] The terminal periodically displays questions to employees using conversational AI. For example, it might ask, "How has your workload been lately?" and retrieve the answers.

[0041] The user (employee) enters their answers to questions displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[0042] 2. Analysis of the dialogue content

[0043] The terminal sends employee response data to the server.

[0044] The server analyzes the received response data using natural language processing (NLP). For example, it evaluates the employee's emotions and stress level based on keywords and context in the response.

[0045] Based on the analyzed data, evaluation items (e.g., stress level, happiness level, etc.) are scored. For example, based on the response, "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[0046] 3. Prompting evaluation items

[0047] The server generates prompts for input to the image generation AI based on the scored evaluation items. For example, if the stress level is 8 points, the prompt will be set to "High Stress".

[0048] 4. Generation of facial expression images

[0049] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression based on the prompt (e.g., sad face, tired face, etc.).

[0050] The server receives the generated facial expression image.

[0051] 5. Saving and displaying time-series data

[0052] The server stores the generated facial expression images chronologically. In this way, it can accumulate facial expression images from the past to the present.

[0053] The server visualizes the saved facial expression images and displays them in the form of a dashboard. This dashboard allows HR personnel and managers to quickly grasp the mental health status of employees.

[0054] Specific example

[0055] 1. Interview Session

[0056] The terminal displays the employee with the question, "How is the progress on your recent projects?"

[0057] The user (employee) responded, "Things are progressing smoothly, but I'm feeling a little stressed."

[0058] 2. Analysis and scoring of response data

[0059] The server analyzes this response data using natural language processing and extracts keywords related to stress.

[0060] The server evaluates the stress level based on the extracted data and assigns a score, for example, 5 points.

[0061] 3. Prompt generation

[0062] The server generates a prompt indicating "moderate stress" for a stress level of 5.

[0063] 4. Generation of facial expression images

[0064] The server inputs the prompt "moderate stress" and an image of the employee's face into the image generation AI, which then generates an image of the employee's facial expression that reflects their stress level.

[0065] 5. Saving and displaying time-series data

[0066] The server saves the generated facial expression images to a database and stores them along with past data.

[0067] The server displays this data on a dashboard, allowing HR personnel to view employees' mental health status over time.

[0068] In this way, the system of the present invention, by combining dialogue generation AI and image generation AI, can intuitively grasp the mental health status of employees and quickly provide appropriate support.

[0069] The following describes the processing flow.

[0070] Step 1:

[0071] The device displays questions randomly selected from a pre-configured list to the employee. For example, it might display the question, "How has your workload been recently?"

[0072] Step 2:

[0073] The user (employee) enters their answer to the question displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[0074] Step 3:

[0075] The terminal sends the entered response data and the employee's facial image to the server.

[0076] Step 4:

[0077] The server analyzes the received response data using a natural language processing (NLP) engine. Here, it extracts keywords indicating negative emotions (e.g., "very difficult," "stressful," "tired," etc.).

[0078] Step 5:

[0079] The server scores evaluation items (e.g., stress level, happiness level) based on the frequency and context of extracted keywords. For example, for the response "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[0080] Step 6:

[0081] The server generates prompts for input to the image generation AI based on the scored evaluation items. For example, if the stress level is 8 points, the prompt will be set to "High Stress".

[0082] Step 7:

[0083] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression based on the prompt (e.g., sad face, tired face, etc.).

[0084] Step 8:

[0085] The server receives facial expression images generated by an image generation AI. These generated facial expression images reflect the employee's current mental state.

[0086] Step 9:

[0087] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[0088] Step 10:

[0089] The server displays the saved facial expression images in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can identify an employee's stress level as rising and consider providing additional support.

[0090] This series of processes allows the system to effectively understand the mental health status of employees and take appropriate measures.

[0091] (Example 1)

[0092] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0093] There is a challenge in continuously and efficiently monitoring employees' mental health and providing appropriate support. A decline in employee mental health can lead to a wide range of problems, including decreased productivity and increased turnover, thus necessitating effective monitoring methods.

[0094] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0095] In this invention, the server includes means for asking interactive questions to employees, means for receiving employee response data, means for analyzing the response data and scoring evaluation items, means for generating prompts based on the scored evaluation items, image generation means for generating facial expression images using the prompts and the employee's facial image, and means for saving and displaying the facial expression images in chronological order. This makes it possible to evaluate the mental health status of employees in real time and monitor its changes in chronological order.

[0096] A "terminal" refers to an electronic device that interacts with users (employees), presents questions, and allows them to input answers.

[0097] A "server" refers to a computer system that receives, analyzes, and stores data sent from a terminal.

[0098] "Response data" refers to data that includes the content of responses entered by users (employees) via their devices.

[0099] "Natural language processing technology" refers to techniques for analyzing text data to understand and interpret its meaning and emotions.

[0100] "Evaluation items" refer to indicators and criteria set based on information extracted from response data.

[0101] "Scoring" refers to the process of assigning numerical values ​​based on evaluation criteria.

[0102] A "prompt" refers to a command or instruction given to an image generation device.

[0103] "Image generation means" refers to a device or algorithm that generates a new facial expression image based on a prompt and a user's (employee's) facial image.

[0104] "Facial expression image" refers to an image showing the facial expression of an employee, generated by an image generation device.

[0105] "Storing in chronological order" refers to the process of saving generated images and data sequentially according to the passage of time.

[0106] "Means for saving and displaying" refers to a system or device for accumulating generated facial expression images and presenting them in a way that allows for visual confirmation.

[0107] A "dashboard" refers to an interface that visually displays the mental health status of employees.

[0108] This invention relates to a system for regularly monitoring the mental health status of employees. This system utilizes a dialogue generation AI to conduct conversations and score evaluation items, and then uses these scored evaluation items as prompts to input into an image generation AI, thereby visualizing changes in facial expressions over time.

[0109] System Configuration

[0110] terminal

[0111] These are devices such as PCs and smartphones used by employees, and they present questions through conversational AI. For example, OpenAI's GPT-3 (registered trademark) can be used as the conversational AI.

[0112] server

[0113] This is a computer system that receives, analyzes, and stores response data sent from terminals. The server is equipped with software that utilizes natural language processing technologies (e.g., spaCy and NLTK) and image generation AI (e.g., DALL-E 2).

[0114] Natural Language Processing (NLP) Module

[0115] It is installed on a server, analyzes employee response data, extracts evaluation items, and assigns scores. For example, it can numerically evaluate stress and emotions included in employee responses.

[0116] Image generation AI

[0117] Based on scored evaluation items, prompts are generated, and by inputting these prompts along with the employee's facial image, a new facial expression image is created. For example, DALL-E 2 can be used here.

[0118] database

[0119] It is configured as part of the server and stores the generated facial expression images chronologically. Database systems such as MySQL® or MongoDB are used.

[0120] Dashboard

[0121] This is an interface for HR personnel to visually check the mental health status of employees. It is built using HTML5 and JavaScript (registered trademark).

[0122] Specific operation of the system

[0123] 1. Generating dialogue with employees

[0124] The terminal periodically displays questions to employees via conversational AI. For example, it might generate and display a question such as, "How has your workload been lately?"

[0125] The user (employee) enters their answer to this question into the terminal. For example, they might enter, "It's very difficult and stressful."

[0126] 2. Analysis of the dialogue content

[0127] The device sends the user's response data to the server.

[0128] The server analyzes the received response data using natural language processing technology to assess stress and emotions. For example, it might assign a stress level of 8 out of 10 to the sentence "I'm feeling stressed."

[0129] 3. Prompt generation

[0130] The server generates prompts for input to the image generation AI based on the scored evaluation items. For example, if the stress level is 8 points, it will generate the prompt "high-stress facial expression".

[0131] 4. Generation of facial expression images

[0132] The server inputs the generated prompt text and the employee's facial image into the image generation AI. The image generation AI, for example, uses DALL-E 2 to generate images of high-stress facial expressions.

[0133] The server saves the generated facial expression images to a database.

[0134] 5. Saving and displaying time-series data

[0135] The server organizes the saved facial expression images chronologically and displays them on a dashboard. This allows HR personnel to intuitively understand employees' mental health status over time.

[0136] Specific example

[0137] 1. Interview Session

[0138] Terminal: "How is the project progressing recently?"

[0139] User: "Things are progressing smoothly, but I'm feeling a little stressed."

[0140] 2. Analysis and scoring of response data

[0141] The server analyzes this response data using natural language processing technology and extracts keywords related to stress.

[0142] Based on responses such as "I feel a little stressed," the stress level is rated on a scale of 1 to 5.

[0143] 3. Prompt generation

[0144] Server: Generates the prompt message "Moderate stress expression".

[0145] 4. Generation of facial expression images

[0146] The server inputs the prompt "moderately stressed facial expression" and an image of the employee's face into the image generation AI, which then generates an image of the facial expression that reflects the stress.

[0147] 5. Saving and displaying time-series data

[0148] The server saves the generated facial expression images to a database and displays them on the dashboard.

[0149] HR personnel can see at a glance how stress levels have changed over the past month and the generated facial expressions.

[0150] This system makes it possible to intuitively understand the mental health status of employees and provide prompt and accurate support.

[0151] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0152] Step 1:

[0153] The terminal periodically displays questions to employees via conversational AI. The input is a list of questions pre-configured in the system, and the output is the question text that is displayed. For example, it generates and displays the question, "How has your workload been recently?"

[0154] Step 2:

[0155] The user (employee) enters their answer to the question displayed on the terminal. The input is the user's answer, and the output is the answer data entered on the terminal. For example, the user might enter, "It's very difficult and stressful."

[0156] Step 3:

[0157] The terminal sends the user's entered response data to the server. The input is the user's response data, and the output is the response data sent to the server. This communication is conducted via the HTTPS protocol.

[0158] Step 4:

[0159] The server receives response data sent from the terminal. The input is the response data sent from the terminal, and the output is the response data stored on the server. For example, the data is received in JSON format.

[0160] Step 5:

[0161] The server analyzes the received response data using natural language processing (NLP) techniques. The input is the response data stored on the server, and the output is the analyzed evaluation items. For example, keywords and emotions are extracted using an NLP toolkit (e.g., spaCy or NLTK). Based on the analysis, the stress level is evaluated as 8 points from the response, "It's very difficult and stressful."

[0162] Step 6:

[0163] The server generates prompts based on scored evaluation items. The input is the scored evaluation item, and the output is the prompt text that is input to the image generation AI. For example, if the stress level is 8 points, the server generates the prompt "High stress expression".

[0164] Step 7:

[0165] The server inputs the generated prompt text and the employee's face image into the image generation AI. The input is the prompt text and the employee's face image, and the output is the generated facial expression image. For example, DALL-E 2 is used to generate a high-stress facial expression image.

[0166] Step 8:

[0167] The server receives facial expression images generated by an image generation AI. The input is facial expression image data received from the image generation AI, and the output is facial expression images stored on the server. For example, it decodes Base64 encoded image data.

[0168] Step 9:

[0169] The server stores the received facial expression images in a database in chronological order. The input is the generated facial expression images, and the output is the chronological data stored in the database. For example, it is stored in a database system (e.g., MySQL or MongoDB).

[0170] Step 10:

[0171] The server displays facial expression images stored in the database on a dashboard. Inputs are the facial expression images and evaluation items stored in the database, while output is the visual information displayed on the dashboard. For example, time-series graphs and timelines can be displayed using HTML5 or JavaScript. HR personnel can view this to quickly grasp the mental health status of employees.

[0172] (Application Example 1)

[0173] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0174] In modern factory environments, properly monitoring employees' mental health is crucial. However, conventional methods do not provide real-time visibility into employees' mental health, potentially leading to the accumulation of overwork and stress. This can result in decreased productivity and increased safety risks. This invention aims to provide a robot-based employee mental health monitoring system within a factory, enabling real-time and time-series tracking of employees' mental health.

[0175] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0176] In this invention, the server includes terminal means for asking interactive questions to employees, means for receiving employee response data, means for analyzing the response data and scoring evaluation items, means for generating prompts based on the scored evaluation items, image generation AI means for generating facial expression images using the prompts and the employee's facial image, means for saving and displaying the facial expression images in chronological order, in-factory robot means for periodically displaying questions to employees and collecting responses, means for generating facial expression images using the image generation AI based on the evaluation items, and means for saving the generated facial expression images in chronological order in a database. This makes it possible to grasp the mental health status of employees in the factory in real time and provide support at the appropriate time.

[0177] A "terminal" is a device used to ask employees interactive questions.

[0178] A "server" is a system that receives, analyzes, and stores employee response data.

[0179] "Response data" refers to the information that employees enter in response to interactive questions.

[0180] "Methods for quantifying evaluation items" refer to technical means for analyzing employee response data and expressing the results as scores.

[0181] A "prompt" is a phrase or command used to give instructions to an image generation AI.

[0182] "Image generation AI" is artificial intelligence that generates facial expression images based on prompts and images of employees' faces.

[0183] A "facial expression image" is an image representing an employee's facial expression, created by a generative AI model based on prompts.

[0184] "Means for saving and displaying in chronological order" refers to technical means for recording generated facial expression images in chronological order and displaying them as needed.

[0185] A "robot in a factory" is an automated device that displays interactive questions to employees in a factory environment and collects their responses.

[0186] "Natural language processing technology" refers to language processing techniques used to analyze employee response data.

[0187] A "dashboard" is a system that visually displays scored evaluation items, allowing for a quick overview of employees' mental health status.

[0188] A "database" is a system for storing and managing generated facial expression images and evaluation data.

[0189] The following system is configured as an embodiment of this invention. The system aims to monitor the mental health status of employees in a factory environment in real time and provide support at the appropriate time.

[0190] First, the terminal periodically displays interactive questions to the employee. This terminal can be used, for example, by a robot in the factory approaching an employee and asking, "How is the progress on your recent work?" The employee then enters their answer to the question.

[0191] The terminal receives employee response data and sends it to the server. This response data is analyzed on the server using natural language processing technology. Specifically, keywords and context are extracted from the responses using transformer models, and sentiment analysis is performed. As a result of this analysis, evaluation items (e.g., stress level) can be scored.

[0192] The server generates prompts based on scored evaluation items. For example, if the stress level is 8 points, it generates a prompt that says "High Stress." This prompt and the employee's facial image are then input into an image generation AI to generate an expression image. Based on this prompt, the image generation AI reflects the stress expression on the employee's facial image.

[0193] The generated facial expression images are stored chronologically in a database by the server. This stored data can later be visualized by HR personnel and managers through a dashboard, allowing them to quickly grasp the mental health status of employees.

[0194] Hardware and software to be used

[0195] Hardware: Factory robots, interactive terminals, servers, databases

[0196] Software: Natural language processing technology (transformer models), image generation AI, database management systems, dashboard display software

[0197] Examples of specific cases and prompt statements

[0198] For example, suppose an employee answers the question, "How is your recent work progressing?" with, "It's extremely difficult and stressful." Based on this answer, a prompt "High Stress" is generated. As a result, the image generation AI receives a prompt message like the following:

[0199] Example of a prompt:

[0200] Prompt: "High stress level, employee face image: face_image.jpg"

[0201] Based on this prompt, the image generation AI generates high-stress facial expression images corresponding to the employee's face image, and the server stores these in a database chronologically. HR personnel can then view the stored data on a dashboard, gaining a clear overview of employees' mental health status. In this way, the system enables real-time monitoring of employee mental health within the factory and the provision of appropriate support.

[0202] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0203] Step 1:

[0204] A terminal (a robot in the factory) periodically displays interactive questions to employees. For example, it might ask, "How is the progress of your recent work?" The user (employee) might respond, "It's very difficult and stressful." In this step, the terminal displays the question, and the user enters the answer. The input data is the user's response.

[0205] Step 2:

[0206] The terminal receives the user's response data and sends it to the server. The server stores the received data for analysis. In this step, the input is the user's response data, and the output is the data sent to the server.

[0207] Step 3:

[0208] The server analyzes the received response data using natural language processing techniques to evaluate emotions and stress levels. This process utilizes a transformer model. Specifically, it analyzes keywords and context from the responses to perform sentiment analysis. The input is the received response data, and the output is the evaluated items, such as the stress level.

[0209] Step 4:

[0210] The server assigns scores to the evaluation items. For example, based on the analyzed emotion response "It's very difficult and stressful," the stress level is scored as 8 out of 10. The input is the evaluation items obtained in the previous step, and the output is the scored stress level.

[0211] Step 5:

[0212] The server generates prompts based on the scored evaluation items. For example, if the stress level is 8, it will generate the prompt "High Stress." The input is the scored stress level, and the output is the generated prompt.

[0213] Step 6:

[0214] The server inputs the generated prompt and the employee's face image into the image generation AI, which then generates an expression image that reflects stress. Specifically, the image generation AI changes the facial expression based on the prompt and the face image. The input for this step is the generated prompt and the employee's face image, and the output is the generated expression image.

[0215] Step 7:

[0216] The server stores the generated facial expression images in a database in chronological order. The stored facial expression images are later used as data for HR personnel and managers to review through a dashboard. The input for this step is the generated facial expression images, and the output is the data stored in the database.

[0217] Step 8:

[0218] The server uses a dashboard to visualize stored facial expression images for HR personnel and managers. This allows them to quickly understand the mental health status of employees. The input is facial expression images stored in the database, and the output is visualized data displayed on the dashboard.

[0219] This clearly explains what data processing and calculations are performed at each step, and what inputs and outputs are generated. Specific operations include data analysis using natural language processing, prompt generation, use of image generation AI, saving to a database, and dashboard display. Based on this, a system for monitoring the user's mental health status in real time is realized.

[0220] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0221] This invention relates to a system for regularly monitoring the mental health status of employees and providing appropriate support. This system utilizes a combination of a dialogue generation AI, an image generation AI, and an emotion engine. The system acquires response data through dialogue with employees, analyzes it, and scores evaluation items. Furthermore, it extracts emotional information using the emotion engine and generates facial expression images by inputting prompts generated based on the scored evaluation items and emotional information into the image generation AI. The generated facial expression images are saved chronologically and displayed on a dashboard.

[0222] System Overview

[0223] 1. Generating dialogue with employees

[0224] The terminal periodically displays questions to employees using conversational AI. For example, it might ask, "How has your workload been lately?" and retrieve the answers.

[0225] The user (employee) enters their answer to the question displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[0226] 2. Analysis of the dialogue content

[0227] The terminal sends the entered response data and the employee's facial image to the server.

[0228] The server analyzes the received response data using a natural language processing (NLP) engine. Here, it extracts keywords indicating negative emotions (e.g., "very difficult," "stressful," "tired," etc.).

[0229] 3. Extraction of emotional information

[0230] The server uses an emotion engine to extract employee emotional information based on extracted keywords and context. For example, based on the phrase "very difficult" from the response, it extracts the emotional information "high stress."

[0231] 4. Scoring of evaluation items

[0232] The server uses the extracted emotional information to score evaluation items (e.g., stress level, happiness level). For example, in response to the answer, "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[0233] 5. Prompt generation based on evaluation items and sentiment information

[0234] The server generates prompts based on scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the server will set the prompt to "high stress."

[0235] 6. Generation of facial expression images

[0236] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression (e.g., sad face, tired face) based on the prompt.

[0237] The server receives facial expression images generated by an image generation AI. These generated facial expression images reflect the employee's current mental state.

[0238] 7. Saving and displaying time-series data

[0239] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[0240] The server displays stored facial images and emotional information in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can identify an employee's stress level as rising and consider providing additional support.

[0241] Specific example

[0242] 1. Interview Session

[0243] The terminal displays the employee with the question, "How is the progress on your recent projects?"

[0244] The user (employee) responded, "Things are progressing smoothly, but I'm feeling a little stressed."

[0245] 2. Analysis of response data and extraction of sentiment information

[0246] The device sends this response data and facial image to the server.

[0247] The server analyzes this response data using natural language processing and extracts keywords related to stress.

[0248] The server extracts emotional information from keywords extracted using an emotion engine. For example, from the response "I feel a little stressed," it extracts the emotional information "moderate stress."

[0249] 3. Scoring of evaluation items

[0250] The server evaluates the stress level based on emotional information and evaluation criteria, and assigns a score, for example, 5 points.

[0251] 4. Prompt generation

[0252] The server generates a "moderate stress" prompt based on a stress level of 5 and emotional information indicating "moderate stress."

[0253] 5. Generation of facial expression images

[0254] The server inputs the prompt "moderate stress" and an image of the employee's face into the image generation AI, which then generates an image of the employee's facial expression that reflects their stress level.

[0255] The server receives facial expression images generated by the image generation AI.

[0256] 6. Saving and displaying time-series data

[0257] The server saves the generated facial expression images to a database and stores them along with past data.

[0258] The server displays saved facial images on a dashboard, allowing HR personnel to monitor employees' mental health status over time.

[0259] In this way, the system of the present invention, by combining a dialogue generation AI, an image generation AI, and an emotion engine, can intuitively grasp the mental health status of employees and quickly provide appropriate support.

[0260] The following describes the processing flow.

[0261] Step 1:

[0262] The device displays questions randomly selected from a pre-configured list to the employee. For example, it might display the question, "How has your workload been recently?"

[0263] Step 2:

[0264] The user (employee) enters their answer to the question displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[0265] Step 3:

[0266] The terminal sends the entered response data and the employee's facial image to the server.

[0267] Step 4:

[0268] The server analyzes the received response data using a natural language processing (NLP) engine. Here, it extracts keywords indicating negative emotions (e.g., "very difficult," "stressful," "tired," etc.).

[0269] Step 5:

[0270] The server uses an emotion engine to extract employee emotional information based on extracted keywords and context. For example, based on the phrase "very difficult" from the response, it extracts the emotional information "high stress."

[0271] Step 6:

[0272] The server uses the extracted emotional information to score evaluation items (e.g., stress level, happiness level). For example, in response to the answer, "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[0273] Step 7:

[0274] The server generates a prompt for input to the image generation AI based on the quantified evaluation items and sentiment information. For example, when the stress level is 8 points and there is sentiment information of "high stress", a prompt of "high stress" is set.

[0275] Step 8:

[0276] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression (e.g., a sad expression, a tired face, etc.) based on the prompt.

[0277] Step 9:

[0278] The server receives the expression image generated by the image generation AI. This generated expression image reflects the current mental state of the employee.

[0279] Step 10:

[0280] The server stores the generated expression image in the database in chronological order. As a result, the expression images from the past to the present are accumulated.

[0281] Step 11:

[0282] The server displays the stored expression images and sentiment information in a dashboard format. Users (HR staff and managers) can view this dashboard and grasp the mental health status of employees in chronological order. For example, confirm that the stress level of a specific employee is rising and consider providing additional support.

[0283] Through this series of processes, the system can effectively grasp the mental health status of employees and take appropriate measures.

[0284] (Example 2)

[0285] Next, Example 2 will be described. In the following description, the data processing device 12 is referred to as a "server", and the smart device 14 is referred to as a "terminal".

[0286] An employee's mental health state has a great impact on productivity and workplace human relations. However, it is not easy to accurately grasp an employee's mental state and provide support at an appropriate timing. Conventional methods often rely solely on the results of simple questionnaires and the observation of expressions, and subjective judgments are likely to be involved. As a result, it is difficult to quickly and appropriately provide effective support.

[0287] The specific processing by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0288] In this invention, the server includes a terminal that asks interactive questions to an employee, a means for receiving the employee's response data, a means for analyzing the response data and quantifying evaluation items, a means for extracting emotion information based on the quantified evaluation items, a means for generating a prompt based on the quantified evaluation items and the extracted emotion information, an image generation AI that generates an expression image using the prompt and the employee's face image, and a means for storing and displaying the expression images in time series. Thereby, it becomes possible to accurately grasp an employee's mental health state and provide effective support at an appropriate timing.

[0289] "Employee" refers to an individual who works in a company or organization.

[0290] "Interactive question" refers to a series of questions for obtaining information through interaction with a user.

[0291] "Terminal" refers to an electronic device for a user to input or view information.

[0292] "Response data" refers to information including the answers input by an employee to interactive questions.

[0293] A "server" refers to a computer system on a network that processes and manages data.

[0294] "Analysis" refers to the process of transforming data into a meaningful form and revealing its structure.

[0295] "Evaluation items" refer to checkpoints or criteria set up to understand the status of employees.

[0296] "Scoring" refers to the process of assigning numerical values ​​to evaluation items.

[0297] "Emotional information" refers to data that indicates the emotional state of employees.

[0298] A "prompt" refers to instructions or hints that a system uses to prompt the user to take the next action.

[0299] "Image generation AI" refers to software that uses artificial intelligence technology to generate images based on specific conditions.

[0300] "Facial expression images" refer to images that visually represent the emotional state of employees.

[0301] "Storing data chronologically" refers to recording data in order, following the flow of time.

[0302] A "dashboard" refers to a visual interface that displays diverse data in a centralized manner, making it easier to understand the situation.

[0303] This invention relates to a system for regularly monitoring the mental health status of employees and providing appropriate support. This system acquires response data through dialogue with employees, analyzes and scores it, extracts emotional information, and generates prompts. Furthermore, it uses image generation AI to generate images of employees' facial expressions, which are then stored and displayed chronologically.

[0304] Specific names of the hardware and software to be used and methods of data processing and data calculation

[0305] The terminal is used for generating interactions with employees. The terminal periodically displays interactive AI-based questions to employees. Specific examples of questions include "How is your recent workload?" etc. Through this, the answers of employees are obtained.

[0306] The server receives the answer data and face images of employees. Furthermore, it uses a natural language processing (NLP) engine to analyze the answer data and extracts keywords indicating negative emotions (e.g., "terrible", "stress", "tired", etc.).

[0307] The emotion engine extracts the emotional information of employees based on the extracted keywords and context. In this process, emotional information such as "high stress" is extracted based on the phrase "very terrible".

[0308] The evaluation engine scores evaluation items (e.g., "stress level", "happiness level", etc.) based on the extracted emotional information. For example, for an answer like "It's very terrible and stress is building up", the stress level is evaluated as 8 out of 10.

[0309] The prompt generation module generates prompts based on the scored evaluation items and emotional information. For example, when the stress level is 8 and there is emotional information of "high stress", a prompt of "high stress" is set.

[0310] The image generation AI takes the generated prompt and the face image of the employee as inputs and generates an expression image. This image reflects the current mental state of the employee. As a specific example, based on the prompt of "high stress", a face image with a sad expression is generated.

[0311] The database stores the generated expression images in chronological order. Thereby, expression images from the past to the present are accumulated.

[0312] The dashboard displays saved facial images and emotional information. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time.

[0313] Specific example

[0314] 1. Interview Session

[0315] The terminal displays the employee with the question, "How is the progress on your recent projects?"

[0316] The user (employee) responded, "Things are progressing smoothly, but I'm feeling a little stressed."

[0317] 2. Analysis of response data and extraction of sentiment information

[0318] The device sends this response data and facial image to the server.

[0319] The server analyzes this response data using natural language processing and extracts keywords related to stress.

[0320] The server extracts emotional information from keywords extracted using an emotion engine. For example, from the response "I feel a little stressed," it extracts the emotional information "moderate stress."

[0321] 3. Scoring of evaluation items

[0322] The server evaluates the stress level based on emotional information and evaluation criteria, and assigns a score, for example, 5 points.

[0323] 4. Prompt generation

[0324] The server generates a "moderate stress" prompt based on a stress level of 5 and emotional information indicating "moderate stress."

[0325] 5. Generation of facial expression images

[0326] The server inputs the prompt "moderate stress" and an image of the employee's face into the image generation AI, which then generates an image of the employee's facial expression that reflects their stress level.

[0327] The server receives facial expression images generated by the image generation AI.

[0328] 6. Saving and displaying time-series data

[0329] The server saves the generated facial expression images to a database and stores them along with past data.

[0330] The server displays saved facial images on a dashboard, allowing HR personnel to monitor employees' mental health status over time.

[0331] In this way, the system of the present invention, by combining a dialogue generation AI, an image generation AI, and an emotion engine, can intuitively grasp the mental health status of employees and quickly provide appropriate support.

[0332] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0333] Step 1:

[0334] Generating dialogue with employees

[0335] The server periodically generates questions for employees using conversational AI. The questions are randomly selected from a pre-configured list of questions.

[0336] The device periodically displays questions sent from the server as pop-ups. This display continues until the user responds.

[0337] Input: Interactive questions selected by the server

[0338] Output: Question displayed on the terminal

[0339] Step 2:

[0340] Employee response input

[0341] The user (employee) enters their answers to the questions displayed on the terminal. For example, in response to "How has your workload been recently?", they might answer "It's been very difficult and stressful."

[0342] Input: Employee's response to a question displayed on the terminal.

[0343] Output: Response data entered into the terminal

[0344] Step 3:

[0345] Sending response data

[0346] The terminal captures employee response data and facial images, and sends this data to the server.

[0347] Input: Employee response data and facial image

[0348] Output: Response data and facial image sent to the server

[0349] Step 4:

[0350] Analysis of response data

[0351] The server analyzes the received response data using a natural language processing (NLP) engine. During this analysis, it extracts keywords that indicate negative emotions (e.g., "difficult," "stressful," "tired," etc.).

[0352] Input: Response data sent to the server

[0353] Output: Analyzed keywords

[0354] Step 5:

[0355] Extraction of emotional information

[0356] The server uses an emotion engine to extract employee emotional information based on extracted keywords and context. For example, based on the phrase "very difficult," it extracts the emotional information "high stress."

[0357] Input: Analyzed keywords

[0358] Output: Extracted emotional information

[0359] Step 6:

[0360] Scoring of evaluation items

[0361] The server uses the extracted emotional information to score evaluation items (e.g., "stress level," "happiness level," etc.). For example, based on the response, "It's very difficult and I'm feeling stressed," the stress level might be rated at 8 out of 10.

[0362] Input: Extracted emotional information

[0363] Output: Scored evaluation items

[0364] Step 7:

[0365] Prompt generation based on evaluation items and sentiment information

[0366] The server generates prompts based on scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the server will set the prompt to "high stress."

[0367] Input: Scored evaluation items and emotional information

[0368] Output: Generated prompt

[0369] Step 8:

[0370] Generation of facial expression images

[0371] The server inputs the generated prompt and the employee's facial image into the image generation AI, which then generates an expression image that reflects the facial image. For example, a prompt "high stress" combined with a facial image will generate a facial image with a sad expression.

[0372] Input: Prompt and facial image

[0373] Output: Generated facial image

[0374] Step 9:

[0375] Saving time-series data

[0376] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[0377] Input: Generated facial image

[0378] Output: Facial expression images stored in the database

[0379] Step 10:

[0380] Displaying saved data

[0381] The server displays stored facial images and emotional information in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can see that a particular employee has shown high stress levels over the past few weeks.

[0382] Input: Facial expression images and emotion information stored in the database

[0383] Output: Data displayed on the dashboard

[0384] (Application Example 2)

[0385] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0386] Managing the mental health of drivers and staff in food delivery companies is a crucial issue for improving employee performance and preventing employee turnover. However, current systems make it difficult to accurately and quickly grasp employees' stress and fatigue levels. Furthermore, there is a lack of efficient and effective systems for visualizing individual employees' emotional states and continuously monitoring changes in them. Thus, there is a need for technological means to grasp employees' mental health status in a timely and accurate manner and provide appropriate support.

[0387] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for analyzing employee response data and scoring evaluation items, means for generating prompts based on evaluation items and emotional information, and for generating facial expression images using image generation AI, and means for saving and displaying the generated facial expression images in chronological order. This makes it possible to accurately and continuously monitor the mental health status of employees and provide prompt and appropriate support.

[0388] An "employee" is a person who engages in work within a company or organization.

[0389] "Interactive questioning" is a method of asking questions to employees in an interactive format and obtaining their answers.

[0390] A "terminal" refers to a device used by employees to answer interactive questions, such as a smartphone or tablet.

[0391] "Response data" refers to information resulting from employees' answers to interactive questions.

[0392] A "server" is a central system that receives response data and performs analysis processing.

[0393] "Means of analyzing and quantifying evaluation items" refers to systems or algorithms that analyze response data and quantify evaluation items based on the results.

[0394] A "prompt" is input information used to instruct an image generation AI to generate specific facial expressions or states.

[0395] "Facial images" refer to image data of employees' faces.

[0396] "Image generation AI" is artificial intelligence that generates new facial expression images based on prompts and facial images.

[0397] "Facial expression images" are facial images generated by image generation AI that reflect the emotional state of employees.

[0398] "Means of saving and displaying in chronological order" refers to a system that saves generated facial expression images as time progresses and displays them on a dashboard or similar.

[0399] "Natural language processing technology" is a technology that uses computers to analyze and understand human language.

[0400] "Emotional information" refers to information about the emotional state of employees extracted from the response data.

[0401] A "dashboard" is an interface that visually displays analysis results, evaluation items, and generated facial expression images.

[0402] This invention relates to a system for monitoring the mental health of drivers and staff in food delivery companies and providing appropriate support. The system asks employees interactive questions, analyzes the response data, extracts emotional information, scores evaluation items, and generates and manages facial expression images based on those scores.

[0403] System Configuration

[0404] This system consists of the following components:

[0405] 1. Terminal

[0406] These are devices such as smartphones and tablets used by employees (drivers). The devices periodically display questions from an interactive AI and retrieve the answers.

[0407] 2. Server

[0408] This is a cloud system that receives response data and facial images and processes them for analysis. It utilizes Amazon Web Services (AWS®) cloud services, specifically DynamoDB, S3, and Lambda.

[0409] The server analyzes the response data using a natural language processing (NLP) engine (such as HuggingFace's Transformer model) to extract emotional information.

[0410] The emotional information extracted by the emotion engine is scored according to evaluation criteria.

[0411] Prompts are generated based on scored evaluation items and emotional information.

[0412] Using a prompt and a facial image, an image generation AI (such as StyleGAN) generates an image of facial expressions.

[0413] The generated facial expression images are saved to AWS S3, and metadata is recorded in DynamoDB.

[0414] Saved facial expression images and evaluation criteria are displayed chronologically on the administrator dashboard (using React and Node.js).

[0415] Details of data processing

[0416] Analysis of employee response data

[0417] Employees answer questions from an interactive AI using an app on their devices. For example, in response to a question like, "Have you been feeling stressed about your recent delivery work?", they might answer, "I'm feeling very stressed."

[0418] The device sends this response data to the cloud server.

[0419] The server analyzes the received response data using natural language processing technology, and the emotion engine extracts emotional information such as "high stress."

[0420] Scoring of evaluation items and prompt generation

[0421] Based on emotional information, evaluation items (e.g., stress level) are scored. For example, if the emotional information is "high stress," the stress level might be rated at 8 out of 10.

[0422] Prompts are generated based on scored evaluation items and emotional information. For example, a prompt such as "high stress" is generated.

[0423] Generation and saving of facial expression images

[0424] The server inputs prompts and employee facial images into an image generation AI, which then generates facial expression images that reflect stress levels.

[0425] The server saves the generated facial expression images to AWS S3 and records metadata in DynamoDB.

[0426] For example, based on the "high stress" prompt, an image with a dark, tired expression is generated and saved as "driver123 / 2023-10-01 / image.png".

[0427] Dashboard display

[0428] Administrators can use the dashboard to monitor employees' mental health status over time. For example, they can identify if a particular employee's stress level has increased compared to past data and make decisions such as providing additional support.

[0429] Example of a prompt

[0430] Prompt: Generate a facial expression that reflects a high stress level.

[0431] Input text: "High stress"

[0432] In this way, the system of the present invention can accurately and continuously monitor the mental health of employees and provide prompt and appropriate support.

[0433] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0434] Step 1:

[0435] The user answers questions from the conversational AI using a smartphone application. For example, in response to the question, "Have you been feeling stressed about your recent delivery work?", the user might input, "I feel very stressed." This generates a text response as input data. The input data is, "I feel very stressed."

[0436] Step 2:

[0437] The device sends the user's text response data and facial image to the cloud server. At the same time, the device captures the facial image with its camera and uploads it to the cloud server along with the response data. The input data consists of the text response and the facial image file, and the output is the completion of the transmission to the cloud server.

[0438] Step 3:

[0439] The server inputs the received text response data into a natural language processing (NLP) engine to analyze the response content. The tool used is the HuggingFace transformer model. Emotional keywords and emotion labels are extracted from the analyzed data. The input data is text responses, and the output is extracted emotion information (e.g., "high stress").

[0440] Step 4:

[0441] The server scores evaluation items (e.g., stress level) based on emotional information extracted by the emotion engine. For example, a response like "I feel very stressed" would result in a stress level score of 8 out of 10. The input data is emotional information, and the output is the scored evaluation item (e.g., stress level 8 points).

[0442] Step 5:

[0443] The server generates prompts based on scored evaluation items and emotional information. For example, for a "stress level of 8 points" and emotional information of "high stress," it generates the prompt "high stress." The input data consists of scored evaluation items and emotional information, and the output is the generated prompt (e.g., "high stress").

[0444] Step 6:

[0445] The server inputs a prompt and a user's face image into an image generation AI, which then generates an expression image based on the prompt. The AI ​​model used is StyleGAN, among others. The input data consists of the prompt and the face image, while the output is an expression image reflecting stress.

[0446] Step 7:

[0447] The server saves the generated facial expression image to AWS S3 and records the metadata in AWS DynamoDB. The input data is the generated facial expression image, and the output is the completion of saving to cloud storage. For example, the saved image is stored as "driver123 / 2023-10-01 / image.png".

[0448] Step 8:

[0449] The server displays saved facial expression images and evaluation items in chronological order on an administrator dashboard. Administrators can check the mental health status of each employee through the dashboard. Input data consists of saved facial expression images and evaluation items, while output is a visualized dashboard display.

[0450] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0451] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search)<url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0452] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0453] [Second Embodiment]

[0454] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0455] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0456] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0457] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0458] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0459] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0460] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0461] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0462] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0463] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0464] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0465] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0466] This invention relates to a system for regularly monitoring the mental health status of employees. This system utilizes a dialogue generation AI to conduct conversations and score evaluation items, and then uses these scored evaluation items as prompts to input into an image generation AI, thereby visualizing changes in facial expressions over time.

[0467] System Overview

[0468] 1. Generating dialogue with employees

[0469] The terminal periodically displays questions to employees using conversational AI. For example, it might ask, "How has your workload been lately?" and retrieve the answers.

[0470] The user (employee) enters their answers to questions displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[0471] 2. Analysis of the dialogue content

[0472] The terminal sends employee response data to the server.

[0473] The server analyzes the received response data using natural language processing (NLP). For example, it evaluates the employee's emotions and stress level based on keywords and context in the response.

[0474] Based on the analyzed data, evaluation items (e.g., stress level, happiness level, etc.) are scored. For example, based on the response, "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[0475] 3. Prompting evaluation items

[0476] The server generates prompts for input to the image generation AI based on the scored evaluation items. For example, if the stress level is 8 points, the prompt will be set to "High Stress".

[0477] 4. Generation of facial expression images

[0478] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression based on the prompt (e.g., sad face, tired face, etc.).

[0479] The server receives the generated facial expression image.

[0480] 5. Saving and displaying time-series data

[0481] The server stores the generated facial expression images chronologically. In this way, it can accumulate facial expression images from the past to the present.

[0482] The server visualizes the saved facial expression images and displays them in the form of a dashboard. This dashboard allows HR personnel and managers to quickly grasp the mental health status of employees.

[0483] Specific example

[0484] 1. Interview Session

[0485] The terminal displays the employee with the question, "How is the progress on your recent projects?"

[0486] The user (employee) responded, "Things are progressing smoothly, but I'm feeling a little stressed."

[0487] 2. Analysis and scoring of response data

[0488] The server analyzes this response data using natural language processing and extracts keywords related to stress.

[0489] The server evaluates the stress level based on the extracted data and assigns a score, for example, 5 points.

[0490] 3. Prompt generation

[0491] The server generates a prompt indicating "moderate stress" for a stress level of 5.

[0492] 4. Generation of facial expression images

[0493] The server inputs the prompt "moderate stress" and an image of the employee's face into the image generation AI, which then generates an image of the employee's facial expression that reflects their stress level.

[0494] 5. Saving and displaying time-series data

[0495] The server saves the generated facial expression images to a database and stores them along with past data.

[0496] The server displays this data on a dashboard, allowing HR personnel to view employees' mental health status over time.

[0497] In this way, the system of the present invention, by combining dialogue generation AI and image generation AI, can intuitively grasp the mental health status of employees and quickly provide appropriate support.

[0498] The following describes the processing flow.

[0499] Step 1:

[0500] The device displays questions randomly selected from a pre-configured list to the employee. For example, it might display the question, "How has your workload been recently?"

[0501] Step 2:

[0502] The user (employee) enters their answer to the question displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[0503] Step 3:

[0504] The terminal sends the entered response data and the employee's facial image to the server.

[0505] Step 4:

[0506] The server analyzes the received response data using a natural language processing (NLP) engine. Here, it extracts keywords indicating negative emotions (e.g., "very difficult," "stressful," "tired," etc.).

[0507] Step 5:

[0508] The server scores evaluation items (e.g., stress level, happiness level) based on the frequency and context of extracted keywords. For example, for the response "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[0509] Step 6:

[0510] The server generates prompts for input to the image generation AI based on the scored evaluation items. For example, if the stress level is 8 points, the prompt will be set to "High Stress".

[0511] Step 7:

[0512] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression based on the prompt (e.g., sad face, tired face, etc.).

[0513] Step 8:

[0514] The server receives facial expression images generated by an image generation AI. These generated facial expression images reflect the employee's current mental state.

[0515] Step 9:

[0516] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[0517] Step 10:

[0518] The server displays the saved facial expression images in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can identify an employee's stress level as rising and consider providing additional support.

[0519] This series of processes allows the system to effectively understand the mental health status of employees and take appropriate measures.

[0520] (Example 1)

[0521] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0522] There is a challenge in continuously and efficiently monitoring employees' mental health and providing appropriate support. A decline in employee mental health can lead to a wide range of problems, including decreased productivity and increased turnover, thus necessitating effective monitoring methods.

[0523] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0524] In this invention, the server includes means for asking interactive questions to employees, means for receiving employee response data, means for analyzing the response data and scoring evaluation items, means for generating prompts based on the scored evaluation items, image generation means for generating facial expression images using the prompts and the employee's facial image, and means for saving and displaying the facial expression images in chronological order. This makes it possible to evaluate the mental health status of employees in real time and monitor its changes in chronological order.

[0525] A "terminal" refers to an electronic device that interacts with users (employees), presents questions, and allows them to input answers.

[0526] A "server" refers to a computer system that receives, analyzes, and stores data sent from a terminal.

[0527] "Response data" refers to data that includes the content of responses entered by users (employees) via their devices.

[0528] "Natural language processing technology" refers to techniques for analyzing text data to understand and interpret its meaning and emotions.

[0529] "Evaluation items" refer to indicators and criteria set based on information extracted from response data.

[0530] "Scoring" refers to the process of assigning numerical values ​​based on evaluation criteria.

[0531] A "prompt" refers to a command or instruction given to an image generation device.

[0532] "Image generation means" refers to a device or algorithm that generates a new facial expression image based on a prompt and a user's (employee's) facial image.

[0533] "Facial expression image" refers to an image showing the facial expression of an employee, generated by an image generation device.

[0534] "Storing in chronological order" refers to the process of saving generated images and data sequentially according to the passage of time.

[0535] "Means for saving and displaying" refers to a system or device for accumulating generated facial expression images and presenting them in a way that allows for visual confirmation.

[0536] A "dashboard" refers to an interface that visually displays the mental health status of employees.

[0537] This invention relates to a system for regularly monitoring the mental health status of employees. This system utilizes a dialogue generation AI to conduct conversations and score evaluation items, and then uses these scored evaluation items as prompts to input into an image generation AI, thereby visualizing changes in facial expressions over time.

[0538] System Configuration

[0539] terminal

[0540] These are devices such as PCs and smartphones used by employees, and they present questions through conversational AI. For example, OpenAI's GPT-3 can be used as the conversational AI.

[0541] server

[0542] This is a computer system that receives, analyzes, and stores response data sent from terminals. The server is equipped with software that utilizes natural language processing technologies (e.g., spaCy and NLTK) and image generation AI (e.g., DALL-E 2).

[0543] Natural Language Processing (NLP) Module

[0544] It is installed on a server, analyzes employee response data, extracts evaluation items, and assigns scores. For example, it can numerically evaluate stress and emotions included in employee responses.

[0545] Image generation AI

[0546] Based on scored evaluation items, prompts are generated, and by inputting these prompts along with the employee's facial image, a new facial expression image is created. For example, DALL-E 2 can be used here.

[0547] database

[0548] It is configured as part of the server and stores the generated facial expression images chronologically. Database systems such as MySQL or MongoDB are used.

[0549] Dashboard

[0550] This is an interface for HR personnel to visually check the mental health status of employees. It is built using HTML5 and JavaScript.

[0551] Specific operation of the system

[0552] 1. Generating dialogue with employees

[0553] The terminal periodically displays questions to employees via conversational AI. For example, it might generate and display a question such as, "How has your workload been lately?"

[0554] The user (employee) enters their answer to this question into the terminal. For example, they might enter, "It's very difficult and stressful."

[0555] 2. Analysis of the dialogue content

[0556] The device sends the user's response data to the server.

[0557] The server analyzes the received response data using natural language processing technology to assess stress and emotions. For example, it might assign a stress level of 8 out of 10 to the sentence "I'm feeling stressed."

[0558] 3. Prompt generation

[0559] The server generates prompts for input to the image generation AI based on the scored evaluation items. For example, if the stress level is 8 points, it will generate the prompt "high-stress facial expression".

[0560] 4. Generation of facial expression images

[0561] The server inputs the generated prompt text and the employee's facial image into the image generation AI. The image generation AI, for example, uses DALL-E 2 to generate images of high-stress facial expressions.

[0562] The server saves the generated facial expression images to a database.

[0563] 5. Saving and displaying time-series data

[0564] The server organizes the saved facial expression images chronologically and displays them on a dashboard. This allows HR personnel to intuitively understand employees' mental health status over time.

[0565] Specific example

[0566] 1. Interview Session

[0567] Terminal: "How is the project progressing recently?"

[0568] User: "Things are progressing smoothly, but I'm feeling a little stressed."

[0569] 2. Analysis and scoring of response data

[0570] The server analyzes this response data using natural language processing technology and extracts keywords related to stress.

[0571] Based on responses such as "I feel a little stressed," the stress level is rated on a scale of 1 to 5.

[0572] 3. Prompt generation

[0573] Server: Generates the prompt message "Moderate stress expression".

[0574] 4. Generation of facial expression images

[0575] The server inputs the prompt "moderately stressed facial expression" and an image of the employee's face into the image generation AI, which then generates an image of the facial expression that reflects the stress.

[0576] 5. Saving and displaying time-series data

[0577] The server saves the generated facial expression images to a database and displays them on the dashboard.

[0578] HR personnel can see at a glance how stress levels have changed over the past month and the generated facial expressions.

[0579] This system makes it possible to intuitively understand the mental health status of employees and provide prompt and accurate support.

[0580] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0581] Step 1:

[0582] The terminal periodically displays questions to employees via conversational AI. The input is a list of questions pre-configured in the system, and the output is the question text that is displayed. For example, it generates and displays the question, "How has your workload been recently?"

[0583] Step 2:

[0584] The user (employee) enters their answer to the question displayed on the terminal. The input is the user's answer, and the output is the answer data entered on the terminal. For example, the user might enter, "It's very difficult and stressful."

[0585] Step 3:

[0586] The terminal sends the user's entered response data to the server. The input is the user's response data, and the output is the response data sent to the server. This communication is conducted via the HTTPS protocol.

[0587] Step 4:

[0588] The server receives response data sent from the terminal. The input is the response data sent from the terminal, and the output is the response data stored on the server. For example, the data is received in JSON format.

[0589] Step 5:

[0590] The server analyzes the received response data using natural language processing (NLP) techniques. The input is the response data stored on the server, and the output is the analyzed evaluation items. For example, keywords and emotions are extracted using an NLP toolkit (e.g., spaCy or NLTK). Based on the analysis, the stress level is evaluated as 8 points from the response, "It's very difficult and stressful."

[0591] Step 6:

[0592] The server generates prompts based on scored evaluation items. The input is the scored evaluation item, and the output is the prompt text that is input to the image generation AI. For example, if the stress level is 8 points, the server generates the prompt "High stress expression".

[0593] Step 7:

[0594] The server inputs the generated prompt text and the employee's face image into the image generation AI. The input is the prompt text and the employee's face image, and the output is the generated facial expression image. For example, DALL-E 2 is used to generate a high-stress facial expression image.

[0595] Step 8:

[0596] The server receives facial expression images generated by an image generation AI. The input is facial expression image data received from the image generation AI, and the output is facial expression images stored on the server. For example, it decodes Base64 encoded image data.

[0597] Step 9:

[0598] The server stores the received facial expression images in a database in chronological order. The input is the generated facial expression images, and the output is the chronological data stored in the database. For example, it is stored in a database system (e.g., MySQL or MongoDB).

[0599] Step 10:

[0600] The server displays facial expression images stored in the database on a dashboard. Inputs are the facial expression images and evaluation items stored in the database, while output is the visual information displayed on the dashboard. For example, time-series graphs and timelines can be displayed using HTML5 or JavaScript. HR personnel can view this to quickly grasp the mental health status of employees.

[0601] (Application Example 1)

[0602] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0603] In modern factory environments, properly monitoring employees' mental health is crucial. However, conventional methods do not provide real-time visibility into employees' mental health, potentially leading to the accumulation of overwork and stress. This can result in decreased productivity and increased safety risks. This invention aims to provide a robot-based employee mental health monitoring system within a factory, enabling real-time and time-series tracking of employees' mental health.

[0604] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0605] In this invention, the server includes terminal means for asking interactive questions to employees, means for receiving employee response data, means for analyzing the response data and scoring evaluation items, means for generating prompts based on the scored evaluation items, image generation AI means for generating facial expression images using the prompts and the employee's facial image, means for saving and displaying the facial expression images in chronological order, in-factory robot means for periodically displaying questions to employees and collecting responses, means for generating facial expression images using the image generation AI based on the evaluation items, and means for saving the generated facial expression images in chronological order in a database. This makes it possible to grasp the mental health status of employees in the factory in real time and provide support at the appropriate time.

[0606] A "terminal" is a device used to ask employees interactive questions.

[0607] A "server" is a system that receives, analyzes, and stores employee response data.

[0608] "Response data" refers to the information that employees enter in response to interactive questions.

[0609] "Methods for quantifying evaluation items" refer to technical means for analyzing employee response data and expressing the results as scores.

[0610] A "prompt" is a phrase or command used to give instructions to an image generation AI.

[0611] "Image generation AI" is artificial intelligence that generates facial expression images based on prompts and images of employees' faces.

[0612] A "facial expression image" is an image representing an employee's facial expression, created by a generative AI model based on prompts.

[0613] "Means for saving and displaying in chronological order" refers to technical means for recording generated facial expression images in chronological order and displaying them as needed.

[0614] A "robot in a factory" is an automated device that displays interactive questions to employees in a factory environment and collects their responses.

[0615] "Natural language processing technology" refers to language processing techniques used to analyze employee response data.

[0616] A "dashboard" is a system that visually displays scored evaluation items, allowing for a quick overview of employees' mental health status.

[0617] A "database" is a system for storing and managing generated facial expression images and evaluation data.

[0618] The following system is configured as an embodiment of this invention. The system aims to monitor the mental health status of employees in a factory environment in real time and provide support at the appropriate time.

[0619] First, the terminal periodically displays interactive questions to the employee. This terminal can be used, for example, by a robot in the factory approaching an employee and asking, "How is the progress on your recent work?" The employee then enters their answer to the question.

[0620] The terminal receives employee response data and sends it to the server. This response data is analyzed on the server using natural language processing technology. Specifically, keywords and context are extracted from the responses using transformer models, and sentiment analysis is performed. As a result of this analysis, evaluation items (e.g., stress level) can be scored.

[0621] The server generates prompts based on scored evaluation items. For example, if the stress level is 8 points, it generates a prompt that says "High Stress." This prompt and the employee's facial image are then input into an image generation AI to generate an expression image. Based on this prompt, the image generation AI reflects the stress expression on the employee's facial image.

[0622] The generated facial expression images are stored chronologically in a database by the server. This stored data can later be visualized by HR personnel and managers through a dashboard, allowing them to quickly grasp the mental health status of employees.

[0623] Hardware and software to be used

[0624] Hardware: Factory robots, interactive terminals, servers, databases

[0625] Software: Natural language processing technology (transformer models), image generation AI, database management systems, dashboard display software

[0626] Examples of specific cases and prompt statements

[0627] For example, suppose an employee answers the question, "How is your recent work progressing?" with, "It's extremely difficult and stressful." Based on this answer, a prompt "High Stress" is generated. As a result, the image generation AI receives a prompt message like the following:

[0628] Example of a prompt:

[0629] Prompt: "High stress level, employee face image: face_image.jpg"

[0630] Based on this prompt, the image generation AI generates high-stress facial expression images corresponding to the employee's face image, and the server stores these in a database chronologically. HR personnel can then view the stored data on a dashboard, gaining a clear overview of employees' mental health status. In this way, the system enables real-time monitoring of employee mental health within the factory and the provision of appropriate support.

[0631] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0632] Step 1:

[0633] A terminal (a robot in the factory) periodically displays interactive questions to employees. For example, it might ask, "How is the progress of your recent work?" The user (employee) might respond, "It's very difficult and stressful." In this step, the terminal displays the question, and the user enters the answer. The input data is the user's response.

[0634] Step 2:

[0635] The terminal receives the user's response data and sends it to the server. The server stores the received data for analysis. In this step, the input is the user's response data, and the output is the data sent to the server.

[0636] Step 3:

[0637] The server analyzes the received response data using natural language processing techniques to evaluate emotions and stress levels. This process utilizes a transformer model. Specifically, it analyzes keywords and context from the responses to perform sentiment analysis. The input is the received response data, and the output is the evaluated items, such as the stress level.

[0638] Step 4:

[0639] The server assigns scores to the evaluation items. For example, based on the analyzed emotion response "It's very difficult and stressful," the stress level is scored as 8 out of 10. The input is the evaluation items obtained in the previous step, and the output is the scored stress level.

[0640] Step 5:

[0641] The server generates prompts based on the scored evaluation items. For example, if the stress level is 8, it will generate the prompt "High Stress." The input is the scored stress level, and the output is the generated prompt.

[0642] Step 6:

[0643] The server inputs the generated prompt and the employee's face image into the image generation AI, which then generates an expression image that reflects stress. Specifically, the image generation AI changes the facial expression based on the prompt and the face image. The input for this step is the generated prompt and the employee's face image, and the output is the generated expression image.

[0644] Step 7:

[0645] The server stores the generated facial expression images in a database in chronological order. The stored facial expression images are later used as data for HR personnel and managers to review through a dashboard. The input for this step is the generated facial expression images, and the output is the data stored in the database.

[0646] Step 8:

[0647] The server uses a dashboard to visualize stored facial expression images for HR personnel and managers. This allows them to quickly understand the mental health status of employees. The input is facial expression images stored in the database, and the output is visualized data displayed on the dashboard.

[0648] This clearly explains what data processing and calculations are performed at each step, and what inputs and outputs are generated. Specific operations include data analysis using natural language processing, prompt generation, use of image generation AI, saving to a database, and dashboard display. Based on this, a system for monitoring the user's mental health status in real time is realized.

[0649] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0650] This invention relates to a system for regularly monitoring the mental health status of employees and providing appropriate support. This system utilizes a combination of a dialogue generation AI, an image generation AI, and an emotion engine. The system acquires response data through dialogue with employees, analyzes it, and scores evaluation items. Furthermore, it extracts emotional information using the emotion engine and generates facial expression images by inputting prompts generated based on the scored evaluation items and emotional information into the image generation AI. The generated facial expression images are saved chronologically and displayed on a dashboard.

[0651] System Overview

[0652] 1. Generating dialogue with employees

[0653] The terminal periodically displays questions to employees using conversational AI. For example, it might ask, "How has your workload been lately?" and retrieve the answers.

[0654] The user (employee) enters their answer to the question displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[0655] 2. Analysis of the dialogue content

[0656] The terminal sends the entered response data and the employee's facial image to the server.

[0657] The server analyzes the received response data using a natural language processing (NLP) engine. Here, it extracts keywords indicating negative emotions (e.g., "very difficult," "stressful," "tired," etc.).

[0658] 3. Extraction of emotional information

[0659] The server uses an emotion engine to extract employee emotional information based on extracted keywords and context. For example, based on the phrase "very difficult" from the response, it extracts the emotional information "high stress."

[0660] 4. Scoring of evaluation items

[0661] The server uses the extracted emotional information to score evaluation items (e.g., stress level, happiness level). For example, in response to the answer, "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[0662] 5. Prompt generation based on evaluation items and sentiment information

[0663] The server generates prompts based on scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the server will set the prompt to "high stress."

[0664] 6. Generation of facial expression images

[0665] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression (e.g., sad face, tired face) based on the prompt.

[0666] The server receives facial expression images generated by an image generation AI. These generated facial expression images reflect the employee's current mental state.

[0667] 7. Saving and displaying time-series data

[0668] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[0669] The server displays stored facial images and emotional information in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can identify an employee's stress level as rising and consider providing additional support.

[0670] Specific example

[0671] 1. Interview Session

[0672] The terminal displays the employee with the question, "How is the progress on your recent projects?"

[0673] The user (employee) responded, "Things are progressing smoothly, but I'm feeling a little stressed."

[0674] 2. Analysis of response data and extraction of sentiment information

[0675] The device sends this response data and facial image to the server.

[0676] The server analyzes this response data using natural language processing and extracts keywords related to stress.

[0677] The server extracts emotional information from keywords extracted using an emotion engine. For example, from the response "I feel a little stressed," it extracts the emotional information "moderate stress."

[0678] 3. Scoring of evaluation items

[0679] The server evaluates the stress level based on emotional information and evaluation criteria, and assigns a score, for example, 5 points.

[0680] 4. Prompt generation

[0681] The server generates a "moderate stress" prompt based on a stress level of 5 and emotional information indicating "moderate stress."

[0682] 5. Generation of facial expression images

[0683] The server inputs the prompt "moderate stress" and an image of the employee's face into the image generation AI, which then generates an image of the employee's facial expression that reflects their stress level.

[0684] The server receives facial expression images generated by the image generation AI.

[0685] 6. Saving and displaying time-series data

[0686] The server saves the generated facial expression images to a database and stores them along with past data.

[0687] The server displays saved facial images on a dashboard, allowing HR personnel to monitor employees' mental health status over time.

[0688] In this way, the system of the present invention, by combining a dialogue generation AI, an image generation AI, and an emotion engine, can intuitively grasp the mental health status of employees and quickly provide appropriate support.

[0689] The following describes the processing flow.

[0690] Step 1:

[0691] The device displays questions randomly selected from a pre-configured list to the employee. For example, it might display the question, "How has your workload been recently?"

[0692] Step 2:

[0693] The user (employee) enters their answer to the question displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[0694] Step 3:

[0695] The terminal sends the entered response data and the employee's facial image to the server.

[0696] Step 4:

[0697] The server analyzes the received response data using a natural language processing (NLP) engine. Here, it extracts keywords indicating negative emotions (e.g., "very difficult," "stressful," "tired," etc.).

[0698] Step 5:

[0699] The server uses an emotion engine to extract employee emotional information based on extracted keywords and context. For example, based on the phrase "very difficult" from the response, it extracts the emotional information "high stress."

[0700] Step 6:

[0701] The server uses the extracted emotional information to score evaluation items (e.g., stress level, happiness level). For example, in response to the answer, "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[0702] Step 7:

[0703] The server generates prompts to input into the image generation AI based on the scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the server sets the prompt to "high stress."

[0704] Step 8:

[0705] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression based on the prompt (e.g., sad face, tired face, etc.).

[0706] Step 9:

[0707] The server receives facial expression images generated by an image generation AI. These generated facial expression images reflect the employee's current mental state.

[0708] Step 10:

[0709] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[0710] Step 11:

[0711] The server displays stored facial images and emotional information in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can identify an employee's stress level as rising and consider providing additional support.

[0712] This series of processes allows the system to effectively understand the mental health status of employees and take appropriate measures.

[0713] (Example 2)

[0714] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0715] Employee mental health significantly impacts productivity and workplace relationships. However, accurately understanding employees' mental state and providing timely support is not easy. Traditional methods often rely on simple questionnaires or observations of facial expressions, which are prone to subjective judgments. This makes it difficult to provide effective support quickly and appropriately.

[0716] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0717] In this invention, the server includes a terminal that asks interactive questions to employees, means for receiving employee response data, means for analyzing the response data and scoring evaluation items, means for extracting emotional information based on the scored evaluation items, means for generating prompts based on the scored evaluation items and extracted emotional information, an image generation AI that generates facial expression images using the prompts and the employee's facial image, and means for saving and displaying the facial expression images in chronological order. This makes it possible to accurately grasp the mental health status of employees and provide effective support at the appropriate time.

[0718] "Employee" refers to an individual who works for a company or organization.

[0719] "Interactive questioning" refers to a series of questions designed to obtain information through dialogue with the user.

[0720] A "terminal" refers to an electronic device used by users to input or view information.

[0721] "Response data" refers to information that includes the answers that employees entered in response to interactive questions.

[0722] A "server" refers to a computer system on a network that processes and manages data.

[0723] "Analysis" refers to the process of transforming data into a meaningful form and revealing its structure.

[0724] "Evaluation items" refer to checkpoints or criteria set up to understand the status of employees.

[0725] "Scoring" refers to the process of assigning numerical values ​​to evaluation items.

[0726] "Emotional information" refers to data that indicates the emotional state of employees.

[0727] A "prompt" refers to instructions or hints that a system uses to prompt the user to take the next action.

[0728] "Image generation AI" refers to software that uses artificial intelligence technology to generate images based on specific conditions.

[0729] "Facial expression images" refer to images that visually represent the emotional state of employees.

[0730] "Storing data chronologically" refers to recording data in order, following the flow of time.

[0731] A "dashboard" refers to a visual interface that displays diverse data in a centralized manner, making it easier to understand the situation.

[0732] This invention relates to a system for regularly monitoring the mental health status of employees and providing appropriate support. This system acquires response data through dialogue with employees, analyzes and scores it, extracts emotional information, and generates prompts. Furthermore, it uses image generation AI to generate images of employees' facial expressions, which are then stored and displayed chronologically.

[0733] Specific names of the hardware and software to be used, and the methods for data processing and calculation.

[0734] The terminal is used to generate conversations with employees. The terminal periodically displays questions from an interactive AI to employees. A specific example of a question is, "How has your workload been recently?" The system then obtains the employee's response.

[0735] The server receives employee response data and facial images. Furthermore, it uses a natural language processing (NLP) engine to analyze the response data and extract keywords indicating negative emotions (e.g., "difficult," "stressful," "tired," etc.).

[0736] The emotion engine extracts employee emotional information based on extracted keywords and context. In this process, it extracts emotional information such as "high stress" based on the phrase "very difficult."

[0737] The evaluation engine scores evaluation items (e.g., "stress level," "happiness level") based on the extracted emotional information. For example, in response to the answer, "It's very difficult and I'm feeling stressed," the stress level is rated at 8 out of 10.

[0738] The prompt generation module generates prompts based on scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the prompt will be set to "high stress."

[0739] The image generation AI takes a generated prompt and an employee's facial image as input to generate an expression image. This image reflects the employee's current mental state. For example, based on the prompt "high stress," an image of a sad-looking face is generated.

[0740] The database stores generated facial expression images chronologically. This allows for the accumulation of facial expression images from the past to the present.

[0741] The dashboard displays saved facial images and emotional information. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time.

[0742] Specific example

[0743] 1. Interview Session

[0744] The terminal displays the employee with the question, "How is the progress on your recent projects?"

[0745] The user (employee) responded, "Things are progressing smoothly, but I'm feeling a little stressed."

[0746] 2. Analysis of response data and extraction of sentiment information

[0747] The device sends this response data and facial image to the server.

[0748] The server analyzes this response data using natural language processing and extracts keywords related to stress.

[0749] The server extracts emotional information from keywords extracted using an emotion engine. For example, from the response "I feel a little stressed," it extracts the emotional information "moderate stress."

[0750] 3. Scoring of evaluation items

[0751] The server evaluates the stress level based on emotional information and evaluation criteria, and assigns a score, for example, 5 points.

[0752] 4. Prompt generation

[0753] The server generates a "moderate stress" prompt based on a stress level of 5 and emotional information indicating "moderate stress."

[0754] 5. Generation of facial expression images

[0755] The server inputs the prompt "moderate stress" and an image of the employee's face into the image generation AI, which then generates an image of the employee's facial expression that reflects their stress level.

[0756] The server receives facial expression images generated by the image generation AI.

[0757] 6. Saving and displaying time-series data

[0758] The server saves the generated facial expression images to a database and stores them along with past data.

[0759] The server displays saved facial images on a dashboard, allowing HR personnel to monitor employees' mental health status over time.

[0760] In this way, the system of the present invention, by combining a dialogue generation AI, an image generation AI, and an emotion engine, can intuitively grasp the mental health status of employees and quickly provide appropriate support.

[0761] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0762] Step 1:

[0763] Generating dialogue with employees

[0764] The server periodically generates questions for employees using conversational AI. The questions are randomly selected from a pre-configured list of questions.

[0765] The device periodically displays questions sent from the server as pop-ups. This display continues until the user responds.

[0766] Input: Interactive questions selected by the server

[0767] Output: Question displayed on the terminal

[0768] Step 2:

[0769] Employee response input

[0770] The user (employee) enters their answers to the questions displayed on the terminal. For example, in response to "How has your workload been recently?", they might answer "It's been very difficult and stressful."

[0771] Input: Employee's response to a question displayed on the terminal.

[0772] Output: Response data entered into the terminal

[0773] Step 3:

[0774] Sending response data

[0775] The terminal captures employee response data and facial images, and sends this data to the server.

[0776] Input: Employee response data and facial image

[0777] Output: Response data and facial image sent to the server

[0778] Step 4:

[0779] Analysis of response data

[0780] The server analyzes the received response data using a natural language processing (NLP) engine. During this analysis, it extracts keywords that indicate negative emotions (e.g., "difficult," "stressful," "tired," etc.).

[0781] Input: Response data sent to the server

[0782] Output: Analyzed keywords

[0783] Step 5:

[0784] Extraction of emotional information

[0785] The server uses an emotion engine to extract employee emotional information based on extracted keywords and context. For example, based on the phrase "very difficult," it extracts the emotional information "high stress."

[0786] Input: Analyzed keywords

[0787] Output: Extracted emotional information

[0788] Step 6:

[0789] Scoring of evaluation items

[0790] The server uses the extracted emotional information to score evaluation items (e.g., "stress level," "happiness level," etc.). For example, based on the response, "It's very difficult and I'm feeling stressed," the stress level might be rated at 8 out of 10.

[0791] Input: Extracted emotional information

[0792] Output: Scored evaluation items

[0793] Step 7:

[0794] Prompt generation based on evaluation items and sentiment information

[0795] The server generates prompts based on scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the server will set the prompt to "high stress."

[0796] Input: Scored evaluation items and emotional information

[0797] Output: Generated prompt

[0798] Step 8:

[0799] Generation of facial expression images

[0800] The server inputs the generated prompt and the employee's facial image into the image generation AI, which then generates an expression image that reflects the facial image. For example, a prompt "high stress" combined with a facial image will generate a facial image with a sad expression.

[0801] Input: Prompt and facial image

[0802] Output: Generated facial image

[0803] Step 9:

[0804] Saving time-series data

[0805] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[0806] Input: Generated facial image

[0807] Output: Facial expression images stored in the database

[0808] Step 10:

[0809] Displaying saved data

[0810] The server displays stored facial images and emotional information in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can see that a particular employee has shown high stress levels over the past few weeks.

[0811] Input: Facial expression images and emotion information stored in the database

[0812] Output: Data displayed on the dashboard

[0813] (Application Example 2)

[0814] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0815] Managing the mental health of drivers and staff in food delivery companies is a crucial issue for improving employee performance and preventing employee turnover. However, current systems make it difficult to accurately and quickly grasp employees' stress and fatigue levels. Furthermore, there is a lack of efficient and effective systems for visualizing individual employees' emotional states and continuously monitoring changes in them. Thus, there is a need for technological means to grasp employees' mental health status in a timely and accurate manner and provide appropriate support.

[0816] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for analyzing employee response data and scoring evaluation items, means for generating prompts based on evaluation items and emotional information, and for generating facial expression images using image generation AI, and means for saving and displaying the generated facial expression images in chronological order. This makes it possible to accurately and continuously monitor the mental health status of employees and provide prompt and appropriate support.

[0817] An "employee" is a person who engages in work within a company or organization.

[0818] "Interactive questioning" is a method of asking questions to employees in an interactive format and obtaining their answers.

[0819] A "terminal" refers to a device used by employees to answer interactive questions, such as a smartphone or tablet.

[0820] "Response data" refers to information resulting from employees' answers to interactive questions.

[0821] A "server" is a central system that receives response data and performs analysis processing.

[0822] "Means of analyzing and quantifying evaluation items" refers to systems or algorithms that analyze response data and quantify evaluation items based on the results.

[0823] A "prompt" is input information used to instruct an image generation AI to generate specific facial expressions or states.

[0824] "Facial images" refer to image data of employees' faces.

[0825] "Image generation AI" is artificial intelligence that generates new facial expression images based on prompts and facial images.

[0826] "Facial expression images" are facial images generated by image generation AI that reflect the emotional state of employees.

[0827] "Means of saving and displaying in chronological order" refers to a system that saves generated facial expression images as time progresses and displays them on a dashboard or similar.

[0828] "Natural language processing technology" is a technology that uses computers to analyze and understand human language.

[0829] "Emotional information" refers to information about the emotional state of employees extracted from the response data.

[0830] A "dashboard" is an interface that visually displays analysis results, evaluation items, and generated facial expression images.

[0831] This invention relates to a system for monitoring the mental health of drivers and staff in food delivery companies and providing appropriate support. The system asks employees interactive questions, analyzes the response data, extracts emotional information, scores evaluation items, and generates and manages facial expression images based on those scores.

[0832] System Configuration

[0833] This system consists of the following components:

[0834] 1. Terminal

[0835] These are devices such as smartphones and tablets used by employees (drivers). The devices periodically display questions from an interactive AI and retrieve the answers.

[0836] 2. Server

[0837] This is a cloud system that receives response data and facial images and processes them for analysis. It utilizes Amazon Web Services (AWS) cloud services, specifically DynamoDB, S3, and Lambda.

[0838] The server analyzes the response data using a natural language processing (NLP) engine (such as HuggingFace's Transformer model) to extract emotional information.

[0839] The emotional information extracted by the emotion engine is scored according to evaluation criteria.

[0840] Prompts are generated based on scored evaluation items and emotional information.

[0841] Using a prompt and a facial image, an image generation AI (such as StyleGAN) generates an image of facial expressions.

[0842] The generated facial expression images are saved to AWS S3, and metadata is recorded in DynamoDB.

[0843] Saved facial expression images and evaluation criteria are displayed chronologically on the administrator dashboard (using React and Node.js).

[0844] Details of data processing

[0845] Analysis of employee response data

[0846] Employees answer questions from an interactive AI using an app on their devices. For example, in response to a question like, "Have you been feeling stressed about your recent delivery work?", they might answer, "I'm feeling very stressed."

[0847] The device sends this response data to the cloud server.

[0848] The server analyzes the received response data using natural language processing technology, and the emotion engine extracts emotional information such as "high stress."

[0849] Scoring of evaluation items and prompt generation

[0850] Based on emotional information, evaluation items (e.g., stress level) are scored. For example, if the emotional information is "high stress," the stress level might be rated at 8 out of 10.

[0851] Prompts are generated based on scored evaluation items and emotional information. For example, a prompt such as "high stress" is generated.

[0852] Generation and saving of facial expression images

[0853] The server inputs prompts and employee facial images into an image generation AI, which then generates facial expression images that reflect stress levels.

[0854] The server saves the generated facial expression images to AWS S3 and records metadata in DynamoDB.

[0855] For example, based on the "high stress" prompt, an image with a dark, tired expression is generated and saved as "driver123 / 2023-10-01 / image.png".

[0856] Dashboard display

[0857] Administrators can use the dashboard to monitor employees' mental health status over time. For example, they can identify if a particular employee's stress level has increased compared to past data and make decisions such as providing additional support.

[0858] Example of a prompt

[0859] Prompt: Generate a facial expression that reflects a high stress level.

[0860] Input text: "High stress"

[0861] In this way, the system of the present invention can accurately and continuously monitor the mental health of employees and provide prompt and appropriate support.

[0862] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0863] Step 1:

[0864] The user answers questions from the conversational AI using a smartphone application. For example, in response to the question, "Have you been feeling stressed about your recent delivery work?", the user might input, "I feel very stressed." This generates a text response as input data. The input data is, "I feel very stressed."

[0865] Step 2:

[0866] The device sends the user's text response data and facial image to the cloud server. At the same time, the device captures the facial image with its camera and uploads it to the cloud server along with the response data. The input data consists of the text response and the facial image file, and the output is the completion of the transmission to the cloud server.

[0867] Step 3:

[0868] The server inputs the received text response data into a natural language processing (NLP) engine to analyze the response content. The tool used is the HuggingFace transformer model. Emotional keywords and emotion labels are extracted from the analyzed data. The input data is text responses, and the output is extracted emotion information (e.g., "high stress").

[0869] Step 4:

[0870] The server scores evaluation items (e.g., stress level) based on emotional information extracted by the emotion engine. For example, a response like "I feel very stressed" would result in a stress level score of 8 out of 10. The input data is emotional information, and the output is the scored evaluation item (e.g., stress level 8 points).

[0871] Step 5:

[0872] The server generates prompts based on scored evaluation items and emotional information. For example, for a "stress level of 8 points" and emotional information of "high stress," it generates the prompt "high stress." The input data consists of scored evaluation items and emotional information, and the output is the generated prompt (e.g., "high stress").

[0873] Step 6:

[0874] The server inputs a prompt and a user's face image into an image generation AI, which then generates an expression image based on the prompt. The AI ​​model used is StyleGAN, among others. The input data consists of the prompt and the face image, while the output is an expression image reflecting stress.

[0875] Step 7:

[0876] The server saves the generated facial expression image to AWS S3 and records the metadata in AWS DynamoDB. The input data is the generated facial expression image, and the output is the completion of saving to cloud storage. For example, the saved image is stored as "driver123 / 2023-10-01 / image.png".

[0877] Step 8:

[0878] The server displays saved facial expression images and evaluation items in chronological order on an administrator dashboard. Administrators can check the mental health status of each employee through the dashboard. Input data consists of saved facial expression images and evaluation items, while output is a visualized dashboard display.

[0879] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0880] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0881] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0882] [Third Embodiment]

[0883] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0884] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0885] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0886] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0887] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0888] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0889] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0890] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0891] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0892] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0893] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0894] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0895] This invention relates to a system for regularly monitoring the mental health status of employees. This system utilizes a dialogue generation AI to conduct conversations and score evaluation items, and then uses these scored evaluation items as prompts to input into an image generation AI, thereby visualizing changes in facial expressions over time.

[0896] System Overview

[0897] 1. Generating dialogue with employees

[0898] The terminal periodically displays questions to employees using conversational AI. For example, it might ask, "How has your workload been lately?" and retrieve the answers.

[0899] The user (employee) enters their answers to questions displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[0900] 2. Analysis of the dialogue content

[0901] The terminal sends employee response data to the server.

[0902] The server analyzes the received response data using natural language processing (NLP). For example, it evaluates the employee's emotions and stress level based on keywords and context in the response.

[0903] Based on the analyzed data, evaluation items (e.g., stress level, happiness level, etc.) are scored. For example, based on the response, "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[0904] 3. Prompting evaluation items

[0905] The server generates prompts for input to the image generation AI based on the scored evaluation items. For example, if the stress level is 8 points, the prompt will be set to "High Stress".

[0906] 4. Generation of facial expression images

[0907] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression based on the prompt (e.g., sad face, tired face, etc.).

[0908] The server receives the generated facial expression image.

[0909] 5. Saving and displaying time-series data

[0910] The server stores the generated facial expression images chronologically. In this way, it can accumulate facial expression images from the past to the present.

[0911] The server visualizes the saved facial expression images and displays them in the form of a dashboard. This dashboard allows HR personnel and managers to quickly grasp the mental health status of employees.

[0912] Specific example

[0913] 1. Interview Session

[0914] The terminal displays the employee with the question, "How is the progress on your recent projects?"

[0915] The user (employee) responded, "Things are progressing smoothly, but I'm feeling a little stressed."

[0916] 2. Analysis and scoring of response data

[0917] The server analyzes this response data using natural language processing and extracts keywords related to stress.

[0918] The server evaluates the stress level based on the extracted data and assigns a score, for example, 5 points.

[0919] 3. Prompt generation

[0920] The server generates a prompt indicating "moderate stress" for a stress level of 5.

[0921] 4. Generation of facial expression images

[0922] The server inputs the prompt "moderate stress" and an image of the employee's face into the image generation AI, which then generates an image of the employee's facial expression that reflects their stress level.

[0923] 5. Saving and displaying time-series data

[0924] The server saves the generated facial expression images to a database and stores them along with past data.

[0925] The server displays this data on a dashboard, allowing HR personnel to view employees' mental health status over time.

[0926] In this way, the system of the present invention, by combining dialogue generation AI and image generation AI, can intuitively grasp the mental health status of employees and quickly provide appropriate support.

[0927] The following describes the processing flow.

[0928] Step 1:

[0929] The device displays questions randomly selected from a pre-configured list to the employee. For example, it might display the question, "How has your workload been recently?"

[0930] Step 2:

[0931] The user (employee) enters their answer to the question displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[0932] Step 3:

[0933] The terminal sends the entered response data and the employee's facial image to the server.

[0934] Step 4:

[0935] The server analyzes the received response data using a natural language processing (NLP) engine. Here, it extracts keywords indicating negative emotions (e.g., "very difficult," "stressful," "tired," etc.).

[0936] Step 5:

[0937] The server scores evaluation items (e.g., stress level, happiness level) based on the frequency and context of extracted keywords. For example, for the response "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[0938] Step 6:

[0939] The server generates prompts for input to the image generation AI based on the scored evaluation items. For example, if the stress level is 8 points, the prompt will be set to "High Stress".

[0940] Step 7:

[0941] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression based on the prompt (e.g., sad face, tired face, etc.).

[0942] Step 8:

[0943] The server receives facial expression images generated by an image generation AI. These generated facial expression images reflect the employee's current mental state.

[0944] Step 9:

[0945] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[0946] Step 10:

[0947] The server displays the saved facial expression images in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can identify an employee's stress level as rising and consider providing additional support.

[0948] This series of processes allows the system to effectively understand the mental health status of employees and take appropriate measures.

[0949] (Example 1)

[0950] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0951] There is a challenge in continuously and efficiently monitoring employees' mental health and providing appropriate support. A decline in employee mental health can lead to a wide range of problems, including decreased productivity and increased turnover, thus necessitating effective monitoring methods.

[0952] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0953] In this invention, the server includes means for asking interactive questions to employees, means for receiving employee response data, means for analyzing the response data and scoring evaluation items, means for generating prompts based on the scored evaluation items, image generation means for generating facial expression images using the prompts and the employee's facial image, and means for saving and displaying the facial expression images in chronological order. This makes it possible to evaluate the mental health status of employees in real time and monitor its changes in chronological order.

[0954] A "terminal" refers to an electronic device that interacts with users (employees), presents questions, and allows them to input answers.

[0955] A "server" refers to a computer system that receives, analyzes, and stores data sent from a terminal.

[0956] "Response data" refers to data that includes the content of responses entered by users (employees) via their devices.

[0957] "Natural language processing technology" refers to techniques for analyzing text data to understand and interpret its meaning and emotions.

[0958] "Evaluation items" refer to indicators and criteria set based on information extracted from response data.

[0959] "Scoring" refers to the process of assigning numerical values ​​based on evaluation criteria.

[0960] A "prompt" refers to a command or instruction given to an image generation device.

[0961] "Image generation means" refers to a device or algorithm that generates a new facial expression image based on a prompt and a user's (employee's) facial image.

[0962] "Facial expression image" refers to an image showing the facial expression of an employee, generated by an image generation device.

[0963] "Storing in chronological order" refers to the process of saving generated images and data sequentially according to the passage of time.

[0964] "Means for saving and displaying" refers to a system or device for accumulating generated facial expression images and presenting them in a way that allows for visual confirmation.

[0965] A "dashboard" refers to an interface that visually displays the mental health status of employees.

[0966] This invention relates to a system for regularly monitoring the mental health status of employees. This system utilizes a dialogue generation AI to conduct conversations and score evaluation items, and then uses these scored evaluation items as prompts to input into an image generation AI, thereby visualizing changes in facial expressions over time.

[0967] System Configuration

[0968] terminal

[0969] These are devices such as PCs and smartphones used by employees, and they present questions through conversational AI. For example, OpenAI's GPT-3 can be used as the conversational AI.

[0970] server

[0971] This is a computer system that receives, analyzes, and stores response data sent from terminals. The server is equipped with software that utilizes natural language processing technologies (e.g., spaCy and NLTK) and image generation AI (e.g., DALL-E 2).

[0972] Natural Language Processing (NLP) Module

[0973] It is installed on a server, analyzes employee response data, extracts evaluation items, and assigns scores. For example, it can numerically evaluate stress and emotions included in employee responses.

[0974] Image generation AI

[0975] Based on scored evaluation items, prompts are generated, and by inputting these prompts along with the employee's facial image, a new facial expression image is created. For example, DALL-E 2 can be used here.

[0976] database

[0977] It is configured as part of the server and stores the generated facial expression images chronologically. Database systems such as MySQL or MongoDB are used.

[0978] Dashboard

[0979] This is an interface for HR personnel to visually check the mental health status of employees. It is built using HTML5 and JavaScript.

[0980] Specific operation of the system

[0981] 1. Generating dialogue with employees

[0982] The terminal periodically displays questions to employees via conversational AI. For example, it might generate and display a question such as, "How has your workload been lately?"

[0983] The user (employee) enters their answer to this question into the terminal. For example, they might enter, "It's very difficult and stressful."

[0984] 2. Analysis of the dialogue content

[0985] The device sends the user's response data to the server.

[0986] The server analyzes the received response data using natural language processing technology to assess stress and emotions. For example, it might assign a stress level of 8 out of 10 to the sentence "I'm feeling stressed."

[0987] 3. Prompt generation

[0988] The server generates prompts for input to the image generation AI based on the scored evaluation items. For example, if the stress level is 8 points, it will generate the prompt "high-stress facial expression".

[0989] 4. Generation of facial expression images

[0990] The server inputs the generated prompt text and the employee's facial image into the image generation AI. The image generation AI, for example, uses DALL-E 2 to generate images of high-stress facial expressions.

[0991] The server saves the generated facial expression images to a database.

[0992] 5. Saving and displaying time-series data

[0993] The server organizes the saved facial expression images chronologically and displays them on a dashboard. This allows HR personnel to intuitively understand employees' mental health status over time.

[0994] Specific example

[0995] 1. Interview Session

[0996] Terminal: "How is the project progressing recently?"

[0997] User: "Things are progressing smoothly, but I'm feeling a little stressed."

[0998] 2. Analysis and scoring of response data

[0999] The server analyzes this response data using natural language processing technology and extracts keywords related to stress.

[1000] Based on responses such as "I feel a little stressed," the stress level is rated on a scale of 1 to 5.

[1001] 3. Prompt generation

[1002] Server: Generates the prompt message "Moderate stress expression".

[1003] 4. Generation of facial expression images

[1004] The server inputs the prompt "moderately stressed facial expression" and an image of the employee's face into the image generation AI, which then generates an image of the facial expression that reflects the stress.

[1005] 5. Saving and displaying time-series data

[1006] The server saves the generated facial expression images to a database and displays them on the dashboard.

[1007] HR personnel can see at a glance how stress levels have changed over the past month and the generated facial expressions.

[1008] This system makes it possible to intuitively understand the mental health status of employees and provide prompt and accurate support.

[1009] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1010] Step 1:

[1011] The terminal periodically displays questions to employees via conversational AI. The input is a list of questions pre-configured in the system, and the output is the question text that is displayed. For example, it generates and displays the question, "How has your workload been recently?"

[1012] Step 2:

[1013] The user (employee) enters their answer to the question displayed on the terminal. The input is the user's answer, and the output is the answer data entered on the terminal. For example, the user might enter, "It's very difficult and stressful."

[1014] Step 3:

[1015] The terminal sends the user's entered response data to the server. The input is the user's response data, and the output is the response data sent to the server. This communication is conducted via the HTTPS protocol.

[1016] Step 4:

[1017] The server receives response data sent from the terminal. The input is the response data sent from the terminal, and the output is the response data stored on the server. For example, the data is received in JSON format.

[1018] Step 5:

[1019] The server analyzes the received response data using natural language processing (NLP) techniques. The input is the response data stored on the server, and the output is the analyzed evaluation items. For example, keywords and emotions are extracted using an NLP toolkit (e.g., spaCy or NLTK). Based on the analysis, the stress level is evaluated as 8 points from the response, "It's very difficult and stressful."

[1020] Step 6:

[1021] The server generates prompts based on scored evaluation items. The input is the scored evaluation item, and the output is the prompt text that is input to the image generation AI. For example, if the stress level is 8 points, the server generates the prompt "High stress expression".

[1022] Step 7:

[1023] The server inputs the generated prompt text and the employee's face image into the image generation AI. The input is the prompt text and the employee's face image, and the output is the generated facial expression image. For example, DALL-E 2 is used to generate a high-stress facial expression image.

[1024] Step 8:

[1025] The server receives facial expression images generated by an image generation AI. The input is facial expression image data received from the image generation AI, and the output is facial expression images stored on the server. For example, it decodes Base64 encoded image data.

[1026] Step 9:

[1027] The server stores the received facial expression images in a database in chronological order. The input is the generated facial expression images, and the output is the chronological data stored in the database. For example, it is stored in a database system (e.g., MySQL or MongoDB).

[1028] Step 10:

[1029] The server displays facial expression images stored in the database on a dashboard. Inputs are the facial expression images and evaluation items stored in the database, while output is the visual information displayed on the dashboard. For example, time-series graphs and timelines can be displayed using HTML5 or JavaScript. HR personnel can view this to quickly grasp the mental health status of employees.

[1030] (Application Example 1)

[1031] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1032] In modern factory environments, properly monitoring employees' mental health is crucial. However, conventional methods do not provide real-time visibility into employees' mental health, potentially leading to the accumulation of overwork and stress. This can result in decreased productivity and increased safety risks. This invention aims to provide a robot-based employee mental health monitoring system within a factory, enabling real-time and time-series tracking of employees' mental health.

[1033] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1034] In this invention, the server includes terminal means for asking interactive questions to employees, means for receiving employee response data, means for analyzing the response data and scoring evaluation items, means for generating prompts based on the scored evaluation items, image generation AI means for generating facial expression images using the prompts and the employee's facial image, means for saving and displaying the facial expression images in chronological order, in-factory robot means for periodically displaying questions to employees and collecting responses, means for generating facial expression images using the image generation AI based on the evaluation items, and means for saving the generated facial expression images in chronological order in a database. This makes it possible to grasp the mental health status of employees in the factory in real time and provide support at the appropriate time.

[1035] A "terminal" is a device used to ask employees interactive questions.

[1036] A "server" is a system that receives, analyzes, and stores employee response data.

[1037] "Response data" refers to the information that employees enter in response to interactive questions.

[1038] "Methods for quantifying evaluation items" refer to technical means for analyzing employee response data and expressing the results as scores.

[1039] A "prompt" is a phrase or command used to give instructions to an image generation AI.

[1040] "Image generation AI" is artificial intelligence that generates facial expression images based on prompts and images of employees' faces.

[1041] A "facial expression image" is an image representing an employee's facial expression, created by a generative AI model based on prompts.

[1042] "Means for saving and displaying in chronological order" refers to technical means for recording generated facial expression images in chronological order and displaying them as needed.

[1043] A "robot in a factory" is an automated device that displays interactive questions to employees in a factory environment and collects their responses.

[1044] "Natural language processing technology" refers to language processing techniques used to analyze employee response data.

[1045] A "dashboard" is a system that visually displays scored evaluation items, allowing for a quick overview of employees' mental health status.

[1046] A "database" is a system for storing and managing generated facial expression images and evaluation data.

[1047] The following system is configured as an embodiment of this invention. The system aims to monitor the mental health status of employees in a factory environment in real time and provide support at the appropriate time.

[1048] First, the terminal periodically displays interactive questions to the employee. This terminal can be used, for example, by a robot in the factory approaching an employee and asking, "How is the progress on your recent work?" The employee then enters their answer to the question.

[1049] The terminal receives employee response data and sends it to the server. This response data is analyzed on the server using natural language processing technology. Specifically, keywords and context are extracted from the responses using transformer models, and sentiment analysis is performed. As a result of this analysis, evaluation items (e.g., stress level) can be scored.

[1050] The server generates prompts based on scored evaluation items. For example, if the stress level is 8 points, it generates a prompt that says "High Stress." This prompt and the employee's facial image are then input into an image generation AI to generate an expression image. Based on this prompt, the image generation AI reflects the stress expression on the employee's facial image.

[1051] The generated facial expression images are stored chronologically in a database by the server. This stored data can later be visualized by HR personnel and managers through a dashboard, allowing them to quickly grasp the mental health status of employees.

[1052] Hardware and software to be used

[1053] Hardware: Factory robots, interactive terminals, servers, databases

[1054] Software: Natural language processing technology (transformer models), image generation AI, database management systems, dashboard display software

[1055] Examples of specific cases and prompt statements

[1056] For example, suppose an employee answers the question, "How is your recent work progressing?" with, "It's extremely difficult and stressful." Based on this answer, a prompt "High Stress" is generated. As a result, the image generation AI receives a prompt message like the following:

[1057] Example of a prompt:

[1058] Prompt: "High stress level, employee face image: face_image.jpg"

[1059] Based on this prompt, the image generation AI generates high-stress facial expression images corresponding to the employee's face image, and the server stores these in a database chronologically. HR personnel can then view the stored data on a dashboard, gaining a clear overview of employees' mental health status. In this way, the system enables real-time monitoring of employee mental health within the factory and the provision of appropriate support.

[1060] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1061] Step 1:

[1062] A terminal (a robot in the factory) periodically displays interactive questions to employees. For example, it might ask, "How is the progress of your recent work?" The user (employee) might respond, "It's very difficult and stressful." In this step, the terminal displays the question, and the user enters the answer. The input data is the user's response.

[1063] Step 2:

[1064] The terminal receives the user's response data and sends it to the server. The server stores the received data for analysis. In this step, the input is the user's response data, and the output is the data sent to the server.

[1065] Step 3:

[1066] The server analyzes the received response data using natural language processing techniques to evaluate emotions and stress levels. This process utilizes a transformer model. Specifically, it analyzes keywords and context from the responses to perform sentiment analysis. The input is the received response data, and the output is the evaluated items, such as the stress level.

[1067] Step 4:

[1068] The server assigns scores to the evaluation items. For example, based on the analyzed emotion response "It's very difficult and stressful," the stress level is scored as 8 out of 10. The input is the evaluation items obtained in the previous step, and the output is the scored stress level.

[1069] Step 5:

[1070] The server generates prompts based on the scored evaluation items. For example, if the stress level is 8, it will generate the prompt "High Stress." The input is the scored stress level, and the output is the generated prompt.

[1071] Step 6:

[1072] The server inputs the generated prompt and the employee's face image into the image generation AI, which then generates an expression image that reflects stress. Specifically, the image generation AI changes the facial expression based on the prompt and the face image. The input for this step is the generated prompt and the employee's face image, and the output is the generated expression image.

[1073] Step 7:

[1074] The server stores the generated facial expression images in a database in chronological order. The stored facial expression images are later used as data for HR personnel and managers to review through a dashboard. The input for this step is the generated facial expression images, and the output is the data stored in the database.

[1075] Step 8:

[1076] The server uses a dashboard to visualize stored facial expression images for HR personnel and managers. This allows them to quickly understand the mental health status of employees. The input is facial expression images stored in the database, and the output is visualized data displayed on the dashboard.

[1077] This clearly explains what data processing and calculations are performed at each step, and what inputs and outputs are generated. Specific operations include data analysis using natural language processing, prompt generation, use of image generation AI, saving to a database, and dashboard display. Based on this, a system for monitoring the user's mental health status in real time is realized.

[1078] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1079] This invention relates to a system for regularly monitoring the mental health status of employees and providing appropriate support. This system utilizes a combination of a dialogue generation AI, an image generation AI, and an emotion engine. The system acquires response data through dialogue with employees, analyzes it, and scores evaluation items. Furthermore, it extracts emotional information using the emotion engine and generates facial expression images by inputting prompts generated based on the scored evaluation items and emotional information into the image generation AI. The generated facial expression images are saved chronologically and displayed on a dashboard.

[1080] System Overview

[1081] 1. Generating dialogue with employees

[1082] The terminal periodically displays questions to employees using conversational AI. For example, it might ask, "How has your workload been lately?" and retrieve the answers.

[1083] The user (employee) enters their answer to the question displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[1084] 2. Analysis of the dialogue content

[1085] The terminal sends the entered response data and the employee's facial image to the server.

[1086] The server analyzes the received response data using a natural language processing (NLP) engine. Here, it extracts keywords indicating negative emotions (e.g., "very difficult," "stressful," "tired," etc.).

[1087] 3. Extraction of emotional information

[1088] The server uses an emotion engine to extract employee emotional information based on extracted keywords and context. For example, based on the phrase "very difficult" from the response, it extracts the emotional information "high stress."

[1089] 4. Scoring of evaluation items

[1090] The server uses the extracted emotional information to score evaluation items (e.g., stress level, happiness level). For example, in response to the answer, "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[1091] 5. Prompt generation based on evaluation items and sentiment information

[1092] The server generates prompts based on scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the server will set the prompt to "high stress."

[1093] 6. Generation of facial expression images

[1094] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression (e.g., sad face, tired face) based on the prompt.

[1095] The server receives facial expression images generated by an image generation AI. These generated facial expression images reflect the employee's current mental state.

[1096] 7. Saving and displaying time-series data

[1097] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[1098] The server displays stored facial images and emotional information in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can identify an employee's stress level as rising and consider providing additional support.

[1099] Specific example

[1100] 1. Interview Session

[1101] The terminal displays the employee with the question, "How is the progress on your recent projects?"

[1102] The user (employee) responded, "Things are progressing smoothly, but I'm feeling a little stressed."

[1103] 2. Analysis of response data and extraction of sentiment information

[1104] The device sends this response data and facial image to the server.

[1105] The server analyzes this response data using natural language processing and extracts keywords related to stress.

[1106] The server extracts emotional information from keywords extracted using an emotion engine. For example, from the response "I feel a little stressed," it extracts the emotional information "moderate stress."

[1107] 3. Scoring of evaluation items

[1108] The server evaluates the stress level based on emotional information and evaluation criteria, and assigns a score, for example, 5 points.

[1109] 4. Prompt generation

[1110] The server generates a "moderate stress" prompt based on a stress level of 5 and emotional information indicating "moderate stress."

[1111] 5. Generation of facial expression images

[1112] The server inputs the prompt "moderate stress" and an image of the employee's face into the image generation AI, which then generates an image of the employee's facial expression that reflects their stress level.

[1113] The server receives facial expression images generated by the image generation AI.

[1114] 6. Saving and displaying time-series data

[1115] The server saves the generated facial expression images to a database and stores them along with past data.

[1116] The server displays saved facial images on a dashboard, allowing HR personnel to monitor employees' mental health status over time.

[1117] In this way, the system of the present invention, by combining a dialogue generation AI, an image generation AI, and an emotion engine, can intuitively grasp the mental health status of employees and quickly provide appropriate support.

[1118] The following describes the processing flow.

[1119] Step 1:

[1120] The device displays questions randomly selected from a pre-configured list to the employee. For example, it might display the question, "How has your workload been recently?"

[1121] Step 2:

[1122] The user (employee) enters their answer to the question displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[1123] Step 3:

[1124] The terminal sends the entered response data and the employee's facial image to the server.

[1125] Step 4:

[1126] The server analyzes the received response data using a natural language processing (NLP) engine. Here, it extracts keywords indicating negative emotions (e.g., "very difficult," "stressful," "tired," etc.).

[1127] Step 5:

[1128] The server uses an emotion engine to extract employee emotional information based on extracted keywords and context. For example, based on the phrase "very difficult" from the response, it extracts the emotional information "high stress."

[1129] Step 6:

[1130] The server uses the extracted emotional information to score evaluation items (e.g., stress level, happiness level). For example, in response to the answer, "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[1131] Step 7:

[1132] The server generates prompts to input into the image generation AI based on the scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the server sets the prompt to "high stress."

[1133] Step 8:

[1134] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression based on the prompt (e.g., sad face, tired face, etc.).

[1135] Step 9:

[1136] The server receives facial expression images generated by an image generation AI. These generated facial expression images reflect the employee's current mental state.

[1137] Step 10:

[1138] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[1139] Step 11:

[1140] The server displays stored facial images and emotional information in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can identify an employee's stress level as rising and consider providing additional support.

[1141] This series of processes allows the system to effectively understand the mental health status of employees and take appropriate measures.

[1142] (Example 2)

[1143] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1144] Employee mental health significantly impacts productivity and workplace relationships. However, accurately understanding employees' mental state and providing timely support is not easy. Traditional methods often rely on simple questionnaires or observations of facial expressions, which are prone to subjective judgments. This makes it difficult to provide effective support quickly and appropriately.

[1145] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1146] In this invention, the server includes a terminal that asks interactive questions to employees, means for receiving employee response data, means for analyzing the response data and scoring evaluation items, means for extracting emotional information based on the scored evaluation items, means for generating prompts based on the scored evaluation items and extracted emotional information, an image generation AI that generates facial expression images using the prompts and the employee's facial image, and means for saving and displaying the facial expression images in chronological order. This makes it possible to accurately grasp the mental health status of employees and provide effective support at the appropriate time.

[1147] "Employee" refers to an individual who works for a company or organization.

[1148] "Interactive questioning" refers to a series of questions designed to obtain information through dialogue with the user.

[1149] A "terminal" refers to an electronic device used by users to input or view information.

[1150] "Response data" refers to information that includes the answers that employees entered in response to interactive questions.

[1151] A "server" refers to a computer system on a network that processes and manages data.

[1152] "Analysis" refers to the process of transforming data into a meaningful form and revealing its structure.

[1153] "Evaluation items" refer to checkpoints or criteria set up to understand the status of employees.

[1154] "Scoring" refers to the process of assigning numerical values ​​to evaluation items.

[1155] "Emotional information" refers to data that indicates the emotional state of employees.

[1156] A "prompt" refers to instructions or hints that a system uses to prompt the user to take the next action.

[1157] "Image generation AI" refers to software that uses artificial intelligence technology to generate images based on specific conditions.

[1158] "Facial expression images" refer to images that visually represent the emotional state of employees.

[1159] "Storing data chronologically" refers to recording data in order, following the flow of time.

[1160] A "dashboard" refers to a visual interface that displays diverse data in a centralized manner, making it easier to understand the situation.

[1161] This invention relates to a system for regularly monitoring the mental health status of employees and providing appropriate support. This system acquires response data through dialogue with employees, analyzes and scores it, extracts emotional information, and generates prompts. Furthermore, it uses image generation AI to generate images of employees' facial expressions, which are then stored and displayed chronologically.

[1162] Specific names of the hardware and software to be used, and the methods for data processing and calculation.

[1163] The terminal is used to generate conversations with employees. The terminal periodically displays questions from an interactive AI to employees. A specific example of a question is, "How has your workload been recently?" The system then obtains the employee's response.

[1164] The server receives employee response data and facial images. Furthermore, it uses a natural language processing (NLP) engine to analyze the response data and extract keywords indicating negative emotions (e.g., "difficult," "stressful," "tired," etc.).

[1165] The emotion engine extracts employee emotional information based on extracted keywords and context. In this process, it extracts emotional information such as "high stress" based on the phrase "very difficult."

[1166] The evaluation engine scores evaluation items (e.g., "stress level," "happiness level") based on the extracted emotional information. For example, in response to the answer, "It's very difficult and I'm feeling stressed," the stress level is rated at 8 out of 10.

[1167] The prompt generation module generates prompts based on scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the prompt will be set to "high stress."

[1168] The image generation AI takes a generated prompt and an employee's facial image as input to generate an expression image. This image reflects the employee's current mental state. For example, based on the prompt "high stress," an image of a sad-looking face is generated.

[1169] The database stores generated facial expression images chronologically. This allows for the accumulation of facial expression images from the past to the present.

[1170] The dashboard displays saved facial images and emotional information. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time.

[1171] Specific example

[1172] 1. Interview Session

[1173] The terminal displays the employee with the question, "How is the progress on your recent projects?"

[1174] The user (employee) responded, "Things are progressing smoothly, but I'm feeling a little stressed."

[1175] 2. Analysis of response data and extraction of sentiment information

[1176] The device sends this response data and facial image to the server.

[1177] The server analyzes this response data using natural language processing and extracts keywords related to stress.

[1178] The server extracts emotional information from keywords extracted using an emotion engine. For example, from the response "I feel a little stressed," it extracts the emotional information "moderate stress."

[1179] 3. Scoring of evaluation items

[1180] The server evaluates the stress level based on emotional information and evaluation criteria, and assigns a score, for example, 5 points.

[1181] 4. Prompt generation

[1182] The server generates a "moderate stress" prompt based on a stress level of 5 and emotional information indicating "moderate stress."

[1183] 5. Generation of facial expression images

[1184] The server inputs the prompt "moderate stress" and an image of the employee's face into the image generation AI, which then generates an image of the employee's facial expression that reflects their stress level.

[1185] The server receives facial expression images generated by the image generation AI.

[1186] 6. Saving and displaying time-series data

[1187] The server saves the generated facial expression images to a database and stores them along with past data.

[1188] The server displays saved facial images on a dashboard, allowing HR personnel to monitor employees' mental health status over time.

[1189] In this way, the system of the present invention, by combining a dialogue generation AI, an image generation AI, and an emotion engine, can intuitively grasp the mental health status of employees and quickly provide appropriate support.

[1190] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1191] Step 1:

[1192] Generating dialogue with employees

[1193] The server periodically generates questions for employees using conversational AI. The questions are randomly selected from a pre-configured list of questions.

[1194] The device periodically displays questions sent from the server as pop-ups. This display continues until the user responds.

[1195] Input: Interactive questions selected by the server

[1196] Output: Question displayed on the terminal

[1197] Step 2:

[1198] Employee response input

[1199] The user (employee) enters their answers to the questions displayed on the terminal. For example, in response to "How has your workload been recently?", they might answer "It's been very difficult and stressful."

[1200] Input: Employee's response to a question displayed on the terminal.

[1201] Output: Response data entered into the terminal

[1202] Step 3:

[1203] Sending response data

[1204] The terminal captures employee response data and facial images, and sends this data to the server.

[1205] Input: Employee response data and facial image

[1206] Output: Response data and facial image sent to the server

[1207] Step 4:

[1208] Analysis of response data

[1209] The server analyzes the received response data using a natural language processing (NLP) engine. During this analysis, it extracts keywords that indicate negative emotions (e.g., "difficult," "stressful," "tired," etc.).

[1210] Input: Response data sent to the server

[1211] Output: Analyzed keywords

[1212] Step 5:

[1213] Extraction of emotional information

[1214] The server uses an emotion engine to extract employee emotional information based on extracted keywords and context. For example, based on the phrase "very difficult," it extracts the emotional information "high stress."

[1215] Input: Analyzed keywords

[1216] Output: Extracted emotional information

[1217] Step 6:

[1218] Scoring of evaluation items

[1219] The server uses the extracted emotional information to score evaluation items (e.g., "stress level," "happiness level," etc.). For example, based on the response, "It's very difficult and I'm feeling stressed," the stress level might be rated at 8 out of 10.

[1220] Input: Extracted emotional information

[1221] Output: Scored evaluation items

[1222] Step 7:

[1223] Prompt generation based on evaluation items and sentiment information

[1224] The server generates prompts based on scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the server will set the prompt to "high stress."

[1225] Input: Scored evaluation items and emotional information

[1226] Output: Generated prompt

[1227] Step 8:

[1228] Generation of facial expression images

[1229] The server inputs the generated prompt and the employee's facial image into the image generation AI, which then generates an expression image that reflects the facial image. For example, a prompt "high stress" combined with a facial image will generate a facial image with a sad expression.

[1230] Input: Prompt and facial image

[1231] Output: Generated facial image

[1232] Step 9:

[1233] Saving time-series data

[1234] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[1235] Input: Generated facial image

[1236] Output: Facial expression images stored in the database

[1237] Step 10:

[1238] Displaying saved data

[1239] The server displays stored facial images and emotional information in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can see that a particular employee has shown high stress levels over the past few weeks.

[1240] Input: Facial expression images and emotion information stored in the database

[1241] Output: Data displayed on the dashboard

[1242] (Application Example 2)

[1243] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1244] Managing the mental health of drivers and staff in food delivery companies is a crucial issue for improving employee performance and preventing employee turnover. However, current systems make it difficult to accurately and quickly grasp employees' stress and fatigue levels. Furthermore, there is a lack of efficient and effective systems for visualizing individual employees' emotional states and continuously monitoring changes in them. Thus, there is a need for technological means to grasp employees' mental health status in a timely and accurate manner and provide appropriate support.

[1245] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for analyzing employee response data and scoring evaluation items, means for generating prompts based on evaluation items and emotional information, and for generating facial expression images using image generation AI, and means for saving and displaying the generated facial expression images in chronological order. This makes it possible to accurately and continuously monitor the mental health status of employees and provide prompt and appropriate support.

[1246] An "employee" is a person who engages in work within a company or organization.

[1247] "Interactive questioning" is a method of asking questions to employees in an interactive format and obtaining their answers.

[1248] A "terminal" refers to a device used by employees to answer interactive questions, such as a smartphone or tablet.

[1249] "Response data" refers to information resulting from employees' answers to interactive questions.

[1250] A "server" is a central system that receives response data and performs analysis processing.

[1251] "Means of analyzing and quantifying evaluation items" refers to systems or algorithms that analyze response data and quantify evaluation items based on the results.

[1252] A "prompt" is input information used to instruct an image generation AI to generate specific facial expressions or states.

[1253] "Facial images" refer to image data of employees' faces.

[1254] "Image generation AI" is artificial intelligence that generates new facial expression images based on prompts and facial images.

[1255] "Facial expression images" are facial images generated by image generation AI that reflect the emotional state of employees.

[1256] "Means of saving and displaying in chronological order" refers to a system that saves generated facial expression images as time progresses and displays them on a dashboard or similar.

[1257] "Natural language processing technology" is a technology that uses computers to analyze and understand human language.

[1258] "Emotional information" refers to information about the emotional state of employees extracted from the response data.

[1259] A "dashboard" is an interface that visually displays analysis results, evaluation items, and generated facial expression images.

[1260] This invention relates to a system for monitoring the mental health of drivers and staff in food delivery companies and providing appropriate support. The system asks employees interactive questions, analyzes the response data, extracts emotional information, scores evaluation items, and generates and manages facial expression images based on those scores.

[1261] System Configuration

[1262] This system consists of the following components:

[1263] 1. Terminal

[1264] These are devices such as smartphones and tablets used by employees (drivers). The devices periodically display questions from an interactive AI and retrieve the answers.

[1265] 2. Server

[1266] This is a cloud system that receives response data and facial images and processes them for analysis. It utilizes Amazon Web Services (AWS) cloud services, specifically DynamoDB, S3, and Lambda.

[1267] The server analyzes the response data using a natural language processing (NLP) engine (such as HuggingFace's Transformer model) to extract emotional information.

[1268] The emotional information extracted by the emotion engine is scored according to evaluation criteria.

[1269] Prompts are generated based on scored evaluation items and emotional information.

[1270] Using a prompt and a facial image, an image generation AI (such as StyleGAN) generates an image of facial expressions.

[1271] The generated facial expression images are saved to AWS S3, and metadata is recorded in DynamoDB.

[1272] Saved facial expression images and evaluation criteria are displayed chronologically on the administrator dashboard (using React and Node.js).

[1273] Details of data processing

[1274] Analysis of employee response data

[1275] Employees answer questions from an interactive AI using an app on their devices. For example, in response to a question like, "Have you been feeling stressed about your recent delivery work?", they might answer, "I'm feeling very stressed."

[1276] The device sends this response data to the cloud server.

[1277] The server analyzes the received response data using natural language processing technology, and the emotion engine extracts emotional information such as "high stress."

[1278] Scoring of evaluation items and prompt generation

[1279] Based on emotional information, evaluation items (e.g., stress level) are scored. For example, if the emotional information is "high stress," the stress level might be rated at 8 out of 10.

[1280] Prompts are generated based on scored evaluation items and emotional information. For example, a prompt such as "high stress" is generated.

[1281] Generation and saving of facial expression images

[1282] The server inputs prompts and employee facial images into an image generation AI, which then generates facial expression images that reflect stress levels.

[1283] The server saves the generated facial expression images to AWS S3 and records metadata in DynamoDB.

[1284] For example, based on the "high stress" prompt, an image with a dark, tired expression is generated and saved as "driver123 / 2023-10-01 / image.png".

[1285] Dashboard display

[1286] Administrators can use the dashboard to monitor employees' mental health status over time. For example, they can identify if a particular employee's stress level has increased compared to past data and make decisions such as providing additional support.

[1287] Example of a prompt

[1288] Prompt: Generate a facial expression that reflects a high stress level.

[1289] Input text: "High stress"

[1290] In this way, the system of the present invention can accurately and continuously monitor the mental health of employees and provide prompt and appropriate support.

[1291] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1292] Step 1:

[1293] The user answers questions from the conversational AI using a smartphone application. For example, in response to the question, "Have you been feeling stressed about your recent delivery work?", the user might input, "I feel very stressed." This generates a text response as input data. The input data is, "I feel very stressed."

[1294] Step 2:

[1295] The device sends the user's text response data and facial image to the cloud server. At the same time, the device captures the facial image with its camera and uploads it to the cloud server along with the response data. The input data consists of the text response and the facial image file, and the output is the completion of the transmission to the cloud server.

[1296] Step 3:

[1297] The server inputs the received text response data into a natural language processing (NLP) engine to analyze the response content. The tool used is the HuggingFace transformer model. Emotional keywords and emotion labels are extracted from the analyzed data. The input data is text responses, and the output is extracted emotion information (e.g., "high stress").

[1298] Step 4:

[1299] The server scores evaluation items (e.g., stress level) based on emotional information extracted by the emotion engine. For example, a response like "I feel very stressed" would result in a stress level score of 8 out of 10. The input data is emotional information, and the output is the scored evaluation item (e.g., stress level 8 points).

[1300] Step 5:

[1301] The server generates prompts based on scored evaluation items and emotional information. For example, for a "stress level of 8 points" and emotional information of "high stress," it generates the prompt "high stress." The input data consists of scored evaluation items and emotional information, and the output is the generated prompt (e.g., "high stress").

[1302] Step 6:

[1303] The server inputs a prompt and a user's face image into an image generation AI, which then generates an expression image based on the prompt. The AI ​​model used is StyleGAN, among others. The input data consists of the prompt and the face image, while the output is an expression image reflecting stress.

[1304] Step 7:

[1305] The server saves the generated facial expression image to AWS S3 and records the metadata in AWS DynamoDB. The input data is the generated facial expression image, and the output is the completion of saving to cloud storage. For example, the saved image is stored as "driver123 / 2023-10-01 / image.png".

[1306] Step 8:

[1307] The server displays saved facial expression images and evaluation items in chronological order on an administrator dashboard. Administrators can check the mental health status of each employee through the dashboard. Input data consists of saved facial expression images and evaluation items, while output is a visualized dashboard display.

[1308] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1309] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1310] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[1311] [Fourth Embodiment]

[1312] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1313] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1314] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1315] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[1316] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1317] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1318] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1319] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[1320] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1321] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1322] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1323] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1324] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1325] This invention relates to a system for regularly monitoring the mental health status of employees. This system utilizes a dialogue generation AI to conduct conversations and score evaluation items, and then uses these scored evaluation items as prompts to input into an image generation AI, thereby visualizing changes in facial expressions over time.

[1326] System Overview

[1327] 1. Generating dialogue with employees

[1328] The terminal periodically displays questions to employees using conversational AI. For example, it might ask, "How has your workload been lately?" and retrieve the answers.

[1329] The user (employee) enters their answers to questions displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[1330] 2. Analysis of the dialogue content

[1331] The terminal sends employee response data to the server.

[1332] The server analyzes the received response data using natural language processing (NLP). For example, it evaluates the employee's emotions and stress level based on keywords and context in the response.

[1333] Based on the analyzed data, evaluation items (e.g., stress level, happiness level, etc.) are scored. For example, based on the response, "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[1334] 3. Prompting evaluation items

[1335] The server generates prompts for input to the image generation AI based on the scored evaluation items. For example, if the stress level is 8 points, the prompt will be set to "High Stress".

[1336] 4. Generation of facial expression images

[1337] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression based on the prompt (e.g., sad face, tired face, etc.).

[1338] The server receives the generated facial expression image.

[1339] 5. Saving and displaying time-series data

[1340] The server stores the generated facial expression images chronologically. In this way, it can accumulate facial expression images from the past to the present.

[1341] The server visualizes the saved facial expression images and displays them in the form of a dashboard. This dashboard allows HR personnel and managers to quickly grasp the mental health status of employees.

[1342] Specific example

[1343] 1. Interview Session

[1344] The terminal displays the employee with the question, "How is the progress on your recent projects?"

[1345] The user (employee) responded, "Things are progressing smoothly, but I'm feeling a little stressed."

[1346] 2. Analysis and scoring of response data

[1347] The server analyzes this response data using natural language processing and extracts keywords related to stress.

[1348] The server evaluates the stress level based on the extracted data and assigns a score, for example, 5 points.

[1349] 3. Prompt generation

[1350] The server generates a prompt indicating "moderate stress" for a stress level of 5.

[1351] 4. Generation of facial expression images

[1352] The server inputs the prompt "moderate stress" and an image of the employee's face into the image generation AI, which then generates an image of the employee's facial expression that reflects their stress level.

[1353] 5. Saving and displaying time-series data

[1354] The server saves the generated facial expression images to a database and stores them along with past data.

[1355] The server displays this data on a dashboard, allowing HR personnel to view employees' mental health status over time.

[1356] In this way, the system of the present invention, by combining dialogue generation AI and image generation AI, can intuitively grasp the mental health status of employees and quickly provide appropriate support.

[1357] The following describes the processing flow.

[1358] Step 1:

[1359] The device displays questions randomly selected from a pre-configured list to the employee. For example, it might display the question, "How has your workload been recently?"

[1360] Step 2:

[1361] The user (employee) enters their answer to the question displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[1362] Step 3:

[1363] The terminal sends the entered response data and the employee's facial image to the server.

[1364] Step 4:

[1365] The server analyzes the received response data using a natural language processing (NLP) engine. Here, it extracts keywords indicating negative emotions (e.g., "very difficult," "stressful," "tired," etc.).

[1366] Step 5:

[1367] The server scores evaluation items (e.g., stress level, happiness level) based on the frequency and context of extracted keywords. For example, for the response "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[1368] Step 6:

[1369] The server generates prompts for input to the image generation AI based on the scored evaluation items. For example, if the stress level is 8 points, the prompt will be set to "High Stress".

[1370] Step 7:

[1371] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression based on the prompt (e.g., sad face, tired face, etc.).

[1372] Step 8:

[1373] The server receives facial expression images generated by an image generation AI. These generated facial expression images reflect the employee's current mental state.

[1374] Step 9:

[1375] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[1376] Step 10:

[1377] The server displays the saved facial expression images in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can identify an employee's stress level as rising and consider providing additional support.

[1378] This series of processes allows the system to effectively understand the mental health status of employees and take appropriate measures.

[1379] (Example 1)

[1380] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1381] There is a challenge in continuously and efficiently monitoring employees' mental health and providing appropriate support. A decline in employee mental health can lead to a wide range of problems, including decreased productivity and increased turnover, thus necessitating effective monitoring methods.

[1382] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1383] In this invention, the server includes means for asking interactive questions to employees, means for receiving employee response data, means for analyzing the response data and scoring evaluation items, means for generating prompts based on the scored evaluation items, image generation means for generating facial expression images using the prompts and the employee's facial image, and means for saving and displaying the facial expression images in chronological order. This makes it possible to evaluate the mental health status of employees in real time and monitor its changes in chronological order.

[1384] A "terminal" refers to an electronic device that interacts with users (employees), presents questions, and allows them to input answers.

[1385] A "server" refers to a computer system that receives, analyzes, and stores data sent from a terminal.

[1386] "Response data" refers to data that includes the content of responses entered by users (employees) via their devices.

[1387] "Natural language processing technology" refers to techniques for analyzing text data to understand and interpret its meaning and emotions.

[1388] "Evaluation items" refer to indicators and criteria set based on information extracted from response data.

[1389] "Scoring" refers to the process of assigning numerical values ​​based on evaluation criteria.

[1390] A "prompt" refers to a command or instruction given to an image generation device.

[1391] "Image generation means" refers to a device or algorithm that generates a new facial expression image based on a prompt and a user's (employee's) facial image.

[1392] "Facial expression image" refers to an image showing the facial expression of an employee, generated by an image generation device.

[1393] "Storing in chronological order" refers to the process of saving generated images and data sequentially according to the passage of time.

[1394] "Means for saving and displaying" refers to a system or device for accumulating generated facial expression images and presenting them in a way that allows for visual confirmation.

[1395] A "dashboard" refers to an interface that visually displays the mental health status of employees.

[1396] This invention relates to a system for regularly monitoring the mental health status of employees. This system utilizes a dialogue generation AI to conduct conversations and score evaluation items, and then uses these scored evaluation items as prompts to input into an image generation AI, thereby visualizing changes in facial expressions over time.

[1397] System Configuration

[1398] terminal

[1399] These are devices such as PCs and smartphones used by employees, and they present questions through conversational AI. For example, OpenAI's GPT-3 can be used as the conversational AI.

[1400] server

[1401] This is a computer system that receives, analyzes, and stores response data sent from terminals. The server is equipped with software that utilizes natural language processing technologies (e.g., spaCy and NLTK) and image generation AI (e.g., DALL-E 2).

[1402] Natural Language Processing (NLP) Module

[1403] It is installed on a server, analyzes employee response data, extracts evaluation items, and assigns scores. For example, it can numerically evaluate stress and emotions included in employee responses.

[1404] Image generation AI

[1405] Based on scored evaluation items, prompts are generated, and by inputting these prompts along with the employee's facial image, a new facial expression image is created. For example, DALL-E 2 can be used here.

[1406] database

[1407] It is configured as part of the server and stores the generated facial expression images chronologically. Database systems such as MySQL or MongoDB are used.

[1408] Dashboard

[1409] This is an interface for HR personnel to visually check the mental health status of employees. It is built using HTML5 and JavaScript.

[1410] Specific operation of the system

[1411] 1. Generating dialogue with employees

[1412] The terminal periodically displays questions to employees via conversational AI. For example, it might generate and display a question such as, "How has your workload been lately?"

[1413] The user (employee) enters their answer to this question into the terminal. For example, they might enter, "It's very difficult and stressful."

[1414] 2. Analysis of the dialogue content

[1415] The device sends the user's response data to the server.

[1416] The server analyzes the received response data using natural language processing technology to assess stress and emotions. For example, it might assign a stress level of 8 out of 10 to the sentence "I'm feeling stressed."

[1417] 3. Prompt generation

[1418] The server generates prompts for input to the image generation AI based on the scored evaluation items. For example, if the stress level is 8 points, it will generate the prompt "high-stress facial expression".

[1419] 4. Generation of facial expression images

[1420] The server inputs the generated prompt text and the employee's facial image into the image generation AI. The image generation AI, for example, uses DALL-E 2 to generate images of high-stress facial expressions.

[1421] The server saves the generated facial expression images to a database.

[1422] 5. Saving and displaying time-series data

[1423] The server organizes the saved facial expression images chronologically and displays them on a dashboard. This allows HR personnel to intuitively understand employees' mental health status over time.

[1424] Specific example

[1425] 1. Interview Session

[1426] Terminal: "How is the project progressing recently?"

[1427] User: "Things are progressing smoothly, but I'm feeling a little stressed."

[1428] 2. Analysis and scoring of response data

[1429] The server analyzes this response data using natural language processing technology and extracts keywords related to stress.

[1430] Based on responses such as "I feel a little stressed," the stress level is rated on a scale of 1 to 5.

[1431] 3. Prompt generation

[1432] Server: Generates the prompt message "Moderate stress expression".

[1433] 4. Generation of facial expression images

[1434] The server inputs the prompt "moderately stressed facial expression" and an image of the employee's face into the image generation AI, which then generates an image of the facial expression that reflects the stress.

[1435] 5. Saving and displaying time-series data

[1436] The server saves the generated facial expression images to a database and displays them on the dashboard.

[1437] HR personnel can see at a glance how stress levels have changed over the past month and the generated facial expressions.

[1438] This system makes it possible to intuitively understand the mental health status of employees and provide prompt and accurate support.

[1439] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1440] Step 1:

[1441] The terminal periodically displays questions to employees via conversational AI. The input is a list of questions pre-configured in the system, and the output is the question text that is displayed. For example, it generates and displays the question, "How has your workload been recently?"

[1442] Step 2:

[1443] The user (employee) enters their answer to the question displayed on the terminal. The input is the user's answer, and the output is the answer data entered on the terminal. For example, the user might enter, "It's very difficult and stressful."

[1444] Step 3:

[1445] The terminal sends the user's entered response data to the server. The input is the user's response data, and the output is the response data sent to the server. This communication is conducted via the HTTPS protocol.

[1446] Step 4:

[1447] The server receives response data sent from the terminal. The input is the response data sent from the terminal, and the output is the response data stored on the server. For example, the data is received in JSON format.

[1448] Step 5:

[1449] The server analyzes the received response data using natural language processing (NLP) techniques. The input is the response data stored on the server, and the output is the analyzed evaluation items. For example, keywords and emotions are extracted using an NLP toolkit (e.g., spaCy or NLTK). Based on the analysis, the stress level is evaluated as 8 points from the response, "It's very difficult and stressful."

[1450] Step 6:

[1451] The server generates prompts based on scored evaluation items. The input is the scored evaluation item, and the output is the prompt text that is input to the image generation AI. For example, if the stress level is 8 points, the server generates the prompt "High stress expression".

[1452] Step 7:

[1453] The server inputs the generated prompt text and the employee's face image into the image generation AI. The input is the prompt text and the employee's face image, and the output is the generated facial expression image. For example, DALL-E 2 is used to generate a high-stress facial expression image.

[1454] Step 8:

[1455] The server receives facial expression images generated by an image generation AI. The input is facial expression image data received from the image generation AI, and the output is facial expression images stored on the server. For example, it decodes Base64 encoded image data.

[1456] Step 9:

[1457] The server stores the received facial expression images in a database in chronological order. The input is the generated facial expression images, and the output is the chronological data stored in the database. For example, it is stored in a database system (e.g., MySQL or MongoDB).

[1458] Step 10:

[1459] The server displays facial expression images stored in the database on a dashboard. Inputs are the facial expression images and evaluation items stored in the database, while output is the visual information displayed on the dashboard. For example, time-series graphs and timelines can be displayed using HTML5 or JavaScript. HR personnel can view this to quickly grasp the mental health status of employees.

[1460] (Application Example 1)

[1461] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1462] In modern factory environments, properly monitoring employees' mental health is crucial. However, conventional methods do not provide real-time visibility into employees' mental health, potentially leading to the accumulation of overwork and stress. This can result in decreased productivity and increased safety risks. This invention aims to provide a robot-based employee mental health monitoring system within a factory, enabling real-time and time-series tracking of employees' mental health.

[1463] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1464] In this invention, the server includes terminal means for asking interactive questions to employees, means for receiving employee response data, means for analyzing the response data and scoring evaluation items, means for generating prompts based on the scored evaluation items, image generation AI means for generating facial expression images using the prompts and the employee's facial image, means for saving and displaying the facial expression images in chronological order, in-factory robot means for periodically displaying questions to employees and collecting responses, means for generating facial expression images using the image generation AI based on the evaluation items, and means for saving the generated facial expression images in chronological order in a database. This makes it possible to grasp the mental health status of employees in the factory in real time and provide support at the appropriate time.

[1465] A "terminal" is a device used to ask employees interactive questions.

[1466] A "server" is a system that receives, analyzes, and stores employee response data.

[1467] "Response data" refers to the information that employees enter in response to interactive questions.

[1468] "Methods for quantifying evaluation items" refer to technical means for analyzing employee response data and expressing the results as scores.

[1469] A "prompt" is a phrase or command used to give instructions to an image generation AI.

[1470] "Image generation AI" is artificial intelligence that generates facial expression images based on prompts and images of employees' faces.

[1471] A "facial expression image" is an image representing an employee's facial expression, created by a generative AI model based on prompts.

[1472] "Means for saving and displaying in chronological order" refers to technical means for recording generated facial expression images in chronological order and displaying them as needed.

[1473] A "robot in a factory" is an automated device that displays interactive questions to employees in a factory environment and collects their responses.

[1474] "Natural language processing technology" refers to language processing techniques used to analyze employee response data.

[1475] A "dashboard" is a system that visually displays scored evaluation items, allowing for a quick overview of employees' mental health status.

[1476] A "database" is a system for storing and managing generated facial expression images and evaluation data.

[1477] The following system is configured as an embodiment of this invention. The system aims to monitor the mental health status of employees in a factory environment in real time and provide support at the appropriate time.

[1478] First, the terminal periodically displays interactive questions to the employee. This terminal can be used, for example, by a robot in the factory approaching an employee and asking, "How is the progress on your recent work?" The employee then enters their answer to the question.

[1479] The terminal receives employee response data and sends it to the server. This response data is analyzed on the server using natural language processing technology. Specifically, keywords and context are extracted from the responses using transformer models, and sentiment analysis is performed. As a result of this analysis, evaluation items (e.g., stress level) can be scored.

[1480] The server generates prompts based on scored evaluation items. For example, if the stress level is 8 points, it generates a prompt that says "High Stress." This prompt and the employee's facial image are then input into an image generation AI to generate an expression image. Based on this prompt, the image generation AI reflects the stress expression on the employee's facial image.

[1481] The generated facial expression images are stored chronologically in a database by the server. This stored data can later be visualized by HR personnel and managers through a dashboard, allowing them to quickly grasp the mental health status of employees.

[1482] Hardware and software to be used

[1483] Hardware: Factory robots, interactive terminals, servers, databases

[1484] Software: Natural language processing technology (transformer models), image generation AI, database management systems, dashboard display software

[1485] Examples of specific cases and prompt statements

[1486] For example, suppose an employee answers the question, "How is your recent work progressing?" with, "It's extremely difficult and stressful." Based on this answer, a prompt "High Stress" is generated. As a result, the image generation AI receives a prompt message like the following:

[1487] Example of a prompt:

[1488] Prompt: "High stress level, employee face image: face_image.jpg"

[1489] Based on this prompt, the image generation AI generates high-stress facial expression images corresponding to the employee's face image, and the server stores these in a database chronologically. HR personnel can then view the stored data on a dashboard, gaining a clear overview of employees' mental health status. In this way, the system enables real-time monitoring of employee mental health within the factory and the provision of appropriate support.

[1490] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1491] Step 1:

[1492] A terminal (a robot in the factory) periodically displays interactive questions to employees. For example, it might ask, "How is the progress of your recent work?" The user (employee) might respond, "It's very difficult and stressful." In this step, the terminal displays the question, and the user enters the answer. The input data is the user's response.

[1493] Step 2:

[1494] The terminal receives the user's response data and sends it to the server. The server stores the received data for analysis. In this step, the input is the user's response data, and the output is the data sent to the server.

[1495] Step 3:

[1496] The server analyzes the received response data using natural language processing techniques to evaluate emotions and stress levels. This process utilizes a transformer model. Specifically, it analyzes keywords and context from the responses to perform sentiment analysis. The input is the received response data, and the output is the evaluated items, such as the stress level.

[1497] Step 4:

[1498] The server assigns scores to the evaluation items. For example, based on the analyzed emotion response "It's very difficult and stressful," the stress level is scored as 8 out of 10. The input is the evaluation items obtained in the previous step, and the output is the scored stress level.

[1499] Step 5:

[1500] The server generates prompts based on the scored evaluation items. For example, if the stress level is 8, it will generate the prompt "High Stress." The input is the scored stress level, and the output is the generated prompt.

[1501] Step 6:

[1502] The server inputs the generated prompt and the employee's face image into the image generation AI, which then generates an expression image that reflects stress. Specifically, the image generation AI changes the facial expression based on the prompt and the face image. The input for this step is the generated prompt and the employee's face image, and the output is the generated expression image.

[1503] Step 7:

[1504] The server stores the generated facial expression images in a database in chronological order. The stored facial expression images are later used as data for HR personnel and managers to review through a dashboard. The input for this step is the generated facial expression images, and the output is the data stored in the database.

[1505] Step 8:

[1506] The server uses a dashboard to visualize stored facial expression images for HR personnel and managers. This allows them to quickly understand the mental health status of employees. The input is facial expression images stored in the database, and the output is visualized data displayed on the dashboard.

[1507] This clearly explains what data processing and calculations are performed at each step, and what inputs and outputs are generated. Specific operations include data analysis using natural language processing, prompt generation, use of image generation AI, saving to a database, and dashboard display. Based on this, a system for monitoring the user's mental health status in real time is realized.

[1508] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1509] This invention relates to a system for regularly monitoring the mental health status of employees and providing appropriate support. This system utilizes a combination of a dialogue generation AI, an image generation AI, and an emotion engine. The system acquires response data through dialogue with employees, analyzes it, and scores evaluation items. Furthermore, it extracts emotional information using the emotion engine and generates facial expression images by inputting prompts generated based on the scored evaluation items and emotional information into the image generation AI. The generated facial expression images are saved chronologically and displayed on a dashboard.

[1510] System Overview

[1511] 1. Generating dialogue with employees

[1512] The terminal periodically displays questions to employees using conversational AI. For example, it might ask, "How has your workload been lately?" and retrieve the answers.

[1513] The user (employee) enters their answer to the question displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[1514] 2. Analysis of the dialogue content

[1515] The terminal sends the entered response data and the employee's facial image to the server.

[1516] The server analyzes the received response data using a natural language processing (NLP) engine. Here, it extracts keywords indicating negative emotions (e.g., "very difficult," "stressful," "tired," etc.).

[1517] 3. Extraction of emotional information

[1518] The server uses an emotion engine to extract employee emotional information based on extracted keywords and context. For example, based on the phrase "very difficult" from the response, it extracts the emotional information "high stress."

[1519] 4. Scoring of evaluation items

[1520] The server uses the extracted emotional information to score evaluation items (e.g., stress level, happiness level). For example, in response to the answer, "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[1521] 5. Prompt generation based on evaluation items and sentiment information

[1522] The server generates prompts based on scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the server will set the prompt to "high stress."

[1523] 6. Generation of facial expression images

[1524] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression (e.g., sad face, tired face) based on the prompt.

[1525] The server receives facial expression images generated by an image generation AI. These generated facial expression images reflect the employee's current mental state.

[1526] 7. Saving and displaying time-series data

[1527] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[1528] The server displays stored facial images and emotional information in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can identify an employee's stress level as rising and consider providing additional support.

[1529] Specific example

[1530] 1. Interview Session

[1531] The terminal displays the employee with the question, "How is the progress on your recent projects?"

[1532] The user (employee) responded, "Things are progressing smoothly, but I'm feeling a little stressed."

[1533] 2. Analysis of response data and extraction of sentiment information

[1534] The device sends this response data and facial image to the server.

[1535] The server analyzes this response data using natural language processing and extracts keywords related to stress.

[1536] The server extracts emotional information from keywords extracted using an emotion engine. For example, from the response "I feel a little stressed," it extracts the emotional information "moderate stress."

[1537] 3. Scoring of evaluation items

[1538] The server evaluates the stress level based on emotional information and evaluation criteria, and assigns a score, for example, 5 points.

[1539] 4. Prompt generation

[1540] The server generates a "moderate stress" prompt based on a stress level of 5 and emotional information indicating "moderate stress."

[1541] 5. Generation of facial expression images

[1542] The server inputs the prompt "moderate stress" and an image of the employee's face into the image generation AI, which then generates an image of the employee's facial expression that reflects their stress level.

[1543] The server receives facial expression images generated by the image generation AI.

[1544] 6. Saving and displaying time-series data

[1545] The server saves the generated facial expression images to a database and stores them along with past data.

[1546] The server displays saved facial images on a dashboard, allowing HR personnel to monitor employees' mental health status over time.

[1547] In this way, the system of the present invention, by combining a dialogue generation AI, an image generation AI, and an emotion engine, can intuitively grasp the mental health status of employees and quickly provide appropriate support.

[1548] The following describes the processing flow.

[1549] Step 1:

[1550] The device displays questions randomly selected from a pre-configured list to the employee. For example, it might display the question, "How has your workload been recently?"

[1551] Step 2:

[1552] The user (employee) enters their answer to the question displayed on the terminal. For example, they might answer, "It's very difficult and stressful."

[1553] Step 3:

[1554] The terminal sends the entered response data and the employee's facial image to the server.

[1555] Step 4:

[1556] The server analyzes the received response data using a natural language processing (NLP) engine. Here, it extracts keywords indicating negative emotions (e.g., "very difficult," "stressful," "tired," etc.).

[1557] Step 5:

[1558] The server uses an emotion engine to extract employee emotional information based on extracted keywords and context. For example, based on the phrase "very difficult" from the response, it extracts the emotional information "high stress."

[1559] Step 6:

[1560] The server uses the extracted emotional information to score evaluation items (e.g., stress level, happiness level). For example, in response to the answer, "It's very difficult and stressful," the stress level is rated at 8 out of 10.

[1561] Step 7:

[1562] The server generates prompts to input into the image generation AI based on the scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the server sets the prompt to "high stress."

[1563] Step 8:

[1564] The server inputs the generated prompt and the employee's face image into the image generation AI. The image generation AI generates an expression based on the prompt (e.g., sad face, tired face, etc.).

[1565] Step 9:

[1566] The server receives facial expression images generated by an image generation AI. These generated facial expression images reflect the employee's current mental state.

[1567] Step 10:

[1568] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[1569] Step 11:

[1570] The server displays stored facial images and emotional information in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can identify an employee's stress level as rising and consider providing additional support.

[1571] This series of processes allows the system to effectively understand the mental health status of employees and take appropriate measures.

[1572] (Example 2)

[1573] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1574] Employee mental health significantly impacts productivity and workplace relationships. However, accurately understanding employees' mental state and providing timely support is not easy. Traditional methods often rely on simple questionnaires or observations of facial expressions, which are prone to subjective judgments. This makes it difficult to provide effective support quickly and appropriately.

[1575] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1576] In this invention, the server includes a terminal that asks interactive questions to employees, means for receiving employee response data, means for analyzing the response data and scoring evaluation items, means for extracting emotional information based on the scored evaluation items, means for generating prompts based on the scored evaluation items and extracted emotional information, an image generation AI that generates facial expression images using the prompts and the employee's facial image, and means for saving and displaying the facial expression images in chronological order. This makes it possible to accurately grasp the mental health status of employees and provide effective support at the appropriate time.

[1577] "Employee" refers to an individual who works for a company or organization.

[1578] "Interactive questioning" refers to a series of questions designed to obtain information through dialogue with the user.

[1579] A "terminal" refers to an electronic device used by users to input or view information.

[1580] "Response data" refers to information that includes the answers that employees entered in response to interactive questions.

[1581] A "server" refers to a computer system on a network that processes and manages data.

[1582] "Analysis" refers to the process of transforming data into a meaningful form and revealing its structure.

[1583] "Evaluation items" refer to checkpoints or criteria set up to understand the status of employees.

[1584] "Scoring" refers to the process of assigning numerical values ​​to evaluation items.

[1585] "Emotional information" refers to data that indicates the emotional state of employees.

[1586] A "prompt" refers to instructions or hints that a system uses to prompt the user to take the next action.

[1587] "Image generation AI" refers to software that uses artificial intelligence technology to generate images based on specific conditions.

[1588] "Facial expression images" refer to images that visually represent the emotional state of employees.

[1589] "Storing data chronologically" refers to recording data in order, following the flow of time.

[1590] A "dashboard" refers to a visual interface that displays diverse data in a centralized manner, making it easier to understand the situation.

[1591] This invention relates to a system for regularly monitoring the mental health status of employees and providing appropriate support. This system acquires response data through dialogue with employees, analyzes and scores it, extracts emotional information, and generates prompts. Furthermore, it uses image generation AI to generate images of employees' facial expressions, which are then stored and displayed chronologically.

[1592] Specific names of the hardware and software to be used, and the methods for data processing and calculation.

[1593] The terminal is used to generate conversations with employees. The terminal periodically displays questions from an interactive AI to employees. A specific example of a question is, "How has your workload been recently?" The system then obtains the employee's response.

[1594] The server receives employee response data and facial images. Furthermore, it uses a natural language processing (NLP) engine to analyze the response data and extract keywords indicating negative emotions (e.g., "difficult," "stressful," "tired," etc.).

[1595] The emotion engine extracts employee emotional information based on extracted keywords and context. In this process, it extracts emotional information such as "high stress" based on the phrase "very difficult."

[1596] The evaluation engine scores evaluation items (e.g., "stress level," "happiness level") based on the extracted emotional information. For example, in response to the answer, "It's very difficult and I'm feeling stressed," the stress level is rated at 8 out of 10.

[1597] The prompt generation module generates prompts based on scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the prompt will be set to "high stress."

[1598] The image generation AI takes a generated prompt and an employee's facial image as input to generate an expression image. This image reflects the employee's current mental state. For example, based on the prompt "high stress," an image of a sad-looking face is generated.

[1599] The database stores generated facial expression images chronologically. This allows for the accumulation of facial expression images from the past to the present.

[1600] The dashboard displays saved facial images and emotional information. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time.

[1601] Specific example

[1602] 1. Interview Session

[1603] The terminal displays the employee with the question, "How is the progress on your recent projects?"

[1604] The user (employee) responded, "Things are progressing smoothly, but I'm feeling a little stressed."

[1605] 2. Analysis of response data and extraction of sentiment information

[1606] The device sends this response data and facial image to the server.

[1607] The server analyzes this response data using natural language processing and extracts keywords related to stress.

[1608] The server extracts emotional information from keywords extracted using an emotion engine. For example, from the response "I feel a little stressed," it extracts the emotional information "moderate stress."

[1609] 3. Scoring of evaluation items

[1610] The server evaluates the stress level based on emotional information and evaluation criteria, and assigns a score, for example, 5 points.

[1611] 4. Prompt generation

[1612] The server generates a "moderate stress" prompt based on a stress level of 5 and emotional information indicating "moderate stress."

[1613] 5. Generation of facial expression images

[1614] The server inputs the prompt "moderate stress" and an image of the employee's face into the image generation AI, which then generates an image of the employee's facial expression that reflects their stress level.

[1615] The server receives facial expression images generated by the image generation AI.

[1616] 6. Saving and displaying time-series data

[1617] The server saves the generated facial expression images to a database and stores them along with past data.

[1618] The server displays saved facial images on a dashboard, allowing HR personnel to monitor employees' mental health status over time.

[1619] In this way, the system of the present invention, by combining a dialogue generation AI, an image generation AI, and an emotion engine, can intuitively grasp the mental health status of employees and quickly provide appropriate support.

[1620] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1621] Step 1:

[1622] Generating dialogue with employees

[1623] The server periodically generates questions for employees using conversational AI. The questions are randomly selected from a pre-configured list of questions.

[1624] The device periodically displays questions sent from the server as pop-ups. This display continues until the user responds.

[1625] Input: Interactive questions selected by the server

[1626] Output: Question displayed on the terminal

[1627] Step 2:

[1628] Employee response input

[1629] The user (employee) enters their answers to the questions displayed on the terminal. For example, in response to "How has your workload been recently?", they might answer "It's been very difficult and stressful."

[1630] Input: Employee's response to a question displayed on the terminal.

[1631] Output: Response data entered into the terminal

[1632] Step 3:

[1633] Sending response data

[1634] The terminal captures employee response data and facial images, and sends this data to the server.

[1635] Input: Employee response data and facial image

[1636] Output: Response data and facial image sent to the server

[1637] Step 4:

[1638] Analysis of response data

[1639] The server analyzes the received response data using a natural language processing (NLP) engine. During this analysis, it extracts keywords that indicate negative emotions (e.g., "difficult," "stressful," "tired," etc.).

[1640] Input: Response data sent to the server

[1641] Output: Analyzed keywords

[1642] Step 5:

[1643] Extraction of emotional information

[1644] The server uses an emotion engine to extract employee emotional information based on extracted keywords and context. For example, based on the phrase "very difficult," it extracts the emotional information "high stress."

[1645] Input: Analyzed keywords

[1646] Output: Extracted emotional information

[1647] Step 6:

[1648] Scoring of evaluation items

[1649] The server uses the extracted emotional information to score evaluation items (e.g., "stress level," "happiness level," etc.). For example, based on the response, "It's very difficult and I'm feeling stressed," the stress level might be rated at 8 out of 10.

[1650] Input: Extracted emotional information

[1651] Output: Scored evaluation items

[1652] Step 7:

[1653] Prompt generation based on evaluation items and sentiment information

[1654] The server generates prompts based on scored evaluation items and emotional information. For example, if the stress level is 8 points and the emotional information is "high stress," the server will set the prompt to "high stress."

[1655] Input: Scored evaluation items and emotional information

[1656] Output: Generated prompt

[1657] Step 8:

[1658] Generation of facial expression images

[1659] The server inputs the generated prompt and the employee's facial image into the image generation AI, which then generates an expression image that reflects the facial image. For example, a prompt "high stress" combined with a facial image will generate a facial image with a sad expression.

[1660] Input: Prompt and facial image

[1661] Output: Generated facial image

[1662] Step 9:

[1663] Saving time-series data

[1664] The server stores the generated facial expression images in a database in chronological order. This allows for the accumulation of facial expression images from the past to the present.

[1665] Input: Generated facial image

[1666] Output: Facial expression images stored in the database

[1667] Step 10:

[1668] Displaying saved data

[1669] The server displays stored facial images and emotional information in a dashboard format. Users (HR personnel and managers) can view this dashboard to understand the mental health status of employees over time. For example, they can see that a particular employee has shown high stress levels over the past few weeks.

[1670] Input: Facial expression images and emotion information stored in the database

[1671] Output: Data displayed on the dashboard

[1672] (Application Example 2)

[1673] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1674] Managing the mental health of drivers and staff in food delivery companies is a crucial issue for improving employee performance and preventing employee turnover. However, current systems make it difficult to accurately and quickly grasp employees' stress and fatigue levels. Furthermore, there is a lack of efficient and effective systems for visualizing individual employees' emotional states and continuously monitoring changes in them. Thus, there is a need for technological means to grasp employees' mental health status in a timely and accurate manner and provide appropriate support.

[1675] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for analyzing employee response data and scoring evaluation items, means for generating prompts based on evaluation items and emotional information, and for generating facial expression images using image generation AI, and means for saving and displaying the generated facial expression images in chronological order. This makes it possible to accurately and continuously monitor the mental health status of employees and provide prompt and appropriate support.

[1676] An "employee" is a person who engages in work within a company or organization.

[1677] "Interactive questioning" is a method of asking questions to employees in an interactive format and obtaining their answers.

[1678] A "terminal" refers to a device used by employees to answer interactive questions, such as a smartphone or tablet.

[1679] "Response data" refers to information resulting from employees' answers to interactive questions.

[1680] A "server" is a central system that receives response data and performs analysis processing.

[1681] "Means of analyzing and quantifying evaluation items" refers to systems or algorithms that analyze response data and quantify evaluation items based on the results.

[1682] A "prompt" is input information used to instruct an image generation AI to generate specific facial expressions or states.

[1683] "Facial images" refer to image data of employees' faces.

[1684] "Image generation AI" is artificial intelligence that generates new facial expression images based on prompts and facial images.

[1685] "Facial expression images" are facial images generated by image generation AI that reflect the emotional state of employees.

[1686] "Means of saving and displaying in chronological order" refers to a system that saves generated facial expression images as time progresses and displays them on a dashboard or similar.

[1687] "Natural language processing technology" is a technology that uses computers to analyze and understand human language.

[1688] "Emotional information" refers to information about the emotional state of employees extracted from the response data.

[1689] A "dashboard" is an interface that visually displays analysis results, evaluation items, and generated facial expression images.

[1690] This invention relates to a system for monitoring the mental health of drivers and staff in food delivery companies and providing appropriate support. The system asks employees interactive questions, analyzes the response data, extracts emotional information, scores evaluation items, and generates and manages facial expression images based on those scores.

[1691] System Configuration

[1692] This system consists of the following components:

[1693] 1. Terminal

[1694] These are devices such as smartphones and tablets used by employees (drivers). The devices periodically display questions from an interactive AI and retrieve the answers.

[1695] 2. Server

[1696] This is a cloud system that receives response data and facial images and processes them for analysis. It utilizes Amazon Web Services (AWS) cloud services, specifically DynamoDB, S3, and Lambda.

[1697] The server analyzes the response data using a natural language processing (NLP) engine (such as HuggingFace's Transformer model) to extract emotional information.

[1698] The emotional information extracted by the emotion engine is scored according to evaluation criteria.

[1699] Prompts are generated based on scored evaluation items and emotional information.

[1700] Using a prompt and a facial image, an image generation AI (such as StyleGAN) generates an image of facial expressions.

[1701] The generated facial expression images are saved to AWS S3, and metadata is recorded in DynamoDB.

[1702] Saved facial expression images and evaluation criteria are displayed chronologically on the administrator dashboard (using React and Node.js).

[1703] Details of data processing

[1704] Analysis of employee response data

[1705] Employees answer questions from an interactive AI using an app on their devices. For example, in response to a question like, "Have you been feeling stressed about your recent delivery work?", they might answer, "I'm feeling very stressed."

[1706] The device sends this response data to the cloud server.

[1707] The server analyzes the received response data using natural language processing technology, and the emotion engine extracts emotional information such as "high stress."

[1708] Scoring of evaluation items and prompt generation

[1709] Based on emotional information, evaluation items (e.g., stress level) are scored. For example, if the emotional information is "high stress," the stress level might be rated at 8 out of 10.

[1710] Prompts are generated based on scored evaluation items and emotional information. For example, a prompt such as "high stress" is generated.

[1711] Generation and saving of facial expression images

[1712] The server inputs prompts and employee facial images into an image generation AI, which then generates facial expression images that reflect stress levels.

[1713] The server saves the generated facial expression images to AWS S3 and records metadata in DynamoDB.

[1714] For example, based on the "high stress" prompt, an image with a dark, tired expression is generated and saved as "driver123 / 2023-10-01 / image.png".

[1715] Dashboard display

[1716] Administrators can use the dashboard to monitor employees' mental health status over time. For example, they can identify if a particular employee's stress level has increased compared to past data and make decisions such as providing additional support.

[1717] Example of a prompt

[1718] Prompt: Generate a facial expression that reflects a high stress level.

[1719] Input text: "High stress"

[1720] In this way, the system of the present invention can accurately and continuously monitor the mental health of employees and provide prompt and appropriate support.

[1721] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1722] Step 1:

[1723] The user answers questions from the conversational AI using a smartphone application. For example, in response to the question, "Have you been feeling stressed about your recent delivery work?", the user might input, "I feel very stressed." This generates a text response as input data. The input data is, "I feel very stressed."

[1724] Step 2:

[1725] The device sends the user's text response data and facial image to the cloud server. At the same time, the device captures the facial image with its camera and uploads it to the cloud server along with the response data. The input data consists of the text response and the facial image file, and the output is the completion of the transmission to the cloud server.

[1726] Step 3:

[1727] The server inputs the received text response data into a natural language processing (NLP) engine to analyze the response content. The tool used is the HuggingFace transformer model. Emotional keywords and emotion labels are extracted from the analyzed data. The input data is text responses, and the output is extracted emotion information (e.g., "high stress").

[1728] Step 4:

[1729] The server scores evaluation items (e.g., stress level) based on emotional information extracted by the emotion engine. For example, a response like "I feel very stressed" would result in a stress level score of 8 out of 10. The input data is emotional information, and the output is the scored evaluation item (e.g., stress level 8 points).

[1730] Step 5:

[1731] The server generates prompts based on scored evaluation items and emotional information. For example, for a "stress level of 8 points" and emotional information of "high stress," it generates the prompt "high stress." The input data consists of scored evaluation items and emotional information, and the output is the generated prompt (e.g., "high stress").

[1732] Step 6:

[1733] The server inputs a prompt and a user's face image into an image generation AI, which then generates an expression image based on the prompt. The AI ​​model used is StyleGAN, among others. The input data consists of the prompt and the face image, while the output is an expression image reflecting stress.

[1734] Step 7:

[1735] The server saves the generated facial expression image to AWS S3 and records the metadata in AWS DynamoDB. The input data is the generated facial expression image, and the output is the completion of saving to cloud storage. For example, the saved image is stored as "driver123 / 2023-10-01 / image.png".

[1736] Step 8:

[1737] The server displays saved facial expression images and evaluation items in chronological order on an administrator dashboard. Administrators can check the mental health status of each employee through the dashboard. Input data consists of saved facial expression images and evaluation items, while output is a visualized dashboard display.

[1738] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1739] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1740] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1741] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1742] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. In the upper and lower directions of the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. Also, the upper side of the concentric circles is where "pleasant" emotions are located, and the lower side is where "unpleasant" emotions are located. In this way, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1743] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1744] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1745] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1746] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1747] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[1748] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[1749] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[1750] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[1751] 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.

[1752] Furthermore, it is not necessary to store the entirety 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 the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[1753] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[1754] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of 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). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[1755] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[1756] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[1757] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[1758] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[1759] The following is further disclosed regarding the embodiments described above.

[1760] (Claim 1)

[1761] A terminal for asking employees interactive questions,

[1762] A server that receives employee response data,

[1763] A means for analyzing the aforementioned response data and assigning scores to the evaluation items,

[1764] means for generating a prompt based on the scored evaluation items,

[1765] An image generation AI that generates an expression image using the aforementioned prompt and the employee's facial image,

[1766] Means for saving and displaying the aforementioned facial expression images in chronological order,

[1767] A system that includes this.

[1768] (Claim 2)

[1769] The system according to claim 1, which analyzes employee response data using natural language processing technology.

[1770] (Claim 3)

[1771] The system according to claim 1, which visualizes and displays the mental health status of employees on a dashboard based on the aforementioned scored evaluation items.

[1772] "Example 1"

[1773] (Claim 1)

[1774] A terminal for asking employees interactive questions,

[1775] A server that receives employee response data,

[1776] A means for analyzing the aforementioned response data and assigning scores to the evaluation items,

[1777] means for generating a prompt based on the scored evaluation items,

[1778] Image generation means for generating an expression image using the aforementioned prompt and the employee's facial image,

[1779] Means for saving and displaying the aforementioned facial expression images in chronological order,

[1780] A system that includes this.

[1781] (Claim 2)

[1782] The system according to claim 1, which analyzes employee response data using natural language processing technology.

[1783] (Claim 3)

[1784] The system according to claim 1, which visualizes and displays the mental health status of employees on a dashboard based on the aforementioned scored evaluation items.

[1785] "Application Example 1"

[1786] (Claim 1)

[1787] A terminal for asking employees interactive questions,

[1788] A server that receives employee response data,

[1789] A means for analyzing the aforementioned response data and assigning scores to the evaluation items,

[1790] means for generating a prompt based on the scored evaluation items,

[1791] An image generation AI that generates an expression image using the aforementioned prompt and the employee's facial image,

[1792] Means for saving and displaying the aforementioned facial expression images in chronological order,

[1793] A robotic device in the factory that periodically displays questions to employees and collects their answers,

[1794] A means for generating facial expression images using an image generation AI based on the aforementioned evaluation items,

[1795] A system that includes means for storing generated facial expression images in a database in chronological order.

[1796] (Claim 2)

[1797] The system according to claim 1, which analyzes employee response data using natural language processing technology.

[1798] (Claim 3)

[1799] The system according to claim 1, which visualizes and displays the mental health status of employees on a dashboard based on the aforementioned scored evaluation items.

[1800] "Example 2 of combining an emotion engine"

[1801] (Claim 1)

[1802] A terminal for asking employees interactive questions,

[1803] A server that receives employee response data,

[1804] A means for analyzing the aforementioned response data and assigning scores to the evaluation items,

[1805] A means for extracting emotional information based on the aforementioned scored evaluation items,

[1806] Means for generating prompts based on the scored evaluation items and extracted sentiment information,

[1807] An image generation AI that generates an expression image using the aforementioned prompt and the employee's facial image,

[1808] Means for saving and displaying the aforementioned facial expression images in chronological order,

[1809] A system that includes this.

[1810] (Claim 2)

[1811] The system according to claim 1, which analyzes employee response data using natural language processing technology.

[1812] (Claim 3)

[1813] The system according to claim 1, which sets prompts based on the scored evaluation items and extracted emotional information, and cooperates with an image generation AI that generates facial expression images using employee facial images.

[1814] "Application example 2 when combining with an emotional engine"

[1815] (Claim 1)

[1816] A terminal for asking employees interactive questions,

[1817] A server that receives employee response data,

[1818] A means for analyzing the aforementioned response data and assigning scores to the evaluation items,

[1819] means for generating a prompt based on the scored evaluation items,

[1820] An image generation AI that generates an expression image using the aforementioned prompt and the employee's facial image,

[1821] Means for saving and displaying the aforementioned facial expression images in chronological order,

[1822] A means of generating prompts based on evaluation items and emotional information, and generating facial expression images using an image generation AI,

[1823] A system that includes this.

[1824] (Claim 2)

[1825] The system according to claim 1, which analyzes employee response data using natural language processing technology and extracts emotional information using an emotion engine.

[1826] (Claim 3)

[1827] The system according to claim 1, which visualizes and displays the mental health status of employees on a dashboard based on the scored evaluation items and emotional information. [Explanation of symbols]

[1828] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A terminal for asking employees interactive questions, A server that receives employee response data, A means for analyzing the aforementioned response data and assigning scores to the evaluation items, means for generating a prompt based on the scored evaluation items, An image generation AI that generates an expression image using the aforementioned prompt and the employee's facial image, Means for saving and displaying the aforementioned facial expression images in chronological order, A system that includes this.

2. The system according to claim 1, which analyzes employee response data using natural language processing technology.

3. The system according to claim 1, which visualizes and displays the mental health status of employees on a dashboard based on the scored evaluation items.

Images