Data processing system

Abstract

The system according to the present embodiment includes a reception unit, an analysis unit, a determination unit, and a proposal unit. The reception unit is configured to receive a genetic test result. The analysis unit is configured to analyze the genetic test result and questionnaire information received by the reception unit. The determination unit is configured to determine a risk of onset on the basis of the information analyzed by the analysis unit. The proposal unit is configured to propose a dietary kit for reducing the risk of onset determined by the determination unit.

Description

Technical Field

[0001] The technology disclosed herein relates to a system. Background Technology

[0002] Patent Document 1 discloses a personalized chatbot control method executed by at least one processor, the method comprising: receiving user speech; adding the user speech to a prompt containing instructions related to a chatbot role; encoding the prompt; and inputting the encoded prompt into a language model to generate chatbot speech in response to the user speech.

[0003] Patent document 1: Japanese Patent Application Publication No. 2022-180282.

[0004] In the existing technology, there is a problem that the risk assessment of disease based on gene testing results has not been fully carried out, and specific countermeasures and suggestions have not been put forward based on the assessment results. Summary of the Invention

[0005] The system described in this embodiment includes a receiving department, an analysis department, a judgment department, and a proposal department. The receiving department uploads genetic testing results. The analysis department analyzes the genetic testing results and questionnaire information uploaded by the receiving department. The judgment department determines the risk of disease based on the information analyzed by the analysis department. The proposal department proposes dietary kits to reduce the risk of disease determined by the judgment department. Attached Figure Description

[0006] Figure 1 This is a conceptual diagram illustrating an example of the configuration of a data processing system according to the first embodiment.

[0007] Figure 2 This is a conceptual diagram illustrating an example of the main functions of the data processing apparatus and smart device according to the first embodiment.

[0008] Figure 3 This is a conceptual diagram illustrating an example of the data processing system configuration in the second embodiment.

[0009] Figure 4 This is a conceptual diagram illustrating an example of the main functions of the data processing device and smart glasses according to the second embodiment.

[0010] Figure 5 This is a conceptual diagram illustrating an example of the data processing system configuration in the third embodiment.

[0011] Figure 6 This is a conceptual diagram illustrating an example of the main functions of the data processing device and head-mounted terminal according to the third embodiment.

[0012] Figure 7This is a conceptual diagram illustrating an example of the data processing system configuration in the fourth embodiment.

[0013] Figure 8 This is a conceptual diagram illustrating an example of the functions of the main parts of the data processing device and robot according to the fourth embodiment.

[0014] Figure 9 It represents an emotion graph that maps multiple emotions.

[0015] Figure 10 It represents an emotion graph that maps multiple emotions.

[0016] Explanation of reference numerals in the attached figures: Data processing systems 10, 210, 310, and 410 12 Data processing devices 14 Smart devices 214 Smart Glasses 314 Head-mounted terminal 414 Robot. Detailed Implementation

[0017] Hereinafter, an example of an implementation of the system involved in this disclosure will be described with reference to the accompanying drawings.

[0018] First, let's explain the terms used in the following description.

[0019] In the following embodiments, the processor (hereinafter referred to as "processor") can be a single computing device or a combination of multiple computing devices. Furthermore, a processor can be a single computing device or a combination of multiple computing devices. Examples of computing devices include CPU (Central Processing Unit), GPU (Graphics Processing Unit), GPGPU (General-Purpose computing on Graphics Processing Units), APU (Accelerated Processing Unit), or TPU (Tensor Processing Unit), etc.

[0020] In the following implementation, the labeled RAM (Random Access Memory) is a memory that temporarily stores information and is used by the processor as working memory.

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

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

[0023] In the following implementation, "A and / or B" is synonymous with "at least one of A and B". That is, "A and / or B" means that it can be only A, only B, or a combination of A and B. Furthermore, in this specification, when "and / or" connects more than three items, the same approach as "A and / or B" applies.

[0024] [First Implementation] Figure 1 An example of the configuration of the data processing system 10 according to the first embodiment is shown.

[0025] like Figure 1 As shown, the data processing system 10 includes a data processing device 12 and an intelligent device 14. An example of the data processing device 12 is a server.

[0026] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, RAM 30, and memory 32. The processor 28, RAM 30, and memory 32 are connected to a bus 34. Furthermore, the database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. An example of the network 54 includes a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0027] The smart device 14 includes a computer 36, a receiver 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, RAM 48, and memory 50. The processor 46, RAM 48, and memory 50 are connected to a bus 52. In addition, the receiver 38, output device 40, and camera 42 are also connected to the bus 52.

[0028] The receiving device 38 includes a touchscreen 38A and a microphone 38B, etc., for receiving user input. The touchscreen 38A receives user input generated by contact with an indicator (e.g., a pen or finger). The microphone 38B receives user input generated by sound by detecting the user's voice. The control unit 46A sends data representing user input received via the touchscreen 38A and microphone 38B to the data processing unit 12. In the data processing unit 12, a specific processing unit 290 (see...) Figure 2 Get the data that represents user input.

[0029] The output device 40 includes a display 40A and a speaker 40B, etc., and presents data to the user by outputting data in a user-perceptible form (e.g., sound and / or text). The display 40A displays visual information such as text and images according to instructions from the processor 46. The speaker 40B outputs sound 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 imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0030] Communication I / F 44 is connected to network 54. Communication I / F 44 and 26 are responsible for sending and receiving various information between processor 46 and processor 28 via network 54.

[0031] Figure 2 An example of the main functions of the data processing device 12 and the smart device 14 is shown.

[0032] like Figure 2 As shown, in the data processing apparatus 12, specific processing is performed by the processor 28. A specific processing program 56 is stored in the memory 32. The specific processing program 56 is an example of a "program" as understood in this disclosure. The processor 28 reads the specific processing program 56 from the memory 32 and executes the read specific processing program 56 on the RAM 30. Specific processing is implemented by the processor 28 operating as a specific processing unit 290 based on the specific processing program 56 executed on the RAM 30.

[0033] The memory 32 stores a data generation model 58 and an emotion-specific model 59. The data generation model 58 and the emotion-specific model 59 are used by the specific processing unit 290. The specific processing unit 290 can use the emotion-specific model 59 to infer the user's emotions and perform specific processing based on the user's emotions. The emotion inference function using the emotion-specific model 59 (emotion-specific function) includes various inferences and predictions related to the user's emotions, but is not limited to this example. Furthermore, emotion inference and prediction also include, for example, emotion analysis (analysis).

[0034] In the smart device 14, specific processing is performed by the processor 46. A specific processing program 60 is stored in the memory 50. The specific processing program 60 is used in conjunction with the data processing system 10. The processor 46 reads the specific processing program 60 from the memory 50 and executes the read specific processing program 60 on the RAM 48. Specific processing is implemented by the processor 46 operating as a control unit 46A based on the specific processing program 60 executed on the RAM 48. Furthermore, the smart device 14 may also have the same data generation model and emotion-specific model as the data generation model 58 and the emotion-specific model 59, and use these models to perform the same processing as the specific processing unit 290.

[0035] Furthermore, other devices besides the data processing device 12 may also have the data generation model 58. For example, a server device (e.g., a generation server) may have the data generation model 58. In this case, the data processing device 12 obtains the processing results (prediction results, etc.) using the data generation model 58 by communicating with the server device that has the data generation model 58. In addition, the data processing device 12 may be a server device or a user-owned terminal device (e.g., a mobile phone, robot, home appliance, etc.). Next, an example of the processing performed by the data processing system 10 of the first embodiment will be described.

[0036] (Example) The system described in this invention involves a user undergoing human genetic testing and uploading the results to a generative AI, which then regularly delivers dietary kits designed to reduce the risk of disease. The system involves the user undergoing human genetic testing, uploading the results to the generative AI, and inputting a simple questionnaire about the user's current health status and lifestyle habits. The system then proposes and regularly delivers dietary kits designed to reduce the risk of disease. For example, the user uses a genetic testing kit to collect saliva or blood samples and sends them to a testing institution. The testing institution analyzes the user's approximately 21,000 gene loci and provides the results. The user then uploads the genetic testing results to the generative AI. The generative AI analyzes the uploaded genetic testing results to understand the user's genetic information. In addition, the user must input a simple questionnaire about their current health status and lifestyle habits. The generative AI analyzes the questionnaire information to understand the user's lifestyle habits. The generative AI analyzes the uploaded genetic testing results and questionnaire information to determine the user's risk of disease. For example, if a specific gene increases the risk of disease, it can propose dietary or lifestyle improvement suggestions to reduce that risk. Finally, the generative AI proposes dietary kits to reduce the determined risk of disease and delivers them regularly. The dietary suite includes a nutritionally balanced diet based on the user's genetic information and lifestyle habits. This allows users to personalize their health management based on their own genetic information, thereby reducing the risk of disease. Specifically, the system uses human genetic testing results obtained by the user (e.g., CSV or JSON format data containing 22,000 SNP sequence information and allele information) as input data. Users input their health status and lifestyle habits (e.g., daily steps, sleep duration, smoking / drinking habits, eating frequency, stress level, etc.) in a questionnaire format. This data is preprocessed into numerical vectors or categorical data. As the generative AI, it can employ large-scale language models based on Transformers, multilayer perceptrons dedicated to genetic data analysis, or multimodal neural networks. The AI ​​model integrates genetic sequence information (e.g., allele vectors for each rsID, 0 / 1 / 2 numerical data), lifestyle habit vectors (e.g., sleep duration = 7.5, steps = 8000, alcohol consumption = 1, etc.), and past disease history or family history into an input tensor for processing. Based on these inputs, the AI ​​model outputs risk scores (probability values ​​between 0 and 1, such as 0.85, 0.32, etc.) for various diseases (e.g., type 2 diabetes, dyslipidemia, hypertension, etc.) and the contribution of each risk factor (e.g., gene A = 0.12, lifestyle habit B = 0.08, etc.). Output examples include "Type 2 diabetes risk 0.78, main causes: rs1234567 (T / T), high-fat diet, insufficient sleep" or "Hypertension risk 0.45, main causes: rs9876543 (C / G), insufficient exercise," etc.Based on these risk scores, the AI ​​model automatically generates dietary packages (such as low-carb menus, Omega-3 fortified diets, and low-sodium diets), and optimizes them by combining user preferences and allergy information. In subsequent processing, the AI ​​output is used for threshold determination (e.g., targeted intervention when risk ≥ 0.7), data linkage with the proposal department, and delivery scheduling instructions. In terms of technical effectiveness, this system differs from previous unified health guidance or manual dietary proposals. It can integrate and analyze massive amounts of genetic and lifestyle data in a high-dimensional space, automatically generating personalized optimal intervention measures in real time, significantly improving the accuracy and efficiency of health management. Furthermore, the AI ​​model uses cross-entropy or MSE as the loss function during training and optimizes weights through gradient descent, achieving continuous improvement in judgment accuracy. Applicable areas include personal health management services, insurance company risk assessment, corporate employee health support, and local government public health policies.

[0037] The system described in this embodiment includes a receiving department, an analysis department, a judgment department, and a proposal department. The receiving department uploads the genetic testing results received by the user. For example, the user uses a genetic testing kit to collect samples such as saliva or blood and sends them to a testing institution. The testing institution analyzes the user's approximately 21,000 gene loci and provides the test results. The receiving department uploads the genetic testing results received by the user to a generative AI. The analysis department analyzes the uploaded genetic testing results and the questionnaire information entered by the user. For example, the analysis department analyzes whether a specific gene increases the risk of disease. The analysis department also analyzes questionnaire information about the user's current health status and lifestyle habits to understand the user's lifestyle. The judgment department determines the risk of disease based on the information analyzed by the analysis department. For example, when a specific gene increases the risk of disease, the judgment department determines the dietary or lifestyle improvement measures needed to reduce that risk. The proposal department proposes dietary kits to reduce the risk of disease determined by the judgment department. For example, the proposal department can propose a nutritionally balanced diet based on the user's genetic information and lifestyle habits. The proposal department can also propose dietary kits that can be customized according to the user's preferences. Therefore, the system described in this embodiment can provide personalized health management based on the user's genetic information and reduce the risk of disease. Specifically, the system's receiving department receives genetic testing data (such as SNP sequences and allele information containing 22,000 loci) in CSV or JSON format from user terminals and stores it in a database. The receiving department performs data consistency verification (such as rsID duplication checks, missing value completion, and outlier detection) and ensures secure data transmission via encrypted communication. The parsing department vectorizes the received data, for example, generating allele vectors represented by 0 / 1 / 2 for each rsID, and numerical vectors from the lifestyle questionnaire (such as sleep duration = 7.0, steps = 9000, alcohol consumption = 0, etc.). The parsing department inputs these vectors into a large-scale language model based on Transformer, a multilayer perceptron, or a multimodal neural network. The AI ​​model integrates genetic and lifestyle information in a high-dimensional space, outputting risk scores for various diseases (such as type 2 diabetes risk 0.82, dyslipidemia risk 0.41, etc.) and the contribution of each risk factor (such as rs1234567 = 0.13, insufficient sleep = 0.09, etc.). The judgment department uses threshold judgments (e.g., ≥0.7 indicates high risk) or rule-based judgments based on combinations of risk factors (e.g., specific genotypes + high-fat diets increase risk). The judgment department can also apply multivariate logistic regression or decision tree algorithms that consider user attributes (age, gender, family history, etc.). The proposal department automatically generates dietary packages based on the judgment results (e.g., low-carb menus, Omega-3 fortified diets, low-salt diets, etc.), and customizes them by incorporating user preferences and allergy information. The proposal department can also implement food combination optimization algorithms (e.g., linear programming, genetic algorithms) and a recommendation system utilizing historical user satisfaction data.In terms of technical effectiveness, this system differs from previous manual, standardized health guidance or dietary proposals. It integrates and analyzes massive amounts of genetic and lifestyle data in a high-dimensional space, automatically generating personalized optimal intervention measures in real time, significantly improving the accuracy and efficiency of health management. The AI ​​model training employs cross-entropy loss or MSE, and uses gradient descent to optimize weights, achieving continuous improvement in judgment accuracy. Applicable areas include personal health management services, insurance company risk assessment, corporate employee health support, and local government public health policies.

[0038] The system includes a voice input unit for setting questionnaire input as voice input. The voice input unit enables users to input questionnaires via voice. For example, the voice input unit supports users inputting questionnaire answers by voice using a microphone. The voice input unit utilizes speech recognition technology to convert user speech into text data. For example, the voice input unit enables users to answer questions such as "Please tell me what I ate today" by voice. The voice input unit parses user speech in real time and automatically inputs questionnaire answers. Thus, the voice input unit allows users to easily input questionnaires. Specifically, this system, as a voice input unit, receives audio waveform data (such as 16kHz, 16-bit PCM format, single-channel timing array) from the user's terminal microphone as input data. The voice input unit first performs preprocessing such as noise suppression and volume normalization, and then converts the audio waveform into labeled sequences or string data using a speech recognition engine (such as an end-to-end speech recognition model based on Transformer or a CTC-based deep neural network). Input examples include natural language speech such as "I ate fish yesterday" and "I walked 8000 steps today". The speech recognition model extracts acoustic features (such as Mel spectrum, MFCC, etc.), learns temporal features at the encoder layer, and outputs text sequences at the decoder layer. The output data is structured as text answers to questionnaire items (such as "diet content = fish", "steps = 8000", etc.). After speech recognition, the natural language processing module extracts intent and entities (such as ingredient names, quantities, frequencies, etc.) and converts them into numerical vectors or categorical data. In subsequent processing, this data is linked with the parsing unit, integrated with genetic information and lifestyle information, and incorporated into the input tensor of the health risk assessment AI model. In terms of technical effectiveness, compared to traditional manual or form input, this system achieves high-speed and intuitive data acquisition through voice, significantly improving the user experience. By introducing speech recognition AI, it can flexibly handle diverse speech patterns and dialects, greatly improving input accuracy and convenience. Furthermore, by accumulating speech input history, applying customized acoustic models tailored to user speech tendencies, and implementing automatic error correction algorithms for misidentification, recognition accuracy can be continuously improved. Applicable fields include health management applications, on-site medical consultation support, elderly health services, and wearable device integration.

[0039] The system includes a visualization department for visualizing the analysis results. The visualization department displays the analysis results visually. For example, it can display the results as charts or graphs. The visualization department uses color differentiation to make the results easier for users to understand. For example, it displays high-risk genes in red and low-risk genes in green. The visualization department can also display the analysis results interactively. For example, it allows users to click on charts to display detailed information. Thus, the visualization department enables users to intuitively understand the analysis results. Specifically, this system, as the visualization department, receives health risk scores and gene contribution data (such as risk probability values ​​for each disease, contribution vectors for each gene locus, and weights of lifestyle factors) output from the analysis department or judgment department as input data. The visualization department uses graphics libraries such as D3.js and WebGL to convert these numerical data into various visualization formats such as bar charts, pie charts, heatmaps, and radar charts. Input examples include structured data such as "Type 2 diabetes risk = 0.78, main cause: rs1234567 (T / T), high-fat diet" or "Hypertension risk = 0.45, main cause: rs9876543 (C / G), insufficient exercise." The visualization department automatically adjusts hue and brightness based on the risk value, highlighting areas exceeding the threshold. It also supports user-clicking or touching chart elements to display detailed explanations of related genetic or lifestyle factors, as well as historical trend charts. Furthermore, the visualization department adaptively optimizes the layout based on user devices (smartphones, tablets, PCs, etc.) and screen size, and provides accessibility features such as color vision diversity and voice reading. In subsequent processing, the visualization department outputs feedback modules to improve user understanding and promote behavioral change, or integrates with medical personnel report generation functions. In terms of technical effectiveness, compared to traditional text-based analysis results, this system can visualize complex, high-dimensional data in an intuitive and interactive way, significantly improving user understanding and acceptance, and increasing the adoption rate of healthy behaviors. Meanwhile, through AI-powered automatic layout optimization and personalized displays based on user operation history, continuous improvement in user experience and enhanced analytical accuracy are achieved. Applicable areas include personal health management dashboards, patient instruction support in medical institutions, risk assessment reports for insurance companies, and public health monitoring by local governments.

[0040] The system includes an Improvement Department for refining proposal content based on user feedback. The Improvement Department collects user feedback and improves proposal content. For example, it analyzes user feedback to identify areas for improvement. The Improvement Department automatically updates proposal content based on user feedback. For instance, when a user provides feedback such as "I hope the meal kits will be more varied and diverse," the Improvement Department updates the proposal, suggesting more diverse meal kits. The Improvement Department accumulates user feedback for improving proposal content. Thus, the Improvement Department can provide proposal content based on user needs. Specifically, this system, acting as the Improvement Department, receives feedback data collected from users (such as free text, 5-point ratings, multiple-choice questionnaires, meal kit satisfaction scores, etc.) as input data. The Improvement Department uses natural language processing AI (such as a large language model based on Transformer) to extract keywords such as needs, dissatisfactions, and improvement points, as well as sentiment scores, from the free text feedback and structures the data. Input examples include text such as "Too few vegetables, hope to increase" or "I like more Japanese-style menus," or numerical data such as "Satisfaction = 3" or "Diversity = 2." The Improvement Department categorizes this data through clustering or topic modeling to identify high-frequency needs or dissatisfactions. Furthermore, by utilizing AI recommendation algorithms (such as collaborative filtering and content recommendation) and combining user attributes with historical feedback records, the system automatically adjusts the parameters generated by the proposal department for the dietary kits (such as ingredient diversity, cooking methods, and nutritional balance). In subsequent processing, the improvement department's output is linked to the proposal department, reflecting in the content of the next dietary kit or proposal text. In terms of technical effectiveness, this system differs from previous one-way proposals or manual feedback summaries. Through AI-automated analysis and real-time reflection, it can continuously provide personalized proposals that respond to users' individual needs and changing preferences, significantly improving user satisfaction and continued usage. In addition, the accumulation of feedback data increases the training data for the AI ​​model, enabling continuous improvement in proposal accuracy. Applicable areas include healthy food subscription services, personalized meal planning systems, corporate welfare support, and local government health promotion policies.

[0041] The receiving department can automatically upload gene testing results. It features an automatic upload function to reduce the need for users to manually upload gene testing results. For example, the receiving department automatically uploads the results to the generative AI when they are saved to the user's device. The receiving department receives a notification when the user receives the gene testing results and automatically begins uploading. This reduces the need for users to manually upload gene testing results, ensuring a smooth parsing process. Specifically, this system, acting as the receiving department, monitors the user terminal's file system or cloud storage, triggering events such as new saves or updates of gene testing result files (e.g., CSV, JSON, VCF, etc., containing 22,000 SNP sequence data). The receiving department automatically verifies the file content (e.g., file format consistency, rsID duplication / deletion checks, virus scanning, etc.) and securely uploads it to the server-side AI parsing platform via encrypted communication (e.g., TLS / SSL). Input examples include files such as "genome_result_20240601.csv" or "user123_snp.json". After the receiving department completes the upload, it sends a notification to the user terminal, automatically displaying progress and providing retransmission instructions in case of errors. In subsequent processing, the uploaded data is linked with the parsing department and integrated with data from lifestyle habit questionnaires or voice input, incorporated into the AI ​​model's input tensor. Technically, this system eliminates the risks of data loss and delays caused by traditional manual uploads or user errors. By automating the data acquisition and parsing process, it significantly reduces the user's burden and improves real-time performance and data quality. Furthermore, the upload history and error occurrences are accumulated as logs, which helps improve system reliability and facilitate rapid root cause identification in case of failures. Applicable areas include personal genetic testing services, automated data linkage in medical institutions, automated risk assessment for insurance companies, and local government health data collection platforms.

[0042] The analysis unit can analyze genetic and questionnaire information to understand users' lifestyle habits. By analyzing these data, the analysis unit gains a detailed understanding of users' lifestyle habits. For example, it can analyze users' dietary patterns, exercise habits, and sleep patterns. Based on users' genetic information, the analysis unit determines whether specific lifestyle habits increase disease risk. For example, it can analyze whether a user's high-fat diet is associated with specific gene mutations. Therefore, the analysis unit can gain a detailed understanding of users' lifestyle habits and provide appropriate health management recommendations. Specifically, this system, acting as the analysis unit, receives genetic testing data (such as 22,000 SNP sequence vectors, 0 / 1 / 2 representations of each rsID) and questionnaire information (such as numerical vectors or categorical data for dietary frequency, exercise volume, sleep duration, and drinking / smoking habits) as input data. The analysis unit preprocesses this data, filling in missing values ​​and standardizing it before inputting it into a multimodal neural network (such as an ensemble model using MLP for genetic information and Transformer for lifestyle habits). The AI ​​model integrates genetic and lifestyle information in a high-dimensional space to estimate the contribution of various lifestyle factors to disease risk, as well as the interaction effect between specific genotypes and lifestyle habits (e.g., increased risk when rs1234567 is T / T and consumes a high-fat diet). Input examples include "rs1234567=2, rs9876543=1, diet=high-fat, steps=8000, sleep=6.5h". The AI ​​model outputs risk scores for various diseases (e.g., type 2 diabetes risk=0.81, dyslipidemia risk=0.39), and contribution vectors for various risk factors (e.g., high-fat diet=0.12, rs1234567=0.09). In subsequent processing, these outputs are linked with the judgment or proposal department as data for personalized health management proposals. In terms of technical effectiveness, this system differs from traditional simple questionnaires or manual lifestyle habit assessments. By integrating and analyzing massive amounts of genetic data and diverse lifestyle habit data through AI, it can automatically generate optimal individualized risk assessments and intervention proposals in real time, significantly improving the accuracy and efficiency of health management. Furthermore, the AI ​​model training employs cross-entropy loss and gradient descent methods to continuously improve judgment accuracy. Applicable areas include personal health management services, insurance company risk assessment, corporate employee health support, and local government public health policies.

[0043] The proposal department can generate customized meal kits based on user preferences. For example, if a user prefers a specific ingredient, the department can propose a meal kit including that ingredient. If a user is allergic to a specific ingredient, the department can propose a meal kit excluding that ingredient. The department combines user preferences and allergy information to generate the optimal meal kit. This improves user satisfaction. Specifically, the system, acting as the proposal department, receives user preference data (such as a list of favorite and disliked ingredients, allergy information, and past meal kit ratings) as input. The department then inputs this information into a meal kit generation algorithm (such as linear programming-based nutritional balance optimization, genetic algorithm-based ingredient combination search, and collaborative filtering recommendation models) to automatically generate the optimal meal kit content for each user. Input examples include "favorite ingredients = chicken, tomato", "allergy = wheat", and "past ratings = 4 points for Japanese food, 2 points for Western food". The proposal department integrates with genetic information and lifestyle data, prioritizing ingredients and cooking methods that help reduce disease risk, and strictly reflecting preferences and allergy restrictions. Output examples include "a low-carb Japanese menu with chicken and tomato as the main ingredients" and "a wheat-free, gluten-free diet kit." In subsequent processing, the proposal content is linked to the delivery scheduler or user notification module for scheduled delivery or proposal text generation. In terms of technical effectiveness, this system differs from traditional standardized diet proposals or manual customization. Through AI multivariate optimization and recommendation algorithms, it achieves personalized optimal proposals that simultaneously meet users' preferences, limitations, and health risks. These proposals are generated automatically in real time, significantly improving user satisfaction and continued usage. Furthermore, it accumulates preference data and evaluation history as training data for the AI ​​model, continuously improving proposal accuracy. Applicable areas include personalized healthy food services, allergy response meal planning systems, corporate welfare support, and local government health promotion policies.

[0044] The receiving department can infer user emotions and adjust the timing of gene testing result uploads based on these inferences. For example, when a user is stressed, the receiving department prompts them to upload during a relaxing time; when a user is busy, it adjusts the upload time to be short; and when a user is relaxed, it provides detailed information and uploads the results. Emotion inference can be achieved through emotion inference functions such as emotion engines or generative AI. Generative AI can be text-generating AI (such as LLM) or multimodal generation AI, but is not limited to these. Therefore, the receiving department can upload gene testing results at the most suitable time for the user's emotions. Specifically, the receiving department receives audio data (such as 16kHz, 16bit PCM format audio waveforms), text input (such as natural language text like "I'm busy today" or "I'm relaxed now"), or facial images (such as facial expression images captured by a camera, 224×224 pixel RGB images) as input data for the emotion inference AI. The sentiment inference AI can employ a multimodal neural network combining Transformer models for speech recognition, CNNs for facial expression recognition, and large-scale language models for text sentiment classification. Input examples include "audio: speaking with a sigh," "text: very busy today," and "image: smiling." The AI ​​model extracts acoustic features (such as MFCC, pitch, and energy), text embedding vectors (such as 768-dimensional vectors based on BERT), and image feature maps (such as 512-dimensional vectors output by ResNet). The integration layer outputs sentiment categories (such as stress, relaxation, excitement, and fatigue) and sentiment scores (such as stress level 0.82 and relaxation level 0.15). Output examples include "stress = 0.75, relaxation = 0.10," "sentiment label = stress," and "sentiment label = relaxation." Based on these sentiment inference results, the processing department executes an upload timing control algorithm (such as notifying users at night or on holidays when stress is high, uploading immediately when relaxation is high, and displaying a one-click upload UI when busy). In subsequent processing, the upload timing decision is linked to the user terminal's notification scheduler, automatic switching of the upload UI, and the explanatory text generation module. In terms of technical effectiveness, this processing department differs from traditional unified upload timing prompts or user-initiated operations. Through AI-powered multimodal sentiment inference and dynamic timing control, it reduces user psychological burden and significantly improves upload completion rates and data quality. Furthermore, the sentiment inference AI training employs cross-entropy loss and data augmentation (such as audio noise addition and facial image rotation / brightness changes) to continuously improve recognition accuracy. Applicable areas include UX optimization for personal genetic testing services, patient stress management in medical institutions, data collection efficiency improvement for insurance companies, and support for local government health data collection.

[0045] The processing department can analyze a user's past upload history and select the optimal upload method when uploading gene testing results. For example, the processing department prioritizes recommending upload methods the user has used in the past (manual, voice input, etc.). The processing department selects the most efficient method from the user's past upload history. The processing department automatically selects methods that the user has used without difficulty in the past. Thus, the processing department can provide the user with the optimal upload method. Specifically, this processing department receives a user upload history database (such as structured data recording upload time, device type, communication method, UI selection history, number of errors, and time required) as input data. The AI ​​model can employ a historical analysis module combining LSTM or Transformer Encoder for time series analysis and clustering algorithms (such as k-means, DBSCAN). Input examples include historical records such as "2024 / 06 / 01 20:15 Manual PC successful", "2024 / 06 / 10 08:30 Voice input smartphone failed", and "2024 / 06 / 15 21:00 Automatic upload successful". The AI ​​model extracts features from these historical records (such as success rate, average time required, error frequency, and user operation tendencies) to calculate efficiency and reliability scores for each upload method (e.g., manual = 0.92, voice = 0.65, automatic = 0.98, etc.). Output examples include "Recommended method = automatic upload", "Recommended method = voice input", and "Recommended method = manual". Based on these AI outputs, the processing department automatically switches the user terminal UI (e.g., defaults to displaying the automatic upload UI, highlights the voice input button), or optimizes the order of upload method options. In subsequent processing, the selected upload method is linked with the upload execution module or user notification module, which helps optimize user experience and reduce errors. In terms of technical effectiveness, this service differs from traditional unified upload methods that rely on prompts or user-selected options. Through AI historical analysis and dynamic UI optimization, it automatically provides the optimal upload experience based on user proficiency and environment, significantly improving upload success rate and user satisfaction. Furthermore, the accumulation of historical data helps the AI ​​model continuously learn and improves the accuracy of personalized upload methods. Applicable areas include personal genetic testing services, data linkage support for medical institutions, automated risk assessment for insurance companies, and local government health data collection platforms.

[0046] The processing department can filter gene testing results based on the user's current health status and lifestyle habits during the upload process. For example, when the user is in good health, the processing department provides detailed upload steps; when the user is in poor health, it provides simplified upload steps. The processing department also recommends the optimal upload time based on the user's lifestyle habits. Therefore, the processing department can provide appropriate upload steps based on the user's current health status and lifestyle habits. Specifically, this processing department receives user-input health status questionnaires (such as numerical vectors or categorical data like body composition score, sleep duration, stress level, exercise volume, and food frequency) and lifestyle habit history (such as time-series data like steps taken in the past week, alcohol / smoking records, and medication use). The AI ​​model can employ a multilayer perceptron for health status classification and a temporal neural network (such as LSTM or GRU) for lifestyle habit pattern extraction. Input examples include "Body Condition = Good, Sleep = 7.5h, Steps = 9000, Alcohol Consumption = 0" and "Body Condition = Poor, Sleep = 4.5h, Steps = 2000, Alcohol Consumption = 2". The AI ​​model outputs health status labels (e.g., Good, Poor, Need Attention) and lifestyle habit clustering (e.g., Active, Low-Activity, Night-Occupancy), and executes an upload step selection algorithm (e.g., detailed steps for good health, one-click simplified steps for poor health, nighttime notification for night-occupancy). Output examples include "Steps = Detailed", "Steps = Simplified", "Timing = Nighttime". Based on these AI outputs, the processing department automatically adjusts the upload UI and notification timing on the user's terminal, minimizing user burden. In subsequent processing, the selected steps and timing are linked with the upload execution module or user notification module. In terms of technical effectiveness, this processing department differs from traditional unified upload step or timing prompts. Through AI health status and lifestyle habit analysis and dynamic step optimization, it achieves flexible data acquisition adapted to the user's body condition and lifestyle rhythm, significantly improving upload completion rate and user satisfaction. Furthermore, the accumulation of health and lifestyle data helps AI models continuously learn and optimize their processes, improving accuracy. Applicable areas include personal genetic testing services, reducing the burden on patients in healthcare institutions, supporting health data collection for insurance companies, and local government public health policies.

[0047] The receiving department can infer user emotions and determine the priority of uploaded genetic testing results based on the inferred emotions. For example, when a user is stressed, the receiving department prioritizes uploading important genetic testing results; when the user is relaxed, it uploads all genetic testing results sequentially; and when the user is in a hurry, it uploads only the most important genetic testing results. Emotion inference can be achieved through emotion inference functions such as emotion engines or generative AI. Generative AI can be text generation AI (such as LLM) or multimodal generation AI, but is not limited to these. Thus, the receiving department can upload genetic testing results with optimal priority based on user emotions. Specifically, this receiving department receives audio data (such as 16kHz, 16bit PCM format), text input (such as "I'm in a hurry today," "I'm relaxed now," etc.), and facial images (such as 224×224 pixel RGB images) obtained from the user's terminal as input data for the emotion inference AI. The emotion inference AI can employ multimodal neural networks of speech, facial expressions, and text (such as Transformer for speech, CNN for facial expressions, and an ensemble of large language models for text). Input examples include "Audio: Fast speaking speed," "Text: Very urgent," and "Image: Frowning." The AI ​​model extracts features from each modality and outputs the emotion category (e.g., stress, relaxation, urgency) and emotion score (e.g., stress level 0.80, urgency level 0.65). Output examples include "Emotion = Stress," "Emotion = Relaxation," and "Emotion = Urgent." The processing department combines these emotion inference results with the importance information of the genetic testing results (e.g., priority labels such as SNPs directly related to disease risk and SNPs related to lifestyle habits), and executes a priority determination algorithm (e.g., only upload high-importance items when stressed, all items when relaxed, and only the most important items when urgent). Output examples include "Priority list = [rs1234567, rs9876543]" and "Upload objects = top 5 items by importance." In subsequent processing, the determined priority is linked to the upload execution module or user notification module and reflected in the upload UI and progress display. In terms of technical effectiveness, this processing department differs from traditional methods that rely on a uniform upload order or user-selected uploads. Through AI sentiment inference and dynamic priority control, it reduces the psychological and time burden on users, enabling rapid acquisition of important data and improving upload completion rates. Furthermore, the sentiment inference AI and priority determination algorithms utilize cross-entropy loss and historical data during training to continuously improve accuracy. Applicable areas include personal genetic testing services, emergency data collection in medical institutions, automated risk assessment for insurance companies, and support for local government health data collection.

[0048] The processing department can consider the user's geographical location information when uploading genetic testing results, prioritizing the upload of highly relevant results. For example, when a user resides in a specific region, the processing department prioritizes uploading genetic testing results related to that region; when a user is traveling, it prioritizes uploading genetic testing results related to their current location; and when a user plans to move, it prioritizes uploading genetic testing results related to their new address. Thus, the processing department can upload the optimal genetic testing results based on the user's geographical location information. Specifically, the processing department receives geographical location information (such as latitude and longitude pairs, prefectural / municipal / town / village codes, country codes, etc.) obtained from the user's terminal GPS or IP address as input data. In addition, the genetic testing result data includes metadata for region-specific disease risk SNPs or environmental factor-related SNPs (such as hay fever risk, UV sensitivity, genes related to region-specific dietary habits, etc.). The AI ​​model can employ graph neural networks or rule-based region mapping algorithms to infer the correlation between geographical and genetic information. Input examples include "Location = Shinjuku Ward, Tokyo", "Location = Sapporo, Hokkaido", "Location = California, USA", etc. The AI ​​model calculates the regional relevance score between the current location or predetermined location and the genetic testing results (e.g., hay fever risk SNP = high, UV sensitivity SNP = moderate), and outputs a priority upload list (e.g., a list of top regionally relevant SNPs). Output examples include "Priority Upload = [rs1234567, rs2345678]" and "Regional Relevance = High", etc. Based on these AI outputs, the processing department automatically determines the data to be uploaded and the order of upload, optimizing the user terminal UI and notification content. In subsequent processing, the priority uploaded data is linked with the analysis or judgment department for regionally specific health risk assessment and intervention proposals. In terms of technical effectiveness, this processing department differs from traditional unified data uploads or data acquisition that ignores regional characteristics. By integrating geographic and genetic information through AI, it achieves priority acquisition of data on regionally specific health risks and environmental factors, significantly improving the accuracy and regional adaptability of health management. In addition, the accumulation of geographic information and genetic data helps the AI ​​model to continuously learn and enhance regionally specialized services. Applicable areas include regional optimization of personal genetic testing services, regional epidemiological support for medical institutions, regional risk assessment for insurance companies, and local government public health policies.

[0049] The processing department can analyze users' social media activities and upload relevant results when uploading genetic testing results. For example, the processing department can upload relevant genetic testing results based on health information shared by users on social media; based on health expert information followed by users on social media; and based on information from health communities participated in by users. Thus, the processing department can upload the optimal genetic testing results based on users' social media activities. Specifically, within the scope of user permission, this processing department receives social media posting data (such as health-related posting text, lists of followed accounts, and information from participating groups in JSON format) as input data. The AI ​​model can employ an analysis module combining large language models for natural language processing (such as those based on Transformer), graph neural networks for network analysis, and clustering algorithms (such as topic modeling and community detection). Input examples include "Post: Recently started losing weight," "Following: Diabetes experts," and "Groups: Hypertension countermeasures community." The AI ​​model extracts health-related topics (such as weight loss, diabetes, and hypertension) and keywords from the posting text, compares them with the information from followed and participated groups, and determines the relevant disease categories. Output examples include "Related Disease = Diabetes", "Related Disease = Hypertension", "Focus Topic = Weight Loss", etc. The processing department matches these AI outputs with disease-related metadata (such as diabetes risk SNPs, hypertension risk SNPs, etc.) from genetic testing results, extracting highly relevant test results as a priority upload list. Output examples include "Priority Upload = [rs1234567, rs2345678]", etc. In subsequent processing, the priority uploaded data is linked with the analysis or judgment department for risk assessment and intervention proposals based on users' health concerns. In terms of technical effectiveness, this processing department differs from traditional unified data uploads or data acquisition that ignores user concerns. Through AI social media activity analysis and correlation inference, it can respond in real time to users' latest health concerns and behavioral changes, achieving priority data acquisition and significantly improving the personalized accuracy of health management and user satisfaction. In addition, the accumulation of social media data helps the AI ​​model to continuously learn and track attention trends. Applicable areas include personalized enhancement of personal genetic testing services, patient attention analysis in medical institutions, enhanced risk assessment by insurance companies, and local government health promotion policies.

[0050] The parsing unit can infer user emotions and adjust its presentation based on these inferences. For example, it provides detailed analysis results when the user is relaxed, concise results when the user is stressed, and visually appealing results when the user is excited. Emotion inference can be achieved through emotion inference functions such as emotion engines or generative AI. Generative AI can be text generation AI (such as LLM) or multimodal generation AI, but is not limited to these. Thus, the parsing unit can provide analysis results in the most optimal way. Specifically, the parsing unit receives audio data (such as 16kHz, 16bit PCM format audio waveforms), text input (such as natural language text like "I'm relaxed today" or "I'm busy now"), or facial images (such as 224×224 pixel RGB images) from the user terminal as input data for the emotion inference AI. The parsing unit can employ a multimodal neural network combining a Transformer model for speech recognition, a CNN for facial expression recognition, and a large language model for text emotion classification. The parsing unit extracts acoustic features (such as MFCC, pitch, and energy), text embedding vectors (such as 768-dimensional vectors based on BERT), and image feature maps (such as 512-dimensional vectors output by ResNet). At the integration layer, it outputs sentiment categories (such as stress, relaxation, and excitement) and sentiment scores (such as stress level 0.82, relaxation level 0.15, etc.). Input examples include "audio: slow speech," "text: feeling relaxed today," and "image: smiling." The parsing unit infers AI output based on sentiment and dynamically switches the presentation of the parsing results. For example, a high relaxation level generates a report containing detailed numerical data, charts, and causal explanations; a high stress level provides a concise list of key points; and a high excitement level offers a richly colored and animated interactive visualization. The parsing unit also considers the user's browsing history and preferences when deciding on the presentation of the parsing results to achieve personalized UI / UX. The parsing unit can automatically generate various output formats such as text summaries, detailed reports, infographics, and dashboards. Output examples include "Type 2 diabetes risk = 0.78, main cause: rs1234567 (T / T), high-fat diet (with detailed explanation)" or "Hypertension risk = 0.45 (key points only)". In subsequent processing, the analysis department's output collaborates with the visualization or proposal department to provide feedback or intervention proposals optimized for the user's emotional state. In terms of technical effectiveness, unlike traditional unified analysis result displays or manual performance adjustments, the analysis department uses AI multimodal sentiment inference and dynamic performance optimization to automatically generate analysis results that adapt to the user's psychological state and context in real time, significantly improving user understanding, identification, and behavioral change rates. Furthermore, the sentiment inference AI training employs cross-entropy loss and data augmentation (such as audio noise addition, facial image rotation / brightness changes) to continuously improve recognition accuracy.Applicable areas include UX optimization for personal health management services, patient instruction support in medical institutions, risk assessment reports for insurance companies, and public health monitoring by local governments.

[0051] The parsing unit can adjust the level of detail in its analysis based on the importance of the genetic information. For example, it provides detailed analysis results for important genetic information and concise results for less important information. The unit prioritizes analyzing high-importance genetic information based on user focus. Thus, the parsing unit can provide detailed analysis of important information to the user. Specifically, the parsing unit receives genetic testing data (such as 22,000 locus SNP sequence vectors, 0 / 1 / 2 representations of each rsID) as input data. It assigns an importance score (such as evidence score based on literature, AI contribution estimate) to each locus or SNP, based on factors like disease risk or lifestyle relevance. The unit performs detailed statistical analysis (such as odds ratios, risk contribution, interaction analysis, etc.) and visualization (such as heatmaps, detailed charts) on high-importance genetic information, while outputting only summary values ​​or simple labels for low-importance genetic information. The analysis department prioritizes extracting high-importance gene information as analysis targets based on user interest areas (such as diabetes, hypertension, obesity, etc.) and past browsing history. The AI ​​model can combine multilayer perceptrons for gene information importance estimation and statistical models (such as logistic regression and decision trees) for disease risk contribution estimation. Input examples include "rs1234567=2 (importance=0.92)" and "rs9876543=1 (importance=0.15)". The analysis department automatically generates detailed and simplified analysis lists based on the importance scores. Output examples include "Detailed analysis: rs1234567 (contribution 0.12, including detailed description), Simplified analysis: rs9876543 (summary only)". In subsequent processing, the detailed analysis results are linked with the judgment or proposal department as data for personalized health management proposals or risk assessments. In terms of technical effectiveness, the analysis department differs from traditional unified genetic information analysis or manual importance determination. Through AI importance scoring and dynamic detail control, it can efficiently and accurately extract and analyze information that is truly important to users from massive genetic data, significantly improving the practicality of health management and user satisfaction. Furthermore, the accumulation of importance scores and analysis history helps the AI ​​model continuously learn and improve analysis accuracy. Applicable areas include automatic generation of personal genetic testing service reports, patient explanation support in medical institutions, risk assessment by insurance companies, and public health policies by local governments.

[0052] The parsing unit can apply different parsing algorithms based on the category of genetic information during parsing. For example, it applies a specific parsing algorithm to genetic information related to health risk; another algorithm to genetic information related to athletic ability; and yet another type of algorithm to genetic information related to dietary habits. Thus, the parsing unit can provide appropriate parsing results based on the category of genetic information. Specifically, the parsing unit receives genetic testing data (such as 22,000 SNP sequence vectors, 0 / 1 / 2 representations of each rsID) and category labels attached to each genetic information (such as health risk, athletic ability, dietary habits, etc.) as input data. The parsing unit automatically selects a different parsing algorithm for each category. For example, the health risk category uses multilayer perceptron or logistic regression models for disease risk estimation; the athletic ability category uses pattern matching or clustering algorithms for muscle strength / endurance-related genes; and the dietary habits category uses network analysis or recommendation algorithms for food preferences or metabolism-related genes. Input examples include "rs1234567=2 (Category = Health Risk)", "rs2345678=1 (Category = Athletic Ability)", and "rs3456789=0 (Category = Dietary Habits)". The parsing department integrates the parsing results for each category, and output examples include "Type 2 Diabetes Risk = 0.78 (Health Risk Algorithm)", "Muscle Strength Type = High (Athletic Ability Algorithm)", and "Lipid Metabolism Type = Low (Dietary Habits Algorithm)". In subsequent processing, the category parsing results are linked with the judgment or proposal department as data to support disease risk assessment or personalized diet / exercise proposals. In terms of technical effectiveness, the parsing department differs from traditional unified genetic information parsing or manual category determination. Through AI-automated category discrimination and algorithm optimization, it can achieve high-precision and high-efficiency parsing for the diverse characteristics of genetic information, significantly improving the accuracy of health management and user satisfaction. In addition, the accumulation of category parsing history and algorithm selection data helps the AI ​​model to continuously learn and improve parsing accuracy. Applicable areas include multi-purpose report generation for personal genetic testing services, patient explanation support in medical institutions, ability assessment in the sports field, and risk assessment for insurance companies.

[0053] The parsing unit can infer user emotions and adjust the parsing length based on the inferred emotions. For example, when the user is in a hurry, the parsing unit provides a concise and key-point-focused parsing result; when the user is relaxed, it provides a detailed parsing result; and when the user is excited, it provides a visually appealing parsing result. Emotion inference can be achieved through emotion inference functions such as emotion engines or generative AI. Generative AI can be text generation AI (such as LLM) or multimodal generation AI, but is not limited to these. Thus, the parsing unit can provide parsing results with optimal length. Specifically, the parsing unit receives audio data (such as 16kHz, 16bit PCM format), text input (such as "I'm in a hurry today," "I'm relaxed now," etc.), and facial images (such as 224×224 pixel RGB images) from the user terminal as input data for the emotion inference AI. The parsing unit can employ a multimodal neural network combining a Transformer model for speech recognition, a CNN for facial expression recognition, and a large language model for text emotion classification. The analysis department extracts acoustic features, text embedding vectors, and image feature maps, and outputs emotion categories (such as stress, relaxation, and urgency) and emotion scores (such as stress level 0.80, urgency level 0.65, etc.) at the integration layer. Input examples include "audio: fast speech," "text: very urgent," and "image: frowning." The analysis department dynamically adjusts the length of the analysis results based on the emotion inference AI output. For example, when the urgency level is high, it provides a concise list of key points; when the relaxation level is high, it generates an analysis report containing detailed numerical data, charts, and causal relationship explanations; and when the excitement level is high, it provides an interactive visualization analysis with rich colors and animations. When determining the length and presentation of the analysis results, the analysis department also considers the user's past browsing history and preferences to achieve personalized UI / UX. Output examples include "Type 2 diabetes risk = 0.78 (key points only)" and "Hypertension risk = 0.45 (including detailed explanation)." In subsequent processing, the analysis department's output is linked with the visualization department or proposal department for feedback or intervention proposals optimized for the user's emotional state. In terms of technical effectiveness, unlike traditional methods of displaying uniform parsing results or manually adjusting length, the parsing department utilizes AI-powered multimodal sentiment inference and dynamic length optimization to automatically generate parsing results that adapt to the user's psychological state and context in real time, significantly improving user comprehension, identification, and behavioral change rates. Furthermore, the sentiment inference AI training employs cross-entropy loss and data augmentation to continuously improve recognition accuracy. Applicable areas include UX optimization for personal health management services, patient explanation support in medical institutions, risk assessment reports for insurance companies, and public health monitoring by local governments.

[0054] The parsing department can determine the parsing priority based on the submission time of the gene information. For example, the parsing department prioritizes parsing the most recently submitted gene information, while processing older submissions later. The parsing department adjusts the parsing schedule based on the submission time. Thus, the parsing department can prioritize parsing the latest information and provide it to the user. Specifically, the parsing department receives metadata gene data as input data, including the submission date or timestamp information of the gene detection results (such as ISO8601 format dates, UNIX timestamps, etc.). The parsing department automatically calculates a priority score based on the submission time, assigning high priority to the latest data and low priority to older data. Based on the priority score, the parsing department automatically determines the parsing order using a parsing job scheduling algorithm (such as priority queues, round-robin, FIFO, etc.). Input examples include "Submission Date = 2024-06-20 10:00" and "Submission Date = 2024-05-15 09:30". The analysis department prioritizes analyzing data submitted at the latest time, with output examples including "Priority Analysis: Data from June 20, 2024" and "Delayed Processing: Data from May 15, 2024." In subsequent processing, the priority analysis results are instantly linked with the judgment or proposal department for rapid user feedback or intervention proposals. In terms of technical effectiveness, unlike traditional unified analysis sequences or manual scheduling, the analysis department uses AI-driven submission timing analysis and dynamic priority control to achieve rapid analysis of the latest data and timely information provision to users, significantly improving the real-time nature of health management and user satisfaction. Furthermore, the accumulation of submission timing data helps the AI ​​model continuously learn and improve scheduling accuracy. Applicable areas include real-time report generation for personal genetic testing services, emergency data analysis in medical institutions, automated risk assessment for insurance companies, and local government health data collection platforms.

[0055] The parsing department can adjust the parsing order based on the relevance of gene information during parsing. For example, the parsing department prioritizes parsing gene information with high relevance, while processing gene information with low relevance is delayed. The parsing department adjusts the parsing schedule based on the relevance of gene information. Thus, the parsing department can prioritize parsing information with high relevance and provide it to the user. Specifically, the parsing department uses gene detection data (e.g., SNP sequence vectors of 22,000 loci) and relevance scores for each gene (e.g., disease risk contribution, lifestyle relevance, network centrality indicators, etc.) as input data. The parsing department uses graph neural networks or clustering algorithms (e.g., community detection, topic modeling) for relevance inference to quantify the relevance between gene information. Input examples include "rs1234567=2 (relevance=0.91)" and "rs9876543=1 (relevance=0.12)". The parsing department prioritizes adding gene information with high relevance scores to the parsing list, while processing information with low relevance is delayed. The parsing department executes parsing order determination algorithms (e.g., descending order of relevance scores, priority queue) to automatically schedule parsing jobs. Output examples include "Priority Parsing: rs1234567" and "Delayed Processing: rs9876543". In subsequent processing, the results of priority parsing are linked to the judgment and proposal departments, serving as the basis for disease risk assessment and individual optimal intervention proposals. The technical benefits are that, unlike previous uniform parsing orders or manual relevance determination, the parsing department, through AI-driven relevance inference and dynamic order optimization, can efficiently and accurately extract and parse information truly important to users from massive amounts of genetic data, thereby significantly improving the practicality of health management and user satisfaction. Furthermore, through the accumulation of relevance scores and parsing history, the AI ​​model can continuously learn and improve parsing accuracy. Applicable areas include automatic generation of personal genetic testing service reports, patient instruction support in medical institutions, risk assessment by insurance companies, and public health policies of local governments.

[0056] The judgment unit can infer the user's emotions and adjust the judgment criteria for disease risk based on the inferred user emotions. For example, when the user is relaxed, the judgment unit applies detailed judgment criteria; when the user is stressed, it applies concise judgment criteria; and when the user is excited, it applies visually appealing judgment criteria. Emotion inference can be achieved through emotion inference functions such as emotion engines or generative AI. Generative AI can be text-generated AI (such as LLM) or multimodal generative AI, but is not limited to these. Thus, the judgment unit can determine the disease risk using the judgment criteria most suitable for the user's emotions. Specifically, this judgment unit uses user genetic testing data received from the parsing unit (such as SNP sequence vectors of 22,000 loci, 0 / 1 / 2 expression of each rsID), lifestyle habit vectors (such as sleep time, step count, drinking / smoking habits, etc.), and the output of the emotion inference AI (such as stress level 0.82, relaxation level 0.15, excitement level 0.03, etc.) as input data. This decision-making unit can employ a multimodal neural network as the AI ​​for sentiment inference, combining a Transformer model for speech recognition, a CNN for facial expression recognition, and a large-scale language model for text sentiment classification. Input examples include "speech: speaking slowly," "text: today is peaceful," and "image: smiling." The sentiment inference AI extracts acoustic features (such as MFCC, pitch, and energy), text embedding vectors (such as a 768-dimensional vector based on BERT), and image feature maps (such as a 512-dimensional vector output by ResNet), and outputs the sentiment category and sentiment score at the ensemble layer. The decision-making unit dynamically switches the parameters and thresholds of the disease risk assessment algorithm based on the sentiment score. For example, when the relaxation level is high, a detailed statistical model (such as multivariate logistic regression, Bayesian inference, risk contribution decomposition, etc.) is applied in the disease risk score calculation to analyze the contribution and causal relationship of explanatory variables in detail. When the stress level is high, only the main risk factors are extracted, and a simple rule-based judgment is applied (such as a risk score ≥0.7 for high risk, <0.3 for low risk, etc.). When the excitement level is high, the judgment results are presented in an interactive visual form such as color and animation. When switching judgment criteria, the judgment department also considers the user's past judgment history and preferences to achieve a personalized judgment experience. The judgment department's output is presented in a structured manner as risk labels for each disease (such as "Type 2 diabetes risk = High", "Hypertension risk = Moderate"), risk scores (such as 0.78, 0.45, etc.), and judgment criteria (such as rs1234567 (T / T), high-fat diet, insufficient sleep, etc.). Output examples include "Type 2 diabetes risk = 0.78 (with detailed explanation)" and "Hypertension risk = 0.45 (key points only)". In subsequent processing, the judgment department's output is linked to the proposal department and visualization department to generate intervention proposals and feedback most suitable for the user's emotional state.The technical advantages are as follows: Unlike previous methods that relied on standardized or manually adjusted judgment criteria, this system utilizes AI-driven multimodal sentiment inference and dynamic judgment criterion optimization to automatically generate judgment results that align with the user's psychological state and context in real time. This significantly improves user comprehension, acceptance, and behavioral change rates. Furthermore, the training of the sentiment inference AI and judgment algorithm can employ cross-entropy loss and data augmentation (such as adding noise to speech, rotating / changing the brightness of facial expression images) to continuously improve recognition and judgment accuracy. Applicable areas include UX optimization for personalized health management services, patient instruction support in medical institutions, risk assessment reports for insurance companies, and public health monitoring by local governments.

[0057] The decision-making unit can improve accuracy by considering the interrelationships of gene information during the decision-making process. For example, the decision-making unit analyzes the interrelationships of gene information to accurately determine the risk of disease. It identifies high-risk genes based on these interrelationships and excludes low-risk genes. Thus, the decision-making unit can improve the accuracy of disease risk determination by considering the interrelationships of gene information. Specifically, this decision-making unit uses gene detection data received from the analysis unit (such as SNP sequence vectors of 22,000 loci, and 0 / 1 / 2 expression of each rsID) as input data. This decision-making unit can use graph neural networks (GNNs), Bayesian networks, or correlation matrix analysis algorithms to infer the interrelationships between genes. Input examples include SNP vectors such as "rs1234567=2, rs2345678=1, rs3456789=0". The decision-making unit uses these gene information as nodes and known biological pathways or disease-related network information as edges to construct a graph structure. The Genetic Neural Network (GNN) propagates and aggregates features from each node (such as allele type, contribution, and disease relevance score), learning higher-order interactions between nodes (such as epistatic interactions and pathway dependence). The decision-making unit uses the GNN output to obtain risk scores for each disease (e.g., type 2 diabetes risk = 0.81, dyslipidemia risk = 0.39, etc.) and contribution vectors for each risk factor (e.g., rs1234567 = 0.12, rs2345678 = 0.09, etc.). The decision-making unit prioritizes gene clusters with high interaction scores as risk assessment targets, excluding genes with weak interactions or low risk contribution. Output examples include "Risk assessment targets: rs1234567, rs2345678 (strong interaction)" and "Exclusion: rs3456789 (low contribution)," etc. In subsequent processing, the decision results are linked to the proposal and visualization departments, serving as the basis for individual optimized intervention proposals or risk descriptions. The technological advantages are as follows: Unlike previous methods that relied solely on single gene evaluation or manual cross-relationship determination, this system utilizes AI-powered high-dimensional graph structure analysis and dynamic optimization to efficiently and accurately extract and determine factors that fundamentally contribute to disease risk from complex gene networks. This significantly improves the practicality and accuracy of health management. Furthermore, the accumulation of cross-relationship scores and historical determination data allows for continuous learning and accuracy enhancement of the AI ​​model. Applicable areas include automated risk assessment for individual genetic testing services, patient information support in healthcare institutions, risk assessment by insurance companies, and public health policies in local governments.

[0058] The judgment unit is able to consider the attribute information of the gene information submitter when making a judgment. For example, the judgment unit determines the risk of disease based on the submitter's age. The judgment unit determines the risk of disease based on the submitter's gender. The judgment unit determines the risk of disease based on the submitter's lifestyle habits. Therefore, the judgment unit can make a more accurate judgment of disease risk by considering the submitter's attribute information. Specifically, this judgment unit uses gene testing data received from the parsing unit (such as SNP sequence vectors of 22,000 loci) and user attribute information (such as age, gender, family history, and lifestyle habit vectors (sleep time, step count, drinking / smoking habits, etc.)) as input data. This judgment unit integrates this attribute information into a multivariate logistic regression model, a decision tree algorithm, or a deep neural network using attribute embedding for judgment processing. Input examples include "age=45, gender=female, rs1234567=2, step count=8000, drinking=1", etc. The assessment department assigns weights to each attribute. For example, older age increases the risk contribution to specific diseases, reflects gender differences in disease susceptibility, and incorporates lifestyle factors (such as insufficient exercise and smoking habits) into the risk score calculation. The assessment department also considers the interaction effects of attribute information and genetic information (e.g., specific genotype × advanced age × smoking leads to increased risk), constructing a complex risk model. Output examples include "Type 2 diabetes risk = 0.78 (reflecting age and lifestyle habits)" and "Hypertension risk = 0.45 (reflecting gender and family history)." In subsequent processing, the assessment results are linked to the proposal and visualization departments, serving as the basis for individual optimized intervention proposals or risk descriptions. The technical benefits are that, unlike previous unified assessments or manual attribute considerations, this assessment department, through AI-powered multivariate analysis and dynamic assessment optimization, can automatically generate high-precision risk assessments that match individual user attributes in real time, significantly improving the practicality of health management and user satisfaction. Furthermore, through the accumulation of attribute data and assessment history, the AI ​​model can continuously learn and improve assessment accuracy. Applicable areas include personalized enhancement of genetic testing services for individuals, patient instruction support in medical institutions, risk assessment by insurance companies, and public health policies of local governments.

[0059] The judgment unit can infer the user's emotions and adjust the display order of the judgment results based on the inferred emotions. For example, when the user is relaxed, the judgment unit prioritizes displaying detailed judgment results; when the user is stressed, it prioritizes displaying concise judgment results; and when the user is excited, it prioritizes displaying visually appealing judgment results. Emotion inference can be achieved through emotion inference functions such as an emotion engine or generative AI. Generative AI can be text-generated AI (such as LLM) or multimodal generative AI, but is not limited to these. Thus, the judgment unit can display the judgment results in the order most suitable for the user's emotions. Specifically, this judgment unit uses judgment result data received from the parsing unit (such as risk scores for various diseases, risk factor labels, detailed description text, etc.) and the output of the emotion inference AI (such as stress level 0.80, relaxation level 0.15, excitement level 0.05, etc.) as input data. This judgment unit can employ a multimodal neural network as the emotion inference AI, combined with a Transformer model for speech recognition, a CNN for facial expression recognition, and a large-scale language model for text emotion classification. Input examples include "voice: fast speaking speed," "text: very anxious," and "image: frowning." The judgment department dynamically adjusts the display order of judgment results based on the emotion score, determining the algorithm (e.g., descending order of detail score, priority of key points, priority of visuals, etc.). For example, when the relaxation level is high, detailed judgment results (such as risk scores, causal relationship explanations, charts, etc.) are displayed first; when the stress level is high, only key point lists are displayed first; when the excitement level is high, interactive judgment results such as colors and animations are placed first. The judgment department also refers to the user's past browsing history and preferences to achieve a personalized display order. Output examples include "display order: detailed → key points → visual" and "display order: key points → detailed → visual." In subsequent processing, the display order is linked to the visualization department or user interface to generate feedback or intervention proposals most suitable for the user's emotional state. The technical effect is that, unlike the previous uniform display order or manual order adjustment, this judgment department, through AI's multimodal emotion inference and dynamic display order optimization, can automatically generate judgment results that match the user's psychological state and context in real time, significantly improving user understanding, identification, and behavior change rate. Furthermore, the training of sentiment inference AI and display order determination algorithms can employ cross-entropy loss and data augmentation to achieve continuous improvement in recognition accuracy and user experience. Applicable areas include UX optimization for personalized health management services, patient instruction support in healthcare institutions, risk assessment reports for insurance companies, and public health monitoring by local governments.

[0060] The determination unit is able to make judgments considering the geographical distribution of genetic information. For example, the determination unit determines the risk of disease based on genetic information related to a specific region. The determination unit identifies high-risk regions based on geographical distribution. The determination unit excludes low-risk regions considering geographical distribution. Thus, the determination unit can perform region-related risk determinations. Specifically, this determination unit uses genetic testing data received from the parsing unit (such as SNP sequence vectors of 22,000 loci) and the user's geographical location information (such as latitude and longitude pairs, prefectural / municipal / township codes, country codes, etc.) as input data. This determination unit can employ graph neural networks or rule-based region mapping algorithms for inferring the correlation between genetic and geographical information. Input examples include "Location = Shinjuku Ward, Tokyo", "Location = Sapporo, Hokkaido", "Location = California, USA", etc. The assessment department utilizes metadata from region-specific disease risk SNPs and environmental factor-related SNPs (such as hay fever risk, UV sensitivity, and genes related to region-specific dietary habits) to calculate the regional relevance score between the current location or predetermined location and the gene testing results (e.g., hay fever risk SNP = high, UV sensitivity SNP = moderate). The assessment department prioritizes gene information with high regional relevance as the risk assessment target, excluding regions or gene information with low relevance. Output examples include "Risk Assessment Target: rs1234567 (High Regional Relevance)" and "Exclusion: rs2345678 (Low Regional Relevance)". In subsequent processing, the assessment results are linked to the proposal and visualization departments for region-specific health risk assessment or intervention proposals. The technical benefits are that this assessment department differs from previous risk assessments that uniformly assessed or ignored regional characteristics. Through AI-integrated analysis of geographic and genetic information, it achieves high-precision assessments adapted to region-specific health risks and environmental factors, significantly improving the regional adaptability of health management and user satisfaction. Furthermore, the accumulation of geographic information and genetic data enables continuous learning of the AI ​​model and the refinement of region-specific services. Applicable areas include regional optimization of genetic testing services for individuals, regional epidemiological support for medical institutions, regional risk assessment for insurance companies, and public health policies of local governments.

[0061] The judgment unit can improve the accuracy of judgment by referring to relevant literature on gene information during the judgment process. For example, the judgment unit accurately determines the risk of disease by referring to relevant literature. The judgment unit identifies high-risk genes based on relevant literature. The judgment unit excludes low-risk genes by referring to relevant literature. Thus, the judgment unit can improve the accuracy of disease risk determination by referring to relevant literature. Specifically, this judgment unit uses gene detection data received from the analysis unit (such as SNP sequence vectors of 22,000 loci) and relevant literature databases associated with each gene information (such as PubMed ID, evidence level, paper abstract text, etc.) as input data. This judgment unit can use large-scale language models for natural language processing (such as Transformer-based models) or literature evidence scoring algorithms to conduct a literature-based evaluation of the correlation between each gene SNP and disease risk. Input examples include "rs1234567=2 (PubMed:12345678, Evidence=High)" and "rs2345678=1 (PubMed:23456789, Evidence=Low)". The judgment department extracts disease risk-related keywords and evidence scores from the literature abstract text, reflecting them in the risk judgment algorithm. The judgment department prioritizes genetic information with high evidence scores as risk judgment targets, excluding genetic information with insufficient evidence. Output examples include "Risk judgment target: rs1234567 (High evidence)" and "Excluded: rs2345678 (Low evidence)". In subsequent processing, the judgment results are linked to the proposal department and visualization department for evidence-based risk assessments or intervention proposals. The technical effect is that, unlike previous unified judgments or manual literature references, this judgment department, through AI-powered automatic literature parsing and evidence scoring, can automatically generate high-precision risk assessments based on scientific evidence in real time, significantly improving the reliability of health management and user acceptance. Furthermore, by accumulating literature data and historical judgment data, AI models can continuously learn and improve their judgment accuracy. Applicable areas include personal genetic testing services with evidence-based report generation, patient instruction support in medical institutions, risk assessment by insurance companies, and public health policies of local governments.

[0062] The proposal department can infer user emotions and adjust the presentation of proposals accordingly. For example, it provides detailed proposals when the user is relaxed, concise proposals when the user is stressed, and visually appealing proposals when the user is excited. Emotion inference can be achieved through emotion inference engines or generative AI, but is not limited to text-generating AI (such as LLM) or multimodal generative AI. This allows the proposal department to provide proposals in a way that best suits the user's emotions. Specifically, the proposal department uses audio data (such as 16kHz, 16bit PCM format audio waveforms), text input (such as natural language sentences like "Today is peaceful" or "I'm busy now"), and facial images (such as 224×224 pixel RGB images) obtained from the user's terminal as input data for the emotion inference AI. The proposal department can employ a multimodal neural network as the emotion inference AI, combined with a Transformer model for speech recognition, a CNN for facial expression recognition, and a large-scale language model for text emotion classification. The proposal department extracts acoustic features (such as MFCC, pitch, and energy), text embedding vectors (such as 768-dimensional vectors based on BERT), and image feature maps (such as 512-dimensional vectors output by ResNet). At the integration layer, it outputs sentiment categories (such as stress, relaxation, and excitement) and sentiment scores (such as stress level 0.82, relaxation level 0.15, etc.). Input examples include "voice: speaking slowly," "text: today is peaceful," and "image: smiling." Based on the output of the sentiment inference AI, the proposal department dynamically switches the presentation of the proposal content. For example, when the relaxation level is high, it generates a proposal report containing a detailed nutritional information table, cooking steps, ingredient origin information, and scientific explanations; when the stress level is high, it only provides key points in a bullet-point format; when the excitement level is high, it provides interactive visual proposals with colors and animations. The proposal department also considers the user's browsing history and preferences when deciding on the presentation of the proposal content to achieve personalized UI / UX. The proposal department can automatically generate proposal result output formats such as text summaries, detailed reports, infographics, and dashboards. Output examples include "A low-sugar Japanese-style menu featuring chicken and tomato as the main ingredients (with detailed description)" and "A wheat-free, gluten-free diet kit (key points only)." In subsequent processing, the proposal department's output is linked to the delivery scheduler or user notification module to generate feedback or intervention proposals best suited to the user's emotional state. The technical effect is that, unlike previous uniform proposal content prompts or manual adjustments, this proposal department, through AI's multimodal sentiment inference and dynamic performance optimization, can automatically generate proposal content that matches the user's psychological state and context in real time, significantly improving user comprehension, acceptance, and behavioral change rates. Furthermore, the sentiment inference AI can be trained using cross-entropy loss and data augmentation (such as adding noise to speech, rotating / changing the brightness of facial expression images) to achieve continuous improvement in recognition accuracy.Applicable areas include UX optimization for personal health management services, patient instruction support for medical institutions, risk assessment reports for insurance companies, and public health monitoring by local governments.

[0063] The proposal department can adjust the level of detail in proposals based on the importance of the dietary packages. For example, the department provides detailed proposals for important dietary packages and concise proposals for less important ones. The department also prioritizes proposals for high-importance dietary packages based on user interests. This allows the department to provide detailed proposals with essential information to users. Specifically, the proposal department uses importance scores assigned to each dietary package (such as contribution to disease risk reduction, nutritional balance score, and consistency with user health goals) and user interest areas (such as blood sugar control, allergy management, and muscle strengthening) as input data. This data is then fed into a multilayer perceptron, decision tree algorithm, or importance scoring AI model to automatically determine the level of detail (e.g., detailed, standard, simple) for each dietary package. Input examples include "Dietary Package A (Importance = 0.92, Interest = High)" and "Dietary Package B (Importance = 0.15, Interest = Low)". This proposal department generates detailed reports for high-importance diet kits, including detailed nutritional information, cooking steps, scientific evidence, ingredient origin information, and explanations of disease risk reduction effects. For low-importance diet kits, only key points are provided. Furthermore, it considers users' past selection history and satisfaction ratings, prioritizing areas of high interest for detailed proposals. Output examples include "Detailed Proposal: Low-Sugar Japanese Style Menu (with detailed description)" and "Simplified Proposal: Gluten-Free Diet Kit (Key Points Only)." In subsequent processing, the proposal content is linked to the delivery scheduler or user notification module, helping to optimize user experience and improve satisfaction. Technically, unlike previous methods that relied on uniform proposal detail or manual importance assessment, this proposal department uses AI-powered importance scoring and dynamic detail control to efficiently and accurately extract and propose information truly important to users, significantly improving the practicality of health management and user satisfaction. Moreover, the accumulation of importance scores and proposal history allows for continuous learning of the AI ​​model and improvement in proposal accuracy. Applicable areas include personalized health food subscription services, patient diet proposals in medical institutions, health intervention projects for insurance companies, and local government health promotion policies.

[0064] The proposal department can apply different proposal algorithms based on the category of the diet kit when making proposals. For example, the department applies a specific algorithm to diet kits related to health risks, another algorithm to diet kits related to athletic ability, and yet another algorithm to diet kits related to dietary habits. Thus, the department can provide appropriate proposals based on the category of the diet kit. Specifically, the proposal department uses the category labels assigned to each diet kit (such as reduced health risk, improved athletic ability, improved dietary habits, etc.) and user health goals and lifestyle data as input data. The department automatically selects different proposal algorithms for each category. For example, the health risk category uses linear programming or multivariate optimization algorithms to maximize the reduction of disease risk; the athletic ability category uses food combination search or pattern matching algorithms that help enhance muscle strength / endurance; and the dietary habits category uses recommendation systems or network analysis algorithms that reflect preferences / allergy limitations. Input examples include "Diet Kit A (Category = Health Risk)", "Diet Kit B (Category = Athletic Ability)", "Diet Kit C (Category = Dietary Habits)", etc. This proposal department integrates proposal results from various categories, outputting examples such as "Low-sugar Japanese-style menu (health risk algorithm)," "High-protein muscle-strengthening menu (exercise ability algorithm)," and "Gluten-free diet kit (dietary habit algorithm)." In subsequent processing, the proposal results by category are linked to the delivery scheduler or user notification module, serving as data for individual optimized diet proposals or health intervention plans. The technical benefits are that, unlike previous unified proposal algorithms or manual category determination, this proposal department uses AI for automatic category identification and algorithm optimization, enabling highly accurate and efficient proposals adapted to the diverse characteristics of diet kits, thereby significantly improving the accuracy of health management and user satisfaction. Furthermore, by accumulating historical proposal data and algorithm selection data for each category, the AI ​​model can continuously learn and improve proposal accuracy. Applicable areas include multi-objective proposals for individual health food services, dietary guidance for patients in medical institutions, ability-enhancing diet proposals in the sports field, and health intervention projects for insurance companies.

[0065] The proposal department can infer user emotions and adjust the proposal length accordingly. For example, when a user is anxious, the department provides a concise and to-the-point proposal; when the user is relaxed, a detailed proposal; and when the user is excited, a visually appealing proposal. Emotion inference can be achieved through emotion inference functions such as an emotion engine or generative AI. Generative AI can be text-generated AI (such as LLM) or multimodal generative AI, but is not limited to these. Thus, the proposal department can provide proposals of the most suitable length for the user's emotion. Specifically, this proposal department uses audio data (such as 16kHz, 16bit PCM format), text input (such as "I'm in a hurry today," "I'm calm now," etc.), and facial images (such as 224×224 pixel RGB images) obtained from the user's terminal as input data for the emotion inference AI. This proposal department can employ a multimodal neural network combining a Transformer model for speech recognition, a CNN for facial expression recognition, and a large-scale language model for text emotion classification. The proposal department extracts acoustic features, text embedding vectors, and image feature maps, and outputs emotion categories (such as stress, relaxation, and anxiety) and emotion scores (such as stress level 0.80, anxiety level 0.65, etc.) at the integration layer. Input examples include "voice: fast speaking speed," "text: very anxious," and "image: frowning." Based on the output of the emotion inference AI, the proposal department dynamically adjusts the length of the proposal content. For example, when the anxiety level is high, only key points are presented; when the relaxation level is high, a proposal report containing a detailed nutritional information table, cooking steps, and scientific explanations is generated; when the excitement level is high, interactive visual proposals with colors and animations are provided. The proposal department also considers the user's past browsing history and preferences when determining the length and presentation of the proposal content to achieve personalized UI / UX. Output examples include "low-sugar Japanese-style menu (key points only)" and "gluten-free diet kit (with detailed description)." In subsequent processing, the output of the proposal department is linked to the delivery scheduler or user notification module to generate feedback or intervention proposals most suitable for the user's emotional state. The technical benefits are significant. Unlike previous approaches that relied on standardized proposal content prompts or manual length adjustments, this proposal system utilizes AI-powered multimodal sentiment inference and dynamic length optimization to automatically generate proposal content that aligns with the user's psychological state and context in real time. This dramatically improves user comprehension, engagement, and behavioral change rates. Furthermore, the sentiment inference AI can be trained using cross-entropy loss and data augmentation to continuously enhance recognition accuracy. Applicable areas include UX optimization for personalized health management services, patient instruction support in healthcare institutions, risk assessment reports for insurance companies, and public health monitoring by local governments.

[0066] The proposal department can determine the priority of proposals based on the submission time of the diet kits. For example, the proposal department prioritizes the most recently submitted diet kits, while older submissions are processed later. The proposal department adjusts the proposal scheduling based on the submission time. This allows the proposal department to prioritize and provide users with the latest information. Specifically, the proposal department uses diet kit data with metadata, including the submission date or timestamp information (such as ISO8601 format dates, UNIX timestamps, etc.), as input data. The proposal department automatically calculates a priority score for each submission time, assigning high priority to the latest data and low priority to older data. Based on the priority score, the proposal department automatically determines the proposal order using a proposal scheduling algorithm (such as priority queue, round-robin, FIFO, etc.). Input examples include "Submission Date = 2024-06-20 10:00" and "Submission Date = 2024-05-15 09:30". This proposal department prioritizes the submission of new dietary kits, with output examples including "Priority Proposal: Dietary Kit for 2024-06-20" and "Delayed Processing: Dietary Kit for 2024-05-15." In subsequent processing, the content of priority proposals will be instantly linked to the delivery scheduler or user notification module for rapid feedback or intervention. The technical benefits are that, unlike previous uniform proposal order or manual scheduling, this proposal department, through AI-powered submission timing analysis and dynamic priority control, achieves rapid submission of the latest data and timely information provision to users, significantly improving the real-time nature of health management and user satisfaction. Furthermore, the accumulation of submission timing data enables continuous learning of the AI ​​model and improvement of scheduling accuracy. Applicable areas include real-time proposals for individual health food subscription services, patient dietary guidance in medical institutions, health intervention projects of insurance companies, and health promotion policies of local governments.

[0067] The proposal department can adjust the proposal order based on the relevance of dietary packages during the proposal process. For example, the department prioritizes proposing dietary packages with high relevance, while delaying the processing of those with low relevance. The proposal department adjusts the proposal scheduling based on the relevance of dietary packages. This allows the department to prioritize and provide users with information that is highly relevant. Specifically, the proposal department uses the relevance scores of each dietary package (such as contribution to disease risk reduction, consistency with user health goals, past satisfaction scores, etc.) and user health status and lifestyle data as input data. The department uses multilayer perceptron or clustering algorithms (such as k-means, topic modeling) to quantify the relevance between dietary packages. Input examples include "Dietary Package A (relevance = 0.91)" and "Dietary Package B (relevance = 0.12)". The department prioritizes adding dietary packages with high relevance scores to the proposal list, while delaying the processing of those with low relevance. The proposal department executes a proposal order determination algorithm (such as descending order of relevance scores, priority queue) to automate proposal scheduling. Output examples include "Priority Proposal: Low-Sugar Japanese Style Menu" and "Deferred Processing: Gluten-Free Diet Kit." In subsequent processing, the content of priority proposals is linked to the delivery scheduler or user notification module, serving as the basis for individual optimized dietary proposals or health intervention plans. The technical benefits are that, unlike previous uniform proposal ordering or manual relevance judgments, this proposal department, through AI relevance inference and dynamic order optimization, can efficiently and accurately extract and propose information that is truly important to users, thereby significantly improving the practicality of health management and user satisfaction. Furthermore, through the accumulation of relevance scores and proposal history, the AI ​​model can continuously learn and improve proposal accuracy. Applicable areas include personalized enhancement of health food subscription services for individuals, patient dietary guidance in medical institutions, health intervention projects of insurance companies, and health promotion policies of local governments.

[0068] The voice input unit can infer the user's emotions and adjust the timing of voice input based on the inferred emotions. For example, when the user is relaxed, the voice input unit prompts detailed voice input; when the user is stressed, it prompts concise voice input; and when the user is anxious, it prompts rapid voice input. Emotion inference can be achieved through emotion inference functions such as an emotion engine or generative AI. Generative AI can be text-generated AI (such as LLM) or multimodal generative AI, but is not limited to these. Thus, the voice input unit can perform voice input at the most suitable time for the user's emotions. Specifically, this voice input unit uses audio data (such as 16kHz, 16bit PCM format audio waveforms), text input (such as natural language sentences like "I am calm now" or "I am anxious"), and facial images (such as 224×224 pixel RGB images) obtained from the user terminal as input data for the emotion inference AI. This voice input unit can employ a multimodal neural network as the emotion inference AI, combined with a Transformer model for speech recognition, a CNN for facial expression recognition, and a large-scale language model for text emotion classification. Input examples include "voice: speak slowly," "text: I'm relaxed today," and "image: smile." This voice input unit extracts acoustic features (such as MFCC, pitch, and energy), text embedding vectors (such as a 768-dimensional vector based on BERT), and image feature maps (such as a 512-dimensional vector output by ResNet). At the ensemble layer, it outputs emotion categories (such as stress, relaxation, and anxiety) and emotion scores (such as stress level 0.80 and relaxation level 0.15). Output examples include "emotion = relaxed," "emotion = stress," and "emotion = anxious." Based on these emotion inferences, this voice input unit executes a voice input timing control algorithm (such as detailed input prompts for relaxation, simplified input prompts for stress, and one-click input UI prompts for anxiety). In subsequent processing, the determined timing and input prompts are linked to the user terminal's UI or notification module, helping to optimize user experience and improve input completion rate. The technical benefits are as follows: Unlike previous methods that relied on uniform input timing prompts or user-defined actions, this voice input system utilizes AI-powered multimodal emotion inference and dynamic timing control. This reduces the psychological and time burden on users, significantly improving voice input completion rates and data quality. Furthermore, the emotion inference AI can be trained using cross-entropy loss and data augmentation (such as adding noise to speech, rotating / changing the brightness of facial expression images) to continuously improve recognition accuracy. Applicable areas include UX optimization for personalized health management services, patient input support in medical institutions, data collection efficiency for insurance companies, and health data collection support for local governments.

[0069] The voice input unit analyzes the user's past voice input history during voice input and selects the optimal input method. For example, it prioritizes suggesting voice input methods the user has used in the past; it selects the most efficient method from the user's past voice input history; and it automatically selects methods that the user has previously been able to input smoothly via voice. Thus, the voice input unit can provide the user with the optimal voice input method. Specifically, this voice input unit uses each user's individual voice input history database (such as structured data like input date, device type, input method (e.g., continuous voice, keyword input, voice commands), number of errors, and time required) as input data. This voice input unit can employ a historical analysis module combining LSTM or Transformer Encoder for time-series analysis and clustering algorithms (such as k-means, DBSCAN). Input examples include historical records such as "2024 / 06 / 01 20:15 Continuous voice input successful", "2024 / 06 / 10 08:30 Keyword input failed", and "2024 / 06 / 15 21:00 Voice command successful". This voice input unit extracts features from these historical records (such as success rate, average time required, error frequency, and user operation tendencies) to calculate efficiency and confidence scores for each input method (e.g., continuous voice = 0.92, keyword = 0.65, voice command = 0.98, etc.). Output examples include "Recommended method = Voice command", "Recommended method = Continuous voice", and "Recommended method = Keyword". Based on these AI outputs, this voice input unit automatically switches the user terminal's UI (e.g., defaulting to displaying the voice command UI, highlighting the continuous voice button) or optimizes the input method selection order. In subsequent processing, the selected input method is linked to the input execution module or user notification module, helping to optimize user experience and reduce errors. The technical benefits are significant. Unlike previous standardized input method prompts or user-selected methods, this voice input system, through AI-powered historical analysis and dynamic UI optimization, automatically provides users with the most suitable voice input experience based on their proficiency and the environment, greatly improving input success rate and user satisfaction. Furthermore, the accumulation of historical data enables continuous learning of the AI ​​model and personalized improvement in input method accuracy. Applicable areas include personal health management services, data integration support for medical institutions, automated risk assessment for insurance companies, and the foundation for local government health data collection.

[0070] The voice input unit can infer the user's emotions and determine the priority of voice input based on the inferred emotions. For example, when the user is stressed, the voice input unit prioritizes important voice input; when the user is relaxed, it processes all voice input sequentially; and when the user is anxious, it processes only the most important voice input. Emotion inference can be achieved through emotion inference functions such as an emotion engine or generative AI. Generative AI can be text-generated AI (such as LLM) or multimodal generative AI, but is not limited to these. Thus, the voice input unit can prioritize voice input according to the most suitable emotion for the user. Specifically, this voice input unit uses audio data (such as 16kHz, 16bit PCM format), text input (such as "I'm anxious today," "I'm calm now," etc.), and facial images (such as 224×224 pixel RGB images) obtained from the user terminal as input data for the emotion inference AI. This voice input unit can employ multimodal neural networks of speech, facial expressions, and text (such as Transformer for speech, CNN for facial expressions, and an ensemble of large-scale language models for text). Input examples include "voice: fast speech," "text: very anxious," and "image: frowning." This voice input unit extracts features from each modality and outputs emotion categories (such as stress, relaxation, anxiety, etc.) and emotion scores (such as stress level 0.80, anxiety level 0.65, etc.). Output examples include "emotion = stress," "emotion = relaxation," and "emotion = anxiety." This voice input unit combines these emotion inference results with the importance information of the voice input items (such as priority tags for health status reports, medication records, and lifestyle changes), and executes a priority determination algorithm (e.g., only the most important items are considered when stressed, all items when relaxed, and only the most important items when anxious). Output examples include "priority list = [health status report, medication record]" and "input object = top 2 items by importance." In subsequent processing, the determined priorities are linked to the input execution module or user notification module, reflected in the input UI or progress display. The technical benefits are as follows: Unlike previous voice input methods that relied on a uniform input order or user-selected choices, this system, through AI-powered sentiment inference and dynamic priority control, reduces the psychological and time burden on users, enabling rapid acquisition of important data and improving input completion rates. Furthermore, the training of the sentiment inference AI and priority determination algorithm can utilize cross-entropy loss and historical data to achieve continuous accuracy improvement. Applicable areas include personal health management services, emergency data collection in medical institutions, automated risk assessment for insurance companies, and support for health data collection by local governments.

[0071] The voice input unit takes into account the user's geographic location information during voice input, prioritizing highly relevant inputs. For example, when the user is located in a specific region, the voice input unit prioritizes voice input related to that region; when the user is traveling, it prioritizes voice input related to the current location; and when the user plans to move, it prioritizes voice input related to the new address. Thus, the voice input unit can provide optimal voice input based on the user's geographic location information. Specifically, this voice input unit uses geographic location information (such as latitude and longitude pairs, prefectural / municipal / town / village codes, country codes, etc.) obtained from the user's terminal's GPS or IP address as input data. In addition, voice input items include metadata related to region-specific health information or environmental factors (such as hay fever symptom reports, UV protection, reports of region-specific dietary habits, etc.). This voice input unit can employ graph neural networks or rule-based region mapping algorithms to infer the relevance between geographic information and input items. Input examples include "Location = Shinjuku Ward, Tokyo", "Location = Sapporo, Hokkaido", and "Location = California, USA". This voice input unit calculates the regional relevance score between the current location or predetermined location and the voice input item (e.g., hay fever report = high, UV protection = moderate), and outputs a priority input list (e.g., a list of top-relevant inputs based on regional relevance). Output examples include "Priority Input = [Hay Fever Symptom Report, UV Protection]" and "Regional Relevance = High". Based on these AI outputs, the voice input unit automatically determines the input data and order, optimizing the user terminal's UI and notification content. In subsequent processing, the priority input data is linked to the analysis and judgment units for regionally specific health risk assessments or intervention proposals. The technical benefits are that, unlike previous uniform data inputs or data collection that ignored regional characteristics, this voice input unit, through AI's integrated analysis of geographic information and input items, achieves priority data collection adapted to regionally specific health risks and environmental factors, significantly improving the accuracy and regional adaptability of health management. Furthermore, through the accumulation of geographic information and input data, continuous learning of the AI ​​model and the refinement of regionally specialized services can be achieved. Applicable areas include regional optimization of health management services for individuals, regional epidemiological support for medical institutions, regional risk assessment for insurance companies, and public health policies of local governments.

[0072] The visualization department can infer user emotions and adjust the display method of the visualization based on the inferred user emotions. For example, the visualization department provides detailed visualizations when the user is relaxed; concise visualizations when the user is stressed; and visually appealing visualizations when the user is excited. Emotion inference can be achieved through emotion inference functions such as emotion engines or generative AI. Generative AI can be text-generated AI (such as LLM) or multimodal generative AI, but is not limited to these. Thus, the visualization department can provide visualizations in a display method most suitable for the user's emotions. Specifically, this visualization department uses audio data (such as 16kHz, 16bit PCM format audio waveforms), text input (such as natural language sentences like "Today is peaceful" or "I'm busy now"), and facial images (such as 224×224 pixel RGB images) obtained from the user's terminal as input data for the emotion inference AI. This visualization department can employ a multimodal neural network as the emotion inference AI, combined with a Transformer model for speech recognition, a CNN for facial expression recognition, and a large-scale language model for text emotion classification. The visualization department extracts acoustic features (such as MFCC, pitch, and energy), text embedding vectors (such as 768-dimensional vectors based on BERT), and image feature maps (such as 512-dimensional vectors output by ResNet). It then outputs sentiment categories (such as stress, relaxation, and excitement) and sentiment scores (such as stress level 0.82 and relaxation level 0.15) at the ensemble layer. Input examples include "voice: speaking slowly," "text: today is peaceful," and "image: smiling." Based on the output of the sentiment inference AI, the visualization department dynamically switches the visualization display method. For example, when the relaxation level is high, it generates a dashboard containing detailed charts, infographics, and causal explanations; when the stress level is high, it only provides key point bullet points; and when the excitement level is high, it provides interactive visual displays such as colors and animations. The visualization department also considers the user's past browsing history and preferences when deciding on the visualization display method to achieve personalized UI / UX. The visualization department can automatically generate visualization output formats such as text summaries, detailed charts, infographics, and dashboards. Output examples include "Type 2 diabetes risk = 0.78, main cause: rs1234567 (T / T), high-fat diet (with detailed charts)" and "Hypertension risk = 0.45 (key points only)". In subsequent processing, the visualization department's output is linked to the user interface or proposal department to generate feedback or intervention proposals best suited to the user's emotional state. The technical effect is that, unlike previous uniform chart displays or manual adjustments, this visualization department, through AI's multimodal sentiment inference and dynamic display optimization, can automatically generate visualization results in real time that match the user's psychological state and context, significantly improving user understanding, identification, and behavioral change rates.Furthermore, the training of sentiment inference AI can employ cross-entropy loss and data augmentation (such as adding noise to speech, rotating / changing the brightness of facial expression images) to achieve continuous improvement in recognition accuracy. Applicable areas include UX optimization for personalized health management services, patient instruction support in healthcare institutions, risk assessment reports for insurance companies, and public health monitoring by local governments.

[0073] The visualization department can reference users' past visualization history to select the optimal display method during visualization. For example, the visualization department prioritizes suggesting visualization methods previously used by the user; it selects the most efficient method from the user's past visualization history; and it automatically selects methods that the user has previously been able to visualize successfully. Thus, the visualization department can provide users with the optimal visualization method. Specifically, this visualization department uses users' individual visualization history databases (such as display dates, device types used, display methods (e.g., charts, infographics, dashboards, etc.), user operation history, satisfaction ratings, and other structured data) as input data. This visualization department can employ a historical analysis module combining LSTM or Transformer Encoder for time series analysis and clustering algorithms (such as k-means, DBSCAN). Input examples include historical records such as "2024 / 06 / 01 20:15 Chart displayed successfully", "2024 / 06 / 10 08:30 Infographic displayed unsuccessfully", and "2024 / 06 / 15 21:00 Dashboard displayed successfully". This visualization department extracts features from these historical records (such as display success rate, average browsing time, user operation preferences, and satisfaction ratings) to calculate efficiency and confidence scores for each display method (e.g., Chart = 0.92, Infographic = 0.65, Dashboard = 0.98, etc.). Output examples include "Recommended method = Dashboard", "Recommended method = Chart", and "Recommended method = Infographic". Based on these AI outputs, this visualization department automatically switches the user terminal's UI (e.g., defaulting to displaying the dashboard UI, highlighting chart buttons) or optimizes the order of display method selection. In subsequent processing, the selected display method is linked to the display execution module or user notification module, helping to optimize user experience and improve comprehension. The technical benefits are significant. Unlike previous uniform display methods or user-selectable options, this visualization system, through AI historical analysis and dynamic UI optimization, automatically provides users with a visualization experience best suited to their operational proficiency and environment, greatly improving display success rate and user satisfaction. Furthermore, by accumulating historical data, the AI ​​model can continuously learn and personalize the display method to enhance accuracy. Applicable areas include personal health management services, data visualization support for medical institutions, risk assessment reports for insurance companies, and public health monitoring by local governments.

[0074] The visualization department can infer user emotions and determine the priority of visualizations based on these inferences. For example, when a user is stressed, the visualization department prioritizes important visualizations; when the user is relaxed, all visualizations are performed sequentially; and when the user is anxious, only the most important visualizations are performed. Emotion inference can be achieved through emotion engines or generative AI, but is not limited to text-generating AI (such as LLM) or multimodal generative AI. Thus, the visualization department can perform visualizations with the most suitable priority for the user's emotions. Specifically, this visualization department uses audio data (such as 16kHz, 16bit PCM format), text input (such as "I'm anxious today," "I'm calm now"), and facial images (such as 224×224 pixel RGB images) obtained from the user's terminal as input data for the emotion inference AI. This visualization department can employ multimodal neural networks (such as Transformer for speech, CNN for facial expressions, and an ensemble of large-scale language models for text) to perform these visualizations. Input examples include "voice: fast speech," "text: very anxious," and "image: frowning." This visualization department extracts features from each modality and outputs emotion categories (such as stress, relaxation, anxiety, etc.) and emotion scores (such as stress level 0.80, anxiety level 0.65, etc.). Output examples include "emotion = stress," "emotion = relaxation," and "emotion = anxiety." This visualization department combines these emotion inference results with the importance information of the visualization items (such as priority labels for disease risk visualization and lifestyle change maps), and executes a priority determination algorithm (e.g., only the most important items are considered when stressed, all items when relaxed, and only the most important items when anxious). Output examples include "priority list = [disease risk map, lifestyle change]" and "display objects = top 2 items by importance." In subsequent processing, the determined priorities are linked to the display execution module or user notification module, reflected in the display UI or progress display. The technical effect is that, unlike previous uniform display orders or user-selected methods, this visualization department, through AI emotion inference and dynamic priority control, can reduce the psychological and time burden on users, achieving rapid visualization and improved understanding of important data. Furthermore, the training of sentiment inference AI and prioritization algorithms can utilize cross-entropy loss and historical data to achieve continuous accuracy improvements. Applicable areas include personalized health management services, emergency data visualization in healthcare institutions, automated risk assessment for insurance companies, and health data monitoring by local governments.

[0075] The visualization department considers the user's device information and selects the optimal display method during visualization. For example, when a user is using a smartphone, the visualization department provides a display method suitable for the screen size; when a user is using a tablet, it provides a display method optimized for large screens; and when a user is using a smartwatch, it provides a simple and highly visual display method. Thus, the visualization department can provide the optimal display method based on the user's device information. Specifically, this visualization department receives device information obtained from the user's terminal (e.g., structured data such as device type, screen resolution, operating system version, and input interface type) as input data. This visualization department can employ rule-based algorithms, decision tree models, or device-adaptive UI generation AI for device information parsing. Input examples include: "Device = Smartphone, Resolution = 1080x2400", "Device = Tablet, Resolution = 2048x1536", "Device = Smartwatch, Resolution = 390x390", etc. This visualization department automatically selects the display method based on device type and screen size (e.g., vertical charts for smartphones, multi-panel dashboards for tablets, icon + numerical display for smartwatches), and optimizes the layout, font size, and interaction methods of UI components. Output examples include: "Display method = charts for smartphones," "Display method = dashboards for tablets," and "Display method = icons for smartwatches." In subsequent processing, the selected display method will be linked with the UI rendering module or user notification module, helping to optimize user experience and improve visibility. In terms of technical effectiveness, this visualization department differs from previous uniform display methods or manual device determination. Through AI-based device information analysis and dynamic UI optimization, it can automatically provide the optimal visualization experience for the user's environment, significantly improving display success rate and user satisfaction. In addition, by accumulating device information and display history, the AI ​​model can continuously learn and personalize the accuracy of the display method. Applicable areas include multi-device support for personal health management services, data visualization support for medical institutions, risk assessment reports for insurance companies, and public health monitoring by local governments.

[0076] The Improvement Department can infer user emotions and adjust improvement methods accordingly. For example, when a user is relaxed, the Improvement Department provides detailed improvement methods; when a user is stressed, it provides concise improvement methods; and when a user is excited, it provides visually appealing improvement methods. Emotion inference can be achieved through emotion inference functions such as emotion engines or generative AI. Generative AI can be text-generated AI (such as LLM) or multimodal generative AI, but is not limited to these. Thus, the Improvement Department can improve in a way that best suits the user's emotions. Specifically, this Improvement Department will use voice data (such as 16kHz, 16bitPCM format audio waveforms), text input (such as natural language sentences like "Today is peaceful" or "I'm busy now"), and facial images (such as 224×224 pixel RGB images) obtained from the user's terminal as input data for the emotion inference AI. This Improvement Department can employ a multimodal neural network, combining a Transformer model for speech recognition, a CNN for facial expression recognition, and a large language model for text emotion classification. This improvement department extracts audio features (such as MFCC, pitch, and energy), text embedding vectors (such as 768-dimensional vectors based on BERT), and image feature maps (such as 512-dimensional vectors output by ResNet). It then outputs sentiment categories (such as stress, relaxation, and excitement) and sentiment scores (such as stress level 0.82, relaxation level 0.15, etc.) at the fusion layer. Input examples include: "Voice: Speak slowly," "Text: Today is peaceful," and "Image: Smile." Based on the output of the sentiment inference AI, this improvement department dynamically switches the performance and content of the improvement methods. For example, when the relaxation level is high, it generates an improvement report containing detailed improvement steps, scientific basis, and examples; when the stress level is high, it only provides a list of key points; when the excitement level is high, it provides interactive visual improvement suggestions using color and animation. When deciding on improvement methods, this improvement department also refers to users' historical feedback records and preferences to achieve personalized UI / UX. This improvement department can automatically generate improvement result output formats such as text summaries, detailed reports, infographics, and dashboards. Output examples include: "Sleep habit improvement plan (with detailed explanation)" and "Exercise habit improvement (key points only)." In subsequent processing, the improvement department's output is linked with the user notification module or progress management module to optimize feedback and intervention suggestions based on the user's emotional state. In terms of technical effectiveness, this improvement department differs from previous approaches that relied on uniform improvement content prompts or manual adjustments. By using AI for multimodal emotion inference and dynamic performance optimization, it can automatically generate improvement content that aligns with the user's psychological state and context in real time, significantly improving user comprehension, acceptance, and behavioral change rates. Furthermore, the emotion inference AI can be trained using cross-entropy loss and data augmentation (such as adding noise to audio, rotating / changing the brightness of facial expression images) to continuously improve recognition accuracy.Applicable areas include UX optimization for personal health management services, patient instruction support in medical institutions, health intervention programs for insurance companies, and health promotion policies of local governments.

[0077] The Improvement Department can refer to users' past feedback history to select the optimal improvement method during the improvement process. For example, the Improvement Department proposes the optimal improvement method based on past user feedback. The Improvement Department selects the most effective improvement method from users' past feedback history. The Improvement Department automatically selects improvement methods that users have previously been able to accept successfully. Therefore, the Improvement Department can provide users with the optimal improvement method. Specifically, this Improvement Department receives each user's feedback history database (such as structured data recording improvement proposal ID, implementation date, user rating score, implementation completion status, relapse rate, satisfaction comments, etc.) as input data. This Improvement Department can employ a historical analysis AI module that combines temporal neural networks (such as LSTM, Transformer Encoder), clustering algorithms (such as k-means, DBSCAN), and large-scale language models for natural language processing of feedback content. This Improvement Department extracts features from historical data (such as the success rate of each improvement method, average satisfaction score, relapse rate, degree of change in user behavior, and the positive / negative judgment of feedback text). Based on these characteristics, this improvement department calculates the efficiency and reliability scores for each improvement method (e.g., sleep habit improvement = 0.92, exercise habit improvement = 0.85, diet habit improvement = 0.78, etc.). Input examples include historical records such as: "2024 / 06 / 01 Sleep habit improvement success satisfaction = 5", "2024 / 06 / 10 Exercise habit improvement failure satisfaction = 2", and "2024 / 06 / 15 Diet habit improvement success satisfaction = 4". As AI output, this improvement department generates "Recommended improvement method = sleep habit improvement", "Recommended improvement method = diet habit improvement", etc., prioritizing improvement methods with high past user satisfaction. Furthermore, this improvement department also combines user attribute information (such as age, gender, and lifestyle habit vectors), current health status, and recent feedback content to achieve personalized automatic selection of improvement methods. In subsequent processing, the selected improvement methods are linked with the user notification module or progress management module, which helps optimize user experience and improve behavior change rates. In terms of technical effectiveness, this improvement department differs from previous uniform improvement method prompts or manual historical references. By utilizing AI for historical analysis and dynamic optimization of improvement methods, it can automatically generate optimal improvement suggestions in real time that align with each user's behavioral tendencies and past acceptance levels, significantly improving implementation rates, satisfaction, and health outcomes. Furthermore, by accumulating historical data and improvement results, the AI ​​model can continuously learn and personalize the accuracy of improvement methods. Applicable areas include behavioral change support for individualized health management services, utilization of patient guidance history in medical institutions, optimization of health intervention programs by insurance companies, and personalization of public health policies by local governments.

[0078] The Improvement Department can infer user emotions and determine the priority of improvements based on the inferred emotions. For example, when a user is stressed, the Improvement Department prioritizes important improvements; when the user is relaxed, all improvements are implemented sequentially; and when the user is anxious, only the most important improvements are implemented. Emotion inference can be achieved through emotion inference functions such as an emotion engine or generative AI. Generative AI can be text-generated AI (such as LLM) or multimodal generative AI, but is not limited to these. Thus, the Improvement Department can prioritize improvements based on the most suitable user emotions. Specifically, this Improvement Department will use voice data (such as 16kHz, 16bit PCM format), text input (such as "I'm anxious today," "I'm calm now," etc.), and facial images (such as 224×224 pixel RGB images) obtained from the user's terminal as input data for the emotion inference AI. This Improvement Department can employ a multimodal neural network, combining a Transformer model for speech recognition, a CNN for facial expression recognition, and a large language model for text emotion classification. This improvement department extracts audio features (such as MFCC, pitch, and energy), text embedding vectors (such as 768-dimensional vectors based on BERT), and image feature maps (such as 512-dimensional vectors output by ResNet). It then outputs sentiment categories (such as stress, relaxation, and anxiety) and sentiment scores (such as stress level 0.80 and anxiety level 0.65) at the fusion layer. Input examples include: "voice: fast speech," "text: anxious," and "image: frowning." This improvement department combines the output of the sentiment inference AI with the importance information of each improvement item (such as sleep habit improvement = 0.92, exercise habit improvement = 0.85, and dietary habit improvement = 0.78), and executes a priority decision algorithm (such as only high-priority items when stressed, all items when relaxed, and only the most important items when anxious). As the AI ​​output, this improvement department generates "priority list = [sleep habit improvement, exercise habit improvement]" and "implementation targets = top 2 items by importance," automatically determining the priority of improvement suggestions that match the user's psychological state and context. In subsequent processing, the determined priority is linked to the user notification module or progress management module, reflected in the improvement implementation UI and progress display. In terms of technical effectiveness, this improvement department differs from previous uniform improvement sequences or manual priority adjustments. By using AI for multimodal sentiment inference and dynamic priority control, it reduces the psychological and time burden on users, enabling rapid implementation of important improvement projects and increasing behavior change rates. Furthermore, the training of the sentiment inference AI and priority decision-making algorithm can utilize cross-entropy loss and historical data to achieve continuous accuracy improvement. Applicable areas include behavior change support for individual health management services, patient guidance prioritization in medical institutions, optimization of health intervention projects by insurance companies, and emergency responses to public health policies by local governments.

[0079] The Improvement Department can consider the user's geographic location information and select the optimal improvement method during the improvement process. For example, when a user is located in a specific region, the Improvement Department prioritizes improvement methods related to that region; when a user is traveling, it prioritizes improvement methods related to their current location; and when a user plans to move, it prioritizes improvement methods related to their new address. Thus, the Improvement Department can provide the optimal improvement method based on the user's geographic location information. Specifically, this Improvement Department receives geographic location information (such as latitude and longitude pairs, prefectural / municipal / town / village codes, country codes, etc.) obtained from the user's terminal's GPS or IP address as input data. In addition, improvement projects include metadata related to improvements related to region-specific health issues or environmental factors (such as hay fever countermeasures, UV countermeasures, improvements to region-specific dietary habits, etc.). This Improvement Department can employ graph neural networks or rule-based region mapping algorithms for inferring the correlation between geographic information and improvement projects. Input examples include: "Location = Shinjuku Ward, Tokyo", "Location = Sapporo, Hokkaido", "Location = California, USA", etc. This improvement department calculates the regional relevance score between the current location or planned location and the improvement project (e.g., hay fever countermeasures = high, UV countermeasures = medium), and outputs a priority improvement list (e.g., a list of improvements with high regional relevance). Output examples include: "Priority Improvement = [Hydroxy Fever Countermeasures, UV Countermeasures]" and "Regional Relevance = High". Based on these AI outputs, this improvement department automatically determines the data and order of improvement targets, optimizing the UI and notification content on user terminals. In subsequent processing, the priority improvement data is linked with the user notification module or progress management module for regionally specific health risk assessment and intervention recommendations. In terms of technical effectiveness, this improvement department differs from previous uniform improvement methods that either suggest or ignore regional interventions. By using AI to comprehensively analyze geographic information and improvement projects, it achieves improvement recommendations that conform to regionally specific health risks and environmental factors, significantly improving the accuracy and regional adaptability of health management. Furthermore, by accumulating geographic information and improvement data, continuous learning of the AI ​​model and the refinement of regionally specialized services can be achieved. Applicable areas include regional optimization of health management services for individuals, regional epidemiological support for medical institutions, regional risk assessment for insurance companies, and public health policies of local governments.

[0080] The system described in this implementation is not limited to the examples above. For instance, various modifications can be made. Specifically, in the modules of gene detection data analysis, judgment, proposal, visualization, and improvement, this system can address the types and algorithm structures of AI models, data flow, input / output specifications, user interface design, database structure, communication protocols, security methods, cloud linkage methods, distributed processing platforms, multi-device support, extended API linkage, external service linkage, learning data expansion methods, personalization methods, feedback loop design, anomaly detection and alarm notification functions, time-series data analysis, real-time inference, batch processing switching, and the linkage of multiple AI models (such as a pipeline structure of sentiment inference AI + risk judgment AI + proposal generation AI). This system integrates multi-dimensional data such as structure, user attributes, geographic information, historical information, and device information; data expansion; transfer learning; continuous learning; federated learning; UI / UX optimization algorithms; addition of an explainable AI (XAI) module; enhanced privacy protection; data anonymization and word segmentation; API integration with external medical institutions, insurance companies, and local government systems; internationalization support (multi-language UI, multi-regional data support); automatic recovery function in case of failure; user group permission management; log auditing and traceability enhancement; automatic version management of AI models; A / B testing; and online learning, enabling multiple technical variations. Therefore, this system can flexibly and scalably respond to changes in the latest AI technologies, cloud platforms, distributed processing technologies, security requirements, laws and regulations, and user needs without relying on a specific structure or algorithm. In terms of technical effectiveness, this system differs from previous fixed health management systems or single AI model-dependent services. It can flexibly apply multi-level technical improvements in module structure, algorithms, data flow, learning methods, UI / UX, and external integration, significantly improving the overall scalability, maintainability, operational efficiency, personalization accuracy, security, reliability, and user satisfaction. Applicable areas include multi-objective expansion of health management services for individuals, patient support platforms for medical institutions, risk assessment and intervention platforms for insurance companies, public health data platforms for local governments, and international health data linkage platforms.

[0081] The processing department can infer a user's emotions and adjust the timing of uploading genetic testing results accordingly. For example, when a user is stressed, the processing department encourages them to upload during a relaxing time; when a user is busy, the upload is scheduled to be completed quickly; and when a user is relaxed, the upload is performed while providing detailed explanations. Thus, the processing department can upload genetic testing results at the optimal time based on the user's emotional state.

[0082] The analysis department can infer user emotions and adjust the presentation of the analysis accordingly. For example, it provides detailed analysis results when the user is relaxed, concise results when the user is stressed, and visually appealing results when the user is excited. Thus, the analysis department can provide analysis results in a way that best suits the user's emotional state.

[0083] The proposal department can anticipate user emotions and adjust the presentation of proposals accordingly. For example, it provides detailed proposals when users are relaxed, concise proposals when users are stressed, and visually appealing proposals when users are excited. Thus, the proposal department can deliver proposals in a way that best suits the user's emotional state.

[0084] The visualization department can infer user emotions and adjust the display of visualizations accordingly. For example, it provides detailed visualizations when users are relaxed, concise visualizations when users are stressed, and visually appealing visualizations when users are excited. Thus, the visualization department can provide visualizations in a way that best suits the user's emotions.

[0085] The Improvement Department can anticipate user emotions and adjust improvement methods accordingly. For example, when users are relaxed, the Improvement Department provides detailed improvement methods; when users are stressed, it provides concise improvement methods; and when users are excited, it provides visually appealing improvement methods. Thus, the Improvement Department can implement improvements in a way that best suits the user's emotional state.

[0086] The processing department can analyze a user's past upload history when uploading genetic testing results to select the optimal upload method. For example, the processing department prioritizes recommending upload methods the user has used in the past (manual, voice input, etc.); the processing department selects the most effective method from the user's past upload history; and the processing department automatically selects methods that the user has successfully uploaded to in the past. Thus, the processing department can provide users with the optimal upload method.

[0087] The parsing unit can adjust the level of detail in the analysis based on the importance of the genetic information. For example, it provides detailed analysis results for important genetic information and concise results for less important genetic information. Furthermore, it can prioritize the analysis of highly important genetic information based on the user's focus. Thus, the parsing unit can provide detailed analysis of important information to the user.

[0088] The proposal department can adjust the level of detail in proposals based on the importance of the dietary packages. For example, it can provide detailed proposals for important dietary packages and concise proposals for less important ones. The department can also prioritize proposals for higher-importance dietary packages based on user concerns. This allows the proposal department to provide users with detailed and relevant information.

[0089] The visualization department is able to consider the user's device information and select the optimal display method during visualization. For example, when a user is using a smartphone, the visualization department provides a display method suitable for the screen size; when a user is using a tablet, it provides a display method optimized for large screens; and when a user is using a smartwatch, it provides a simple and highly visible display method. Thus, the visualization department can provide the optimal display method based on the user's device information.

[0090] The Improvement Department can refer to users' past feedback history to select the optimal improvement method during the improvement process. For example, the Improvement Department proposes the optimal improvement method based on past user feedback; the Improvement Department selects the most effective improvement method from users' past feedback history; the Improvement Department automatically selects improvement methods that users have previously been able to accept. Thus, the Improvement Department is able to provide users with the optimal improvement method.

[0091] The following is a brief description of the processing flow of the implementation method.

[0092] Step 1: The processing department uploads the genetic testing results received by the user. For example, the user uses a genetic testing kit to collect samples such as saliva or blood and sends them to the testing institution. The testing institution analyzes the user's approximately 21,000 genomes and provides the results. The processing department then uploads the genetic testing results received by the user to the generative AI.

[0093] Step 2: The analysis department analyzes the uploaded genetic testing results and the questionnaire information entered by the user. For example, the analysis department analyzes whether a specific gene increases the risk of disease. The analysis department also analyzes questionnaire information about the user's current health status and lifestyle habits to understand the user's lifestyle.

[0094] Step 3: The determination unit determines the risk of disease based on the information analyzed by the analysis unit. For example, when a specific gene increases the risk of disease, the determination unit determines the dietary or lifestyle improvements needed to reduce that risk.

[0095] Step 4: The proposal department proposes dietary kits to reduce the risk of disease determined by the assessment department. For example, based on the user's genetic information and lifestyle habits, the proposal department proposes nutritionally balanced dietary recommendations. The proposal department can also propose dietary kits that can be customized according to the user's preferences.

[0096] The specific processing unit 290 sends 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 voice representing the user's input to the result of the specific processing. The control unit 46A sends the voice data representing the user's 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 voice data.

[0097] Data generation model 58 is what is known as generative AI (Artificial Intelligence). An example of data generation model 58 includes ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> Generative AI, such as data generation model 58, is obtained by deep learning through a neural network. The data generation model 58 is input with a prompt containing instructions, and with inference data such as speech data representing speech, text data representing text, and image data representing images (e.g., data of still images or data of moving images). The data generation model 58 infers the inference data based on the instructions shown in the prompt and outputs the inference result in one or more data forms such as speech data, text data, and image data. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specific processing unit 290 performs the above-described specific processing while using the data generation model 58. The data generation model 58 can also be a fine-tuned model to output inference results from prompts without instructions; in this case, the data generation model 58 can output inference results from prompts without instructions. The data processing device 12, etc., includes multiple data generation models 58, including AI other than generative AI. AI other than generative AI includes, but is not limited to, linear regression, logistic regression, decision trees, random forests, support vector machines (SVM), k-means clustering, convolutional neural networks (CNN), recurrent neural networks (RNN), generative adversarial networks (GAN), or Naive Bayes. Furthermore, AI can also act as an AI agent. Additionally, when the processing described above is performed by AI, this processing may be partially or entirely performed by AI, but is not limited to this example. Moreover, processing performed by AI, including generative AI, can be replaced by rule-based processing, and vice versa.

[0098] Furthermore, the processing performed by the aforementioned data processing system 10 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart device 14, but it can also be executed jointly by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart device 14. Additionally, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the smart device 14 or external devices, and the smart device 14 acquires or collects information required for processing from the data processing device 12 or external devices.

[0099] Each of the aforementioned elements—acceptance department, analysis department, judgment department, proposal department, voice input department, visualization department, and improvement department—can be implemented, for example, in at least one of the smart device 14 and the data processing device 12. For instance, the acceptance department can be implemented by the control unit 46A of the smart device 14, enabling users to upload genetic testing results. The analysis department can be implemented by the specific processing unit 290 of the data processing device 12, analyzing the uploaded genetic testing results and questionnaire information. The judgment department can be implemented by the specific processing unit 290 of the data processing device 12, determining the risk of disease based on the analyzed information. The proposal department can be implemented by the control unit 46A of the smart device 14, proposing dietary kits to reduce the risk of disease. The voice input department can support voice input of the questionnaire via the microphone 38B of the smart device 14. The visualization department can visually display the analysis results via the display 40A of the smart device 14. The improvement department can be implemented by the specific processing unit 290 of the data processing device 12, collecting user feedback and improving the proposal content. The correspondence between the various departments and the device or control unit is not limited to the above example and can be modified in various ways.

[0100] [Second Implementation] Figure 3 An example of the configuration of the data processing system 210 according to the second embodiment is shown.

[0101] like Figure 3 As shown, the data processing system 210 includes a data processing device 12 and smart glasses 214. One example of the data processing device 12 is a server.

[0102] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, RAM 30, and memory 32. The processor 28, RAM 30, and memory 32 are connected to a bus 34. Furthermore, the database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. An example of the network 54 includes a WAN and / or LAN, etc.

[0103] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, RAM 48, and memory 50. The processor 46, RAM 48, and memory 50 are connected to a bus 52. Furthermore, the microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0104] Microphone 238 receives user commands by receiving the user's voice. Microphone 238 captures the user's voice, converts the captured sound into speech data, and outputs it to processor 46. Speaker 240 outputs sound according to commands from processor 46.

[0105] Camera 42 is a small digital camera equipped with an optical system such as a lens, aperture and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, used to capture the user's surroundings (e.g., the shooting range defined by an angle of view equivalent to the field of vision of an average healthy person).

[0106] Communication I / F 44 is connected to network 54. Communication I / F 44 and 26 are responsible for the transmission and reception of various information between processor 46 and processor 28 via network 54. The transmission and reception of various information between processor 46 and processor 28 using communication I / F 44 and 26 is performed in a secure state.

[0107] Figure 4 An example of the main functions of the data processing device 12 and the smart glasses 214 is shown. Figure 4 As shown, in the data processing device 12, specific processing is performed by the processor 28. The memory 32 stores a specific processing program 56.

[0108] The processor 28 reads a specific processing program 56 from the memory 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is implemented by the processor 28 operating as a specific processing unit 290 based on the specific processing program 56 executed on the RAM 30.

[0109] The memory 32 stores a data generation model 58 and an emotion-specific model 59. The data generation model 58 and the emotion-specific model 59 are used by the specific processing unit 290. The specific processing unit 290 can use the emotion-specific model 59 to infer the user's emotions and perform specific processing based on the user's emotions. The emotion inference function using the emotion-specific model 59 (emotion-specific function) includes various inferences and predictions related to the user's emotions, but is not limited to this example. Furthermore, emotion inference and prediction also include, for example, emotion analysis (analysis).

[0110] In the smart glasses 214, specific processing is performed by the processor 46. A specific processing program 60 is stored in the memory 50. The processor 46 reads the specific processing program 60 from the memory 50 and executes the read specific processing program 60 on the RAM 48. Specific processing is implemented by the processor 46 operating as a control unit 46A based on the specific processing program 60 executed on the RAM 48. Furthermore, the smart glasses 214 may also have the same data generation model and emotion-specific model as the data generation model 58 and the emotion-specific model 59, and use these models to perform the same processing as the specific processing unit 290.

[0111] Furthermore, other devices besides the data processing device 12 may also have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 obtains the processing results (prediction results, etc.) using the data generation model 58 by communicating with the server device that has the data generation model 58. In addition, the data processing device 12 may be a server device or a user-owned terminal device (e.g., a mobile phone, robot, home appliance, etc.).

[0112] The specific processing unit 290 sends 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 voice input representing the user's input to the specific processing result. The control unit 46A sends the voice data representing the user's input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[0113] The data generation model 58 is a so-called generative AI. An example of the data generation model 58 includes generative AIs such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with a prompt containing instructions, and also with inference data such as speech data representing speech, text data representing text, and image data representing images (e.g., data of still images or data of moving images). The data generation model 58 infers the inference data based on the instructions shown in the prompt and outputs the inference result in one or more data forms such as speech data, text data, and image data. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specific processing unit 290 performs the aforementioned specific processing while using the data generation model 58. The data generation model 58 can also be a fine-tuned model to output inference results from prompts that do not contain instructions; in this case, the data generation model 58 is able to output inference results from prompts that do not contain instructions. The data processing device 12, etc., includes various data generation models 58, which include AI other than generative AI. These AIs include, for example, linear regression, logistic regression, decision trees, random forests, support vector machines (SVM), k-means clustering, convolutional neural networks (CNN), recurrent neural networks (RNN), generative adversarial networks (GAN), or Naive Bayes, and are capable of various processing methods, but are not limited to these examples. Furthermore, AI can also be an AI agent. Additionally, when the processing described above is performed by AI, this processing may be partially or entirely performed by AI, but is not limited to these examples. Moreover, processing performed by AI including generative AI can be replaced by rule-based processing, and vice versa.

[0114] The data processing system 210 of the second embodiment performs the same processing as the data processing system 10 of the first embodiment. The processing performed by the data processing system 210 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart glasses 214, but it can also be executed jointly by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart glasses 214. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the smart glasses 214 or external devices, and the smart glasses 214 acquires or collects information required for processing from the data processing device 12 or external devices.

[0115] Each of the aforementioned elements—acceptance department, analysis department, judgment department, proposal department, voice input department, visualization department, and improvement department—can be implemented, for example, in at least one of the smart glasses 214 and the data processing device 12. For instance, the acceptance department can be implemented by the control unit 46A of the smart glasses 214, enabling users to upload genetic testing results. The analysis department can be implemented by the specific processing unit 290 of the data processing device 12, analyzing the uploaded genetic testing results and questionnaire information. The judgment department can be implemented by the specific processing unit 290 of the data processing device 12, determining the risk of disease based on the analyzed information. The proposal department can be implemented by the control unit 46A of the smart glasses 214, proposing dietary kits to reduce the risk of disease. The voice input department can support voice input of the questionnaire via the microphone 238 of the smart glasses 214. The visualization department can visually display the analysis results via the display of the smart glasses 214. The improvement department can be implemented by the specific processing unit 290 of the data processing device 12, collecting user feedback and improving the proposal content. The correspondence between each department and the device or control unit is not limited to the above example and can be modified in various ways.

[0116] [Third Implementation] Figure 5 An example of the configuration of the data processing system 310 according to the third embodiment is shown.

[0117] like Figure 5 As shown, the data processing system 310 includes a data processing device 12 and a head-mounted terminal 314. One example of the data processing device 12 is a server.

[0118] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, RAM 30, and memory 32. The processor 28, RAM 30, and memory 32 are connected to a bus 34. Furthermore, the database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. An example of the network 54 includes a WAN and / or LAN, etc.

[0119] The head-mounted terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and memory 50. The processor 46, RAM 48, and memory 50 are connected to a bus 52. Furthermore, the microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0120] Microphone 238 receives user commands by receiving the user's voice. Microphone 238 captures the user's voice, converts the captured sound into speech data, and outputs it to processor 46. Speaker 240 outputs sound according to commands from processor 46.

[0121] Camera 42 is a small digital camera equipped with an optical system such as a lens, aperture and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, used to capture the user's surroundings (e.g., the shooting range defined by an angle of view equivalent to the field of vision of an average healthy person).

[0122] Communication I / F 44 is connected to network 54. Communication I / F 44 and 26 are responsible for the transmission and reception of various information between processor 46 and processor 28 via network 54. The transmission and reception of various information between processor 46 and processor 28 using communication I / F 44 and 26 is performed in a secure state.

[0123] Figure 6 An example of the main functions of the data processing device 12 and the head-mounted terminal 314 is shown. Figure 6 As shown, in the data processing device 12, specific processing is performed by the processor 28. The memory 32 stores a specific processing program 56.

[0124] The processor 28 reads a specific processing program 56 from the memory 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is implemented by the processor 28 operating as a specific processing unit 290 based on the specific processing program 56 executed on the RAM 30.

[0125] The memory 32 stores a data generation model 58 and an emotion-specific model 59. The data generation model 58 and the emotion-specific model 59 are used by the specific processing unit 290. The specific processing unit 290 can use the emotion-specific model 59 to infer the user's emotions and perform specific processing based on the user's emotions. The emotion inference function using the emotion-specific model 59 (emotion-specific function) includes various inferences and predictions related to the user's emotions, but is not limited to this example. Furthermore, emotion inference and prediction also include, for example, emotion analysis (analysis).

[0126] In the head-mounted terminal 314, specific processing is performed by the processor 46. A specific program 60 is stored in the memory 50. The processor 46 reads the specific program 60 from the memory 50 and executes the read specific program 60 on the RAM 48. Specific processing is implemented by the processor 46 operating as a control unit 46A based on the specific program 60 executed on the RAM 48. Furthermore, the head-mounted terminal 314 may also have the same data generation model and emotion-specific model as the data generation model 58 and the emotion-specific model 59, and use these models to perform the same processing as the specific processing unit 290.

[0127] Furthermore, other devices besides the data processing device 12 may also have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 obtains the processing results (prediction results, etc.) using the data generation model 58 by communicating with the server device that has the data generation model 58. In addition, the data processing device 12 may be a server device or a user-owned terminal device (e.g., a mobile phone, robot, home appliance, etc.).

[0128] The specific processing unit 290 sends the result of the specific processing to the head-mounted terminal 314. In the head-mounted 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 voice representing the user's input to the specific processing result. The control unit 46A sends the voice data representing the user's input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[0129] The data generation model 58 is a so-called generative AI. An example of the data generation model 58 includes generative AIs such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with a prompt containing instructions, and also with inference data such as speech data representing speech, text data representing text, and image data representing images (e.g., data of still images or data of moving images). The data generation model 58 infers the inference data based on the instructions shown in the prompt and outputs the inference result in one or more data forms such as speech data, text data, and image data. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specific processing unit 290 performs the aforementioned specific processing while using the data generation model 58. The data generation model 58 can also be a fine-tuned model to output inference results from prompts that do not contain instructions; in this case, the data generation model 58 is able to output inference results from prompts that do not contain instructions. The data processing device 12, etc., includes various data generation models 58, which include AI other than generative AI. These AIs include, for example, linear regression, logistic regression, decision trees, random forests, support vector machines (SVM), k-means clustering, convolutional neural networks (CNN), recurrent neural networks (RNN), generative adversarial networks (GAN), or Naive Bayes, and are capable of various processing methods, but are not limited to these examples. Furthermore, AI can also be an AI agent. Additionally, when the processing described above is performed by AI, this processing may be partially or entirely performed by AI, but is not limited to these examples. Moreover, processing performed by AI including generative AI can be replaced by rule-based processing, and vice versa.

[0130] The data processing system 310 of the third embodiment performs the same processing as the data processing system 10 of the first embodiment. The processing performed by the data processing system 310 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the head-mounted terminal 314, but it can also be executed jointly by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the head-mounted terminal 314. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the head-mounted terminal 314 or external devices, and the head-mounted terminal 314 acquires or collects information required for processing from the data processing device 12 or external devices.

[0131] Each of the aforementioned elements—acceptance, analysis, judgment, proposal, voice input, visualization, and improvement—can be implemented, for example, in at least one of the head-mounted terminal 314 and the data processing device 12. For instance, the acceptance department can be implemented by the control unit 46A of the head-mounted terminal 314, enabling users to upload genetic testing results. The analysis department can be implemented by the specific processing unit 290 of the data processing device 12, analyzing the uploaded genetic testing results and questionnaire information. The judgment department can be implemented by the specific processing unit 290 of the data processing device 12, determining the risk of disease based on the analyzed information. The proposal department can be implemented by the control unit 46A of the head-mounted terminal 314, proposing dietary kits to reduce the risk of disease. The voice input unit can support voice input of the questionnaire via the microphone 238 of the head-mounted terminal 314. The visualization department can visually display the analysis results via the display 343 of the head-mounted terminal 314. The improvement department can be implemented by the specific processing unit 290 of the data processing device 12, collecting user feedback and improving the proposal content. The correspondence between the various parts and the devices or control units is not limited to the above example and can be modified in various ways.

[0132] [Fourth Implementation] Figure 7 An example of the configuration of the data processing system 410 according to the fourth embodiment is shown.

[0133] like Figure 7 As shown, 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.

[0134] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, RAM 30, and memory 32. The processor 28, RAM 30, and memory 32 are connected to a bus 34. Furthermore, the database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. An example of the network 54 includes a WAN and / or LAN, etc.

[0135] Robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control object 443. The computer 36 includes a processor 46, RAM 48, and memory 50. The processor 46, RAM 48, and memory 50 are connected to a bus 52. Furthermore, the microphone 238, speaker 240, camera 42, and control object 443 are also connected to the bus 52.

[0136] Microphone 238 receives user commands by receiving the user's voice. Microphone 238 captures the user's voice, converts the captured sound into speech data, and outputs it to processor 46. Speaker 240 outputs sound according to commands from processor 46.

[0137] The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture and shutter, and an imaging element such as a CMOS image sensor or a CCD image sensor, used to photograph the user's surroundings (e.g., the shooting range defined by an angle of view equivalent to the field of vision of an average healthy person).

[0138] Communication I / F 44 is connected to network 54. Communication I / F 44 and 26 are responsible for the transmission and reception of various information between processor 46 and processor 28 via network 54. The transmission and reception of various information between processor 46 and processor 28 using communication I / F 44 and 26 is performed in a secure state.

[0139] The controlled object 443 includes a display device, LEDs for the eyes, and motors for driving the arms, hands, and feet. The posture and movements of the robot 414 are controlled by controlling the motors for the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. In addition, facial expressions of the robot 414 can also be expressed by controlling the illumination state of the LEDs for the robot 414's eyes.

[0140] Figure 8 An example of the main functions of the data processing device 12 and the robot 414 is shown. For example... Figure 8 As shown, in the data processing device 12, specific processing is performed by the processor 28. The memory 32 stores a specific processing program 56.

[0141] The processor 28 reads a specific processing program 56 from the memory 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is implemented by the processor 28 operating as a specific processing unit 290 based on the specific processing program 56 executed on the RAM 30.

[0142] The memory 32 stores a data generation model 58 and an emotion-specific model 59. The data generation model 58 and the emotion-specific model 59 are used by the specific processing unit 290. The specific processing unit 290 can use the emotion-specific model 59 to infer the user's emotions and perform specific processing based on the user's emotions. The emotion inference function using the emotion-specific model 59 (emotion-specific function) includes various inferences and predictions related to the user's emotions, but is not limited to this example. Furthermore, emotion inference and prediction also include, for example, emotion analysis (analysis).

[0143] In robot 414, specific processing is performed by processor 46. A specific program 60 is stored in memory 50. Processor 46 reads the specific program 60 from memory 50 and executes the read specific program 60 on RAM 48. Specific processing is achieved by processor 46 acting as control unit 46A based on the specific program 60 executed on RAM 48. Furthermore, robot 414 may also have the same data generation model and emotion-specific model as data generation model 58 and emotion-specific model 59, and use these models to perform the same processing as specific processing unit 290.

[0144] Furthermore, other devices besides the data processing device 12 may also have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 obtains the processing results (prediction results, etc.) using the data generation model 58 by communicating with the server device that has the data generation model 58. In addition, the data processing device 12 may be a server device or a user-owned terminal device (e.g., a mobile phone, robot, home appliance, etc.).

[0145] The specific processing unit 290 sends 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 voice representing the user's input regarding the result of the specific processing. The control unit 46A sends the voice data representing the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[0146] The data generation model 58 is a so-called generative AI. An example of the data generation model 58 includes generative AIs such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with a prompt containing instructions, and also with inference data such as speech data representing speech, text data representing text, and image data representing images (e.g., data of still images or data of moving images). The data generation model 58 infers the inference data based on the instructions shown in the prompt and outputs the inference result in one or more data forms such as speech data, text data, and image data. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specific processing unit 290 performs the aforementioned specific processing while using the data generation model 58. The data generation model 58 can also be a fine-tuned model to output inference results from prompts that do not contain instructions; in this case, the data generation model 58 is able to output inference results from prompts that do not contain instructions. The data processing device 12, etc., includes various data generation models 58, which include AI other than generative AI. These AIs include, for example, linear regression, logistic regression, decision trees, random forests, support vector machines (SVM), k-means clustering, convolutional neural networks (CNN), recurrent neural networks (RNN), generative adversarial networks (GAN), or Naive Bayes, and are capable of various processing methods, but are not limited to these examples. Furthermore, AI can also be an AI agent. Additionally, when the processing described above is performed by AI, this processing may be partially or entirely performed by AI, but is not limited to these examples. Moreover, processing performed by AI including generative AI can be replaced by rule-based processing, and vice versa.

[0147] The data processing system 410 of the fourth embodiment performs the same processing as the data processing system 10 of the first embodiment. The processing performed by the data processing system 410 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the robot 414, but it can also be executed jointly by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the robot 414. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the robot 414 or external devices, and the robot 414 acquires or collects information required for processing from the data processing device 12 or external devices.

[0148] Each of the aforementioned elements—acceptance department, analysis department, judgment department, proposal department, voice input department, visualization department, and improvement department—can be implemented, for example, in at least one of the robot 414 and the data processing device 12. For instance, the acceptance department can be implemented by the control unit 46A of the robot 414, enabling users to upload genetic testing results. The analysis department can be implemented by the specific processing unit 290 of the data processing device 12, analyzing the uploaded genetic testing results and questionnaire information. The judgment department can be implemented by the specific processing unit 290 of the data processing device 12, determining the risk of disease based on the analyzed information. The proposal department can be implemented by the control unit 46A of the robot 414, proposing dietary kits to reduce the risk of disease. The voice input department can support voice input of the questionnaire via the microphone 238 of the robot 414. The visualization department can visually display the analysis results via the display of the robot 414. The improvement department can be implemented by the specific processing unit 290 of the data processing device 12, collecting user feedback and improving the proposal content. The correspondence between the various departments and the device or control unit is not limited to the above example and can be modified in various ways.

[0149] Furthermore, the emotion-specific model 59, serving as an emotion engine, can determine the user's emotion based on a specific mapping. Specifically, the emotion-specific model 59 can determine the user's emotion based on an emotion graph that serves as a specific mapping (see...). Figure 9 The system determines the user's emotions. In addition, the emotion-specific model 59 can also determine the robot's emotions in the same way, and the specific processing unit 290 can also perform specific processing using the robot's emotions.

[0150] Figure 9 This is a diagram representing an emotion map 400 that maps various emotions. In the emotion map 400, emotions are arranged radially from the center in concentric circles. The closer to the center of the concentric circles, the more primitive the emotion is. Further out on the concentric circles, emotions are arranged representing states or actions arising from mood. Emotion is a concept that includes both feelings and mental states. To the left of the concentric circles, emotions generated by reactions occurring in the brain are arranged roughly. To the right of the concentric circles, emotions guided by situational judgments are arranged roughly. Above and below the concentric circles, emotions generated by reactions occurring in the brain and guided by situational judgments are arranged roughly. Furthermore, the emotion of "pleasure" is arranged above the concentric circles, and the emotion of "unpleasantness" is arranged below. Thus, in the emotion map 400, various emotions are mapped according to the structure of emotion generation, while easily generated emotions are mapped nearby.

[0151] These emotions are distributed at the 3 o'clock position on the Emotion Chart 400, and usually fluctuate between peace and unease. In the right half of the Emotion Chart 400, because situational awareness is more dominant than internal feelings, it gives a sense of calm.

[0152] The inner side of the emotion diagram 400 represents the mind, and the outer side of the emotion diagram 400 represents actions. Therefore, the further you go to the outer side of the emotion diagram 400, the more the emotion can be seen (manifested in actions).

[0153] Here, human emotions are based on a balance of various factors such as posture and blood sugar levels. When these balances deviate from the ideal, it indicates unhappiness; when they approach the ideal, it indicates pleasure. In robots, cars, and motorcycles, emotions can also be created based on a balance of factors such as posture and remaining battery power. When these balances deviate from the ideal, it indicates unhappiness; when they approach the ideal, it indicates pleasure. Emotion maps can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Speech Emotion Recognition and Brain Physiological Signal Analysis Systems for Emotion, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). In the left half of the emotion map, emotions belonging to the "Reaction" domain, where sensation is dominant, are arranged. Furthermore, in the right half of the emotion map, emotions belonging to the "Situation" domain, where situational cognition is dominant, are arranged.

[0154] The emotion map defines two types of emotions that promote learning. One is a negative emotion located near the middle of "repentance" or "reflection" on the situation side. That is, when the robot experiences negative emotions such as "I never want to feel this way again" or "I never want to be scolded again." The other is a positive emotion located near "desire" on the response side. That is, when the robot experiences positive feelings such as "wanting more" or "wanting to know more."

[0155] The emotion-specific model 59 feeds user input into a pre-learned neural network to obtain emotion values ​​representing each emotion shown in the emotion graph 400, and determines the user's emotion. This neural network is pre-learned based on multiple learning data sets that combine user input with emotion values ​​representing each emotion shown in the emotion graph 400. Furthermore, this neural network is learned to... Figure 10 As shown in sentiment graph 900, sentiment values ​​in nearby configurations are similar to each other. Figure 10 Examples show how emotions such as "peace of mind," "stability," and "reassurance" can result in similar emotional values.

[0156] In the above embodiments, a specific processing is described by a single computer 22, but the technology disclosed herein is not limited to this, and distributed processing by multiple computers, including computer 22, is also possible.

[0157] In the above embodiments, an example of storing a specific processing program 56 in memory 32 is illustrated, but the technology disclosed herein is not limited thereto. For example, the specific processing program 56 may also be stored in a portable computer-readable non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 performs specific processing according to the specific processing program 56.

[0158] Alternatively, the specific processing program 56 can be stored in a storage device such as a server connected to the data processing device 12 via a network 54, and the specific processing program 56 can be downloaded and installed into the computer 22 upon request from the data processing device 12.

[0159] Furthermore, it is not necessary to store the entire specific process 56 in a storage device such as a server connected to the data processing device 12 via the network 54, nor is it necessary to store the entire specific process 56 in the memory 32; a portion of the specific process 56 may also be stored.

[0160] As a hardware resource for performing specific processing, various processors can be used. For example, a CPU is a general-purpose processor that functions as a hardware resource for performing specific processing by executing software, i.e., programs. Additionally, a dedicated circuit can be listed as a processor; it is a processor with a circuit structure specifically designed for performing specific processing, such as a FPGA (Field-Programmable Gate Array), PLD (Programmable Logic Device), or ASIC (Application Specific Integrated Circuit). Every processor has built-in or connected memory, and every processor executes specific processing by using that memory.

[0161] The hardware resources for performing a specific process can consist of one of these various processors, or a combination of two or more processors of the same or different types (e.g., a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Furthermore, the hardware resources for performing a specific process can also be a single processor.

[0162] As an example of a single processor, the first type consists of a combination of one or more CPUs and software, which functions as a hardware resource to perform specific processing. The second type uses a processor, such as a System-on-a-chip (SoC), which implements the entire system functionality, including multiple hardware resources performing specific processing, using a single IC chip. In this case, the specific processing is implemented using one or more of the aforementioned processors that serve as hardware resources.

[0163] Furthermore, as the hardware architecture of these various processors, more specifically, circuits combining semiconductor elements and other circuit components can be used. Moreover, the specific process described above is merely an example. Therefore, it goes without saying that, without departing from the main point, unnecessary steps can be removed, new steps can be added, or the processing order can be changed.

[0164] Furthermore, although the above examples have been described in terms of first to fourth embodiments, some or all of these embodiments can also be combined. Additionally, the smart device 14, smart glasses 214, head-mounted terminal 314, and robot 414 are just examples; they can be combined separately or are other devices.

[0165] The foregoing descriptions and illustrations are detailed explanations of the parts covered by this disclosure and are merely one example of this disclosure. For instance, the descriptions of the above-described structure, function, role, and effect are just one example of the structure, function, role, and effect of the parts covered by this disclosure. Therefore, it goes without saying that, without departing from the spirit of this disclosure, unnecessary parts can be deleted, new elements can be added, or replacements can be made to the foregoing descriptions and illustrations. Furthermore, to avoid confusion and facilitate understanding of the parts covered by this disclosure, explanations of technical common sense that does not require special explanation for implementing this disclosure have been omitted from the foregoing descriptions and illustrations.

[0166] All documents, patent applications and technical standards described in this specification are incorporated herein by reference as if they were specifically and individually described as incorporated by reference.

[0167] (Note 1) A system comprising: The receiving department is used to upload genetic testing results; The analysis unit is used to analyze the gene testing results and questionnaire information uploaded by the receiving unit. The determination unit is used to determine the risk of disease based on the information analyzed by the analysis unit; The proposal department is responsible for proposing dietary kits to reduce the risk of disease determined by the determination department.

[0168] (Note 2) The system as described in Appendix 1 is characterized by further comprising a part for setting the questionnaire input as voice input.

[0169] (Note 3) The system as described in Appendix 1 is characterized by further comprising a part for visualizing the analysis results.

[0170] (Note 4) The system as described in Appendix 1 is characterized by further including a part for improving the content of the proposal based on user feedback.

[0171] (Note 5) The system as described in Appendix 1 is characterized in that the receiving unit includes a unit for automatically uploading gene detection results.

[0172] (Note 6) The system as described in Appendix 1 is characterized in that the analysis unit includes a section for analyzing genetic information and questionnaire information and understanding the user's lifestyle habits.

[0173] (Note 7) The system as described in Appendix 1 is characterized in that the proposal department includes a department for proposing dietary kits that can be customized according to the user's preferences.

[0174] (Note 8) The system as described in Appendix 1 is characterized in that the receiving unit includes a unit for inferring user emotions and adjusting the timing of uploading gene detection results based on the inferred user emotions.

[0175] (Note 9) The system as described in Appendix 1 is characterized in that, when the gene testing results are uploaded, the receiving department analyzes the user's past upload history and selects the upload method.

[0176] (Postscript 10) The system as described in Appendix 1 is characterized in that the receiving unit includes a unit for filtering based on the user's current health status and lifestyle when uploading gene testing results.

[0177] (Postscript 11) The system as described in Appendix 1 is characterized in that the receiving unit includes a unit for inferring user emotions and determining the priority of uploaded gene detection results based on the inferred user emotions.

[0178] (Postscript 12) The system described in Appendix 1 is characterized in that, when uploading gene detection results, the receiving department considers the user's geographical location information and prioritizes uploading results with high relevance.

[0179] (Postscript 13) The system as described in Appendix 1 is characterized in that, when the gene testing results are uploaded, the receiving department analyzes the user's social media activities and uploads the relevant results.

[0180] (Postscript 14) The system as described in Appendix 1 is characterized in that the parsing unit includes a unit for inferring user emotions and adjusting the parsing presentation method according to the inferred user emotions.

[0181] (Postscript 15) The system as described in Appendix 1 is characterized in that, during the analysis, the analysis unit adjusts the level of detail of the analysis according to the importance of the gene information.

[0182] (Postscript 16) The system as described in Appendix 1 is characterized in that the parsing unit applies different parsing algorithms according to the category of gene information during parsing.

[0183] (Postscript 17) The system as described in Appendix 1 is characterized in that the parsing unit includes a unit for inferring user emotions and adjusting the parsing length according to the inferred user emotions.

[0184] (Postscript 18) The system as described in Appendix 1 is characterized in that, during the parsing process, the parsing unit determines the parsing priority based on the timing of the submission of the gene information.

[0185] (Postscript 19) The system as described in Appendix 1 is characterized in that, during the analysis, the analysis unit adjusts the analysis order according to the correlation of gene information.

[0186] (Postscript 20) The system as described in Appendix 1 is characterized in that the determination unit includes a unit for inferring user emotions and adjusting the disease risk determination criterion based on the inferred user emotions.

[0187] (Postscript 21) The system as described in Appendix 1 is characterized in that the determination unit considers the interrelationships of gene information during determination to improve determination accuracy.

[0188] (Postscript 22) The system as described in Appendix 1 is characterized in that the determination unit considers the attribute information of the gene information submitter when making a determination.

[0189] (Postscript 23) The system as described in Appendix 1 is characterized in that the determination unit includes a unit for inferring user emotions and adjusting the display order of determination results based on the inferred user emotions.

[0190] (Postscript 24) The system as described in Appendix 1 is characterized in that the determination unit considers the geographical distribution of gene information when making a determination.

[0191] (Postscript 25) The system as described in Appendix 1 is characterized in that the determination unit refers to relevant literature on gene information to improve the accuracy of the determination.

[0192] (Postscript 26) The system as described in Appendix 1 is characterized in that the proposal department includes a department for inferring user emotions and adjusting the presentation of the proposal based on the inferred user emotions.

[0193] (Postscript 27) The system as described in Appendix 1 is characterized in that, when making a proposal, the proposal department adjusts the level of detail of the proposal according to the importance of the dietary kit.

[0194] (Postscript 28) The system as described in Appendix 1 is characterized in that the proposal department applies different proposal algorithms according to the category of the diet kit when making a proposal.

[0195] (Postscript 29) The system as described in Appendix 1 is characterized in that the proposal department includes a department for inferring user emotions and adjusting the proposal length based on the inferred user emotions.

[0196] (Note 30) The system as described in Appendix 1 is characterized in that the proposal department determines the priority of the proposal based on the timing of the submission of the diet kit.

[0197] (Postscript 31) The system as described in Appendix 1 is characterized in that the proposal department adjusts the order of proposals based on the relevance of the dietary kits when making proposals.

[0198] (Note 32) The system as described in Appendix 2 is characterized in that the voice input unit includes a unit for inferring the user's emotions and adjusting the timing of voice input based on the inferred user emotions.

[0199] (Postscript 33) The system as described in Appendix 2 is characterized in that, when the voice input unit performs voice input, it analyzes the user's past voice input history and selects the optimal input method.

[0200] (Postscript 34) The system as described in Appendix 2 is characterized in that the voice input unit includes a unit for inferring the user's emotions and determining the voice input priority based on the inferred user emotions.

[0201] (Postscript 35) The system described in Appendix 2 is characterized in that, when the voice input unit performs voice input, it considers the user's geographical location information and prioritizes inputs with high relevance.

[0202] (Postscript 36) The system as described in Appendix 3 is characterized in that the visualization unit includes a unit for inferring user emotions and adjusting the visualization display method according to the inferred user emotions.

[0203] (Postscript 37) The system as described in Appendix 3 is characterized in that, when performing visualization, the visualization unit selects the optimal display method by referring to the user's past visualization history.

[0204] (Postscript 38) The system as described in Appendix 3 is characterized in that the visualization unit includes a unit for inferring user emotions and determining visualization priority based on the inferred user emotions.

[0205] (Postscript 39) The system as described in Appendix 3 is characterized in that, when visualizing, the visualization unit considers the user's device information and selects the optimal display method.

[0206] (Postscript 40) The system as described in Appendix 4 is characterized in that the improvement unit includes a unit for inferring user emotions and adjusting the improvement method according to the inferred user emotions.

[0207] (Postscript 41) The system described in Appendix 4 is characterized in that, when making improvements, the improvement unit selects the optimal improvement method by referring to the user's past feedback history.

[0208] (Postscript 42) The system as described in Appendix 4 is characterized in that the improvement unit includes a unit for inferring user emotions and determining improvement priorities based on the inferred user emotions.

[0209] (Postscript 43) The system as described in Appendix 4 is characterized in that, when making improvements, the improvement unit considers the user's geographical location information and selects the optimal improvement method.

Claims

1. A system, characterized in that, include: The receiving department is used to upload genetic testing results; The analysis unit is used to analyze the gene testing results and questionnaire information uploaded by the receiving unit. The determination unit is used to determine the risk of disease based on the information parsed by the analysis unit; The proposal department is responsible for proposing dietary kits to reduce the risk of disease determined by the determination department.

2. The system as described in claim 1, characterized in that, It also includes a section for setting questionnaire input as voice input.

3. The system as described in claim 1, characterized in that, It also includes a section for visualizing the analysis results.

4. The system as described in claim 1, characterized in that, It also includes a department for improving proposals based on user feedback.

5. The system as described in claim 1, characterized in that, The receiving department includes a section for automatically uploading gene testing results.

6. The system as described in claim 1, characterized in that, The analysis unit includes a section for analyzing genetic information and questionnaire information and understanding users' lifestyle habits.

7. The system as described in claim 1, characterized in that, The proposal department includes a section for proposing dietary kits that can be customized according to the user's preferences.

8. The system as described in claim 1, characterized in that, The receiving department includes a section for inferring user emotions and adjusting the timing of uploading gene test results based on the inferred user emotions.

9. The system as described in claim 1, characterized in that, When uploading gene testing results, the receiving department analyzes the user's past upload history and selects the upload method.

10. The system as claimed in claim 1, characterized in that, The receiving department includes a section for filtering data based on the user's current health status and lifestyle when uploading gene testing results.

Images