Information processing device, information processing method, and program

Abstract

This invention provides an information processing device, an information processing method, and a program that allow users to receive personalized advice from a variety of AI characters, thereby increasing their motivation to form health management habits and achieve their goals. [Solution] The information processing device 100 includes a character storage unit 120 that stores multiple AI characters having different personalities, a character selection unit 130 that allows the user to select an AI character, a health data acquisition unit 140 that acquires the user's health data, an advice generation unit 180 that generates advice about the user's health by reflecting the personalities of the AI ​​characters in the generated AI model 161, and an advice provision unit 190 that has the AI ​​character provide feedback to the user in the AI ​​character's language style.

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, and a program.

Background Art

[0002] In recent years, the number of people who are interested in health management by getting into the habit of diet and exercise has been increasing. Instead of going to a gym, they receive advice on exercise and diet from a trainer via an app that operates on a smartphone.

[0003] Here, there is a technology that can record daily diet, exercise, weight, etc. on a smartphone app and consult with an AI character about health management via the app's AI (artificial intelligence) chat function (see Non-Patent Document 1).

Prior Art Documents

Non-Patent Documents

[0004]

Non-Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, with conventional apps, it is not possible to select the app's AI character (AI trainer) according to the user's preference, like choosing a trainer at a gym, and there is a risk that the user will get bored with health management via that app.

[0006] This invention was made in consideration of these circumstances, and aims to provide an information processing device, an information processing method, and a program that allow users to receive advice tailored to their preferences from a variety of AI characters, thereby increasing their motivation to form health management habits and achieve their goals. [Means for solving the problem]

[0007] To solve the above problems, the present invention provides an information processing device in which an AI character provides advice to a user via chat with a user registered in a messaging app, comprising: a character storage means for storing a plurality of AI characters, each having a different personality; a character selection means for causing the user to select an AI character from the plurality of AI characters to provide advice to the user; a health data acquisition means for acquiring the user's health data; a model storage means for storing a generated AI model; a health data analysis means for causing the generated AI model to analyze the health data; an advice generation means for the generated AI model to generate advice regarding the user's health based on the analysis results of the health data; and an advice provision means for the AI ​​character to provide feedback of the advice to the user, wherein the advice generation means generates the advice reflecting the personality of the AI ​​character, and the advice provision means provides feedback of the advice to the user via chat on the messaging app in a language style that reflects the personality of the AI ​​character.

[0008] Furthermore, in order to solve the above problems, the present invention provides an information processing method in which an AI character provides advice to a user via chat with a user registered in a messaging application, comprising: a character selection step in which the information processing device stores a plurality of AI characters, each having a different personality, to be selected from among them; a health data acquisition step in which the health data of the user is acquired; a health data analysis step in which the information processing device stores a generation AI model to analyze the health data; an advice generation step in which the generation AI model generates advice regarding the user's health based on the results of the analysis of the health data; and an advice provision step in which the AI ​​character provides feedback of the advice to the user, wherein the advice generation step generates the advice reflecting the personality of the AI ​​character, and the advice provision step provides feedback of the advice to the user via chat on the messaging application in a language style that reflects the personality of the AI ​​character.

[0009] Furthermore, in order to solve the above problems, the present invention provides a computer-readable program for causing an AI character to provide advice to a user registered in a messaging application via chat, characterized in that the program causes the computer to select an AI character to provide advice to the user from among a plurality of AI characters, each having a different personality, which the computer stores; a health data acquisition step to acquire the user's health data; a health data analysis step to cause a generative AI model stored in the computer to analyze the health data; an advice generation step in which the generative AI model generates advice regarding the user's health based on the results of the health data analysis; and an advice provision step in which the AI ​​character provides feedback of the advice to the user, wherein the advice generation step generates the advice reflecting the personality of the AI ​​character, and the advice provision step provides feedback of the advice to the user via chat on the messaging application in a language style that reflects the personality of the AI ​​character. [Effects of the Invention]

[0010] According to the present invention, users can receive personalized advice from a variety of characters, thereby increasing their motivation to form health management habits and achieve their goals. [Brief explanation of the drawing]

[0011] [Figure 1] This is a block diagram showing the configuration of an information processing system relating to one embodiment of the present invention. [Figure 2] This is a block diagram showing the configuration of an information processing device according to one embodiment of the present invention. [Figure 3] This figure shows an example of the hardware configuration of an information processing device according to one embodiment of the present invention. [Figure 4]This is a flowchart illustrating the information processing procedure for receiving health management advice in an information processing device according to one embodiment of the present invention. [Figure 5] This flowchart provides a detailed explanation of the user settings and AI character selection process (step S10) shown in Figure 4. [Figure 6] This figure shows an example of an AI character selection screen displayed on a user terminal according to one embodiment of the present invention. [Figure 7] This figure shows an example of a selection screen for a specific AI character displayed on a user terminal according to one embodiment of the present invention. [Figure 8] This figure shows an example of a selection screen for a specific AI character displayed on a user terminal according to one embodiment of the present invention. [Figure 9] This figure shows an example of a comprehensive advice feedback screen displayed on a user terminal according to one embodiment of the present invention. [Modes for carrying out the invention]

[0012] Embodiments of the present invention will be described below with reference to the drawings. In all drawings, similar components are denoted by the same reference numerals, and their descriptions are omitted where appropriate. Furthermore, in the following descriptions, unless otherwise specified, each component of each device / system refers to a functional block, not a hardware-level configuration. Each component of each device / system is realized by any combination of hardware and software, centered around the CPU, memory, programs loaded into memory, storage media such as a hard disk that stores those programs, and a network connection interface of any computer. There are various variations in the implementation methods and devices.

[0013] <1. Configuration of Information Processing System 1> FIG. 1 is a block diagram showing the configuration of an information processing system 1 according to an embodiment of the present invention. The information processing system 1 according to this embodiment includes an information processing device 100 that acquires health data from a user and provides advice regarding health management, and a user terminal 200 operated by a user who receives advice regarding health management. The information processing device 100 and the user terminal 200 are connected to each other via a network N such as the Internet, and can communicate with each other data including text and images. Note that in FIG. 1, for simplicity, the case where one user terminal 200 is provided is shown as an example, but two or more user terminals 200 may be provided so that a plurality of users can receive similar advice regarding health management.

[0014] In this embodiment, the network N is described by taking the Internet as an example, but it may be a wired or wireless communication network such as a LAN (Local Area Network) or a WAN (Wide Area Network), and may include communication via a 4G or 5G line.

[0015] The information processing device 100 prepares an AI character as an AI trainer exclusive to the user who operates the user terminal 200, acquires health data from the user, analyzes it with a generated AI model, and uses the analysis result to provide the user with advice regarding health management such as dietary guidance and exercise guidance. It is a server computer.

[0016] In this embodiment, in the information processing system 1, the AI character (AI trainer) provided by the information processing device 100 and the user of the user terminal 200 execute a conversation by text or the like via the chat function of a messaging app (messaging application) such as LINE (registered trademark), WhatsApp (registered trademark), WeChat (registered trademark), Facebook Messenger (registered trademark).

[0017] The information processing device 100 can be introduced, for example, by business operators in various industries that support health management, such as personal training business operators like fitness gyms and training gyms, healthcare service providers, and supplement manufacturers. These business operators provide services to member users as SaaS (Software as a Service).

[0018] In the information processing system 1 in this embodiment, only one information processing device 100 is provided. However, the information processing device 100 may be realized by a single server computer, or may be realized by a plurality of server computers.

[0019] The user terminal 200 is a computer such as a smartphone, a tablet terminal, a notebook PC, or a desktop PC for a user who is a member enjoying the service provided by the information processing device 100 to perform data input and output.

[0020] <2. Configuration of Information Processing Device 100> The information processing device 100 includes a user setting unit 110, a character storage unit 120, a character selection unit 130, a health data acquisition unit 140, a health data recording unit 150, a model storage unit 160 having a generation AI model 161, a health data analysis unit 170, an advice generation unit 180, and an advice providing unit 190, and provides advice (counsel) regarding health management to the user on the messaging app.

[0021] For example, the information processing device 100 conducts a conversation with a user registered (for example, friend registration) on the messaging app from the official account of the administrator who provides the above-described service through the chat function of the messaging app. Then, the information processing device 100 acquires and analyzes health data regarding diet, weight, and exercise from the user, and based on the analysis result, an AI character (AI trainer) gives advice regarding health management to the user.

[0022] The user settings unit 110 sets the user's profile information necessary for the user to receive advice on health management, and sets the user's health management goals. The profile information set in the user settings unit 110 includes, for example, the user's gender, date of birth, and height. The user settings unit 110 also includes information regarding the user's health management goals, such as the user's ideal weight, whether or not to record body fat percentage, and the target date for achieving those goals. Furthermore, the user settings unit 110 allows users to set goals by inputting, for example, text data of 100 characters or less, describing their purpose for using the service for health management and their ideal self. In this way, the user information set in the user setting unit 110 is stored by the information processing device 100 and used as part of the prompt when the generating AI model 161, which will be described later, generates advice.

[0023] The character memory unit 120 stores multiple AI characters (AI trainers), each having a different personality. In this embodiment, the AI ​​character is an anthropomorphic trainer on the application that provides health management advice to the user operating the user terminal 200 via a messaging application, and chats with the user using the generated AI model 161 described later. In this case, "chat" can include not only text-based communication but also images (including videos) and audio.

[0024] Here, we will explain the AI ​​character further. In this embodiment, there are no particular restrictions on the method of creating the AI ​​character stored in the character storage unit 120. For example, the character storage unit 120 may store an AI character that personifies the personality and language style of a fitness trainer who actually exists in a fitness club. That is, the AI ​​character is created to mimic the role and characteristics of an actual trainer who has knowledge and skills in fitness and health management, and such an AI character has the characteristic of being designed to be human-like in terms of the trainer's personality and way of speaking (language style expressed in chat) in order to be approachable to the user and to provide effective support to the user.

[0025] The character selection unit 130 allows the user to select an AI character from among multiple AI characters stored in the character memory unit 120 to provide advice. By allowing the user to select from multiple AI characters, the user can choose an AI character that suits them, taking into account the character's settings and personality.

[0026] The health data acquisition unit 140 acquires health data entered by the user via chat with an AI character on a messaging application using the user terminal 200. In this embodiment, data related to diet, weight, and exercise are used as health data. For meal data, for example, photographs of the food as it was served (meal photos) and text data of the meal's name are used. For weight data, for example, text data of the user's measured weight is used, or when the user types "weight record" in the chat, the app displays multiple selection buttons based on the previous day's weight (for example, if the previous day's weight was 70.0kg, selection buttons in 0.5kg increments within a range of plus or minus 5.0kg, such as "69.5kg", "70.0kg", "70.5kg", etc.), and the user selects the appropriate value. Furthermore, for exercise data, for example, text data of the type and duration of exercise entered by the user in the chat is used, or when the user types "exercise record" in the chat, the app displays selection buttons for the type of exercise (gym training, home training, aerobic exercise, etc.) and the body part and duration of exercise, and the user selects the appropriate option, allowing the data to be obtained.

[0027] The health data recording unit 150 records and stores health data related to diet, weight, and exercise acquired by the health data acquisition unit 140. The health data may be recorded in the health data recording unit 150 or an internal recording medium (not shown) within the information processing device 100, or it may be recorded on an external recording medium (not shown) that can be connected to the information processing device 100.

[0028] The model memory unit 160 stores the generated AI model 161. The generated AI model 161 is a generated AI that uses a large language model (LLM) to generate advice for the user by combining health data acquired from the user with prompts associated with the AI ​​character. Examples of LLMs used in the generated AI model in this embodiment include GPT-3.5, GPT-4, GPT-4 Turbo, and GPT-4o related to OpenAI's "GPT®", and Claude 3, Claude 3 Haike, Claude 3 Sonnet, and Claude 3 Opus related to ANTHROPIC's "CLAUDE®".

[0029] LLM is used as a generative AI that learns from a large amount of text data and generates text in response to specific prompts. In this embodiment, various LLMs are used in the generative AI model 161, but the application of the present invention is not limited to LLMs as the generative AI. For example, in addition to LLMs, other generative AIs such as GANs (Generative Adversarial Networks) for image generation, music generation AI, video generation AI, or multimodal language models that can understand and generate multiple different types of data such as text, images, and audio may be used. A language model that supports multimodal processing is called, for example, an MMLLM (Multi-Modal Large Language Model) or a multimodal model. A multimodal language model may generate responses using text data, or it may generate responses that include image data or audio data.

[0030] Furthermore, the model memory unit 160 is not limited to storing only one generated AI model, but can store multiple generated AI models. The system may store this information and allow users to switch between generated AI models based on the type of health data or instructions from the user or system administrator. Furthermore, the generation AI model 161 is not provided in the information processing device 100, but rather a generation AI model managed outside the information processing device 100 may be used.

[0031] The health data analysis unit 170 has the generating AI model 161 analyze the user's health data stored in the health data recording unit 150. For example, the health data analysis unit 170 provides a photo of a meal (photo data) and the name of the meal (text data) to the generating AI model 161, which then analyzes the name of the dish (e.g., mapo tofu, fried dumplings, etc.), the amount of food (e.g., 400g, etc.), the calories (e.g., 550kcal, etc.), and the PFC balance, which is the ratio of protein, fat, and carbohydrate intake. The health data analysis unit 170 also analyzes, for example, the number of consecutive days of weight recording and changes in weight from the input weight data. Furthermore, the health data analysis unit 170 provides, for example, input exercise data to the generating AI model 161, which then analyzes and calculates the calories burned by that exercise.

[0032] The advice generation unit 180 generates advice about the user's health based on the analysis results obtained when the generated AI model 161 analyzes the user's health data in the health data analysis unit 170. In doing so, the advice generation unit 180 generates advice that reflects the personality of the AI ​​character selected by the user. Specifically, the character memory unit 120 stores information about the AI ​​character's personality as partial prompts, including the AI ​​character's tone of voice, its encouraging attitude towards the user, and information about the experiences or insights the AI ​​character will gain. The advice generation unit 180 then includes these partial prompts in the prompts input to the generated AI model 161, enabling the generated AI model 161 to generate advice based on these partial prompts. As a result, the AI ​​character's personality is provided in part of the prompts input to the generated AI model, and since these prompts are editable, it becomes easy to diversify the variations of the AI ​​character.

[0033] The advice-providing unit 190 provides advice to the user via chat through an AI character selected by the user. Here, the advice-providing unit 190 provides advice to the user in a language style that reflects the personality of the AI ​​character (speaking style, attitude, dialect, honorifics, frankness, friendliness, etc.).

[0034] By equipping the information processing device 100 with the above configuration, users can receive personalized health advice from their favorite AI character, selected from among multiple AI characters, reflecting the character's personality, such as being strict or gentle. Furthermore, the AI ​​character provides feedback on health in the user's preferred language style, allowing users to receive advice tailored to their preferences from a variety of AI characters. This can enhance users' motivation to form health management habits and achieve their goals.

[0035] <3. Hardware configuration of the information processing device 100> Figure 3 shows an example of the hardware configuration of the information processing device 100 described above according to one embodiment of the present invention. The information processing device 100 includes a bus 1010, a processor 1020, a memory 1030, a storage device 1040, an input / output interface 1050, and a network interface 1060.

[0036] Bus 1010 is a data transmission path for the processor 1020, memory 1030, storage device 1040, input / output interface 1050, and network interface 1060 to send and receive data to and from each other. However, the method of connecting the processor 1020 and the other components to each other is not limited to bus connection.

[0037] The 1020 processor is a processor implemented in components such as the CPU (Central Processing Unit) and GPU (Graphics Processing Unit).

[0038] Memory 1030 is a main memory device implemented using RAM (Random Access Memory), etc.

[0039] The storage device 1040 is an auxiliary storage device implemented as a removable media such as an HDD (Hard Disk Drive), SSD (Solid State Drive), or memory card, or as ROM (Read Only Memory), and has a recording medium. The recording medium of the storage device 1040 stores program modules that realize each function of the information processing device 100. The processor 1020 reads these program modules into the memory 1030 and executes them, thereby realizing each function corresponding to that program module.

[0040] The input / output interface 1050 is an interface for connecting the information processing device 100 with various input / output devices.

[0041] The network interface 1060 is an interface for connecting the information processing device 100 to the network N. The method by which the network interface 1060 connects to the network N may be a wireless connection or a wired connection.

[0042] <4. Operation of the information processing device 100> In the following embodiment, the information processing device 100 acquires the user's health data through communication using the chat function of a messaging app from the user terminal 200, and an AI character realized by an AI model generated based on that health data provides the user with health management advice.

[0043] <4.1. User Registration> The user adds the official account of the health management support service provided by the information processing device 100 as a "friend" on the messaging app from their user terminal 200, and registers it in the messaging app's "official accounts" list.

[0044] <4.2. User Settings> Figure 4 is a flowchart illustrating the information processing procedure for receiving health management advice in an information processing device 100 according to one embodiment of the present invention. To record daily health data and receive advice from an AI trainer, the user first performs several settings, such as user settings and selecting an AI trainer (Step S10).

[0045] Figure 5 is a flowchart that provides a detailed explanation of the user settings and AI character (AI trainer) selection process (step S10) in the flowchart shown in Figure 4. The user sets up their basic profile by entering simple profile information such as gender, date of birth, and height via the chat function of the messaging app from the user terminal 200 (step S11).

[0046] Next, the user sets health management goals by similarly entering information such as whether or not they want to record their ideal weight and body fat percentage, and the goal date, as well as entering their purpose for using the service and their ideal self in text data of, for example, 100 characters or less (Step S12). As described above, in the information processing device 100, the user setting unit 110 sets the user's profile information and health-related goals, and in the advice generation unit 180, the generating AI model 161 further considers the user's profile information and goals to generate health-related advice for the user. Therefore, health-related advice can be generated considering the user's profile information such as age, gender, and height entered by the user as user setting information, as well as the goals the user set when receiving training. Consequently, the information processing device 100 can generate more personalized advice, thereby improving the user's motivation and generating advice that is more likely to be implemented.

[0047] <4.3. Selecting an AI Character (AI Trainer)> Then, the user selects an AI character (AI trainer) that will provide health management advice via the chat function of the messaging app (step S13). Figure 6 shows an example of an AI character selection screen 210 displayed on a user terminal 200 according to one embodiment of the present invention. As shown in Figure 6, the AI ​​character selection screen 210 displays a list 211 of AI characters stored in the character storage unit 120 and selectable by the character selection unit 130. In Figure 6, four AI characters are displayed in the list as an example, but if the number of AI characters stored in the character storage unit 120 is greater than that, the list 211 can be displayed in a way that allows the user to view and select all AI characters, for example, by scrolling.

[0048] In this embodiment, the AI ​​character selection screen 210 shown in Figure 6 displays four AI characters: "Trainer Yoshida" 212, "Oni Yoshida" 213, "Teacher Tanaka" 214, and "Mr. X" 215. The display area for each AI character shows the icon and character name assigned to that character. For example, a user can select an AI trainer from these four AI characters to devise a training menu related to health management or to provide health management advice. Here, when the user clicks the selection button 212A within the display area of ​​"Yoshida Trainer" 212 to consider selecting "Yoshida Trainer" 212, a selection confirmation screen 220 for "Yoshida Trainer" 212, as shown in Figure 7, is displayed.

[0049] Figure 7 shows an example of a selection screen 220 for a specific AI character displayed on a user terminal 200 according to one embodiment of the present invention. Figure 7 is a screen that allows the user to choose whether to select "Yoshida Trainer" 212 as the AI ​​character (AI trainer) after the user has selected "Yoshida Trainer" 212 on the AI ​​character selection screen 210 shown in Figure 6. The selection screen 220 shown in Figure 7 displays a trainer description area 221, which includes the icon of "Yoshida Trainer" 212, the character name, and a brief profile of the trainer, allowing the user to confirm the characteristics of the character. The selection screen 220 shown in Figure 7 also includes a "Select" button 222 for the user to decide to select "Yoshida Trainer" as the AI ​​character, and a "Close" button 223 for the user to return to the AI ​​character selection screen 210 shown in Figure 6 and redo the AI ​​character selection.

[0050] In Figure 7, when the user clicks the "Select" button 222, the character selection unit 130 selects the AI ​​character (in this case, "Trainer Yoshida" 212) as the AI ​​trainer. On the other hand, when the user clicks the "Close" button 223, the character selection unit 130 does not perform the AI ​​character selection, and the display screen of the user terminal 200 returns to the AI ​​character selection screen 210 shown in Figure 6. In Figure 6, when a user clicks the selection button 213A within the display area of ​​"Oni Yoshida" 213 to consider selecting "Oni Yoshida" 213, a selection confirmation screen 230 for "Oni Yoshida" 213, as shown in Figure 8, is displayed.

[0051] Figure 8 shows an example of a selection screen 230 for a specific AI character displayed on a user terminal 200 according to one embodiment of the present invention. Figure 8 is a screen that allows the user to choose whether to select "Oni Yoshida" 213 as an AI character (AI trainer) after the user has selected "Oni Yoshida" 213 on the AI ​​character selection screen 210 shown in Figure 6. The selection screen 230 shown in Figure 8 displays a trainer description area 231, which includes the icon of "Oni Yoshida" 213, the character name, and a brief profile of the trainer, allowing the user to confirm the character's characteristics. The selection screen 230 shown in Figure 8 also includes a "Select" button 232 for the user to decide to select "Oni Yoshida Trainer" as an AI character, and a "Close" button 233 for the user to return to the AI ​​character selection screen 210 shown in Figure 6 and redo the AI ​​character selection.

[0052] <4.4. AI Characters> Here, the AI ​​character in this embodiment will be described in detail. As mentioned above, in this embodiment, the AI ​​character stored in the character storage unit 120 may be an AI character that personifies the personality and language style of a fitness trainer who actually exists in the fitness club, or it may be an AI character based on a fictional person, and there are no particular restrictions on the method of creating the AI ​​character.

[0053] On the other hand, the AI ​​characters in this embodiment provide users with an image corresponding to the appearance of a trainer through icons and names, and have a basic model set up that includes knowledge and skills as a trainer. Furthermore, each individual AI character in this embodiment is assigned a separate personality. Thus, each AI character possesses knowledge of fitness and health management, as well as knowledge and skills as a trainer, and each has a unique personality. Differences in the personalities, knowledge, and skills of the AI ​​characters result in differences in the advice generated by the generated AI model 161. In this way, each AI character can be realized as a useful and approachable AI character for the user by combining a basic setting model as a trainer with a setting personality that determines the content of the advice and the language style in chat.

[0054] The configuration model is the basic structure, rules, and foundational design for character formation that determine the role and characteristics of the AI ​​character. Specifically, it defines the character's role (e.g., fitness trainer) and the functions necessary to fulfill that role (e.g., generating training menus, managing learning progress). It also defines the character's appearance on messaging applications (e.g., icon design, color, clothing, etc.) and attributes (e.g., name, background story, etc.). In this way, a model is established for the AI ​​character that possesses expertise in the fitness field and analyzes the user's health data to create training plans.

[0055] Furthermore, the established personality defines the individuality, behavioral patterns, and communication style of the AI ​​character, and is a crucial part that influences the user's perception of "human-likeness" and "approachability." Specifically, it defines the AI ​​character's personality (for example, being cheerful, energetic, and friendly like a friend) and the tone of voice expressed in chat text (for example, using polite language). It also defines emotional expression (for example, expressing joy when the user achieves a goal) and ethical judgment criteria (recommending healthy and manageable fitness goals). In this way, a model is created for a particular AI character, set up with a cheerful and encouraging personality, and providing positive encouragement to users who are on the verge of giving up.

[0056] For example, in the information processing device 100 according to this embodiment, among the AI ​​characters stored in the character storage unit 120, three AI characters, "Trainer Yoshida" 212, "Teacher Tanaka" 214, and "Mr. X" 215, are created by personifying real-life trainers, while the AI ​​character "Demon Yoshida" 213 is an AI character (basic AI character) that is a trainer who provides stricter guidance than "Trainer Yoshida" 212. In this case, "Trainer Yoshida" 212 and "Demon Yoshida" 213 can be generated with the same or similar setting models, but with significantly different setting personalities.

[0057] For example, as the personality of a specific AI character ("Trainer Yoshida" 212) stored in the character memory unit 120, Personality: Energetic and good at encouraging others • Approach style: Positive communication, emphasis on friendliness • Supporting goal achievement: We closely monitor progress and use a flexible approach to prevent feelings of frustration. Set it as follows.

[0058] Furthermore, if "Trainer Yoshida" 212 is used as the base AI character and "Oni Yoshida" 213 is used as a derivative AI character of the base AI character, then its personality would be as follows: Personality: Strict and honest • Support Style: Prioritizing the user's goal achievement and providing uncompromising guidance. • Supporting goal achievement: Focus on results and accelerate progress by challenging users. This involves creating different personalities.

[0059] Furthermore, the setting model for the derivative AI character "Oni Yoshida" 213 is, for example, that it has the same role as the basic AI character "Yoshida Trainer" as a fitness trainer, but the icon design and name are expressed as those of a strict trainer, allowing it to be distinguished as "Oni Yoshida". Furthermore, the number of derived AI characters for a single basic AI character is not limited to one, and there is no particular restriction on the number. That is, in this embodiment, there is a basic AI character that has a basic personality (for example, being gentle in guidance and using polite language). In addition, if the AI ​​character has one or more personalities different from the basic personality (for example, being strict in guidance and using commanding language), then in addition to the basic AI character, there are one or more derived AI characters that have one or more personalities different from the basic personality.

[0060] Thus, because the information processing device 100 has both a basic AI character and derivative AI characters, a user who wants to receive advice from "Trainer Yoshida" can choose "Trainer Yoshida" 212 if they want the same gentle guidance as always, or they can choose "Demon Yoshida" 213 if they want stricter guidance from the usually gentle "Trainer Yoshida". This diversifies the choices of AI characters as health trainers and eliminates content fatigue (monotony) for the user.

[0061] In the process of selecting an AI character (AI trainer) by the user (step S13), as shown in Figure 6, the user may select one AI character from a list of all AI characters stored in the character storage unit 120 on the AI ​​character selection screen 210, or other methods may be used. Alternatively, for example, the AI ​​character selection screen 210 could display basic AI characters as list 211, allowing the user to select a basic AI character (step S13A), and then select one AI character from the selected basic AI character and its derivative AI characters (step S13B). Selecting from the entire list is easier when the number of AI characters is relatively small, as it allows the user to compare characters with each other. On the other hand, selecting AI characters step by step is easier when there are many basic AI characters and their derivative AI characters, as the user can first compare characters with different setting models, and then compare characters with different personality models.

[0062] <4.5. Recording and Feedback of Meal Data> Returning to the flowchart in Figure 4, once the trainer selection (step S10) is complete, the information processing device 100 acquires and analyzes the user's health data from the user terminal 200 via chat on the messaging app and provides the user with health management advice (step S20). In this embodiment, as described above, health data includes data on diet, weight, and exercise, and users typically report daily for each type of data. In the following, data acquisition, analysis, and feedback are performed in the order of diet, weight, and exercise, but this is just one example, and processing may be performed in other orders.

[0063] As described above, the health data acquisition unit 140 acquires meal data by taking photos of meals or entering the names of meals as text on the chat function of the messaging application on the user terminal 200. The acquired meal data is recorded by the health data recording unit 150, while, as mentioned above, the health data analysis unit 170 provides the meal photos and meal names to the generating AI model 161, which analyzes the name of the dish, the amount of food, calories, and the PFC balance (protein, fat, and carbohydrate intake ratio), and records the analysis results in the health data recording unit 150 in association with the meal data. Furthermore, the dietary data (including analysis results) recorded in the health data recording unit 150 can be manually modified by the administrator of the information processing device 100 or the administrator of this service. In addition, when acquiring dietary data by the health data acquisition unit 140, past dietary data recorded in the health data recording unit 150 can also be referenced.

[0064] Next, the advice generation unit 180 generates advice regarding the user's diet based on the analysis results from the health data analysis unit 170. In this process, the advice generation unit 180 generates advice that reflects the personality of the AI ​​character selected by the user. The advice provision unit 190 then provides feedback to the user via chat through the AI ​​character selected by the user, using a language style that reflects the personality of the AI ​​character (step S21). For example, if a user selects "Oni Yoshida" 213, an AI character who provides strict guidance, the AI ​​might analyze the user's dietary data and advise them to restrict their diet further, saying something like, "Fried chicken is high in protein, but also high in fat. You need 12% body fat to reach your goal! A small amount of fried chicken once a week after training is okay. Otherwise, stick to sautéed chicken breast! You're only 6kg away from your goal, no slacking allowed!"

[0065] Next, as described above, the health data acquisition unit 140 acquires weight data when the user inputs their weight as text in the chat of the messaging app on the user terminal 200, or when the user types "record weight" in the chat, by displaying multiple selection buttons on the app based on the previous day's weight and allowing the user to make a selection. The acquired weight data is recorded by the health data recording unit 150, while, as described above, the health data analysis unit 170 analyzes the number of consecutive days of recording and changes in weight based on the input weight data, and records the analysis results in the health data recording unit 150 in association with the weight data.

[0066] Next, the advice generation unit 180 generates advice regarding the user's weight based on the analysis results from the health data analysis unit 170. In this process, the advice generation unit 180 generates advice that reflects the personality of the AI ​​character selected by the user. The advice provision unit 190 may also provide feedback to the user via chat through the AI ​​character selected by the user, using a language style that reflects the personality of the AI ​​character (step S22). For example, if weight data analysis reveals a downward trend in weight, the system can provide visual feedback to the user using image data such as graphs showing weight changes over several days, or provide feedback such as, "You've lost 0.1 kg since the last record! Your BMI is 22.9!"

[0067] Furthermore, as described above, the health data acquisition unit 140 acquires exercise data when the user inputs the type of exercise and exercise time as text in the chat of the messaging app on the user terminal 200, or when the user types "exercise record" or similar, and the user selects from the exercise name and exercise time selection buttons displayed on the app. The acquired exercise data is recorded by the health data recording unit 150, while, as described above, the health data analysis unit 170 provides the input exercise data to the generating AI model 161 to analyze and calculate the calories burned by the exercise, and records the analysis results in the health data recording unit 150 in association with the exercise data.

[0068] Next, the advice generation unit 180 generates exercise advice for the user based on the analysis results from the health data analysis unit 170. In doing so, the advice generation unit 180 generates advice that reflects the personality of the AI ​​character selected by the user. The advice provision unit 190 then provides feedback to the user via chat through the AI ​​character selected by the user, using a language style that reflects the personality of the AI ​​character (step S23). For example, if a user selects "Oni Yoshida" 213, an AI character who provides strict guidance, the AI ​​will analyze the user's exercise data and, if it recognizes that the user is actively exercising, it will offer advice to maintain this pace, such as, "You're training your whole body in a balanced way with squats, chin-ups, dips, and an ab roller. 50 squats is especially impressive!"

[0069] <4.6. Overall Feedback> As described above, after the analysis of each of the multiple health data acquired by the health data acquisition unit 140 is completed and individual advice (suggestions) is provided (step S20), the advice generation unit 180 inputs the analysis results of all the health data into the generating AI model 161 to generate personalized comprehensive advice (comprehensive report) for the user. Furthermore, the advice provision unit 190 provides comprehensive advice to the user via chat through an AI character selected by the user, using a language style that reflects the personality of the AI ​​character (step S30).

[0070] Figure 9 shows an example of a comprehensive advice feedback screen 240 displayed on a user terminal 200 according to one embodiment of the present invention. Figure 9 displays comprehensive advice (comprehensive feedback) regarding health management from "Oni Yoshida," a trainer who is an AI character selected by the user. In this embodiment, the comprehensive advice includes Good Points 241 regarding the user's approach to health management, More Suggestions 242, and an overall review comment 243. Furthermore, the overall review comment 243 may include a score display 244 so that the user can see at a glance how highly they are being evaluated.

[0071] It should be noted that each flowchart described in this embodiment is merely one embodiment. Without altering the spirit of the present invention, there may be processes other than those described in each flowchart, some of the described processes may be omitted, or the order of the processes may be changed. Furthermore, the configuration of the display screen and the display information used to explain the embodiments are also just examples, and some of the display information may be omitted, or other information may be included.

[0072] While embodiments of the present invention have been described above, the embodiments disclosed herein should be considered in all respects to be illustrative and not restrictive. The scope of this invention is indicated not by the above description but by the claims, and all modifications within the meaning and scope of the claims are intended to be included. [Explanation of symbols]

[0073] 1. Information Processing System 100 Information Processing Devices 110 User Settings Section 120 Character Memory Unit 130 Character Selection Section 140 Health Data Acquisition Department 150 Health Data Recording Department 160 Model Memory Unit 161 Generative AI Models 170 Health Data Analysis Department 180 Advice Generation Unit 190 Advice Provider Department 200 user terminals 210 AI Character Selection Screen 220, 230 Selection Confirmation Screen 1010 Bus 1020 processor 1030 memory 1040 Storage Devices 1050 Input / Output Interface 1060 Network Interfaces N Network

Claims

1. An information processing device in which an AI character provides advice to a user via chat with the user registered in a messaging app, A character memory means for storing multiple AI characters, each with a different personality, A character selection means for causing the user to select an AI character from among the plurality of AI characters to provide advice to the user, A health data acquisition means for acquiring the user's health data, A model storage means for storing a generated AI model, A health data analysis means for causing the generated AI model to analyze the health data, The AI ​​model generates advice based on the analysis results of the health data, and the advice generation means generates advice regarding the user's health. The AI ​​character includes an advice-providing means that provides the advice to the user, The advice generation means generates the advice while reflecting the personality of the AI ​​character. The advice-providing means provides feedback to the user via chat on the messaging app, using a language style that reflects the personality of the AI ​​character. An information processing device characterized by the following:

2. An information processing apparatus according to claim 1, The system further comprises user setting means for setting the user's profile information and health-related goals, The advice generation means generates health advice for the user by further considering the user's profile information and goals in the AI ​​model. An information processing device characterized by the following:

3. An information processing apparatus according to claim 1, The AI ​​characters stored in the character storage means include a basic AI character having a basic personality for each AI character, and, if the AI ​​character has one or more personalities different from the basic personality, one or more derived AI characters having one or more personalities different from the basic personality, in addition to the basic AI character. An information processing device characterized by the following:

4. An information processing apparatus according to claim 3, The character selection means causes the user to select one AI character from a list of all the basic AI characters and derived AI characters stored in the character storage means. An information processing device characterized by the following:

5. An information processing apparatus according to claim 3, The character selection means first causes the user to select one basic AI character from all the AI ​​characters stored in the character storage means, and then, if there are one or more derived AI characters for the basic AI character, causes the user to select one of the AI ​​characters from the basic AI character or the one or more derived AI characters. An information processing device characterized by the following:

6. An information processing apparatus according to claim 1, The character memory means stores partial prompts as information relating to the personality of the AI ​​character, including the tone of voice of the AI ​​character, the attitude of encouraging the user, and information relating to the experience or feelings given to the AI ​​character. The advice generation means includes the partial prompt in the prompt input to the generating AI model and generates the advice using the generating AI model. An information processing device characterized by the following:

7. An information processing apparatus according to claim 1, The aforementioned model storage means stores multiple generative AI models of different types, The advice generation means selects a generation AI model to be used from among the plurality of generation AI models according to the type of health data and generates the advice. An information processing device characterized by the following:

8. An information processing method in which an AI character provides advice to a user through a chat with the user registered in a messaging app, A character selection step in which the user selects an AI character from among multiple AI characters, each having a different personality, in which the information processing device stores an AI character to provide advice to the user. A health data acquisition step for acquiring the user's health data, A health data analysis step in which the information processing device stores a generating AI model that analyzes the health data, The generation AI model generates advice regarding the user's health based on the results of the analysis of the health data in an advice generation step, The AI ​​character provides advice to the user as feedback, The advice generation step generates the advice while reflecting the personality of the AI ​​character, The advice-providing step involves providing feedback to the user via chat on the messaging app, using a language style that reflects the personality of the AI ​​character. An information processing method characterized by the following:

9. A computer-readable program for causing an AI character to provide advice to a user via chat with the user registered in a messaging app, A character selection step in which the computer stores an AI character to provide advice to the user, and the user selects from among multiple AI characters, each having a different personality. A health data acquisition step for acquiring the user's health data, A health data analysis step in which the computer has a stored generative AI model analyze the health data, The generation AI model generates advice regarding the user's health based on the results of the analysis of the health data in an advice generation step, The computer is made to perform an advice-providing step in which the AI ​​character provides the advice as feedback to the user, The advice generation step generates the advice while reflecting the personality of the AI ​​character, The advice-providing step involves providing feedback to the user via chat on the messaging app, using a language style that reflects the personality of the AI ​​character. A program characterized by the following features.

Images