system

Abstract

We provide the system. [Solution] Means for collecting past survey data, A means for analyzing the aforementioned survey data and extracting the characteristics of the questions, A means for training and generating a model that predicts the optimal answer to a question using machine learning techniques, A means of automatically generating answers to new survey questions using the generated model, The means for presenting the automatically generated answer, A system that includes this.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a persona chatbot control method performed by at least one processor, the method including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a character of the chatbot, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In modern times, there is a need for technology to improve the efficiency of answering surveys. In particular, users who conduct surveys with the intention of earning points spend time answering similar questions manually repeatedly, which is a burden. Against this background, there is a need for a method to significantly reduce the time and effort that users spend on survey responses.

Means for Solving the Problems

[0005] This invention provides a means for generating a model that predicts the optimal answer using machine learning techniques by collecting past survey data, analyzing that data, and extracting the characteristics of the questions. Furthermore, by providing means for automatically generating and presenting answers to new survey questions using the generated model, the invention enables users to answer surveys quickly and reduces the time required to answer. In addition, it provides an interface that allows users to confirm and correct the answers, and provides means for improving the accuracy of the model through user feedback.

[0006] "Collecting past survey data" refers to the process of recording the questions and answers from surveys that users have previously conducted and incorporating them into the system.

[0007] "Extracting the characteristics of the questions" is the process of analyzing and identifying important linguistic and semantic patterns from the survey questions.

[0008] "Machine learning technology" is a general term for algorithms and methods that enable computers to learn from data and make future predictions and decisions.

[0009] "Predicting the optimal answer" is the process of making inferences based on past data to generate the answer that best fits the current situation.

[0010] "Model generation" refers to establishing a mathematical or statistical structure that learns from collected data and is built to perform a specific task.

[0011] "Automatically generating answers to new survey questions" refers to the process by which a system automatically constructs answers to newly presented questions using existing models.

[0012] "Presentation means" refers to the part that includes the functions and methods for displaying the generated response to the user.

[0013] An "interface" refers to a means of providing input for users to make modifications and to interact with the system.

[0014] "Feedback" is information that users return to the system regarding their evaluation of the automatically generated answers and suggestions for improvement.

[0015] "Improving model accuracy" is a process of continuous adjustment and optimization using user feedback to improve the accuracy and consistency of predictions. [Brief explanation of the drawing]

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

Mode for Carrying Out the Invention

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

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

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

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

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

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

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

[0024] [First Embodiment]

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

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

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

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

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

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

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

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

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

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

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

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

[0037] This invention provides a system for users to conduct research efficiently. In one embodiment, the system has a core technology that collects past research data and generates a machine learning model based on it. The overall system embodiment is described in detail below.

[0038] First, users upload past survey questions and answer data to the system. This data is used as a basis for predicting the most appropriate answers for each survey question. The server receives this data and stores it in a database, thereby efficiently managing data resources.

[0039] Next, the server applies natural language processing techniques to the stored data to analyze the characteristics of the questions. This analysis extracts important keywords and contexts contained in the questions, preparing them for use in subsequent machine learning modeling.

[0040] During model training, the server uses machine learning algorithms to analyze past response data and generate a predictive model. This predictive model forms the basis for automatically generating appropriate answers when users receive similar questions.

[0041] Subsequently, using the generated model, the server builds an agent that can automatically generate answers to new questions. This agent has the ability to quickly derive answers based on the analysis results of the received questions when a user receives a new survey.

[0042] As a concrete example of its use, if a user receives a survey that includes the question "Where did you last visit?", the server automatically generates "Tokyo" as the appropriate answer based on past response patterns, and the terminal presents this to the user. The user can then review this answer, submit it as is if satisfied, and make corrections if necessary.

[0043] Furthermore, the system continuously collects user feedback and uses it to refine its machine learning model, thereby improving the accuracy of its responses. This feedback loop allows the system to improve over time, providing more accurate answers. Specifically, the server can use feedback data to optimize model parameters, further enhancing the accuracy of analysis and predictions.

[0044] The following describes the processing flow.

[0045] Step 1:

[0046] Users prepare past survey questions and answer data, access the system, and upload this data.

[0047] Step 2:

[0048] The server receives the uploaded data, converts each piece of data into a parseable format, and stores it in the database. This stored data is then used as the foundational data for model training.

[0049] Step 3:

[0050] The server analyzes the stored survey data using natural language processing techniques. To understand the characteristics of the questions, it performs sentence tokenization and morphological analysis to extract the meaning and context of the questions.

[0051] Step 4:

[0052] The server applies machine learning algorithms based on past response data to train a model to predict the optimal answer. This provides a statistical understanding of question and answer patterns.

[0053] Step 5:

[0054] The server builds an AI agent based on a pre-trained model. This agent automatically generates appropriate answers when a new survey question is presented.

[0055] Step 6:

[0056] When a user receives a new survey, their device sends the information to the server. The server then uses an AI agent to analyze the received questions.

[0057] Step 7:

[0058] The server automatically generates the optimal answer to a question via an AI agent and presents it to the user by sending it to the terminal.

[0059] Step 8:

[0060] The user reviews the automatically generated response, makes any necessary corrections, and submits the final response to the server.

[0061] Step 9:

[0062] The server receives feedback from users and uses that information to readjust the machine learning model. This feedback improves the accuracy of the model.

[0063] (Example 1)

[0064] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0065] In conventional information processing systems, it was difficult to effectively utilize past data and generate rapid and appropriate responses to new information. Furthermore, user feedback was not adequately utilized to improve the accuracy of the generated responses. As a result, challenges existed in information management and response accuracy.

[0066] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0067] In this invention, the server includes means for collecting past information, means for analyzing the information and extracting its characteristics, means for training and generating a model to predict the optimal response using machine learning techniques, means for automatically generating a response to new information using the generated model, means for displaying the automatically generated response, means for applying natural language processing techniques to analyze the context of the object to be processed and identify necessary elements, means for managing information input via the user and efficiently utilizing the stored information, and means for presenting the machine learning-generated response to the user and providing the user with means to confirm or modify it. This enables efficient information management, advanced response generation, and continuous improvement of prediction accuracy through the use of user feedback.

[0068] "Past information" refers to data and records collected previously.

[0069] "Means of collection" refers to the technologies and methods used to collect, store, and manage information.

[0070] "Analysis methods" refer to techniques for analyzing information and extracting important characteristics and patterns from it.

[0071] "Characteristics" refer to the important elements and features of data revealed through analysis.

[0072] "Machine learning technology" refers to artificial intelligence techniques that train models based on data and perform estimations and classifications based on new data.

[0073] "Response" refers to the reply or output generated in response to user input.

[0074] A "model" refers to a mathematical or statistical representation built on data for a specific purpose.

[0075] "Automatic generation means" refers to a process in which a program automatically generates a response based on the input information.

[0076] "Display means" refers to devices or methods for visualizing and presenting the generated response to the user.

[0077] "Natural language processing technology" refers to computer technology that understands and responds to human language.

[0078] "Feedback" refers to evaluations and correction information provided by users.

[0079] "Improving accuracy" refers to processes and methods for increasing the accuracy of models and responses.

[0080] This invention is a system that efficiently processes information and generates highly accurate responses. First, the user uploads past information to the system. This information can be provided as a file in formats such as CSV or Excel. Information uploading is done through a web interface, which allows the user to operate the system intuitively.

[0081] The server receives this information and stores it in a database system (for example, MySQL® or PostgreSQL). The stored information undergoes duplicate checking and data cleansing before being prepared for analysis. The server uses natural language processing techniques, such as Python's NLTK library, to extract important keywords and context from the information. This clarifies the characteristics of the information and lays the groundwork for the next step.

[0082] Next, the server trains a predictive model using machine learning libraries such as Scikit-learn and TENSORFLOW® based on the analyzed information. This model is designed to generate the optimal response to new information based on past data. By generating a predictive model, the server can be equipped with an agent that automatically creates responses to new questions.

[0083] The terminal presents the user with a generated response. The user can review the presented response, use it if appropriate, and modify it as needed. For example, if the user is asked a question such as "What service did you use last time?", the server will generate a response such as "Online streaming service" and display it to the user through the terminal.

[0084] Furthermore, the server continuously collects user feedback and uses it to improve the model's accuracy. By utilizing the information obtained from the feedback, the server optimizes the model parameters, enabling more accurate predictions. This feedback loop allows the system to improve over time, enabling it to provide higher-performance responses.

[0085] A concrete example of a prompt message might be input such as, "Please suggest the next service based on the services I've recently used." In response to this prompt, the system can generate the optimal response based on past data and meet the user's needs.

[0086] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0087] Step 1:

[0088] Users upload previously collected information to the system via a web interface. Input data is provided in CSV or Excel file format. In this step, users select files and press the upload button to import their digital data into the system. The data is then transferred to the server and becomes available for use in the next step.

[0089] Step 2:

[0090] The server verifies the received data and stores it in the database system. The input data consists of data uploaded by users. Data cleansing is performed to remove duplicates and standardize the format, ensuring data integrity. The output is clean, standardized data stored in the database.

[0091] Step 3:

[0092] The server retrieves data stored in the database and analyzes it using natural language processing techniques. The input is text data from the database. Here, the Python NLTK library is used to extract important keywords and context. The output is a dataset from which characteristics have been extracted. This set is used in the next machine learning step.

[0093] Step 4:

[0094] The server uses the analyzed characteristic data to train a predictive model with a machine learning algorithm. The input data consists of characteristic data and historical response data. Using libraries such as Scikit-learn and TensorFlow, the algorithm learns and generates a model that predicts the optimal response. The output is the trained predictive model. This model serves as the foundation for generating new responses.

[0095] Step 5:

[0096] The server automatically generates responses to newly received information using the generated predictive model. The input is the newly provided question data. The model analyzes this question data and generates an appropriate predicted response. The output is the generated response text. This response is the first step towards being presented to the user.

[0097] Step 6:

[0098] The terminal displays the response received from the server to the user. The user reviews this response on the screen. The input is the response data from the server, and the output is the text information displayed on the screen. The user can check whether the response is appropriate and modify it if necessary.

[0099] Step 7:

[0100] The server collects user feedback and uses it to improve the accuracy of its predictive model. The input is the feedback information provided by the user. Based on this feedback, the model is retrained, and the output is a new predictive model with improved accuracy. This enhances the overall response accuracy and adaptability of the system.

[0101] (Application Example 1)

[0102] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0103] In today's information-saturated world, it is difficult for users to efficiently acquire information and content optimized for their individual interests. Furthermore, there is a need for technology that utilizes collected data to provide automated responses and recommendations tailored to the user. This invention aims to improve the user experience by providing a system that accurately recommends new content based on the user's past behavior and interests.

[0104] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0105] In this invention, the server includes means for collecting past survey information, means for analyzing the survey information to extract the characteristics of the questions, means for training and generating a model that predicts the optimal response to the questions using machine learning technology, and means for recommending new content based on the user's past interests. This enables the automatic suggestion of content optimized for each user's individual interests, making it easier for users to access appropriate information.

[0106] "Survey information" refers to data on past questions and answers collected from users.

[0107] "Characteristics of the questions" refer to the important keywords and contexts hidden within the questions extracted from the survey information.

[0108] "Machine learning technology" is a general term for algorithms and methods used to build predictive models using data and automate specific tasks.

[0109] A "model that predicts the optimal response" is a machine learning model trained to generate appropriate answers to new questions based on collected data.

[0110] "User's past interests" refers to information based on the user's previously indicated interests and browsing history.

[0111] "Means of recommending content" refers to processes or technologies for suggesting new information or media based on users' past behavioral data.

[0112] The system for carrying out this invention comprises a server, a user terminal, and certain communication means. The server collects past research information, analyzes the data, trains machine learning models, and provides customized responses and content recommendations to each user.

[0113] First, past survey information provided by the user's terminal is uploaded to the server. The server receives this information and automatically stores it in a database. Next, the server uses a natural language processing engine (e.g., spaCy or BERT model) to analyze the characteristics of the questions and extract important keywords and context. This process makes the context of the survey data clearer.

[0114] The server uses machine learning frameworks such as TensorFlow to train and generate models that predict optimal responses based on historical data. This response generation model can handle new survey questions and automatically provides accurate answers to newly received questions from the user's device. Furthermore, it suggests new content through a smartphone app based on the user's past interests. Recommended content is determined by server-side data analysis based on the user's viewing history and feedback.

[0115] For example, if a user has recently watched many action movies, the system is likely to recommend "the latest action movies" and related "thriller movies." This recommendation is based on up-to-date data to accurately reflect the user's interests. Examples of prompts include, "What are your recent favorite genres? What are this week's recommendations from that genre?" The server sends the results, combining this information with the model, to the user's terminal, where the user can review the suggestions and choose responses or options as needed.

[0116] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0117] Step 1:

[0118] The terminal collects past survey information and uploads it to the server. The input is data of questions and answers recorded by the user, and the output is raw data stored in a database on the server. This data serves as a foundation for facilitating responses to the user's future questions.

[0119] Step 2:

[0120] The server analyzes the uploaded survey information using a natural language processing engine. The input is raw data stored in a database, and the output is analyzed data with the context of the questions and keywords extracted. The server uses spaCy and BERT models to perform analysis to reveal important features within the data.

[0121] Step 3:

[0122] The server trains a predictive model using a machine learning framework. The input is the previously analyzed data, and the output is a trained generative AI model to respond to new questions. The server leverages TensorFlow to build the model based on past response patterns.

[0123] Step 4:

[0124] When a user enters a new question, the server uses a generative AI model to generate the optimal response. The input is the user's new question, and the output is the generated answer. By applying the model, the server provides answers quickly and accurately.

[0125] Step 5:

[0126] The server analyzes the user's past interests and recommends content. The input is the user's viewing history, and the output is a list of recommended content. The server performs data analysis to select information that best matches the user's interests.

[0127] Step 6:

[0128] The user's terminal displays recommended content and generated responses. Input consists of responses and recommended content sent from the server, while output is the information displayed on the terminal. The user can review the provided information and make evaluations or modifications as needed.

[0129] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0130] This invention provides an automated response generation system that efficiently collects and analyzes survey data, while also considering the user's emotional state. This system utilizes machine learning technology to provide rapid and appropriate responses to survey questions, integrating emotional analysis throughout the process.

[0131] Specifically, it begins with users uploading past survey data to the system. The server receives this data and stores it in a database. The stored data is used to train machine learning models and forms the basis for understanding the characteristics of the questions and the user's past response patterns.

[0132] Next, the server analyzes the question data using natural language processing techniques and extracts important features. Because an emotion engine is integrated into the system, it can also analyze the user's emotional state related to each question. For example, in response to a question like, "How do you feel about this product?", the emotion engine evaluates the user's emotions and adds the result to the data.

[0133] The resulting machine learning model not only predicts the optimal answer to a question, but also takes into account sentiment data analyzed through the sentiment engine. This allows it to provide answers that are tailored to the user's emotions.

[0134] For example, if a user is asked, "How was your recent visit experience?", the server generates a response such as "Very satisfied" based on past data and sentiment analysis results, and presents it to the user via their device. The user can review this response, and if they are satisfied with the accuracy of the automatically generated response, which also depends on their emotions, they can submit it as is.

[0135] Furthermore, user feedback is continuously collected and used to improve machine learning models and sentiment engines. The servers leverage user feedback to continuously improve model accuracy and strengthen the foundation for providing a more personalized experience.

[0136] The following describes the processing flow.

[0137] Step 1:

[0138] Users access the survey platform and upload their past survey questions and answers to the system. This data is then used as the system's knowledge base.

[0139] Step 2:

[0140] The server receives the uploaded data and stores it in a database. The data is structured and organized into a format that will be used later for model training.

[0141] Step 3:

[0142] The server analyzes the question data using natural language processing techniques. This analysis extracts important linguistic features and patterns. Additionally, sentiment data related to each question is collected as supplementary information.

[0143] Step 4:

[0144] The server uses an emotion engine to evaluate the user's emotional state. It performs text analysis to identify emotional tendencies from the user's responses and comments, and then assigns an evaluation such as positive, negative, or neutral.

[0145] Step 5:

[0146] The server uses these analysis results to train a machine learning model. The model also considers sentiment analysis results when predicting the best answer to a question.

[0147] Step 6:

[0148] When a new survey question arrives, the device sends the information to the server. The server analyzes the question using an AI agent and an emotion engine.

[0149] Step 7:

[0150] The server automatically generates responses based on the analysis results, adjusting them to suit the user's emotional state. These automatically generated responses are carefully designed to be emotionally appropriate.

[0151] Step 8:

[0152] The device presents the generated response to the user. The user can review the response and, if satisfied with its content, submit it. If necessary, the user can modify the response.

[0153] Step 9:

[0154] The server receives user feedback and analyzes the data to improve the accuracy of machine learning models and the sentiment engine. Through this feedback, the models are refined, improving the overall system performance.

[0155] (Example 2)

[0156] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0157] Conventional information gathering and analysis systems generated responses without considering the user's emotional state, making it difficult to provide appropriate and personalized responses. Furthermore, they were unable to effectively utilize user feedback on the generated responses, hindering system improvement.

[0158] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0159] In this invention, the server includes means for accumulating past information, means for analyzing the information to extract the characteristics of items, and means for evaluating the user's emotional state related to the items and incorporating the results of the emotional analysis into the response. This enables the generation of automated responses that correspond to the user's emotions and continuous improvement of the system's accuracy using feedback.

[0160] "Information storage" refers to the systematic storage of data collected in the past, making it available for later analysis and processing.

[0161] "Extracting item characteristics" refers to identifying distinctive information from a specific dataset and extracting it in a format useful for analysis.

[0162] "Evaluating the user's emotional state related to an item" means analyzing the emotions a user has towards a particular item and expressing them using numerical values ​​or categories.

[0163] "Incorporating the results of sentiment analysis" means reflecting the results of an analysis of the user's emotional state in the generated responses and the operation of the system.

[0164] "Automated response generation" refers to the process of automatically creating answers to predefined items using machine learning models or similar tools.

[0165] "Continuous system accuracy improvement using feedback" is a process of collecting responses and evaluations from users and using them to improve the system's performance and response accuracy.

[0166] This invention is a system that generates automated responses that reflect the user's emotional state. The following describes a specific implementation of this system.

[0167] Users upload past information to the system in CSV or JSON format. This information includes survey data and questionnaire results. The server receives the uploaded data and stores it in a database such as MySQL or PostgreSQL. The data is then used to train machine learning models.

[0168] The server uses the Python language and libraries such as TensorFlow and PyTorch to extract item characteristics from information and generate foundational data for training machine learning models. Natural language processing techniques are used for preprocessing, such as tokenization and stop word removal. Specifically, sentiment analysis libraries (e.g., VADER and TextBlob) are used to evaluate the emotional state associated with each item.

[0169] Based on these analysis results, the server trains an AI model to generate an automated response optimized for the user. The generated response is also adjusted to take the user's emotional state into consideration. The terminal then presents the generated response to the user. This process uses web application frontend technologies (e.g., React or Vue.js).

[0170] For example, suppose a user asks, "Tell me about your recent experiences." In this case, the server refers to past data and sentiment analysis results and generates a response such as, "I am very happy with my recent experiences." This response is then presented to the user via the terminal.

[0171] Furthermore, users can provide feedback on the responses. The server collects this feedback and stores it in a database. This feedback is used to retrain the model, continuously improving the system's accuracy.

[0172] An example of a prompt is, "Generate a sentiment-sensitive response to a question about past visit experiences." This prompt serves as an instruction for the generative AI model to generate an appropriate response.

[0173] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0174] Step 1:

[0175] Users upload historical data to the system. This data is provided in CSV or JSON format and includes survey results and questionnaire response records. The server receives this data and processes it for storage in the database. Specifically, it verifies the data format and converts it to a different format if necessary. The input is a CSV or JSON file, and the output is a structured database entry.

[0176] Step 2:

[0177] The server begins preprocessing the stored data. The input is the raw data in the database, and the output is data in a format usable for training machine learning models. Specific operations include imputing missing values, normalizing the data, tokenizing text data, and removing stop words. This transforms the data into a format suitable for input to machine learning models.

[0178] Step 3:

[0179] The server performs sentiment analysis using tokenized text data. The input is pre-processed text data, and the output is a sentiment score for each data point. Here, a sentiment analysis library (e.g., VADER or TextBlob) is used to assign sentiment categories such as positive, negative, and neutral. The results will influence subsequent model training.

[0180] Step 4:

[0181] The server begins training a machine learning model. The input is feature data with sentiment scores assigned to it, and the output is the trained model. Deep learning frameworks such as TensorFlow and PyTorch are used to learn patterns within the dataset. Feature extraction and parameter optimization are the main processes here.

[0182] Step 5:

[0183] When the server receives new question data, it begins natural language processing analysis. The input is the new question text, and the output is keywords and grammatical patterns as analysis data. This allows the server to understand the structure and context of the received question and prepares it for response generation.

[0184] Step 6:

[0185] The server uses the analysis results and trained model to generate automated responses. The input is the question's analysis data and the trained model, and the output is the generated response text. Here, the model constructs the optimal answer to the question based on the patterns it has learned. Since the results of sentiment analysis are also considered, the response will be sentiment-appropriate.

[0186] Step 7:

[0187] The terminal presents the user with a response generated from the server. The input here is the response text sent from the server, and the output is the display of the response on the user's screen. A web application framework is used, and the response is displayed via a user interface.

[0188] (Application Example 2)

[0189] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0190] Traditional ad delivery systems have struggled to select the most relevant ads considering users' emotional states, making it difficult to effectively improve click-through rates and conversion rates. Therefore, providing personalized ads that respond to users' diverse emotional states is a key challenge.

[0191] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0192] In this invention, the server includes means for collecting historical data, means for analyzing the data and extracting features, and means for analyzing the user's emotional state. This makes it possible to provide personalized advertisements based on the user's current emotions.

[0193] "Means for collecting past data" refers to technical means used to efficiently collect information about a user's past search and browsing history.

[0194] "Methods for analyzing data and extracting features" refer to technical methods for analyzing collected data and clarifying features that are useful for selecting advertisements.

[0195] "Means for training and generating models that predict answers using machine learning techniques" refers to means for building machine learning models that predict the best answers or advertisements to questions based on collected and analyzed data.

[0196] "Means for automatically generating answers to new questions using generated models" refers to technical means that use trained machine learning models to automatically generate predictive answers and selection results that correspond to new questions and situations.

[0197] "Means of presenting automatically generated answers" refers to interfaces or systems for visually displaying predicted or selected answers or advertisements to users.

[0198] "Means for analyzing a user's emotional state" refers to technical means for analyzing and determining a user's emotional state based on their search and browsing history and current behavior.

[0199] "A means of displaying personalized information based on the user's emotional state" refers to a technological means that selects the most relevant information and advertisements based on the analyzed emotional state of the user and displays them at the appropriate time.

[0200] "Means of providing an interface that allows users to review and make choices as needed" refers to technical means of providing an operating screen or system that allows users to review the information presented to them and make choices of whether to support or oppose it.

[0201] "Means for collecting user feedback and improving the accuracy of the machine learning model and sentiment analysis" refers to technical means for analyzing user reactions and feedback and continuously improving the machine learning model and sentiment analysis algorithm based on that data.

[0202] The system for implementing the present invention consists of a user-operable terminal and a server that performs data processing. The system collects the user's past browsing and search history and performs data analysis based on this data. The server uses this data to operate an emotion analysis engine to analyze the user's emotional state. It also uses machine learning technology to select the most suitable advertisements and information to provide to the user.

[0203] Specifically, the server uses Python and leverages the Google Cloud Natural Language API for sentiment analysis. TensorFlow is used to build the machine learning model, enabling real-time model operation. This allows for the immediate generation of personalized information, such as advertisements, based on the user's emotions, which are then displayed on the device through a user interface developed with React Native.

[0204] For example, if a user frequently views travel-related content, the server associates this data with their emotional state and predicts that their interest in travel is increasing. Based on this, highly relevant advertisements, such as the latest campaign information from travel agencies or recommendations for restaurants in travel destinations, are displayed on the device. In this process, user feedback is collected by the server and contributes to the continuous improvement of the model. An example of a prompt message might be, "Based on this user's recent browsing history, please suggest products that they might be interested in."

[0205] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0206] Step 1:

[0207] The server collects past search history and browsing data from the user's terminal. The input is the user's website visit history and search queries, and the output is a dataset containing this historical information. This dataset is stored in a database for later analysis.

[0208] Step 2:

[0209] The server analyzes the collected dataset and extracts features. The input is a dataset of historical information, and the output is a list of features indicating the user's interests and concerns. The analysis uses Python and includes the detection of keywords and trends using NLP (Natural Language Processing) techniques.

[0210] Step 3:

[0211] The server uses the Google Cloud Natural Language API to analyze the user's emotional state. The input is a list of extracted features, and the output is emotional state label data. The API calculates an emotional score and records the results in a database.

[0212] Step 4:

[0213] The server predicts the most suitable advertisement for the user using a generative AI model trained with TensorFlow. The input is a list of emotional state labels and features, and the output is the content of the selected advertisement. This model prioritizes advertisements based on emotional state.

[0214] Step 5:

[0215] The device displays selected advertisements through an interface built with React Native. The input is the content of the advertisement, and the output is the advertisement presented to the user visually. The device includes actions such as notifying the user of the advertisement and displaying the information in an easy-to-read format.

[0216] Step 6:

[0217] The user reviews the presented advertisements and chooses to click on those that interest them. The input is the presented advertisements, and the output is the user's click data and feedback information. This feedback is used for subsequent data collection.

[0218] Step 7:

[0219] The server collects user feedback and uses it to improve the machine learning model. Inputs include click data and feedback information, while outputs are new training data aimed at improving the model's accuracy. This allows the model to be continuously updated, enabling the delivery of more personalized services.

[0220] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0221] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0222] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0223] [Second Embodiment]

[0224] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0225] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0226] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0227] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0228] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0229] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0230] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0231] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0232] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0233] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0234] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0235] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0236] This invention provides a system for users to conduct research efficiently. In one embodiment, the system has a core technology that collects past research data and generates a machine learning model based on it. The overall system embodiment is described in detail below.

[0237] First, users upload past survey questions and answer data to the system. This data is used as a basis for predicting the most appropriate answers for each survey question. The server receives this data and stores it in a database, thereby efficiently managing data resources.

[0238] Next, the server applies natural language processing techniques to the stored data to analyze the characteristics of the questions. This analysis extracts important keywords and contexts contained in the questions, preparing them for use in subsequent machine learning modeling.

[0239] During model training, the server uses machine learning algorithms to analyze past response data and generate a predictive model. This predictive model forms the basis for automatically generating appropriate answers when users receive similar questions.

[0240] Subsequently, using the generated model, the server builds an agent that can automatically generate answers to new questions. This agent has the ability to quickly derive answers based on the analysis results of the received questions when a user receives a new survey.

[0241] As a concrete example of its use, if a user receives a survey that includes the question "Where did you last visit?", the server automatically generates "Tokyo" as the appropriate answer based on past response patterns, and the terminal presents this to the user. The user can then review this answer, submit it as is if satisfied, and make corrections if necessary.

[0242] Furthermore, the system continuously collects user feedback and uses it to refine its machine learning model, thereby improving the accuracy of its responses. This feedback loop allows the system to improve over time, providing more accurate answers. Specifically, the server can use feedback data to optimize model parameters, further enhancing the accuracy of analysis and predictions.

[0243] The following describes the processing flow.

[0244] Step 1:

[0245] Users prepare past survey questions and answer data, access the system, and upload this data.

[0246] Step 2:

[0247] The server receives the uploaded data, converts each piece of data into a parseable format, and stores it in the database. This stored data is then used as the foundational data for model training.

[0248] Step 3:

[0249] The server analyzes the stored survey data using natural language processing techniques. To understand the characteristics of the questions, it performs sentence tokenization and morphological analysis to extract the meaning and context of the questions.

[0250] Step 4:

[0251] The server applies machine learning algorithms based on past response data to train a model to predict the optimal answer. This provides a statistical understanding of question and answer patterns.

[0252] Step 5:

[0253] The server builds an AI agent based on a pre-trained model. This agent automatically generates appropriate answers when a new survey question is presented.

[0254] Step 6:

[0255] When a user receives a new survey, their device sends the information to the server. The server then uses an AI agent to analyze the received questions.

[0256] Step 7:

[0257] The server automatically generates the optimal answer to a question via an AI agent and presents it to the user by sending it to the terminal.

[0258] Step 8:

[0259] The user reviews the automatically generated response, makes any necessary corrections, and submits the final response to the server.

[0260] Step 9:

[0261] The server receives feedback from users and uses that information to readjust the machine learning model. This feedback improves the accuracy of the model.

[0262] (Example 1)

[0263] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0264] In conventional information processing systems, it was difficult to effectively utilize past data and generate rapid and appropriate responses to new information. Furthermore, user feedback was not adequately utilized to improve the accuracy of the generated responses. As a result, challenges existed in information management and response accuracy.

[0265] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0266] In this invention, the server includes means for collecting past information, means for analyzing the information and extracting its characteristics, means for training and generating a model to predict the optimal response using machine learning techniques, means for automatically generating a response to new information using the generated model, means for displaying the automatically generated response, means for applying natural language processing techniques to analyze the context of the object to be processed and identify necessary elements, means for managing information input via the user and efficiently utilizing the stored information, and means for presenting the machine learning-generated response to the user and providing the user with means to confirm or modify it. This enables efficient information management, advanced response generation, and continuous improvement of prediction accuracy through the use of user feedback.

[0267] "Past information" refers to data and records collected previously.

[0268] "Means of collection" refers to the technologies and methods used to collect, store, and manage information.

[0269] "Analysis methods" refer to techniques for analyzing information and extracting important characteristics and patterns from it.

[0270] "Characteristics" refer to the important elements and features of data revealed through analysis.

[0271] "Machine learning technology" refers to artificial intelligence techniques that train models based on data and perform estimations and classifications based on new data.

[0272] "Response" refers to the reply or output generated in response to user input.

[0273] A "model" refers to a mathematical or statistical representation built on data for a specific purpose.

[0274] "Automatic generation means" refers to a process in which a program automatically generates a response based on the input information.

[0275] "Display means" refers to devices or methods for visualizing and presenting the generated response to the user.

[0276] "Natural language processing technology" refers to computer technology that understands and responds to human language.

[0277] "Feedback" refers to evaluations and correction information provided by users.

[0278] "Improving accuracy" refers to processes and methods for increasing the accuracy of models and responses.

[0279] This invention is a system that efficiently processes information and generates highly accurate responses. First, the user uploads past information to the system. This information can be provided as a file in formats such as CSV or Excel. Information uploading is done through a web interface, which allows the user to operate the system intuitively.

[0280] The server receives this information and stores it in a database system (e.g., MySQL or PostgreSQL). The stored information undergoes duplicate checking and data cleansing before being prepared for analysis. The server uses natural language processing techniques, such as Python's NLTK library, to extract important keywords and context from the information. This clarifies the characteristics of the information and lays the groundwork for the next step.

[0281] Next, the server trains a predictive model using machine learning libraries such as Scikit-learn and TensorFlow based on the analyzed information. This model is designed to generate the optimal response to new information based on past data. By generating a predictive model, the server can be equipped with an agent that automatically creates responses to new questions.

[0282] The terminal presents the generated response to the user. The user can view the presented response, use it if appropriate, and make corrections as needed. As a specific example, when the user receives a question such as "What was the previous service used?", the server generates a response such as "Online streaming service" and displays it to the user through the terminal.

[0283] Furthermore, the server continuously collects feedback from the user and improves the accuracy of the model based on this. By leveraging the information obtained from the feedback, the server optimizes the model parameters, enabling more accurate predictions. Through this feedback loop, the system is improved over time, enabling the provision of higher-performance responses.

[0284] As a specific example of a prompt sentence, an input such as "Please make the following proposal based on the recently used services" can be considered. In response to this prompt, the system can generate an optimal response through past data and meet the user's needs.

[0285] The flow of the specific process in Example 1 will be described using FIG. 11.

[0286] Step 1:

[0287] The user uploads the information collected in the past to the system through the web interface. The input data is provided in CSV or Excel file format. In this step, to let the system import the digital data at hand, the user selects a file and clicks the upload button. As output, the data is transferred to the server and can be used in the next step.

[0288] Step 2:

[0289] The server verifies the received data and stores it in the database system. The input data consists of data uploaded by users. Data cleansing is performed to remove duplicates and standardize the format, ensuring data integrity. The output is clean, standardized data stored in the database.

[0290] Step 3:

[0291] The server retrieves data stored in the database and analyzes it using natural language processing techniques. The input is text data from the database. Here, the Python NLTK library is used to extract important keywords and context. The output is a dataset from which characteristics have been extracted. This set is used in the next machine learning step.

[0292] Step 4:

[0293] The server uses the analyzed characteristic data to train a predictive model with a machine learning algorithm. The input data consists of characteristic data and historical response data. Using libraries such as Scikit-learn and TensorFlow, the algorithm learns and generates a model that predicts the optimal response. The output is the trained predictive model. This model serves as the foundation for generating new responses.

[0294] Step 5:

[0295] The server automatically generates responses to newly received information using the generated predictive model. The input is the newly provided question data. The model analyzes this question data and generates an appropriate predicted response. The output is the generated response text. This response is the first step towards being presented to the user.

[0296] Step 6:

[0297] The terminal displays the response received from the server to the user. The user reviews this response on the screen. The input is the response data from the server, and the output is the text information displayed on the screen. The user can check whether the response is appropriate and modify it if necessary.

[0298] Step 7:

[0299] The server collects user feedback and uses it to improve the accuracy of its predictive model. The input is the feedback information provided by the user. Based on this feedback, the model is retrained, and the output is a new predictive model with improved accuracy. This enhances the overall response accuracy and adaptability of the system.

[0300] (Application Example 1)

[0301] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0302] In today's information-saturated world, it is difficult for users to efficiently acquire information and content optimized for their individual interests. Furthermore, there is a need for technology that utilizes collected data to provide automated responses and recommendations tailored to the user. This invention aims to improve the user experience by providing a system that accurately recommends new content based on the user's past behavior and interests.

[0303] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0304] In this invention, the server includes means for collecting past survey information, means for analyzing the survey information to extract question characteristics, means for training and generating a model for predicting an optimal response to a question using machine learning techniques, and means for recommending new content based on the past interests of the user. As a result, it becomes possible to automatically propose content optimized for each user's interests, enabling the user to more easily access appropriate information.

[0305] "Survey information" refers to data on past questions and answers collected from users.

[0306] "Question characteristics" refer to important keywords and contexts underlying the questions extracted from the survey information.

[0307] "Machine learning techniques" is a general term for algorithms and methods for constructing a prediction model using data and automating specific tasks.

[0308] "Model for predicting an optimal response" is a machine learning model trained to generate an appropriate answer to a new question based on the collected data.

[0309] "Past interests of the user" refers to information based on the interests and browsing history previously shown by the user.

[0310] "Means for recommending content" is a process or technology for proposing new information and media based on the past behavior data of the user.

[0311] A system for implementing this invention includes a server, a user terminal, and certain communication means. The server collects past survey information, analyzes data, trains a machine learning model, and provides responses and content recommendations customized for each user.

[0312] First, past survey information provided by the user's terminal is uploaded to the server. The server receives this information and automatically stores it in a database. Next, the server uses a natural language processing engine (e.g., spaCy or BERT model) to analyze the characteristics of the questions and extract important keywords and context. This process makes the context of the survey data clearer.

[0313] The server uses machine learning frameworks such as TensorFlow to train and generate models that predict optimal responses based on historical data. This response generation model can handle new survey questions and automatically provides accurate answers to newly received questions from the user's device. Furthermore, it suggests new content through a smartphone app based on the user's past interests. Recommended content is determined by server-side data analysis based on the user's viewing history and feedback.

[0314] For example, if a user has recently watched many action movies, the system is likely to recommend "the latest action movies" and related "thriller movies." This recommendation is based on up-to-date data to accurately reflect the user's interests. Examples of prompts include, "What are your recent favorite genres? What are this week's recommendations from that genre?" The server sends the results, combining this information with the model, to the user's terminal, where the user can review the suggestions and choose responses or options as needed.

[0315] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0316] Step 1:

[0317] The terminal collects past survey information and uploads it to the server. The input is data of questions and answers recorded by the user, and the output is raw data stored in a database on the server. This data serves as a foundation for facilitating responses to the user's future questions.

[0318] Step 2:

[0319] The server analyzes the uploaded survey information using a natural language processing engine. The input is raw data stored in a database, and the output is analyzed data with the context of the questions and keywords extracted. The server uses spaCy and BERT models to perform analysis to reveal important features within the data.

[0320] Step 3:

[0321] The server trains a predictive model using a machine learning framework. The input is the previously analyzed data, and the output is a trained generative AI model to respond to new questions. The server leverages TensorFlow to build the model based on past response patterns.

[0322] Step 4:

[0323] When a user enters a new question, the server uses a generative AI model to generate the optimal response. The input is the user's new question, and the output is the generated answer. By applying the model, the server provides answers quickly and accurately.

[0324] Step 5:

[0325] The server analyzes the user's past interests and recommends content. The input is the user's viewing history, and the output is a list of recommended content. The server performs data analysis to select information that best matches the user's interests.

[0326] Step 6:

[0327] The user's terminal displays recommended content and generated responses. Input consists of responses and recommended content sent from the server, while output is the information displayed on the terminal. The user can review the provided information and make evaluations or modifications as needed.

[0328] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0329] This invention provides an automated response generation system that efficiently collects and analyzes survey data, while also considering the user's emotional state. This system utilizes machine learning technology to provide rapid and appropriate responses to survey questions, integrating emotional analysis throughout the process.

[0330] Specifically, it begins with users uploading past survey data to the system. The server receives this data and stores it in a database. The stored data is used to train machine learning models and forms the basis for understanding the characteristics of the questions and the user's past response patterns.

[0331] Next, the server analyzes the question data using natural language processing techniques and extracts important features. Because an emotion engine is integrated into the system, it can also analyze the user's emotional state related to each question. For example, in response to a question like, "How do you feel about this product?", the emotion engine evaluates the user's emotions and adds the result to the data.

[0332] The resulting machine learning model not only predicts the optimal answer to a question, but also takes into account sentiment data analyzed through the sentiment engine. This allows it to provide answers that are tailored to the user's emotions.

[0333] For example, if a user is asked, "How was your recent visit experience?", the server generates a response such as "Very satisfied" based on past data and sentiment analysis results, and presents it to the user via their device. The user can review this response, and if they are satisfied with the accuracy of the automatically generated response, which also depends on their emotions, they can submit it as is.

[0334] Furthermore, user feedback is continuously collected and used to improve machine learning models and sentiment engines. The servers leverage user feedback to continuously improve model accuracy and strengthen the foundation for providing a more personalized experience.

[0335] The following describes the processing flow.

[0336] Step 1:

[0337] Users access the survey platform and upload their past survey questions and answers to the system. This data is then used as the system's knowledge base.

[0338] Step 2:

[0339] The server receives the uploaded data and stores it in a database. The data is structured and organized into a format that will be used later for model training.

[0340] Step 3:

[0341] The server analyzes the question data using natural language processing techniques. This analysis extracts important linguistic features and patterns. Additionally, sentiment data related to each question is collected as supplementary information.

[0342] Step 4:

[0343] The server uses an emotion engine to evaluate the user's emotional state. It performs text analysis to identify emotional tendencies from the user's responses and comments, and then assigns an evaluation such as positive, negative, or neutral.

[0344] Step 5:

[0345] The server uses these analysis results to train a machine learning model. The model also considers sentiment analysis results when predicting the best answer to a question.

[0346] Step 6:

[0347] When a new survey question arrives, the device sends the information to the server. The server analyzes the question using an AI agent and an emotion engine.

[0348] Step 7:

[0349] The server automatically generates responses based on the analysis results, adjusting them to suit the user's emotional state. These automatically generated responses are carefully designed to be emotionally appropriate.

[0350] Step 8:

[0351] The device presents the generated response to the user. The user can review the response and, if satisfied with its content, submit it. If necessary, the user can modify the response.

[0352] Step 9:

[0353] The server receives user feedback and analyzes the data to improve the accuracy of machine learning models and the sentiment engine. Through this feedback, the models are refined, improving the overall system performance.

[0354] (Example 2)

[0355] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0356] Conventional information gathering and analysis systems generated responses without considering the user's emotional state, making it difficult to provide appropriate and personalized responses. Furthermore, they were unable to effectively utilize user feedback on the generated responses, hindering system improvement.

[0357] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0358] In this invention, the server includes means for accumulating past information, means for analyzing the information to extract the characteristics of items, and means for evaluating the user's emotional state related to the items and incorporating the results of the emotional analysis into the response. This enables the generation of automated responses that correspond to the user's emotions and continuous improvement of the system's accuracy using feedback.

[0359] "Information storage" refers to the systematic storage of data collected in the past, making it available for later analysis and processing.

[0360] "Extracting item characteristics" refers to identifying distinctive information from a specific dataset and extracting it in a format useful for analysis.

[0361] "Evaluating the user's emotional state related to an item" means analyzing the emotions a user has towards a particular item and expressing them using numerical values ​​or categories.

[0362] "Incorporating the results of sentiment analysis" means reflecting the results of an analysis of the user's emotional state in the generated responses and the operation of the system.

[0363] "Automated response generation" refers to the process of automatically creating answers to predefined items using machine learning models or similar tools.

[0364] "Continuous system accuracy improvement using feedback" is a process of collecting responses and evaluations from users and using them to improve the system's performance and response accuracy.

[0365] This invention is a system that generates automated responses that reflect the user's emotional state. The following describes a specific implementation of this system.

[0366] Users upload past information to the system in CSV or JSON format. This information includes survey data and questionnaire results. The server receives the uploaded data and stores it in a database such as MySQL or PostgreSQL. The data is then used to train machine learning models.

[0367] The server uses the Python language and libraries such as TensorFlow and PyTorch to extract item characteristics from information and generate foundational data for training machine learning models. Natural language processing techniques are used for preprocessing, such as tokenization and stop word removal. Specifically, sentiment analysis libraries (e.g., VADER and TextBlob) are used to evaluate the emotional state associated with each item.

[0368] Based on these analysis results, the server trains an AI model to generate an automated response optimized for the user. The generated response is also adjusted to take the user's emotional state into consideration. The terminal then presents the generated response to the user. This process uses web application frontend technologies (e.g., React or Vue.js).

[0369] For example, suppose a user asks, "Tell me about your recent experiences." In this case, the server refers to past data and sentiment analysis results and generates a response such as, "I am very happy with my recent experiences." This response is then presented to the user via the terminal.

[0370] Furthermore, users can provide feedback on the responses. The server collects this feedback and stores it in a database. This feedback is used to retrain the model, continuously improving the system's accuracy.

[0371] An example of a prompt is, "Generate a sentiment-sensitive response to a question about past visit experiences." This prompt serves as an instruction for the generative AI model to generate an appropriate response.

[0372] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0373] Step 1:

[0374] Users upload historical data to the system. This data is provided in CSV or JSON format and includes survey results and questionnaire response records. The server receives this data and processes it for storage in the database. Specifically, it verifies the data format and converts it to a different format if necessary. The input is a CSV or JSON file, and the output is a structured database entry.

[0375] Step 2:

[0376] The server begins preprocessing the stored data. The input is the raw data in the database, and the output is data in a format usable for training machine learning models. Specific operations include imputing missing values, normalizing the data, tokenizing text data, and removing stop words. This transforms the data into a format suitable for input to machine learning models.

[0377] Step 3:

[0378] The server performs sentiment analysis using tokenized text data. The input is pre-processed text data, and the output is a sentiment score for each data point. Here, a sentiment analysis library (e.g., VADER or TextBlob) is used to assign sentiment categories such as positive, negative, and neutral. The results will influence subsequent model training.

[0379] Step 4:

[0380] The server begins training a machine learning model. The input is feature data with sentiment scores assigned to it, and the output is the trained model. Deep learning frameworks such as TensorFlow and PyTorch are used to learn patterns within the dataset. Feature extraction and parameter optimization are the main processes here.

[0381] Step 5:

[0382] When the server receives new question data, it begins natural language processing analysis. The input is the new question text, and the output is keywords and grammatical patterns as analysis data. This allows the server to understand the structure and context of the received question and prepares it for response generation.

[0383] Step 6:

[0384] The server uses the analysis results and trained model to generate automated responses. The input is the question's analysis data and the trained model, and the output is the generated response text. Here, the model constructs the optimal answer to the question based on the patterns it has learned. Since the results of sentiment analysis are also considered, the response will be sentiment-appropriate.

[0385] Step 7:

[0386] The terminal presents the user with a response generated from the server. The input here is the response text sent from the server, and the output is the display of the response on the user's screen. A web application framework is used, and the response is displayed via a user interface.

[0387] (Application Example 2)

[0388] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0389] Traditional ad delivery systems have struggled to select the most relevant ads considering users' emotional states, making it difficult to effectively improve click-through rates and conversion rates. Therefore, providing personalized ads that respond to users' diverse emotional states is a key challenge.

[0390] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0391] In this invention, the server includes means for collecting historical data, means for analyzing the data and extracting features, and means for analyzing the user's emotional state. This makes it possible to provide personalized advertisements based on the user's current emotions.

[0392] "Means for collecting past data" refers to technical means used to efficiently collect information about a user's past search and browsing history.

[0393] "Methods for analyzing data and extracting features" refer to technical methods for analyzing collected data and clarifying features that are useful for selecting advertisements.

[0394] "Means for training and generating models that predict answers using machine learning techniques" refers to means for building machine learning models that predict the best answers or advertisements to questions based on collected and analyzed data.

[0395] "Means for automatically generating answers to new questions using generated models" refers to technical means that use trained machine learning models to automatically generate predictive answers and selection results that correspond to new questions and situations.

[0396] "Means of presenting automatically generated answers" refers to interfaces or systems for visually displaying predicted or selected answers or advertisements to users.

[0397] "Means for analyzing a user's emotional state" refers to technical means for analyzing and determining a user's emotional state based on their search and browsing history and current behavior.

[0398] "A means of displaying personalized information based on the user's emotional state" refers to a technological means that selects the most relevant information and advertisements based on the analyzed emotional state of the user and displays them at the appropriate time.

[0399] "Means of providing an interface that allows users to review and make choices as needed" refers to technical means of providing an operating screen or system that allows users to review the information presented to them and make choices of whether to support or oppose it.

[0400] "Means for collecting user feedback and improving the accuracy of the machine learning model and sentiment analysis" refers to technical means for analyzing user reactions and feedback and continuously improving the machine learning model and sentiment analysis algorithm based on that data.

[0401] The system for implementing the present invention consists of a user-operable terminal and a server that performs data processing. The system collects the user's past browsing and search history and performs data analysis based on this data. The server uses this data to operate an emotion analysis engine to analyze the user's emotional state. It also uses machine learning technology to select the most suitable advertisements and information to provide to the user.

[0402] Specifically, the server uses Python and leverages the Google Cloud Natural Language API for sentiment analysis. TensorFlow is used to build the machine learning model, enabling real-time model operation. This allows for the immediate generation of personalized information, such as advertisements, based on the user's emotions, and displays them on the device through a user interface developed with React Native.

[0403] For example, if a user frequently views travel-related content, the server associates this data with their emotional state and predicts that their interest in travel is increasing. Based on this, highly relevant advertisements, such as the latest campaign information from travel agencies or recommendations for restaurants in travel destinations, are displayed on the device. In this process, user feedback is collected by the server and contributes to the continuous improvement of the model. An example of a prompt message might be, "Based on this user's recent browsing history, please suggest products that they might be interested in."

[0404] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0405] Step 1:

[0406] The server collects past search history and browsing data from the user's terminal. The input is the user's website visit history and search queries, and the output is a dataset containing this historical information. This dataset is stored in a database for later analysis.

[0407] Step 2:

[0408] The server analyzes the collected dataset and extracts features. The input is a dataset of historical information, and the output is a list of features indicating the user's interests and concerns. The analysis uses Python and includes the detection of keywords and trends using NLP (Natural Language Processing) techniques.

[0409] Step 3:

[0410] The server uses the Google Cloud Natural Language API to analyze the user's emotional state. The input is a list of extracted features, and the output is emotional state label data. The API calculates an emotional score and records the results in a database.

[0411] Step 4:

[0412] The server predicts the most suitable advertisement for the user using a generative AI model trained with TensorFlow. The input is a list of emotional state labels and features, and the output is the content of the selected advertisement. This model prioritizes advertisements based on emotional state.

[0413] Step 5:

[0414] The device displays selected advertisements through an interface built with React Native. The input is the content of the advertisement, and the output is the advertisement presented to the user visually. The device includes actions such as notifying the user of the advertisement and displaying the information in an easy-to-read format.

[0415] Step 6:

[0416] The user reviews the presented advertisements and chooses to click on those that interest them. The input is the presented advertisements, and the output is the user's click data and feedback information. This feedback is used for subsequent data collection.

[0417] Step 7:

[0418] The server collects user feedback and uses it to improve the machine learning model. Inputs include click data and feedback information, while outputs are new training data aimed at improving the model's accuracy. This allows the model to be continuously updated, enabling the delivery of more personalized services.

[0419] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0420] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0421] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0422] [Third Embodiment]

[0423] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0424] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0425] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0426] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0427] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0428] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0429] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0430] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0431] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0432] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0433] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0434] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0435] This invention provides a system for users to conduct research efficiently. In one embodiment, the system has a core technology that collects past research data and generates a machine learning model based on it. The overall system embodiment is described in detail below.

[0436] First, users upload past survey questions and answer data to the system. This data is used as a basis for predicting the most appropriate answers for each survey question. The server receives this data and stores it in a database, thereby efficiently managing data resources.

[0437] Next, the server applies natural language processing techniques to the stored data to analyze the characteristics of the questions. This analysis extracts important keywords and contexts contained in the questions, preparing them for use in subsequent machine learning modeling.

[0438] During model training, the server uses machine learning algorithms to analyze past response data and generate a predictive model. This predictive model forms the basis for automatically generating appropriate answers when users receive similar questions.

[0439] Subsequently, using the generated model, the server builds an agent that can automatically generate answers to new questions. This agent has the ability to quickly derive answers based on the analysis results of the received questions when a user receives a new survey.

[0440] As a concrete example of its use, if a user receives a survey that includes the question "Where did you last visit?", the server automatically generates "Tokyo" as the appropriate answer based on past response patterns, and the terminal presents this to the user. The user can then review this answer, submit it as is if satisfied, and make corrections if necessary.

[0441] Furthermore, the system continuously collects user feedback and uses it to refine its machine learning model, thereby improving the accuracy of its responses. This feedback loop allows the system to improve over time, providing more accurate answers. Specifically, the server can use feedback data to optimize model parameters, further enhancing the accuracy of analysis and predictions.

[0442] The following describes the processing flow.

[0443] Step 1:

[0444] Users prepare past survey questions and answer data, access the system, and upload this data.

[0445] Step 2:

[0446] The server receives the uploaded data, converts each piece of data into a parseable format, and stores it in the database. This stored data is then used as the foundational data for model training.

[0447] Step 3:

[0448] The server analyzes the stored survey data using natural language processing techniques. To understand the characteristics of the questions, it performs sentence tokenization and morphological analysis to extract the meaning and context of the questions.

[0449] Step 4:

[0450] The server applies machine learning algorithms based on past response data to train a model to predict the optimal answer. This provides a statistical understanding of question and answer patterns.

[0451] Step 5:

[0452] The server builds an AI agent based on a pre-trained model. This agent automatically generates appropriate answers when a new survey question is presented.

[0453] Step 6:

[0454] When a user receives a new survey, their device sends the information to the server. The server then uses an AI agent to analyze the received questions.

[0455] Step 7:

[0456] The server automatically generates the optimal answer to a question via an AI agent and presents it to the user by sending it to the terminal.

[0457] Step 8:

[0458] The user reviews the automatically generated response, makes any necessary corrections, and submits the final response to the server.

[0459] Step 9:

[0460] The server receives feedback from users and uses that information to readjust the machine learning model. This feedback improves the accuracy of the model.

[0461] (Example 1)

[0462] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0463] In conventional information processing systems, it was difficult to effectively utilize past data and generate rapid and appropriate responses to new information. Furthermore, user feedback was not adequately utilized to improve the accuracy of the generated responses. As a result, challenges existed in information management and response accuracy.

[0464] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0465] In this invention, the server includes means for collecting past information, means for analyzing the information and extracting its characteristics, means for training and generating a model to predict the optimal response using machine learning techniques, means for automatically generating a response to new information using the generated model, means for displaying the automatically generated response, means for applying natural language processing techniques to analyze the context of the object to be processed and identify necessary elements, means for managing information input via the user and efficiently utilizing the stored information, and means for presenting the machine learning-generated response to the user and providing the user with means to confirm or modify it. This enables efficient information management, advanced response generation, and continuous improvement of prediction accuracy through the use of user feedback.

[0466] "Past information" refers to data and records collected previously.

[0467] "Means of collection" refers to the technologies and methods used to collect, store, and manage information.

[0468] "Analysis methods" refer to techniques for analyzing information and extracting important characteristics and patterns from it.

[0469] "Characteristics" refer to the important elements and features of data revealed through analysis.

[0470] "Machine learning technology" refers to artificial intelligence techniques that train models based on data and perform estimations and classifications based on new data.

[0471] "Response" refers to the reply or output generated in response to user input.

[0472] A "model" refers to a mathematical or statistical representation built on data for a specific purpose.

[0473] "Automatic generation means" refers to a process in which a program automatically generates a response based on the input information.

[0474] "Display means" refers to devices or methods for visualizing and presenting the generated response to the user.

[0475] "Natural language processing technology" refers to computer technology that understands and responds to human language.

[0476] "Feedback" refers to evaluations and correction information provided by users.

[0477] "Improving accuracy" refers to processes and methods for increasing the accuracy of models and responses.

[0478] This invention is a system that efficiently processes information and generates highly accurate responses. First, the user uploads past information to the system. This information can be provided as a file in formats such as CSV or Excel. Information uploading is done through a web interface, which allows the user to operate the system intuitively.

[0479] The server receives this information and stores it in a database system (e.g., MySQL or PostgreSQL). The stored information undergoes duplicate checking and data cleansing before being prepared for analysis. The server uses natural language processing techniques, such as Python's NLTK library, to extract important keywords and context from the information. This clarifies the characteristics of the information and lays the groundwork for the next step.

[0480] Next, the server trains a predictive model using machine learning libraries such as Scikit-learn and TensorFlow based on the analyzed information. This model is designed to generate the optimal response to new information based on past data. By generating a predictive model, the server can be equipped with an agent that automatically creates responses to new questions.

[0481] The terminal presents the user with a generated response. The user can review the presented response, use it if appropriate, and modify it as needed. For example, if the user is asked a question such as "What service did you use last time?", the server will generate a response such as "Online streaming service" and display it to the user through the terminal.

[0482] Furthermore, the server continuously collects user feedback and uses it to improve the model's accuracy. By utilizing the information obtained from the feedback, the server optimizes the model parameters, enabling more accurate predictions. This feedback loop allows the system to improve over time, enabling it to provide higher-performance responses.

[0483] A concrete example of a prompt message might be input such as, "Please suggest the next service based on the services I've recently used." In response to this prompt, the system can generate the optimal response based on past data and meet the user's needs.

[0484] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0485] Step 1:

[0486] Users upload previously collected information to the system via a web interface. Input data is provided in CSV or Excel file format. In this step, users select files and press the upload button to import their digital data into the system. The data is then transferred to the server and becomes available for use in the next step.

[0487] Step 2:

[0488] The server verifies the received data and stores it in the database system. The input data consists of data uploaded by users. Data cleansing is performed to remove duplicates and standardize the format, ensuring data integrity. The output is clean, standardized data stored in the database.

[0489] Step 3:

[0490] The server retrieves data stored in the database and analyzes it using natural language processing techniques. The input is text data from the database. Here, the Python NLTK library is used to extract important keywords and context. The output is a dataset from which characteristics have been extracted. This set is used in the next machine learning step.

[0491] Step 4:

[0492] The server uses the analyzed characteristic data to train a predictive model with a machine learning algorithm. The input data consists of characteristic data and historical response data. Using libraries such as Scikit-learn and TensorFlow, the algorithm learns and generates a model that predicts the optimal response. The output is the trained predictive model. This model serves as the foundation for generating new responses.

[0493] Step 5:

[0494] The server automatically generates responses to newly received information using the generated predictive model. The input is the newly provided question data. The model analyzes this question data and generates an appropriate predicted response. The output is the generated response text. This response is the first step towards being presented to the user.

[0495] Step 6:

[0496] The terminal displays the response received from the server to the user. The user reviews this response on the screen. The input is the response data from the server, and the output is the text information displayed on the screen. The user can check whether the response is appropriate and modify it if necessary.

[0497] Step 7:

[0498] The server collects user feedback and uses it to improve the accuracy of its predictive model. The input is the feedback information provided by the user. Based on this feedback, the model is retrained, and the output is a new predictive model with improved accuracy. This enhances the overall response accuracy and adaptability of the system.

[0499] (Application Example 1)

[0500] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0501] In today's information-saturated world, it is difficult for users to efficiently acquire information and content optimized for their individual interests. Furthermore, there is a need for technology that utilizes collected data to provide automated responses and recommendations tailored to the user. This invention aims to improve the user experience by providing a system that accurately recommends new content based on the user's past behavior and interests.

[0502] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0503] In this invention, the server includes means for collecting past survey information, means for analyzing the survey information to extract the characteristics of the questions, means for training and generating a model that predicts the optimal response to the questions using machine learning technology, and means for recommending new content based on the user's past interests. This enables the automatic suggestion of content optimized for each user's individual interests, making it easier for users to access appropriate information.

[0504] "Survey information" refers to data on past questions and answers collected from users.

[0505] "Characteristics of the questions" refer to the important keywords and contexts hidden within the questions extracted from the survey information.

[0506] "Machine learning technology" is a general term for algorithms and methods used to build predictive models using data and automate specific tasks.

[0507] A "model that predicts the optimal response" is a machine learning model trained to generate appropriate answers to new questions based on collected data.

[0508] "User's past interests" refers to information based on the user's previously indicated interests and browsing history.

[0509] "Means of recommending content" refers to processes or technologies for suggesting new information or media based on users' past behavioral data.

[0510] The system for carrying out this invention comprises a server, a user terminal, and certain communication means. The server collects past research information, analyzes the data, trains machine learning models, and provides customized responses and content recommendations to each user.

[0511] First, past survey information provided by the user's terminal is uploaded to the server. The server receives this information and automatically stores it in a database. Next, the server uses a natural language processing engine (e.g., spaCy or BERT model) to analyze the characteristics of the questions and extract important keywords and context. This process makes the context of the survey data clearer.

[0512] The server uses machine learning frameworks such as TensorFlow to train and generate models that predict optimal responses based on historical data. This response generation model can handle new survey questions and automatically provides accurate answers to newly received questions from the user's device. Furthermore, it suggests new content through a smartphone app based on the user's past interests. Recommended content is determined by server-side data analysis based on the user's viewing history and feedback.

[0513] For example, if a user has recently watched many action movies, the system is likely to recommend "the latest action movies" and related "thriller movies." This recommendation is based on up-to-date data to accurately reflect the user's interests. Examples of prompts include, "What are your recent favorite genres? What are this week's recommendations from that genre?" The server sends the results, combining this information with the model, to the user's terminal, where the user can review the suggestions and choose responses or options as needed.

[0514] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0515] Step 1:

[0516] The terminal collects past survey information and uploads it to the server. The input is data of questions and answers recorded by the user, and the output is raw data stored in a database on the server. This data serves as a foundation for facilitating responses to the user's future questions.

[0517] Step 2:

[0518] The server analyzes the uploaded survey information using a natural language processing engine. The input is raw data stored in a database, and the output is analyzed data with the context of the questions and keywords extracted. The server uses spaCy and BERT models to perform analysis to reveal important features within the data.

[0519] Step 3:

[0520] The server trains a predictive model using a machine learning framework. The input is the previously analyzed data, and the output is a trained generative AI model to respond to new questions. The server leverages TensorFlow to build the model based on past response patterns.

[0521] Step 4:

[0522] When a user enters a new question, the server uses a generative AI model to generate the optimal response. The input is the user's new question, and the output is the generated answer. By applying the model, the server provides answers quickly and accurately.

[0523] Step 5:

[0524] The server analyzes the user's past interests and recommends content. The input is the user's viewing history, and the output is a list of recommended content. The server performs data analysis to select information that best matches the user's interests.

[0525] Step 6:

[0526] The user's terminal displays recommended content and generated responses. Input consists of responses and recommended content sent from the server, while output is the information displayed on the terminal. The user can review the provided information and make evaluations or modifications as needed.

[0527] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0528] This invention provides an automated response generation system that efficiently collects and analyzes survey data, while also considering the user's emotional state. This system utilizes machine learning technology to provide rapid and appropriate responses to survey questions, integrating emotional analysis throughout the process.

[0529] Specifically, it begins with users uploading past survey data to the system. The server receives this data and stores it in a database. The stored data is used to train machine learning models and forms the basis for understanding the characteristics of the questions and the user's past response patterns.

[0530] Next, the server analyzes the question data using natural language processing techniques and extracts important features. Because an emotion engine is integrated into the system, it can also analyze the user's emotional state related to each question. For example, in response to a question like, "How do you feel about this product?", the emotion engine evaluates the user's emotions and adds the result to the data.

[0531] The resulting machine learning model not only predicts the optimal answer to a question, but also takes into account sentiment data analyzed through the sentiment engine. This allows it to provide answers that are tailored to the user's emotions.

[0532] For example, if a user is asked, "How was your recent visit experience?", the server generates a response such as "Very satisfied" based on past data and sentiment analysis results, and presents it to the user via their device. The user can review this response, and if they are satisfied with the accuracy of the automatically generated response, which also depends on their emotions, they can submit it as is.

[0533] Furthermore, user feedback is continuously collected and used to improve machine learning models and sentiment engines. The servers leverage user feedback to continuously improve model accuracy and strengthen the foundation for providing a more personalized experience.

[0534] The following describes the processing flow.

[0535] Step 1:

[0536] Users access the survey platform and upload their past survey questions and answers to the system. This data is then used as the system's knowledge base.

[0537] Step 2:

[0538] The server receives the uploaded data and stores it in a database. The data is structured and organized into a format that will be used later for model training.

[0539] Step 3:

[0540] The server analyzes the question data using natural language processing techniques. This analysis extracts important linguistic features and patterns. Additionally, sentiment data related to each question is collected as supplementary information.

[0541] Step 4:

[0542] The server uses an emotion engine to evaluate the user's emotional state. It performs text analysis to identify emotional tendencies from the user's responses and comments, and then assigns an evaluation such as positive, negative, or neutral.

[0543] Step 5:

[0544] The server uses these analysis results to train a machine learning model. The model also considers sentiment analysis results when predicting the best answer to a question.

[0545] Step 6:

[0546] When a new survey question arrives, the device sends the information to the server. The server analyzes the question using an AI agent and an emotion engine.

[0547] Step 7:

[0548] The server automatically generates responses based on the analysis results, adjusting them to suit the user's emotional state. These automatically generated responses are carefully designed to be emotionally appropriate.

[0549] Step 8:

[0550] The device presents the generated response to the user. The user can review the response and, if satisfied with its content, submit it. If necessary, the user can modify the response.

[0551] Step 9:

[0552] The server receives user feedback and analyzes the data to improve the accuracy of machine learning models and the sentiment engine. Through this feedback, the models are refined, improving the overall system performance.

[0553] (Example 2)

[0554] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0555] Conventional information gathering and analysis systems generated responses without considering the user's emotional state, making it difficult to provide appropriate and personalized responses. Furthermore, they were unable to effectively utilize user feedback on the generated responses, hindering system improvement.

[0556] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0557] In this invention, the server includes means for accumulating past information, means for analyzing the information to extract the characteristics of items, and means for evaluating the user's emotional state related to the items and incorporating the results of the emotional analysis into the response. This enables the generation of automated responses that correspond to the user's emotions and continuous improvement of the system's accuracy using feedback.

[0558] "Information storage" refers to the systematic storage of data collected in the past, making it available for later analysis and processing.

[0559] "Extracting item characteristics" refers to identifying distinctive information from a specific dataset and extracting it in a format useful for analysis.

[0560] "Evaluating the user's emotional state related to an item" means analyzing the emotions a user has towards a particular item and expressing them using numerical values ​​or categories.

[0561] "Incorporating the results of sentiment analysis" means reflecting the results of an analysis of the user's emotional state in the generated responses and the operation of the system.

[0562] "Automated response generation" refers to the process of automatically creating answers to predefined items using machine learning models or similar tools.

[0563] "Continuous system accuracy improvement using feedback" is a process of collecting responses and evaluations from users and using them to improve the system's performance and response accuracy.

[0564] This invention is a system that generates automated responses that reflect the user's emotional state. The following describes a specific implementation of this system.

[0565] Users upload past information to the system in CSV or JSON format. This information includes survey data and questionnaire results. The server receives the uploaded data and stores it in a database such as MySQL or PostgreSQL. The data is then used to train machine learning models.

[0566] The server uses the Python language and libraries such as TensorFlow and PyTorch to extract item characteristics from information and generate foundational data for training machine learning models. Natural language processing techniques are used for preprocessing, such as tokenization and stop word removal. Specifically, sentiment analysis libraries (e.g., VADER and TextBlob) are used to evaluate the emotional state associated with each item.

[0567] Based on these analysis results, the server trains an AI model to generate an automated response optimized for the user. The generated response is also adjusted to take the user's emotional state into consideration. The terminal then presents the generated response to the user. This process uses web application frontend technologies (e.g., React or Vue.js).

[0568] For example, suppose a user asks, "Tell me about your recent experiences." In this case, the server refers to past data and sentiment analysis results and generates a response such as, "I am very happy with my recent experiences." This response is then presented to the user via the terminal.

[0569] Furthermore, users can provide feedback on the responses. The server collects this feedback and stores it in a database. This feedback is used to retrain the model, continuously improving the system's accuracy.

[0570] An example of a prompt is, "Generate a sentiment-sensitive response to a question about past visit experiences." This prompt serves as an instruction for the generative AI model to generate an appropriate response.

[0571] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0572] Step 1:

[0573] Users upload historical data to the system. This data is provided in CSV or JSON format and includes survey results and questionnaire response records. The server receives this data and processes it for storage in the database. Specifically, it verifies the data format and converts it to a different format if necessary. The input is a CSV or JSON file, and the output is a structured database entry.

[0574] Step 2:

[0575] The server begins preprocessing the stored data. The input is the raw data in the database, and the output is data in a format usable for training machine learning models. Specific operations include imputing missing values, normalizing the data, tokenizing text data, and removing stop words. This transforms the data into a format suitable for input to machine learning models.

[0576] Step 3:

[0577] The server performs sentiment analysis using tokenized text data. The input is pre-processed text data, and the output is a sentiment score for each data point. Here, a sentiment analysis library (e.g., VADER or TextBlob) is used to assign sentiment categories such as positive, negative, and neutral. The results will influence subsequent model training.

[0578] Step 4:

[0579] The server begins training a machine learning model. The input is feature data with sentiment scores assigned to it, and the output is the trained model. Deep learning frameworks such as TensorFlow and PyTorch are used to learn patterns within the dataset. Feature extraction and parameter optimization are the main processes here.

[0580] Step 5:

[0581] When the server receives new question data, it begins natural language processing analysis. The input is the new question text, and the output is keywords and grammatical patterns as analysis data. This allows the server to understand the structure and context of the received question and prepares it for response generation.

[0582] Step 6:

[0583] The server uses the analysis results and trained model to generate automated responses. The input is the question's analysis data and the trained model, and the output is the generated response text. Here, the model constructs the optimal answer to the question based on the patterns it has learned. Since the results of sentiment analysis are also considered, the response will be sentiment-appropriate.

[0584] Step 7:

[0585] The terminal presents the user with a response generated from the server. The input here is the response text sent from the server, and the output is the display of the response on the user's screen. A web application framework is used, and the response is displayed via a user interface.

[0586] (Application Example 2)

[0587] Next, we will explain Application Example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0588] Traditional ad delivery systems have struggled to select the most relevant ads considering users' emotional states, making it difficult to effectively improve click-through rates and conversion rates. Therefore, providing personalized ads that respond to users' diverse emotional states is a key challenge.

[0589] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0590] In this invention, the server includes means for collecting historical data, means for analyzing the data and extracting features, and means for analyzing the user's emotional state. This makes it possible to provide personalized advertisements based on the user's current emotions.

[0591] "Means for collecting past data" refers to technical means used to efficiently collect information about a user's past search and browsing history.

[0592] "Methods for analyzing data and extracting features" refer to technical methods for analyzing collected data and clarifying features that are useful for selecting advertisements.

[0593] "Means for training and generating models that predict answers using machine learning techniques" refers to means for building machine learning models that predict the best answers or advertisements to questions based on collected and analyzed data.

[0594] "Means for automatically generating answers to new questions using generated models" refers to technical means that use trained machine learning models to automatically generate predictive answers and selection results that correspond to new questions and situations.

[0595] "Means of presenting automatically generated answers" refers to interfaces or systems for visually displaying predicted or selected answers or advertisements to users.

[0596] "Means for analyzing a user's emotional state" refers to technical means for analyzing and determining a user's emotional state based on their search and browsing history and current behavior.

[0597] "A means of displaying personalized information based on the user's emotional state" refers to a technological means that selects the most relevant information and advertisements based on the analyzed emotional state of the user and displays them at the appropriate time.

[0598] "Means of providing an interface that allows users to review and make choices as needed" refers to technical means of providing an operating screen or system that allows users to review the information presented to them and make choices of whether to support or oppose it.

[0599] "Means for collecting user feedback and improving the accuracy of the machine learning model and sentiment analysis" refers to technical means for analyzing user reactions and feedback and continuously improving the machine learning model and sentiment analysis algorithm based on that data.

[0600] The system for implementing the present invention consists of a user-operable terminal and a server that performs data processing. The system collects the user's past browsing and search history and performs data analysis based on this data. The server uses this data to operate an emotion analysis engine to analyze the user's emotional state. It also uses machine learning technology to select the most suitable advertisements and information to provide to the user.

[0601] Specifically, the server uses Python and leverages the Google Cloud Natural Language API for sentiment analysis. TensorFlow is used to build the machine learning model, enabling real-time model operation. This allows for the immediate generation of personalized information, such as advertisements, based on the user's emotions, and displays them on the device through a user interface developed with React Native.

[0602] For example, if a user frequently views travel-related content, the server associates this data with their emotional state and predicts that their interest in travel is increasing. Based on this, highly relevant advertisements, such as the latest campaign information from travel agencies or recommendations for restaurants in travel destinations, are displayed on the device. In this process, user feedback is collected by the server and contributes to the continuous improvement of the model. An example of a prompt message might be, "Based on this user's recent browsing history, please suggest products that they might be interested in."

[0603] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0604] Step 1:

[0605] The server collects past search history and browsing data from the user's terminal. The input is the user's website visit history and search queries, and the output is a dataset containing this historical information. This dataset is stored in a database for later analysis.

[0606] Step 2:

[0607] The server analyzes the collected dataset and extracts features. The input is a dataset of historical information, and the output is a list of features indicating the user's interests and concerns. The analysis uses Python and includes the detection of keywords and trends using NLP (Natural Language Processing) techniques.

[0608] Step 3:

[0609] The server uses the Google Cloud Natural Language API to analyze the user's emotional state. The input is a list of extracted features, and the output is emotional state label data. The API calculates an emotional score and records the results in a database.

[0610] Step 4:

[0611] The server predicts the most suitable advertisement for the user using a generative AI model trained with TensorFlow. The input is a list of emotional state labels and features, and the output is the content of the selected advertisement. This model prioritizes advertisements based on emotional state.

[0612] Step 5:

[0613] The device displays selected advertisements through an interface built with React Native. The input is the content of the advertisement, and the output is the advertisement presented to the user visually. The device includes actions such as notifying the user of the advertisement and displaying the information in an easy-to-read format.

[0614] Step 6:

[0615] The user reviews the presented advertisements and chooses to click on those that interest them. The input is the presented advertisements, and the output is the user's click data and feedback information. This feedback is used for subsequent data collection.

[0616] Step 7:

[0617] The server collects user feedback and uses it to improve the machine learning model. Inputs include click data and feedback information, while outputs are new training data aimed at improving the model's accuracy. This allows the model to be continuously updated, enabling the delivery of more personalized services.

[0618] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0619] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0620] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0621] [Fourth Embodiment]

[0622] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0623] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0624] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0625] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0626] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0627] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0628] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0629] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0630] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0631] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0632] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0633] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0634] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0635] This invention provides a system for users to conduct research efficiently. In one embodiment, the system has a core technology that collects past research data and generates a machine learning model based on it. The overall system embodiment is described in detail below.

[0636] First, users upload past survey questions and answer data to the system. This data is used as a basis for predicting the most appropriate answers for each survey question. The server receives this data and stores it in a database, thereby efficiently managing data resources.

[0637] Next, the server applies natural language processing techniques to the stored data to analyze the characteristics of the questions. This analysis extracts important keywords and contexts contained in the questions, preparing them for use in subsequent machine learning modeling.

[0638] During model training, the server uses machine learning algorithms to analyze past response data and generate a predictive model. This predictive model forms the basis for automatically generating appropriate answers when users receive similar questions.

[0639] Subsequently, using the generated model, the server builds an agent that can automatically generate answers to new questions. This agent has the ability to quickly derive answers based on the analysis results of the received questions when a user receives a new survey.

[0640] As a concrete example of its use, if a user receives a survey that includes the question "Where did you last visit?", the server automatically generates "Tokyo" as the appropriate answer based on past response patterns, and the terminal presents this to the user. The user can then review this answer, submit it as is if satisfied, and make corrections if necessary.

[0641] Furthermore, the system continuously collects user feedback and uses it to refine its machine learning model, thereby improving the accuracy of its responses. This feedback loop allows the system to improve over time, providing more accurate answers. Specifically, the server can use feedback data to optimize model parameters, further enhancing the accuracy of analysis and predictions.

[0642] The following describes the processing flow.

[0643] Step 1:

[0644] Users prepare past survey questions and answer data, access the system, and upload this data.

[0645] Step 2:

[0646] The server receives the uploaded data, converts each piece of data into a parseable format, and stores it in the database. This stored data is then used as the foundational data for model training.

[0647] Step 3:

[0648] The server analyzes the stored survey data using natural language processing techniques. To understand the characteristics of the questions, it performs sentence tokenization and morphological analysis to extract the meaning and context of the questions.

[0649] Step 4:

[0650] The server applies machine learning algorithms based on past response data to train a model to predict the optimal answer. This provides a statistical understanding of question and answer patterns.

[0651] Step 5:

[0652] The server builds an AI agent based on a pre-trained model. This agent automatically generates appropriate answers when a new survey question is presented.

[0653] Step 6:

[0654] When a user receives a new survey, their device sends the information to the server. The server then uses an AI agent to analyze the received questions.

[0655] Step 7:

[0656] The server automatically generates the optimal answer to a question via an AI agent and presents it to the user by sending it to the terminal.

[0657] Step 8:

[0658] The user reviews the automatically generated response, makes any necessary corrections, and submits the final response to the server.

[0659] Step 9:

[0660] The server receives feedback from users and uses that information to readjust the machine learning model. This feedback improves the accuracy of the model.

[0661] (Example 1)

[0662] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0663] In conventional information processing systems, it was difficult to effectively utilize past data and generate rapid and appropriate responses to new information. Furthermore, user feedback was not adequately utilized to improve the accuracy of the generated responses. As a result, challenges existed in information management and response accuracy.

[0664] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0665] In this invention, the server includes means for collecting past information, means for analyzing the information and extracting its characteristics, means for training and generating a model to predict the optimal response using machine learning techniques, means for automatically generating a response to new information using the generated model, means for displaying the automatically generated response, means for applying natural language processing techniques to analyze the context of the object to be processed and identify necessary elements, means for managing information input via the user and efficiently utilizing the stored information, and means for presenting the machine learning-generated response to the user and providing the user with means to confirm or modify it. This enables efficient information management, advanced response generation, and continuous improvement of prediction accuracy through the use of user feedback.

[0666] "Past information" refers to data and records collected previously.

[0667] "Means of collection" refers to the technologies and methods used to collect, store, and manage information.

[0668] "Analysis methods" refer to techniques for analyzing information and extracting important characteristics and patterns from it.

[0669] "Characteristics" refer to the important elements and features of data revealed through analysis.

[0670] "Machine learning technology" refers to artificial intelligence techniques that train models based on data and perform estimations and classifications based on new data.

[0671] "Response" refers to the reply or output generated in response to user input.

[0672] A "model" refers to a mathematical or statistical representation built on data for a specific purpose.

[0673] "Automatic generation means" refers to a process in which a program automatically generates a response based on the input information.

[0674] "Display means" refers to devices or methods for visualizing and presenting the generated response to the user.

[0675] "Natural language processing technology" refers to computer technology that understands and responds to human language.

[0676] "Feedback" refers to evaluations and correction information provided by users.

[0677] "Improving accuracy" refers to processes and methods for increasing the accuracy of models and responses.

[0678] This invention is a system that efficiently processes information and generates highly accurate responses. First, the user uploads past information to the system. This information can be provided as a file in formats such as CSV or Excel. Information uploading is done through a web interface, which allows the user to operate the system intuitively.

[0679] The server receives this information and stores it in a database system (e.g., MySQL or PostgreSQL). The stored information undergoes duplicate checking and data cleansing before being prepared for analysis. The server uses natural language processing techniques, such as Python's NLTK library, to extract important keywords and context from the information. This clarifies the characteristics of the information and lays the groundwork for the next step.

[0680] Next, the server trains a predictive model using machine learning libraries such as Scikit-learn and TensorFlow based on the analyzed information. This model is designed to generate the optimal response to new information based on past data. By generating a predictive model, the server can be equipped with an agent that automatically creates responses to new questions.

[0681] The terminal presents the user with a generated response. The user can review the presented response, use it if appropriate, and modify it as needed. For example, if the user is asked a question such as "What service did you use last time?", the server will generate a response such as "Online streaming service" and display it to the user through the terminal.

[0682] Furthermore, the server continuously collects user feedback and uses it to improve the model's accuracy. By utilizing the information obtained from the feedback, the server optimizes the model parameters, enabling more accurate predictions. This feedback loop allows the system to improve over time, enabling it to provide higher-performance responses.

[0683] A concrete example of a prompt message might be input such as, "Please suggest the next service based on the services I've recently used." In response to this prompt, the system can generate the optimal response based on past data and meet the user's needs.

[0684] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0685] Step 1:

[0686] Users upload previously collected information to the system via a web interface. Input data is provided in CSV or Excel file format. In this step, users select files and press the upload button to import their digital data into the system. The data is then transferred to the server and becomes available for use in the next step.

[0687] Step 2:

[0688] The server verifies the received data and stores it in the database system. The input data consists of data uploaded by users. Data cleansing is performed to remove duplicates and standardize the format, ensuring data integrity. The output is clean, standardized data stored in the database.

[0689] Step 3:

[0690] The server retrieves data stored in the database and analyzes it using natural language processing techniques. The input is text data from the database. Here, the Python NLTK library is used to extract important keywords and context. The output is a dataset from which characteristics have been extracted. This set is used in the next machine learning step.

[0691] Step 4:

[0692] The server uses the analyzed characteristic data to train a predictive model with a machine learning algorithm. The input data consists of characteristic data and historical response data. Using libraries such as Scikit-learn and TensorFlow, the algorithm learns and generates a model that predicts the optimal response. The output is the trained predictive model. This model serves as the foundation for generating new responses.

[0693] Step 5:

[0694] The server automatically generates responses to newly received information using the generated predictive model. The input is the newly provided question data. The model analyzes this question data and generates an appropriate predicted response. The output is the generated response text. This response is the first step towards being presented to the user.

[0695] Step 6:

[0696] The terminal displays the response received from the server to the user. The user reviews this response on the screen. The input is the response data from the server, and the output is the text information displayed on the screen. The user can check whether the response is appropriate and modify it if necessary.

[0697] Step 7:

[0698] The server collects user feedback and uses it to improve the accuracy of its predictive model. The input is the feedback information provided by the user. Based on this feedback, the model is retrained, and the output is a new predictive model with improved accuracy. This enhances the overall response accuracy and adaptability of the system.

[0699] (Application Example 1)

[0700] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0701] In today's information-saturated world, it is difficult for users to efficiently acquire information and content optimized for their individual interests. Furthermore, there is a need for technology that utilizes collected data to provide automated responses and recommendations tailored to the user. This invention aims to improve the user experience by providing a system that accurately recommends new content based on the user's past behavior and interests.

[0702] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0703] In this invention, the server includes means for collecting past survey information, means for analyzing the survey information to extract the characteristics of the questions, means for training and generating a model that predicts the optimal response to the questions using machine learning technology, and means for recommending new content based on the user's past interests. This enables the automatic suggestion of content optimized for each user's individual interests, making it easier for users to access appropriate information.

[0704] "Survey information" refers to data on past questions and answers collected from users.

[0705] "Characteristics of the questions" refer to the important keywords and contexts hidden within the questions extracted from the survey information.

[0706] "Machine learning technology" is a general term for algorithms and methods used to build predictive models using data and automate specific tasks.

[0707] A "model that predicts the optimal response" is a machine learning model trained to generate appropriate answers to new questions based on collected data.

[0708] "User's past interests" refers to information based on the user's previously indicated interests and browsing history.

[0709] "Means of recommending content" refers to processes or technologies for suggesting new information or media based on users' past behavioral data.

[0710] The system for carrying out this invention comprises a server, a user terminal, and certain communication means. The server collects past research information, analyzes the data, trains machine learning models, and provides customized responses and content recommendations to each user.

[0711] First, past survey information provided by the user's terminal is uploaded to the server. The server receives this information and automatically stores it in a database. Next, the server uses a natural language processing engine (e.g., spaCy or BERT model) to analyze the characteristics of the questions and extract important keywords and context. This process makes the context of the survey data clearer.

[0712] The server uses machine learning frameworks such as TensorFlow to train and generate models that predict optimal responses based on historical data. This response generation model can handle new survey questions and automatically provides accurate answers to newly received questions from the user's device. Furthermore, it suggests new content through a smartphone app based on the user's past interests. Recommended content is determined by server-side data analysis based on the user's viewing history and feedback.

[0713] For example, if a user has recently watched many action movies, the system is likely to recommend "the latest action movies" and related "thriller movies." This recommendation is based on up-to-date data to accurately reflect the user's interests. Examples of prompts include, "What are your recent favorite genres? What are this week's recommendations from that genre?" The server sends the results, combining this information with the model, to the user's terminal, where the user can review the suggestions and choose responses or options as needed.

[0714] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0715] Step 1:

[0716] The terminal collects past survey information and uploads it to the server. The input is data of questions and answers recorded by the user, and the output is raw data stored in a database on the server. This data serves as a foundation for facilitating responses to the user's future questions.

[0717] Step 2:

[0718] The server analyzes the uploaded survey information using a natural language processing engine. The input is raw data stored in a database, and the output is analyzed data with the context of the questions and keywords extracted. The server uses spaCy and BERT models to perform analysis to reveal important features within the data.

[0719] Step 3:

[0720] The server trains a predictive model using a machine learning framework. The input is the previously analyzed data, and the output is a trained generative AI model to respond to new questions. The server leverages TensorFlow to build the model based on past response patterns.

[0721] Step 4:

[0722] When a user enters a new question, the server uses a generative AI model to generate the optimal response. The input is the user's new question, and the output is the generated answer. By applying the model, the server provides answers quickly and accurately.

[0723] Step 5:

[0724] The server analyzes the user's past interests and recommends content. The input is the user's viewing history, and the output is a list of recommended content. The server performs data analysis to select information that best matches the user's interests.

[0725] Step 6:

[0726] The user's terminal displays recommended content and generated responses. Input consists of responses and recommended content sent from the server, while output is the information displayed on the terminal. The user can review the provided information and make evaluations or modifications as needed.

[0727] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0728] This invention provides an automated response generation system that efficiently collects and analyzes survey data, while also considering the user's emotional state. This system utilizes machine learning technology to provide rapid and appropriate responses to survey questions, integrating emotional analysis throughout the process.

[0729] Specifically, it begins with users uploading past survey data to the system. The server receives this data and stores it in a database. The stored data is used to train machine learning models and forms the basis for understanding the characteristics of the questions and the user's past response patterns.

[0730] Next, the server analyzes the question data using natural language processing techniques and extracts important features. Because an emotion engine is integrated into the system, it can also analyze the user's emotional state related to each question. For example, in response to a question like, "How do you feel about this product?", the emotion engine evaluates the user's emotions and adds the result to the data.

[0731] The resulting machine learning model not only predicts the optimal answer to a question, but also takes into account sentiment data analyzed through the sentiment engine. This allows it to provide answers that are tailored to the user's emotions.

[0732] For example, if a user is asked, "How was your recent visit experience?", the server generates a response such as "Very satisfied" based on past data and sentiment analysis results, and presents it to the user via their device. The user can review this response, and if they are satisfied with the accuracy of the automatically generated response, which also depends on their emotions, they can submit it as is.

[0733] Furthermore, user feedback is continuously collected and used to improve machine learning models and sentiment engines. The servers leverage user feedback to continuously improve model accuracy and strengthen the foundation for providing a more personalized experience.

[0734] The following describes the processing flow.

[0735] Step 1:

[0736] Users access the survey platform and upload their past survey questions and answers to the system. This data is then used as the system's knowledge base.

[0737] Step 2:

[0738] The server receives the uploaded data and stores it in a database. The data is structured and organized into a format that will be used later for model training.

[0739] Step 3:

[0740] The server analyzes the question data using natural language processing techniques. This analysis extracts important linguistic features and patterns. Additionally, sentiment data related to each question is collected as supplementary information.

[0741] Step 4:

[0742] The server uses an emotion engine to evaluate the user's emotional state. It performs text analysis to identify emotional tendencies from the user's responses and comments, and then assigns an evaluation such as positive, negative, or neutral.

[0743] Step 5:

[0744] The server uses these analysis results to train a machine learning model. The model also considers sentiment analysis results when predicting the best answer to a question.

[0745] Step 6:

[0746] When a new survey question arrives, the device sends the information to the server. The server analyzes the question using an AI agent and an emotion engine.

[0747] Step 7:

[0748] The server automatically generates responses based on the analysis results, adjusting them to suit the user's emotional state. These automatically generated responses are carefully designed to be emotionally appropriate.

[0749] Step 8:

[0750] The device presents the generated response to the user. The user can review the response and, if satisfied with its content, submit it. If necessary, the user can modify the response.

[0751] Step 9:

[0752] The server receives user feedback and analyzes the data to improve the accuracy of machine learning models and the sentiment engine. Through this feedback, the models are refined, improving the overall system performance.

[0753] (Example 2)

[0754] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0755] Conventional information gathering and analysis systems generated responses without considering the user's emotional state, making it difficult to provide appropriate and personalized responses. Furthermore, they were unable to effectively utilize user feedback on the generated responses, hindering system improvement.

[0756] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0757] In this invention, the server includes means for accumulating past information, means for analyzing the information to extract the characteristics of items, and means for evaluating the user's emotional state related to the items and incorporating the results of the emotional analysis into the response. This enables the generation of automated responses that correspond to the user's emotions and continuous improvement of the system's accuracy using feedback.

[0758] "Information storage" refers to the systematic storage of data collected in the past, making it available for later analysis and processing.

[0759] "Extracting item characteristics" refers to identifying distinctive information from a specific dataset and extracting it in a format useful for analysis.

[0760] "Evaluating the user's emotional state related to an item" means analyzing the emotions a user has towards a particular item and expressing them using numerical values ​​or categories.

[0761] "Incorporating the results of sentiment analysis" means reflecting the results of an analysis of the user's emotional state in the generated responses and the operation of the system.

[0762] "Automated response generation" refers to the process of automatically creating answers to predefined items using machine learning models or similar tools.

[0763] "Continuous system accuracy improvement using feedback" is a process of collecting responses and evaluations from users and using them to improve the system's performance and response accuracy.

[0764] This invention is a system that generates automated responses that reflect the user's emotional state. The following describes a specific implementation of this system.

[0765] Users upload past information to the system in CSV or JSON format. This information includes survey data and questionnaire results. The server receives the uploaded data and stores it in a database such as MySQL or PostgreSQL. The data is then used to train machine learning models.

[0766] The server uses the Python language and libraries such as TensorFlow and PyTorch to extract item characteristics from information and generate foundational data for training machine learning models. Natural language processing techniques are used for preprocessing, such as tokenization and stop word removal. Specifically, sentiment analysis libraries (e.g., VADER and TextBlob) are used to evaluate the emotional state associated with each item.

[0767] Based on these analysis results, the server trains an AI model to generate an automated response optimized for the user. The generated response is also adjusted to take the user's emotional state into consideration. The terminal then presents the generated response to the user. This process uses web application frontend technologies (e.g., React or Vue.js).

[0768] For example, suppose a user asks, "Tell me about your recent experiences." In this case, the server refers to past data and sentiment analysis results and generates a response such as, "I am very happy with my recent experiences." This response is then presented to the user via the terminal.

[0769] Furthermore, users can provide feedback on the responses. The server collects this feedback and stores it in a database. This feedback is used to retrain the model, continuously improving the system's accuracy.

[0770] An example of a prompt is, "Generate a sentiment-sensitive response to a question about past visit experiences." This prompt serves as an instruction for the generative AI model to generate an appropriate response.

[0771] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0772] Step 1:

[0773] Users upload historical data to the system. This data is provided in CSV or JSON format and includes survey results and questionnaire response records. The server receives this data and processes it for storage in the database. Specifically, it verifies the data format and converts it to a different format if necessary. The input is a CSV or JSON file, and the output is a structured database entry.

[0774] Step 2:

[0775] The server begins preprocessing the stored data. The input is the raw data in the database, and the output is data in a format usable for training machine learning models. Specific operations include imputing missing values, normalizing the data, tokenizing text data, and removing stop words. This transforms the data into a format suitable for input to machine learning models.

[0776] Step 3:

[0777] The server performs sentiment analysis using tokenized text data. The input is pre-processed text data, and the output is a sentiment score for each data point. Here, a sentiment analysis library (e.g., VADER or TextBlob) is used to assign sentiment categories such as positive, negative, and neutral. The results will influence subsequent model training.

[0778] Step 4:

[0779] The server begins training a machine learning model. The input is feature data with sentiment scores assigned to it, and the output is the trained model. Deep learning frameworks such as TensorFlow and PyTorch are used to learn patterns within the dataset. Feature extraction and parameter optimization are the main processes here.

[0780] Step 5:

[0781] When the server receives new question data, it begins natural language processing analysis. The input is the new question text, and the output is keywords and grammatical patterns as analysis data. This allows the server to understand the structure and context of the received question and prepares it for response generation.

[0782] Step 6:

[0783] The server uses the analysis results and trained model to generate automated responses. The input is the question's analysis data and the trained model, and the output is the generated response text. Here, the model constructs the optimal answer to the question based on the patterns it has learned. Since the results of sentiment analysis are also considered, the response will be sentiment-appropriate.

[0784] Step 7:

[0785] The terminal presents the user with a response generated from the server. The input here is the response text sent from the server, and the output is the display of the response on the user's screen. A web application framework is used, and the response is displayed via a user interface.

[0786] (Application Example 2)

[0787] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0788] Traditional ad delivery systems have struggled to select the most relevant ads considering users' emotional states, making it difficult to effectively improve click-through rates and conversion rates. Therefore, providing personalized ads that respond to users' diverse emotional states is a key challenge.

[0789] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0790] In this invention, the server includes means for collecting historical data, means for analyzing the data and extracting features, and means for analyzing the user's emotional state. This makes it possible to provide personalized advertisements based on the user's current emotions.

[0791] "Means for collecting past data" refers to technical means used to efficiently collect information about a user's past search and browsing history.

[0792] "Methods for analyzing data and extracting features" refer to technical methods for analyzing collected data and clarifying features that are useful for selecting advertisements.

[0793] "Means for training and generating models that predict answers using machine learning techniques" refers to means for building machine learning models that predict the best answers or advertisements to questions based on collected and analyzed data.

[0794] "Means for automatically generating answers to new questions using generated models" refers to technical means that use trained machine learning models to automatically generate predictive answers and selection results that correspond to new questions and situations.

[0795] "Means of presenting automatically generated answers" refers to interfaces or systems for visually displaying predicted or selected answers or advertisements to users.

[0796] "Means for analyzing a user's emotional state" refers to technical means for analyzing and determining a user's emotional state based on their search and browsing history and current behavior.

[0797] "A means of displaying personalized information based on the user's emotional state" refers to a technological means that selects the most relevant information and advertisements based on the analyzed emotional state of the user and displays them at the appropriate time.

[0798] "Means of providing an interface that allows users to review and make choices as needed" refers to technical means of providing an operating screen or system that allows users to review the information presented to them and make choices of whether to support or oppose it.

[0799] "Means for collecting user feedback and improving the accuracy of the machine learning model and sentiment analysis" refers to technical means for analyzing user reactions and feedback and continuously improving the machine learning model and sentiment analysis algorithm based on that data.

[0800] The system for implementing the present invention consists of a user-operable terminal and a server that performs data processing. The system collects the user's past browsing and search history and performs data analysis based on this data. The server uses this data to operate an emotion analysis engine to analyze the user's emotional state. It also uses machine learning technology to select the most suitable advertisements and information to provide to the user.

[0801] Specifically, the server uses Python and leverages the Google Cloud Natural Language API for sentiment analysis. TensorFlow is used to build the machine learning model, enabling real-time model operation. This allows for the immediate generation of personalized information, such as advertisements, based on the user's emotions, and displays them on the device through a user interface developed with React Native.

[0802] For example, if a user frequently views travel-related content, the server associates this data with their emotional state and predicts that their interest in travel is increasing. Based on this, highly relevant advertisements, such as the latest campaign information from travel agencies or recommendations for restaurants in travel destinations, are displayed on the device. In this process, user feedback is collected by the server and contributes to the continuous improvement of the model. An example of a prompt message might be, "Based on this user's recent browsing history, please suggest products that they might be interested in."

[0803] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0804] Step 1:

[0805] The server collects past search history and browsing data from the user's terminal. The input is the user's website visit history and search queries, and the output is a dataset containing this historical information. This dataset is stored in a database for later analysis.

[0806] Step 2:

[0807] The server analyzes the collected dataset and extracts features. The input is a dataset of historical information, and the output is a list of features indicating the user's interests and concerns. The analysis uses Python and includes the detection of keywords and trends using NLP (Natural Language Processing) techniques.

[0808] Step 3:

[0809] The server uses the Google Cloud Natural Language API to analyze the user's emotional state. The input is a list of extracted features, and the output is emotional state label data. The API calculates an emotional score and records the results in a database.

[0810] Step 4:

[0811] The server predicts the most suitable advertisement for the user using a generative AI model trained with TensorFlow. The input is a list of emotional state labels and features, and the output is the content of the selected advertisement. This model prioritizes advertisements based on emotional state.

[0812] Step 5:

[0813] The device displays selected advertisements through an interface built with React Native. The input is the content of the advertisement, and the output is the advertisement presented to the user visually. The device includes actions such as notifying the user of the advertisement and displaying the information in an easy-to-read format.

[0814] Step 6:

[0815] The user reviews the presented advertisements and chooses to click on those that interest them. The input is the presented advertisements, and the output is the user's click data and feedback information. This feedback is used for subsequent data collection.

[0816] Step 7:

[0817] The server collects user feedback and uses it to improve the machine learning model. Inputs include click data and feedback information, while outputs are new training data aimed at improving the model's accuracy. This allows the model to be continuously updated, enabling the delivery of more personalized services.

[0818] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0819] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0820] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0821] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0822] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0823] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0824] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0825] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0826] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0827] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0828] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0829] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[0830] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0831] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[0832] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0833] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0834] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0835] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0836] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0837] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0838] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[0839] The following is further disclosed regarding the embodiments described above.

[0840] (Claim 1)

[0841] Means for collecting past survey data,

[0842] A means for analyzing the aforementioned survey data and extracting the characteristics of the questions,

[0843] A means for training and generating a model that predicts the optimal answer to a question using machine learning techniques,

[0844] A means of automatically generating answers to new survey questions using the generated model,

[0845] The means for presenting the automatically generated answer,

[0846] A system that includes this.

[0847] (Claim 2)

[0848] The system according to claim 1, further comprising means for providing an interface that allows the user to review the response generated by the means described above and modify it as necessary.

[0849] (Claim 3)

[0850] The system according to claim 1, further comprising means for collecting user feedback and improving the accuracy of the machine learning model.

[0851] "Example 1"

[0852] (Claim 1)

[0853] Means for collecting past information,

[0854] A means for analyzing the aforementioned information and extracting characteristics,

[0855] A means for training and generating a model that predicts the optimal response using machine learning techniques,

[0856] A means of automatically generating responses to new information using the generated model,

[0857] means for displaying the automatically generated response,

[0858] A means of applying natural language processing techniques to analyze the context of the object to be processed and identify the necessary elements,

[0859] A means of managing information entered through users and efficiently utilizing stored information,

[0860] A means of presenting a response generated by machine learning to the user and providing a means for the user to review or modify it,

[0861] A system that includes this.

[0862] (Claim 2)

[0863] The system according to claim 1, further comprising means for collecting feedback on responses confirmed by the user and improving the accuracy of the model.

[0864] (Claim 3)

[0865] The system according to claim 1, further comprising means for retraining a predictive model based on received user information to improve predictive accuracy.

[0866] "Application Example 1"

[0867] (Claim 1)

[0868] Means for collecting past survey information,

[0869] A means for analyzing the aforementioned survey information and extracting the characteristics of the questions,

[0870] A means for training and generating a model that predicts the optimal response to a question using machine learning techniques,

[0871] A means for automatically generating responses to new survey questions using the generated model,

[0872] The means for presenting the automatically generated response,

[0873] A means of recommending new content based on the user's past interests,

[0874] A system that includes this.

[0875] (Claim 2)

[0876] The system according to claim 1, further comprising means for providing an interactive interface that allows a user to review the response generated by the means and modify it as necessary.

[0877] (Claim 3)

[0878] The system according to claim 1, further comprising means for collecting user feedback and improving the accuracy of the machine learning model.

[0879] "Example 2 of combining an emotion engine"

[0880] (Claim 1)

[0881] Means for accumulating past information,

[0882] A means for analyzing the aforementioned information and extracting the characteristics of the items,

[0883] A means for training and generating a model that predicts the optimal response to an item using machine learning techniques,

[0884] A means of automatically generating responses to new items using the generated model,

[0885] The means for presenting the automatically generated response,

[0886] A means for evaluating the user's emotional state related to the aforementioned items and incorporating the results of the emotional analysis into the response,

[0887] A system that includes this.

[0888] (Claim 2)

[0889] The system according to claim 1, further comprising means for providing a communication device that allows a user to confirm the response generated by the means and modify it as necessary.

[0890] (Claim 3)

[0891] The system according to claim 1, further comprising means for collecting user feedback and improving the accuracy of the machine learning model.

[0892] "Application example 2 when combining with an emotional engine"

[0893] (Claim 1)

[0894] Means for collecting historical data,

[0895] A means for analyzing the aforementioned data and extracting features,

[0896] A means of training and generating a model that predicts answers using machine learning techniques,

[0897] A means of automatically generating answers to new questions using the generated model,

[0898] The means for presenting the automatically generated answer,

[0899] A means of analyzing the user's emotional state,

[0900] A means of displaying personalized information based on the user's emotional state,

[0901] A system that includes this.

[0902] (Claim 2)

[0903] The system according to claim 1, further comprising means for providing an interface that allows a user to review the information generated by the means and make selections as necessary.

[0904] (Claim 3)

[0905] The system according to claim 1, further comprising means for collecting user feedback and improving the accuracy of the machine learning model and sentiment analysis. [Explanation of symbols]

[0906] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. Means for collecting past survey data, A means for analyzing the aforementioned survey data and extracting the characteristics of the questions, A means for training and generating a model that predicts the optimal answer to a question using machine learning techniques, A means of automatically generating answers to new survey questions using the generated model, The means for presenting the automatically generated answer, A system that includes this.

2. The system according to claim 1, further comprising means for providing an interface that allows the user to review the answer generated by the means described above and modify it as necessary.

3. The system according to claim 1, further comprising means for collecting user feedback and improving the accuracy of the machine learning model.

Images