system

Abstract

We provide the system. [Solution] A means for collecting past inquiry information, preprocessing the information, and providing it as training data, A method for training an artificial intelligence model, analyzing questions, and searching for the most relevant answers, A means for receiving questions from users and providing answers to those questions, A means for obtaining feedback on the answer and updating the artificial intelligence model, A means of collecting past viewing history information, analyzing the user's viewing trends, and recommending relevant content, 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, including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In an enterprise's support department, it is a problem that dealing with repetitive similar inquiries places a great burden on the staff and makes it difficult to respond promptly. Also, it is required to efficiently provide answers to users while maintaining the consistency of support quality.

Means for Solving the Problems

[0005] This invention provides a system that collects past inquiry information, preprocesses it, and provides it as training data for an artificial intelligence model. This allows the AI ​​model to analyze user questions using natural language processing techniques and search for the most relevant answers. Furthermore, by obtaining user feedback and updating the AI ​​model based on that feedback, continuous accuracy improvement is achieved. This process enables more efficient support operations and faster user response.

[0006] "Past inquiry information" refers to data on past questions asked by users to the support department and the answers provided.

[0007] "Preprocessing" is the process of removing unnecessary information from acquired data, correcting typos, and standardizing formatting.

[0008] An "artificial intelligence model" is an algorithm that learns patterns in data and has the ability to predict appropriate answers to input questions.

[0009] "Natural language processing technology" refers to the technology that enables computers to understand, interpret, and generate human language.

[0010] "Question analysis" involves breaking down user questions and identifying important keywords and context.

[0011] A "highly relevant answer" is an answer from the past database that contains appropriate content for the user's question.

[0012] "Feedback" refers to users' evaluations and opinions on the answers provided.

[0013] "Model updating" is the process of retraining an artificial intelligence model based on acquired feedback to improve the accuracy of responses. [Brief explanation of the drawing]

[0014] [Figure 1] It is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] It is a conceptual diagram showing an example of the main functions of a data processing device and a smart device according to the first embodiment. [Figure 3] It is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] It 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] It is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] It 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] It is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] It 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] It shows an emotion map to which a plurality of emotions are mapped. [Figure 10] It shows an emotion map to which a plurality of emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when an 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 an emotion engine is combined.

Embodiments for Carrying Out the Invention

[0015] An example of an embodiment of the system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0016] First, the terms used in the following description will be explained.

[0017] 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.

[0018] 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.

[0019] In the following embodiments, the numbered storage is one or more non-volatile storage devices that store various programs, various parameters, and the like. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, and the like.

[0020] 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).

[0021] 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."

[0022] [First Embodiment]

[0023] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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.

[0030] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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".

[0035] The system of this invention consists of a server, a terminal, and a user. The server connects to the API of a communication tool and collects past inquiry information. This data is preprocessed within the server and supplied as training data for an artificial intelligence model.

[0036] The artificial intelligence model is built using natural language processing technology and, by learning from collected data, acquires the ability to analyze user questions and identify the most relevant answers. The server uses this model to analyze new questions received from users and searches past answers in the database.

[0037] For example, if a user sends a question via their device such as "What should I do if my VPN connection fails?", the server receives this question. The artificial intelligence model analyzes this question and extracts important keywords such as "VPN," "connection," and "failure." Then, the server selects the most relevant answer from its past database and prepares the response.

[0038] The server sends answers to the user via the terminal, helping the user resolve the problem based on that information. If the user provides feedback on the answers, the server feeds this feedback back into the artificial intelligence model for additional learning to improve the model's accuracy. In this way, the system enables more efficient support operations and faster user response.

[0039] The following describes the processing flow.

[0040] Step 1:

[0041] The server collects past inquiry information through the communication tool's API. This includes past questions and their answers.

[0042] Step 2:

[0043] The server preprocesses the collected query information. This preprocessing involves removing noise from the data, standardizing the format, correcting typos, and converting it into a format suitable for the next learning stage.

[0044] Step 3:

[0045] The server supplies pre-processed data to the artificial intelligence model, allowing the model to train. Here, the model learns question-and-answer pairs and recognizes their relationships.

[0046] Step 4:

[0047] Users use their devices to post new questions to the system. These questions are sent to the server via a communication tool.

[0048] Step 5:

[0049] The server receives user questions and passes them to an artificial intelligence model for analysis. This model uses natural language processing techniques to analyze the questions and identify important keywords and context.

[0050] Step 6:

[0051] The server uses an artificial intelligence model to search a historical database and select the most relevant answer. This answer is then appropriately mapped based on the analysis of the question.

[0052] Step 7:

[0053] The server transfers the selected answer to the terminal and presents it to the user. The user then attempts to solve the problem based on that answer.

[0054] Step 8:

[0055] When a user provides feedback on the answers they have provided, the server receives that feedback and feeds it back into the artificial intelligence model as training data. This improves the system's accuracy.

[0056] (Example 1)

[0057] 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."

[0058] In today's information and communication landscape, where rapid and accurate information delivery is essential, efficiently providing optimal answers to user inquiries is a challenge. Traditional systems often struggle with rapid response times and provide insufficient accuracy. As a result, user satisfaction declines, and the efficiency of support operations decreases.

[0059] 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.

[0060] In this invention, the server includes means for collecting past inquiry information, preprocessing the information, and providing it as training data; means for analyzing questions and searching for the most relevant answers based on machine learning techniques using natural language processing; and means for receiving questions from users and providing answers to those questions. This enables the provision of answers quickly and accurately.

[0061] "Past inquiry information" refers to data about questions and problems previously submitted by users, which can be referenced in future inquiry processing.

[0062] "Preprocessing" refers to data processing operations such as cleaning and formatting performed on collected data, preparing it into a format that can be used by artificial intelligence models for learning.

[0063] "Training data" refers to a dataset used by artificial intelligence models to learn patterns and rules, and contributes to improving the accuracy of the model.

[0064] "Natural language processing" refers to the technology that enables computers to understand, analyze, and generate human language, and includes the analysis of grammatical structure and the understanding of meaning.

[0065] "Machine learning technology" is a technique that automatically learns rules and patterns from data and uses that learned information to make predictions and perform analyses on new data.

[0066] A "database" is an information management system built to efficiently manage, search, and update organized information.

[0067] A "generated prompt sentence" is a guiding sentence that an artificial intelligence model generates to analyze a user's question and guide them to an appropriate answer.

[0068] "Information and communication technology" refers to all technologies used to process and transmit information, and includes hardware, software, and network technologies.

[0069] To implement this invention, the involvement of a server, a terminal, and a user is required. The server first collects past inquiry information from external systems (e.g., email systems and chat platforms) via communication means. Since this information may contain noise in its raw state, the server performs data cleaning using natural language processing libraries (e.g., NLTK or spaCy) and preprocessing such as tokenization.

[0070] The server feeds pre-processed data to a machine learning model, which then uses a generative AI model (such as BERT or GPT) to learn how to generate the best possible answers to queries. This enables the server to analyze user questions.

[0071] The user enters a question through a terminal, and the server receives this input. For example, if the user enters "What should I do if the VPN connection fails?", the server uses an artificial intelligence model to analyze this input. The model extracts keywords such as "VPN," "connection," and "failure," and searches a database based on these keywords. This database stores past inquiry response data, enabling efficient searching.

[0072] The server generates relevant answers from search results and provides them to the user. For example, it might provide an answer such as, "Please try resetting your VPN settings and then reconfiguring them," to help the user solve their problem.

[0073] Furthermore, if a user provides feedback on the response they have received, the server feeds that feedback back into the artificial intelligence model, thereby improving the model's accuracy. This gradually improves the overall response accuracy and speed of the system.

[0074] As a concrete example of a prompt, it is envisioned that the user will directly input a question in the form of, "What should I do if I forget my password?". This type of input allows the server to respond immediately.

[0075] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0076] Step 1:

[0077] The server uses communication methods to collect past query information from external systems.

[0078] Input: Original data from emails or chat messages.

[0079] The server retrieves message data via the API and stores it in a local database.

[0080] Output: Saved raw data.

[0081] Step 2:

[0082] The server performs preprocessing on the collected raw data.

[0083] Input: Raw data from past inquiry information.

[0084] The server performs data cleaning, removing unnecessary characters and special symbols. Furthermore, it tokenizes the data using a natural language processing library.

[0085] Output: Pre-processed query information.

[0086] Step 3:

[0087] The server supplies pre-processed data as training data for the generating AI model.

[0088] Input: Pre-processed query information.

[0089] The server inputs this data into models such as BERT and GPT, allowing them to learn patterns and rules.

[0090] Output: A pre-trained AI model.

[0091] Step 4:

[0092] The user enters the question through the terminal.

[0093] Input: User's question text (e.g., "What should I do if my VPN connection fails?").

[0094] The user sends a question from their device to the server.

[0095] Output: Question data sent to the server.

[0096] Step 5:

[0097] The server analyzes the questions received from the user.

[0098] Input: User's question data.

[0099] The server uses a generative AI model to analyze the questions and extract important keywords.

[0100] Output: Extracted keyword list.

[0101] Step 6:

[0102] The server searches the database based on the extracted keywords to find the most relevant answer.

[0103] Input: Keyword list.

[0104] The server searches the database for relevant information using SQL queries and selects the appropriate answer.

[0105] Output: Selected answer.

[0106] Step 7:

[0107] The server sends the selected answer to the user.

[0108] Input: Selected answer.

[0109] The server sends the response to the terminal, allowing the user to view it on the screen.

[0110] Output: The answer displayed on the user's terminal.

[0111] Step 8:

[0112] Users provide feedback on the answers.

[0113] Input: Feedback on the answer (e.g., "It was helpful").

[0114] The user sends feedback information to the server.

[0115] Output: Feedback data sent to the server.

[0116] Step 9:

[0117] The server updates the AI ​​model using feedback.

[0118] Input: Feedback data.

[0119] The server analyzes the feedback and uses it as further training data for the AI ​​model.

[0120] Output: An improved AI model.

[0121] (Application Example 1)

[0122] 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."

[0123] In recent years, content distribution services have faced the challenge of users finding it difficult to select appropriate content from a vast amount of information. Furthermore, establishing a system for promptly responding to user inquiries is crucial for improving service quality. However, current systems have difficulty fully utilizing past viewing history and inquiry information to provide personalized recommendations to individual users.

[0124] 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.

[0125] In this invention, the server includes means for collecting past inquiry information, preprocessing the information, and providing it as training data; means for training an artificial intelligence model, analyzing questions, and searching for the most relevant answers; means for receiving questions from users and providing answers to those questions; means for obtaining feedback on the answers and updating the artificial intelligence model; and means for collecting past viewing history information, analyzing the user's viewing trends, and recommending relevant content. This enables personalized content recommendations to users and quick and appropriate responses to inquiries.

[0126] "Past inquiry information" refers to a collection of data related to questions and requests previously submitted by users.

[0127] "Preprocessing" refers to the initial stages of processing to convert collected data into a format suitable for analysis and learning.

[0128] "Training data" refers to the dataset used to train an artificial intelligence model.

[0129] An "artificial intelligence model" is a part of a program that automates specific tasks based on large amounts of data.

[0130] "Natural language processing technology" refers to the technology that allows computers to analyze and understand human speech.

[0131] "Analyzing a question" means understanding the content of the question submitted by the user and extracting important information.

[0132] A "highly relevant answer" is a response that contains the most accurate and useful information in response to a user's question.

[0133] "Feedback" refers to user feedback, such as impressions of using the product or suggestions for improvement.

[0134] "Viewing history information" refers to a record of content that a user has viewed in the past.

[0135] "Analyzing viewing trends" means analyzing a user's past viewing history to identify patterns in content they prefer.

[0136] "Personalized recommendations" refer to presenting the most suitable content based on the user's individual preferences and history.

[0137] The system that realizes this invention consists of a server, a terminal, and a user. The server first obtains past inquiry information from an external system using communication means, preprocesses this data, and provides it as training data. Next, it uses an artificial intelligence model and natural language processing technology to analyze questions from the user and searches the database for the most relevant answers to those questions. Furthermore, it also collects past viewing history information and analyzes the user's viewing trends to recommend relevant content in a personalized manner.

[0138] The server applies a generative AI model to analyze user preferences and past behavioral patterns based on collected inquiry and viewing history information. This enables the rapid delivery of the information and content users are looking for. User feedback is then fed back into the AI ​​model by the server and used for further learning to improve accuracy. In this way, a system is created that provides users with more appropriate answers and recommendations.

[0139] For example, if a user has a viewing history indicating they "like science fiction and action movies," the server will use that information to recommend new movies and other relevant content that match their preferences. Furthermore, based on a prompt like, "I'm looking for a romantic comedy to relax in on my day off, do you have any recommendations?", the system will generate answers tailored to the user's new questions. By using this prompt to power the AI ​​model, the system can quickly suggest the most suitable content for the user.

[0140] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0141] Step 1:

[0142] The server uses communication methods to retrieve past query information from an external system. The input is query data from the external system, and the output is pre-processable query data stored on the server. The server converts this data into an initial format so that it can be supplied to the artificial intelligence model.

[0143] Step 2:

[0144] The server preprocesses the collected query information and prepares it as training data for the generating AI model. The input is the query data obtained in the previous step, and the output is a dataset optimized for training. The server cleans up the data, extracts important keywords, and constructs the dataset.

[0145] Step 3:

[0146] The server receives new questions from users and analyzes them using natural language processing techniques. The input is the user's question, and the output is information identifying important keywords and their semantic relationships. The server uses a generative AI model to analyze the question content and understand its structure and context.

[0147] Step 4:

[0148] The server searches past answers in the database to identify the most relevant answers. The input is the analysis results obtained in step 3, and the output is a list of candidate answers to the user's question. The server queries the existing database and selects answers based on their relevance scores.

[0149] Step 5:

[0150] The server collects user viewing history information, analyzes viewing trends, and recommends relevant content. The input is user viewing history data, and the output is a personalized list of content. The server uses a generative AI model to analyze viewing history and extract content that matches the user's preferences.

[0151] Step 6:

[0152] The server provides the user with answers and recommended content, and receives feedback from the user. The input is the information obtained in steps 4 and 5, and the output is the answers and related content displayed to the user. The server collects user feedback and records it to improve the system.

[0153] Step 7:

[0154] The server updates its artificial intelligence model using the feedback it receives, improving the system's accuracy. The input is user feedback, and the output is an updated version of the generative AI model that reflects that feedback. The server retrains the model to improve the accuracy of its responses based on prompts.

[0155] 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.

[0156] The system of the present invention consists of a server, a terminal, a user, and an emotion engine. The server collects past inquiry information via the API of a communication tool, preprocesses it, and provides it as training data for an artificial intelligence model. The preprocessed data is supplied to the artificial intelligence model through the server, enhancing its ability to analyze questions using natural language processing techniques.

[0157] The artificial intelligence model receives new questions from users, analyzes them, and then searches for highly relevant answers. The server uses a sentiment engine in conjunction with the question analysis to recognize emotions from the user's questions and feedback. The sentiment engine analyzes the user's emotions and adjusts the tone and content of the answers based on the results.

[0158] As a concrete example, consider a scenario where a user sends a question via their device saying, "The system has become slow since the recent update; is there anything that can be done?" The server receives this question, analyzes it using an artificial intelligence model, and recognizes the user's emotions using an emotion engine. In this case, the emotion engine can determine that the user is dissatisfied.

[0159] As a result, the server selects a response that includes a more apologetic tone and suggested solutions to alleviate dissatisfaction, and presents it to the user through the terminal. For example, it might adjust its response to say, "We are very sorry for the inconvenience. As a solution to the slowdown that occurred after the update, please try clearing your cache."

[0160] The server then acquires further feedback and uses that feedback to retrain the artificial intelligence model to improve its accuracy. In this way, the system of the present invention enables responses that take user emotions into account, resulting in the provision of more consistent support quality.

[0161] The following describes the processing flow.

[0162] Step 1:

[0163] The server collects past query information through the communication tool's API and performs preprocessing. This preprocessing cleans the data and prepares it for training.

[0164] Step 2:

[0165] The server feeds pre-processed data to an artificial intelligence model for training. Through this process, the model improves its ability to find the optimal answer to a question.

[0166] Step 3:

[0167] The user uses a device to enter a new question and submit it. The device then forwards the question to the server.

[0168] Step 4:

[0169] The server receives the user's question and passes it to an artificial intelligence model for analysis. Simultaneously, an emotion engine is used to recognize the user's emotions from the question.

[0170] Step 5:

[0171] The emotion engine analyzes the user's emotions, and the server adjusts the tone and content of the response based on the results. At this stage, the most appropriate answer to the question is selected.

[0172] Step 6:

[0173] The server sends a prepared response to the terminal and presents it to the user. This allows the user to obtain information corresponding to the question.

[0174] Step 7:

[0175] The user sends feedback on their response to the server via their device.

[0176] Step 8:

[0177] The server retrains its artificial intelligence model based on user feedback to improve accuracy.

[0178] (Example 2)

[0179] 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 as the "terminal".

[0180] Conventional AI-powered question answering systems often fail to consider user emotions, potentially leading to a degraded user experience. Furthermore, this can result in inconsistent support quality. There is a growing need for flexible responses that take user emotions into account, thereby providing higher-quality support.

[0181] 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.

[0182] In this invention, the server includes means for collecting past inquiry information, preprocessing the information, and providing it as input data for machine learning; means for training a generative AI model, analyzing questions using natural language processing technology, and searching for the most relevant response; and means for receiving questions from users, recognizing emotions, and adjusting the tone and content of the response before providing it. This makes it possible to generate optimal responses tailored to the user's emotions and consistently provide high-quality user support.

[0183] "Past inquiry information" refers to data related to questions and comments submitted by users to date.

[0184] "Preprocessing" refers to the process of preparing collected data into a format suitable for analysis and learning, and includes processing such as correcting typos and normalizing text.

[0185] "Machine learning input data" refers to the basic data that generative AI models use to learn, and the information used to make predictions and decisions.

[0186] A "generative AI model" is a type of artificial intelligence equipped with an algorithm that generates new information and creates responses based on input data.

[0187] "Natural language processing technology" refers to the technology that enables computers to understand, interpret, and generate human language, and is used when dealing with text and audio data.

[0188] "Emotion recognition" is a technology that identifies the emotions contained in a user's text and adjusts the tone and content of the response accordingly.

[0189] "Feedback" refers to the opinions and evaluations received from users regarding the provided answers, and is information that can be used to improve the system.

[0190] The embodiments for carrying out the present invention will be described in detail below.

[0191] This system primarily consists of a server, terminals, users, and a generative AI model equipped with emotion recognition capabilities. First, the server collects past inquiry information from external information processing systems via multiple communication methods. This information is preprocessed to enable accurate analysis and prepared as input data for machine learning. As a result, the generative AI model learns based on the collected data and utilizes natural language processing techniques to respond to new questions from users.

[0192] Specifically, the server receives questions sent by users through their devices and analyzes their content. The generative AI model understands the context and intent of the text during this analysis process and generates highly relevant responses. The server also has the ability to determine the user's emotions using sentiment recognition capabilities and provide answers in an appropriate tone.

[0193] A typical example of this behavior is when a user asks, "My system has become slow since the recent update; is there anything that can be done?" The server uses a generative AI model to generate a response that includes an apology. For example, it might say, "We are very sorry for the inconvenience. As a solution to the slowdown that occurred after the update, please try clearing your cache."

[0194] An example of a prompt message to be input into the generation AI model would be, "Generate an appropriate apology message for when a user is dissatisfied with the system's slowdown."

[0195] Through feedback obtained throughout this process, the server can continuously update the generated AI model, enabling it to provide even more advanced user support. This allows the entire system to be continuously improved through user interaction, ensuring consistent high-quality service.

[0196] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0197] Step 1:

[0198] The server collects past query information from external information processing systems using communication methods. This input data is in text format, and since the server may not be able to analyze it directly, a preprocessing step is necessary. Specifically, this involves normalization, removing spelling errors and redundant information to maintain data consistency, and formatting the text data into a unified format. The output is data formatted in a format that can be used by machine learning models.

[0199] Step 2:

[0200] The server supplies pre-processed data to a generative AI model and runs the learning process. This input data consists of past queries and their responses. The server feeds this to the generative AI model, and the model's algorithm uses natural language processing techniques to analyze it and improve its ability to generate more relevant responses. The output is a computational model that produces more accurate and relevant responses.

[0201] Step 3:

[0202] The user submits a question via their device. This input data, the question, is in natural language text format. The server receives this data and sends it to the AI ​​model for response generation. Specifically, the server analyzes the context and intent of the question, classifies the question content appropriately, and prepares for response generation. The output is the analyzed question data.

[0203] Step 4:

[0204] The server generates appropriate responses to questions analyzed using a generative AI model. During this process, emotion recognition is employed to determine the user's emotions, and therefore emotional information is included in the input. For example, if emotion recognition determines that the user is feeling dissatisfied, a response with appropriate tone will be generated. The output will be a response text with an appropriate tone corresponding to the emotion.

[0205] Step 5:

[0206] The server sends the generated response to the user via the terminal. The input here is a response text corresponding to the emotion. The server formats it and presents it to the user in an appropriate format. The output is the final response presented to the user.

[0207] Step 6:

[0208] Users send feedback on their answers to the server via their device. This feedback arrives at the server as input data and is then used for the continuous improvement of the generated AI model. Specifically, the server systematically analyzes this feedback and uses it as data for retraining the model. As output, feedback data for improvement is obtained.

[0209] (Application Example 2)

[0210] 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".

[0211] Modern e-commerce sites are required to respond to user inquiries quickly and appropriately. However, it is difficult to respond while considering the user's feelings, and in some cases, this can lead to dissatisfaction. Furthermore, if the answers provided do not match the user's feelings, it can result in a decline in the quality of customer support. Solving these challenges is essential.

[0212] 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.

[0213] In this invention, the server includes means for collecting past query data and preprocessing the data to provide it as learning information; means for analyzing questions using natural language processing techniques with an artificial intelligence model to search for the most appropriate response; and means for receiving questions from users and providing responses to those questions in a manner that recognizes the user's emotions and adjusts the tone accordingly. This enables a quick and appropriate response while taking the user's emotions into consideration.

[0214] "Past inquiry data" refers to records of questions and requests previously made by users.

[0215] "Preprocessing" refers to the process of converting raw query data into a format suitable for analysis and learning.

[0216] "Learning information" refers to information used as training data to improve artificial intelligence models.

[0217] An "artificial intelligence model" is a computer program designed to understand and analyze language like a human.

[0218] "Natural language processing technology" is a technology that enables computers to understand and generate human language.

[0219] A "response" is an answer or reaction to a user's question or request.

[0220] "Evaluation" refers to data used to measure user satisfaction and reactions to the responses provided.

[0221] An "emotion recognition engine" is an algorithm or application that identifies emotions from a user's words and actions.

[0222] "Adjusting the tone" means changing the expression and attitude of the response according to the user's emotions.

[0223] To implement this invention, the following hardware and software are used. The server uses a dedicated data processing module to collect past query data and preprocess it. This data is analyzed by an artificial intelligence model (generative AI model).

[0224] The server uses natural language processing technology to analyze user questions. For this analysis, it utilizes the Google® Cloud Natural Language API, a commercial natural language processing API.

[0225] The server also features an emotion recognition engine that identifies the user's emotions. This uses an emotion analysis API to analyze the sentiment of messages sent by the user.

[0226] Based on the analysis results, the tone of the response to the user is adjusted. The adjusted response is then delivered to the user via the device. This enables customer support that is sensitive to the user's feelings.

[0227] As a concrete example, if a user sends a message expressing concern such as, "My ordered item hasn't arrived yet," the server can recognize this concern and quickly generate a response such as, "We apologize for your concern. We are checking the delivery status, so please wait a moment."

[0228] In situations where a generative AI model is used, an example of a prompt message would be: "For the user's inquiry, 'User's Question,' we request an answer that combines appropriate emotions and relevant information."

[0229] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0230] Step 1:

[0231] The server collects past query data. This data collection is performed via an API from an external system. The collected data is then preprocessed to remove unnecessary information and extract necessary information. The input is raw query data, and the output is preprocessed training data.

[0232] Step 2:

[0233] The server trains a generative AI model using pre-processed training data. Natural language processing techniques are used to improve its ability to find the most relevant responses to queries. The input is pre-processed query data, and the output is the optimized generative AI model.

[0234] Step 3:

[0235] The user submits a query using a terminal. The input from the terminal is the user's question, and this information is transmitted to the server. The output is data containing the query details.

[0236] Step 4:

[0237] The server uses a generative AI model to analyze questions received from users. It extracts relevant information related to the questions and searches for the most relevant responses. The input is the user's question, and the output is the relevant response information.

[0238] Step 5:

[0239] The server uses an emotion recognition engine to analyze the emotions expressed in the user's questions. Based on the analysis results, it adjusts the tone of its responses. The input is the user's question data, and the output is a set of adjusted response options.

[0240] Step 6:

[0241] The server determines the final response and delivers it to the user via the terminal. Here, it selects the most appropriate response from several generated options and sends it to the terminal. The input consists of response options adjusted based on emotion, and the output is the selected response message.

[0242] Step 7:

[0243] The user reviews the provided response and may send feedback to the server. Based on this feedback, the generative AI model continuously learns and improves response quality. The input is the feedback data, and the output is the adjustment and improvement of the model.

[0244] 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.

[0245] 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.

[0246] 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.

[0247] [Second Embodiment]

[0248] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0249] 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.

[0250] 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).

[0251] 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.

[0252] 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.

[0253] 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).

[0254] 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.

[0255] 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.

[0256] 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.

[0257] 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.

[0258] 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.

[0259] 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".

[0260] The system of this invention consists of a server, a terminal, and a user. The server connects to the API of a communication tool and collects past inquiry information. This data is preprocessed within the server and supplied as training data for an artificial intelligence model.

[0261] The artificial intelligence model is built using natural language processing technology and, by learning from collected data, acquires the ability to analyze user questions and identify the most relevant answers. The server uses this model to analyze new questions received from users and searches past answers in the database.

[0262] For example, if a user sends a question via their device such as "What should I do if my VPN connection fails?", the server receives this question. The artificial intelligence model analyzes this question and extracts important keywords such as "VPN," "connection," and "failure." Then, the server selects the most relevant answer from its past database and prepares the response.

[0263] The server sends answers to the user via the terminal, helping the user resolve the problem based on that information. If the user provides feedback on the answers, the server feeds this feedback back into the artificial intelligence model for additional learning to improve the model's accuracy. In this way, the system enables more efficient support operations and faster user response.

[0264] The following describes the processing flow.

[0265] Step 1:

[0266] The server collects past inquiry information through the communication tool's API. This includes past questions and their answers.

[0267] Step 2:

[0268] The server preprocesses the collected query information. This preprocessing involves removing noise from the data, standardizing the format, correcting typos, and converting it into a format suitable for the next learning stage.

[0269] Step 3:

[0270] The server supplies pre-processed data to the artificial intelligence model, allowing the model to train. Here, the model learns question-and-answer pairs and recognizes their relationships.

[0271] Step 4:

[0272] Users use their devices to post new questions to the system. These questions are sent to the server via a communication tool.

[0273] Step 5:

[0274] The server receives user questions and passes them to an artificial intelligence model for analysis. This model uses natural language processing techniques to analyze the questions and identify important keywords and context.

[0275] Step 6:

[0276] The server uses an artificial intelligence model to search a historical database and select the most relevant answer. This answer is then appropriately mapped based on the analysis of the question.

[0277] Step 7:

[0278] The server transfers the selected answer to the terminal and presents it to the user. The user then attempts to solve the problem based on that answer.

[0279] Step 8:

[0280] When a user provides feedback on the answers they have provided, the server receives that feedback and feeds it back into the artificial intelligence model as training data. This improves the system's accuracy.

[0281] (Example 1)

[0282] 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."

[0283] In modern information communication, while rapid and accurate information provision is required, it is an issue to efficiently provide an optimal answer to inquiries from users. In conventional systems, it is difficult to respond promptly, and the accuracy of the provided answers is often insufficient. As a result, there is a problem that user satisfaction decreases and the efficiency of support operations also decreases.

[0284] The specific processing by the specific processing unit 290 of the data processing apparatus 12 in Example 1 is realized by the following respective means.

[0285] In this invention, the server includes means for collecting past inquiry information, preprocessing the information to provide it as learning data, means for analyzing a question and searching for the most relevant answer based on a machine learning technique using natural language processing, and means for receiving a question from a user and providing an answer to the question. Thereby, it becomes possible to provide a prompt and highly accurate answer.

[0286] "Past inquiry information" is data regarding questions or problems previously provided by users and is information that can be referred to in future inquiry processing.

[0287] "Preprocessing" refers to processing operations such as cleaning and format conversion performed on the collected data, and is to prepare the data in a form that can be learned by an artificial intelligence model.

[0288] "Learning data" is a data set used for an artificial intelligence model to learn patterns and rules and contributes to improving the accuracy of the model.

[0289] "Natural language processing" refers to a technology for a computer to understand, analyze, and generate human language and includes analysis of grammar structures and understanding of meanings.

[0290] "Machine learning technology" is a technology that automatically learns rules and patterns from data and uses the learned content to perform predictions and analyzes on new data.

[0291] A "database" is an information management system built to efficiently manage, search, and update organized information.

[0292] A "generated prompt sentence" is a guiding sentence that an artificial intelligence model generates to analyze a user's question and guide them to an appropriate answer.

[0293] "Information and communication technology" refers to all technologies used to process and transmit information, and includes hardware, software, and network technologies.

[0294] To implement this invention, the involvement of a server, a terminal, and a user is required. The server first collects past inquiry information from external systems (e.g., email systems and chat platforms) via communication means. Since this information may contain noise in its raw state, the server performs data cleaning using natural language processing libraries (e.g., NLTK or spaCy) and preprocessing such as tokenization.

[0295] The server feeds pre-processed data to a machine learning model, which then uses a generative AI model (such as BERT or GPT) to learn how to generate the best possible answers to queries. This enables the server to analyze user questions.

[0296] The user enters a question through a terminal, and the server receives this input. For example, if the user enters "What should I do if the VPN connection fails?", the server uses an artificial intelligence model to analyze this input. The model extracts keywords such as "VPN," "connection," and "failure," and searches a database based on these keywords. This database stores past inquiry response data, enabling efficient searching.

[0297] The server generates highly relevant answers from the search results and provides them to the user. For example, by providing an answer such as "Please reset the VPN settings once and try to reconfigure them", it assists the user in solving the problem.

[0298] Also, when the user provides feedback on the answer, the server supplies the feedback to the artificial intelligence model again to improve the accuracy of the model. As a result, the response accuracy and response speed of the entire system can be gradually improved.

[0299] As a specific example of the prompt sentence, it is assumed that it is directly input into the model in the form of "What should I do if I forget my password?". By inputting in this form, the server can respond immediately.

[0300] The flow of the specific process in Example 1 will be described using FIG. 11.

[0301] Step 1:

[0302] The server collects past inquiry information from an external system using communication means.

[0303] Input: Original data of emails and chat messages.

[0304] The server obtains message data via an API and stores it in a local database.

[0305] Output: The stored raw data.

[0306] Step 2:

[0307] The server performs preprocessing on the collected raw data.

[0308] Input: Raw data of past inquiry information.

[0309] The server performs data cleaning, removing unnecessary characters and special symbols. Furthermore, it tokenizes the data using a natural language processing library.

[0310] Output: Pre-processed query information.

[0311] Step 3:

[0312] The server supplies pre-processed data as training data for the generating AI model.

[0313] Input: Pre-processed query information.

[0314] The server inputs this data into models such as BERT and GPT, allowing them to learn patterns and rules.

[0315] Output: A pre-trained AI model.

[0316] Step 4:

[0317] The user enters the question through the terminal.

[0318] Input: User's question text (e.g., "What should I do if my VPN connection fails?").

[0319] The user sends a question from their device to the server.

[0320] Output: Question data sent to the server.

[0321] Step 5:

[0322] The server analyzes the questions received from the user.

[0323] Input: User's question data.

[0324] The server uses a generative AI model to analyze the questions and extract important keywords.

[0325] Output: Extracted keyword list.

[0326] Step 6:

[0327] The server searches the database based on the extracted keywords to find the most relevant answer.

[0328] Input: Keyword list.

[0329] The server searches the database for relevant information using SQL queries and selects the appropriate answer.

[0330] Output: Selected answer.

[0331] Step 7:

[0332] The server sends the selected answer to the user.

[0333] Input: Selected answer.

[0334] The server sends the response to the terminal, allowing the user to view it on the screen.

[0335] Output: The answer displayed on the user's terminal.

[0336] Step 8:

[0337] Users provide feedback on the answers.

[0338] Input: Feedback on the answer (e.g., "It was helpful").

[0339] The user sends feedback information to the server.

[0340] Output: Feedback data sent to the server.

[0341] Step 9:

[0342] The server updates the AI ​​model using feedback.

[0343] Input: Feedback data.

[0344] The server analyzes the feedback and uses it as further training data for the AI ​​model.

[0345] Output: An improved AI model.

[0346] (Application Example 1)

[0347] 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."

[0348] In recent years, content distribution services have faced the challenge of users finding it difficult to select appropriate content from a vast amount of information. Furthermore, establishing a system for promptly responding to user inquiries is crucial for improving service quality. However, current systems have difficulty fully utilizing past viewing history and inquiry information to provide personalized recommendations to individual users.

[0349] 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.

[0350] In this invention, the server includes means for collecting past inquiry information, preprocessing the information, and providing it as training data; means for training an artificial intelligence model, analyzing questions, and searching for the most relevant answers; means for receiving questions from users and providing answers to those questions; means for obtaining feedback on the answers and updating the artificial intelligence model; and means for collecting past viewing history information, analyzing the user's viewing trends, and recommending relevant content. This enables personalized content recommendations to users and quick and appropriate responses to inquiries.

[0351] "Past inquiry information" refers to a collection of data related to questions and requests previously submitted by users.

[0352] "Preprocessing" refers to the initial stages of processing to convert collected data into a format suitable for analysis and learning.

[0353] "Training data" refers to the dataset used to train an artificial intelligence model.

[0354] An "artificial intelligence model" is a part of a program that automates specific tasks based on large amounts of data.

[0355] "Natural language processing technology" refers to the technology that allows computers to analyze and understand human speech.

[0356] "Analyzing a question" means understanding the content of the question submitted by the user and extracting important information.

[0357] A "highly relevant answer" is a response that contains the most accurate and useful information in response to a user's question.

[0358] "Feedback" refers to user feedback, such as impressions of using the product or suggestions for improvement.

[0359] "Viewing history information" refers to a record of content that a user has viewed in the past.

[0360] "Analyzing viewing trends" means analyzing a user's past viewing history to identify patterns in content they prefer.

[0361] "Personalized recommendations" refer to presenting the most suitable content based on the user's individual preferences and history.

[0362] The system that realizes this invention consists of a server, a terminal, and a user. The server first obtains past inquiry information from an external system using communication means, preprocesses this data, and provides it as training data. Next, it uses an artificial intelligence model and natural language processing technology to analyze questions from the user and searches the database for the most relevant answers to those questions. Furthermore, it also collects past viewing history information and analyzes the user's viewing trends to recommend relevant content in a personalized manner.

[0363] The server applies a generative AI model to analyze user preferences and past behavioral patterns based on collected inquiry and viewing history information. This enables the rapid delivery of the information and content users are looking for. User feedback is then fed back into the AI ​​model by the server and used for further learning to improve accuracy. In this way, a system is created that provides users with more appropriate answers and recommendations.

[0364] For example, if a user has a viewing history indicating they "like science fiction and action movies," the server will use that information to recommend new movies and other relevant content that match their preferences. Furthermore, based on a prompt like, "I'm looking for a romantic comedy to relax in on my day off, do you have any recommendations?", the system will generate answers tailored to the user's new questions. By using this prompt to power the AI ​​model, the system can quickly suggest the most suitable content for the user.

[0365] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0366] Step 1:

[0367] The server uses communication methods to retrieve past query information from an external system. The input is query data from the external system, and the output is pre-processable query data stored on the server. The server converts this data into an initial format so that it can be supplied to the artificial intelligence model.

[0368] Step 2:

[0369] The server preprocesses the collected query information and prepares it as training data for the generating AI model. The input is the query data obtained in the previous step, and the output is a dataset optimized for training. The server cleans up the data, extracts important keywords, and constructs the dataset.

[0370] Step 3:

[0371] The server receives new questions from users and analyzes them using natural language processing techniques. The input is the user's question, and the output is information identifying important keywords and their semantic relationships. The server uses a generative AI model to analyze the question content and understand its structure and context.

[0372] Step 4:

[0373] The server searches past answers in the database to identify the most relevant answers. The input is the analysis results obtained in step 3, and the output is a list of candidate answers to the user's question. The server queries the existing database and selects answers based on their relevance scores.

[0374] Step 5:

[0375] The server collects user viewing history information, analyzes viewing trends, and recommends relevant content. The input is user viewing history data, and the output is a personalized list of content. The server uses a generative AI model to analyze viewing history and extract content that matches the user's preferences.

[0376] Step 6:

[0377] The server provides the user with answers and recommended content, and receives feedback from the user. The input is the information obtained in steps 4 and 5, and the output is the answers and related content displayed to the user. The server collects user feedback and records it to improve the system.

[0378] Step 7:

[0379] The server updates its artificial intelligence model using the feedback it receives, improving the system's accuracy. The input is user feedback, and the output is an updated version of the generative AI model that reflects that feedback. The server retrains the model to improve the accuracy of its responses based on prompts.

[0380] 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.

[0381] The system of the present invention consists of a server, a terminal, a user, and an emotion engine. The server collects past inquiry information via the API of a communication tool, preprocesses it, and provides it as training data for an artificial intelligence model. The preprocessed data is supplied to the artificial intelligence model through the server, enhancing its ability to analyze questions using natural language processing techniques.

[0382] The artificial intelligence model receives new questions from users, analyzes them, and then searches for highly relevant answers. The server uses a sentiment engine in conjunction with the question analysis to recognize emotions from the user's questions and feedback. The sentiment engine analyzes the user's emotions and adjusts the tone and content of the answers based on the results.

[0383] As a concrete example, consider a scenario where a user sends a question via their device saying, "The system has become slow since the recent update; is there anything that can be done?" The server receives this question, analyzes it using an artificial intelligence model, and recognizes the user's emotions using an emotion engine. In this case, the emotion engine can determine that the user is dissatisfied.

[0384] As a result, the server selects a response that includes a more apologetic tone and suggested solutions to alleviate dissatisfaction, and presents it to the user through the terminal. For example, it might adjust its response to say, "We are very sorry for the inconvenience. As a solution to the slowdown that occurred after the update, please try clearing your cache."

[0385] The server then acquires further feedback and uses that feedback to retrain the artificial intelligence model to improve its accuracy. In this way, the system of the present invention enables responses that take user emotions into account, resulting in the provision of more consistent support quality.

[0386] The following describes the processing flow.

[0387] Step 1:

[0388] The server collects past query information through the communication tool's API and performs preprocessing. This preprocessing cleans the data and prepares it for training.

[0389] Step 2:

[0390] The server feeds pre-processed data to an artificial intelligence model for training. Through this process, the model improves its ability to find the optimal answer to a question.

[0391] Step 3:

[0392] The user uses a device to enter a new question and submit it. The device then forwards the question to the server.

[0393] Step 4:

[0394] The server receives the user's question and passes it to an artificial intelligence model for analysis. Simultaneously, an emotion engine is used to recognize the user's emotions from the question.

[0395] Step 5:

[0396] The emotion engine analyzes the user's emotions, and the server adjusts the tone and content of the response based on the results. At this stage, the most appropriate answer to the question is selected.

[0397] Step 6:

[0398] The server sends a prepared response to the terminal and presents it to the user. This allows the user to obtain information corresponding to the question.

[0399] Step 7:

[0400] The user sends feedback on their response to the server via their device.

[0401] Step 8:

[0402] The server retrains its artificial intelligence model based on user feedback to improve accuracy.

[0403] (Example 2)

[0404] 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".

[0405] Conventional AI-powered question answering systems often fail to consider user emotions, potentially leading to a degraded user experience. Furthermore, this can result in inconsistent support quality. There is a growing need for flexible responses that take user emotions into account, thereby providing higher-quality support.

[0406] 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.

[0407] In this invention, the server includes means for collecting past inquiry information, preprocessing the information, and providing it as input data for machine learning; means for training a generative AI model, analyzing questions using natural language processing technology, and searching for the most relevant response; and means for receiving questions from users, recognizing emotions, and adjusting the tone and content of the response before providing it. This makes it possible to generate optimal responses tailored to the user's emotions and consistently provide high-quality user support.

[0408] "Past inquiry information" refers to data related to questions and comments submitted by users to date.

[0409] "Preprocessing" refers to the process of preparing collected data into a format suitable for analysis and learning, and includes processing such as correcting typos and normalizing text.

[0410] "Machine learning input data" refers to the basic data that generative AI models use to learn, and the information used to make predictions and decisions.

[0411] A "generative AI model" is a type of artificial intelligence equipped with an algorithm that generates new information and creates responses based on input data.

[0412] "Natural language processing technology" refers to the technology that enables computers to understand, interpret, and generate human language, and is used when dealing with text and audio data.

[0413] "Emotion recognition" is a technology that identifies the emotions contained in a user's text and adjusts the tone and content of the response accordingly.

[0414] "Feedback" refers to the opinions and evaluations received from users regarding the provided answers, and is information that can be used to improve the system.

[0415] The embodiments for carrying out the present invention will be described in detail below.

[0416] This system primarily consists of a server, terminals, users, and a generative AI model equipped with emotion recognition capabilities. First, the server collects past inquiry information from external information processing systems via multiple communication methods. This information is preprocessed to enable accurate analysis and prepared as input data for machine learning. As a result, the generative AI model learns based on the collected data and utilizes natural language processing techniques to respond to new questions from users.

[0417] Specifically, the server receives questions sent by users through their devices and analyzes their content. The generative AI model understands the context and intent of the text during this analysis process and generates highly relevant responses. The server also has the ability to determine the user's emotions using sentiment recognition capabilities and provide answers in an appropriate tone.

[0418] A typical example of this behavior is when a user asks, "My system has become slow since the recent update; is there anything that can be done?" The server uses a generative AI model to generate a response that includes an apology. For example, it might say, "We are very sorry for the inconvenience. As a solution to the slowdown that occurred after the update, please try clearing your cache."

[0419] An example of a prompt message to be input into the generation AI model would be, "Generate an appropriate apology message for when a user is dissatisfied with the system's slowdown."

[0420] Through feedback obtained throughout this process, the server can continuously update the generated AI model, enabling it to provide even more advanced user support. This allows the entire system to be continuously improved through user interaction, ensuring consistent high-quality service.

[0421] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0422] Step 1:

[0423] The server collects past query information from external information processing systems using communication methods. This input data is in text format, and since the server may not be able to analyze it directly, a preprocessing step is necessary. Specifically, this involves normalization, removing spelling errors and redundant information to maintain data consistency, and formatting the text data into a unified format. The output is data formatted in a format that can be used by machine learning models.

[0424] Step 2:

[0425] The server supplies pre-processed data to a generative AI model and runs the learning process. This input data consists of past queries and their responses. The server feeds this to the generative AI model, and the model's algorithm uses natural language processing techniques to analyze it and improve its ability to generate more relevant responses. The output is a computational model that produces more accurate and relevant responses.

[0426] Step 3:

[0427] The user submits a question via their device. This input data, the question, is in natural language text format. The server receives this data and sends it to the AI ​​model for response generation. Specifically, the server analyzes the context and intent of the question, classifies the question content appropriately, and prepares for response generation. The output is the analyzed question data.

[0428] Step 4:

[0429] The server generates appropriate responses to questions analyzed using a generative AI model. During this process, emotion recognition is employed to determine the user's emotions, and therefore emotional information is included in the input. For example, if emotion recognition determines that the user is feeling dissatisfied, a response with appropriate tone will be generated. The output will be a response text with an appropriate tone corresponding to the emotion.

[0430] Step 5:

[0431] The server sends the generated response to the user via the terminal. The input here is a response text corresponding to the emotion. The server formats it and presents it to the user in an appropriate format. The output is the final response presented to the user.

[0432] Step 6:

[0433] Users send feedback on their answers to the server via their device. This feedback arrives at the server as input data and is then used for the continuous improvement of the generated AI model. Specifically, the server systematically analyzes this feedback and uses it as data for retraining the model. As output, feedback data for improvement is obtained.

[0434] (Application Example 2)

[0435] 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 as the "terminal".

[0436] Modern e-commerce sites are required to respond to user inquiries quickly and appropriately. However, it is difficult to respond while considering the user's feelings, and in some cases, this can lead to dissatisfaction. Furthermore, if the answers provided do not match the user's feelings, it can result in a decline in the quality of customer support. Solving these challenges is essential.

[0437] 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.

[0438] In this invention, the server includes means for collecting past query data and preprocessing the data to provide it as learning information; means for analyzing questions using natural language processing techniques with an artificial intelligence model to search for the most appropriate response; and means for receiving questions from users and providing responses to those questions in a manner that recognizes the user's emotions and adjusts the tone accordingly. This enables a quick and appropriate response while taking the user's emotions into consideration.

[0439] "Past inquiry data" refers to records of questions and requests previously made by users.

[0440] "Preprocessing" refers to the process of converting raw query data into a format suitable for analysis and learning.

[0441] "Learning information" refers to information used as training data to improve artificial intelligence models.

[0442] An "artificial intelligence model" is a computer program designed to understand and analyze language like a human.

[0443] "Natural language processing technology" is a technology that enables computers to understand and generate human language.

[0444] A "response" is an answer or reaction to a user's question or request.

[0445] "Evaluation" refers to data used to measure user satisfaction and reactions to the responses provided.

[0446] An "emotion recognition engine" is an algorithm or application that identifies emotions from a user's words and actions.

[0447] "Adjusting the tone" means changing the expression and attitude of the response according to the user's emotions.

[0448] To implement this invention, the following hardware and software are used. The server uses a dedicated data processing module to collect past query data and preprocess it. This data is analyzed by an artificial intelligence model (generative AI model).

[0449] The server uses natural language processing technology to analyze user questions. For this analysis, it utilizes the Google Cloud Natural Language API, a commercial natural language processing API.

[0450] The server also features an emotion recognition engine that identifies the user's emotions. This uses an emotion analysis API to analyze the sentiment of messages sent by the user.

[0451] Based on the analysis results, the tone of the response to the user is adjusted. The adjusted response is then delivered to the user via the device. This enables customer support that is sensitive to the user's feelings.

[0452] As a concrete example, if a user sends a message expressing concern such as, "My ordered item hasn't arrived yet," the server can recognize this concern and quickly generate a response such as, "We apologize for your concern. We are checking the delivery status, so please wait a moment."

[0453] In situations where a generative AI model is used, an example of a prompt message would be: "For the user's inquiry, 'User's Question,' we request an answer that combines appropriate emotions and relevant information."

[0454] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0455] Step 1:

[0456] The server collects past query data. This data collection is performed via an API from an external system. The collected data is then preprocessed to remove unnecessary information and extract necessary information. The input is raw query data, and the output is preprocessed training data.

[0457] Step 2:

[0458] The server trains a generative AI model using pre-processed training data. Natural language processing techniques are used to improve its ability to find the most relevant responses to queries. The input is pre-processed query data, and the output is the optimized generative AI model.

[0459] Step 3:

[0460] The user submits a query using a terminal. The input from the terminal is the user's question, and this information is transmitted to the server. The output is data containing the query details.

[0461] Step 4:

[0462] The server uses a generative AI model to analyze questions received from users. It extracts relevant information related to the questions and searches for the most relevant responses. The input is the user's question, and the output is the relevant response information.

[0463] Step 5:

[0464] The server uses an emotion recognition engine to analyze the emotions expressed in the user's questions. Based on the analysis results, it adjusts the tone of its responses. The input is the user's question data, and the output is a set of adjusted response options.

[0465] Step 6:

[0466] The server determines the final response and delivers it to the user via the terminal. Here, it selects the most appropriate response from several generated options and sends it to the terminal. The input consists of response options adjusted based on emotion, and the output is the selected response message.

[0467] Step 7:

[0468] The user reviews the provided response and may send feedback to the server. Based on this feedback, the generative AI model continuously learns and improves response quality. The input is the feedback data, and the output is the adjustment and improvement of the model.

[0469] 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.

[0470] 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.

[0471] 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.

[0472] [Third Embodiment]

[0473] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0474] 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.

[0475] 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).

[0476] 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.

[0477] 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.

[0478] 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).

[0479] 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.

[0480] 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.

[0481] 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.

[0482] 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.

[0483] 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.

[0484] 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".

[0485] The system of this invention consists of a server, a terminal, and a user. The server connects to the API of a communication tool and collects past inquiry information. This data is preprocessed within the server and supplied as training data for an artificial intelligence model.

[0486] The artificial intelligence model is built using natural language processing technology and, by learning from collected data, acquires the ability to analyze user questions and identify the most relevant answers. The server uses this model to analyze new questions received from users and searches past answers in the database.

[0487] For example, if a user sends a question via their device such as "What should I do if my VPN connection fails?", the server receives this question. The artificial intelligence model analyzes this question and extracts important keywords such as "VPN," "connection," and "failure." Then, the server selects the most relevant answer from its past database and prepares the response.

[0488] The server sends answers to the user via the terminal, helping the user resolve the problem based on that information. If the user provides feedback on the answers, the server feeds this feedback back into the artificial intelligence model for additional learning to improve the model's accuracy. In this way, the system enables more efficient support operations and faster user response.

[0489] The following describes the processing flow.

[0490] Step 1:

[0491] The server collects past inquiry information through the communication tool's API. This includes past questions and their answers.

[0492] Step 2:

[0493] The server preprocesses the collected query information. This preprocessing involves removing noise from the data, standardizing the format, correcting typos, and converting it into a format suitable for the next learning stage.

[0494] Step 3:

[0495] The server supplies pre-processed data to the artificial intelligence model, allowing the model to train. Here, the model learns question-and-answer pairs and recognizes their relationships.

[0496] Step 4:

[0497] Users use their devices to post new questions to the system. These questions are sent to the server via a communication tool.

[0498] Step 5:

[0499] The server receives user questions and passes them to an artificial intelligence model for analysis. This model uses natural language processing techniques to analyze the questions and identify important keywords and context.

[0500] Step 6:

[0501] The server uses an artificial intelligence model to search a historical database and select the most relevant answer. This answer is then appropriately mapped based on the analysis of the question.

[0502] Step 7:

[0503] The server transfers the selected answer to the terminal and presents it to the user. The user then attempts to solve the problem based on that answer.

[0504] Step 8:

[0505] When a user provides feedback on the answers they have provided, the server receives that feedback and feeds it back into the artificial intelligence model as training data. This improves the system's accuracy.

[0506] (Example 1)

[0507] 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."

[0508] In today's information and communication landscape, where rapid and accurate information delivery is essential, efficiently providing optimal answers to user inquiries is a challenge. Traditional systems often struggle with rapid response times and provide insufficient accuracy. As a result, user satisfaction declines, and the efficiency of support operations decreases.

[0509] 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.

[0510] In this invention, the server includes means for collecting past inquiry information, preprocessing the information, and providing it as training data; means for analyzing questions and searching for the most relevant answers based on machine learning techniques using natural language processing; and means for receiving questions from users and providing answers to those questions. This enables the provision of answers quickly and accurately.

[0511] "Past inquiry information" refers to data about questions and problems previously submitted by users, which can be referenced in future inquiry processing.

[0512] "Preprocessing" refers to data processing operations such as cleaning and formatting performed on collected data, preparing it into a format that can be used by artificial intelligence models for learning.

[0513] "Training data" refers to a dataset used by artificial intelligence models to learn patterns and rules, and contributes to improving the accuracy of the model.

[0514] "Natural language processing" refers to the technology that enables computers to understand, analyze, and generate human language, and includes the analysis of grammatical structure and the understanding of meaning.

[0515] "Machine learning technology" is a technique that automatically learns rules and patterns from data and uses that learned information to make predictions and perform analyses on new data.

[0516] A "database" is an information management system built to efficiently manage, search, and update organized information.

[0517] A "generated prompt sentence" is a guiding sentence that an artificial intelligence model generates to analyze a user's question and guide them to an appropriate answer.

[0518] "Information and communication technology" refers to all technologies used to process and transmit information, and includes hardware, software, and network technologies.

[0519] To implement this invention, the involvement of a server, a terminal, and a user is required. The server first collects past inquiry information from external systems (e.g., email systems and chat platforms) via communication means. Since this information may contain noise in its raw state, the server performs data cleaning using natural language processing libraries (e.g., NLTK or spaCy) and preprocessing such as tokenization.

[0520] The server feeds pre-processed data to a machine learning model, which then uses a generative AI model (such as BERT or GPT) to learn how to generate the best possible answers to queries. This enables the server to analyze user questions.

[0521] The user enters a question through a terminal, and the server receives this input. For example, if the user enters "What should I do if the VPN connection fails?", the server uses an artificial intelligence model to analyze this input. The model extracts keywords such as "VPN," "connection," and "failure," and searches a database based on these keywords. This database stores past inquiry response data, enabling efficient searching.

[0522] The server generates relevant answers from search results and provides them to the user. For example, it might provide an answer such as, "Please try resetting your VPN settings and then reconfiguring them," to help the user solve their problem.

[0523] Furthermore, if a user provides feedback on the response they have received, the server feeds that feedback back into the artificial intelligence model, thereby improving the model's accuracy. This gradually improves the overall response accuracy and speed of the system.

[0524] As a concrete example of a prompt, it is envisioned that the user will directly input a question in the form of, "What should I do if I forget my password?". This type of input allows the server to respond immediately.

[0525] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0526] Step 1:

[0527] The server uses communication methods to collect past query information from external systems.

[0528] Input: Original data from emails or chat messages.

[0529] The server retrieves message data via the API and stores it in a local database.

[0530] Output: Saved raw data.

[0531] Step 2:

[0532] The server performs preprocessing on the collected raw data.

[0533] Input: Raw data from past inquiry information.

[0534] The server performs data cleaning, removing unnecessary characters and special symbols. Furthermore, it tokenizes the data using a natural language processing library.

[0535] Output: Pre-processed query information.

[0536] Step 3:

[0537] The server supplies pre-processed data as training data for the generating AI model.

[0538] Input: Pre-processed query information.

[0539] The server inputs this data into models such as BERT and GPT, allowing them to learn patterns and rules.

[0540] Output: A pre-trained AI model.

[0541] Step 4:

[0542] The user enters the question through the terminal.

[0543] Input: User's question text (e.g., "What should I do if my VPN connection fails?").

[0544] The user sends a question from their device to the server.

[0545] Output: Question data sent to the server.

[0546] Step 5:

[0547] The server analyzes the questions received from the user.

[0548] Input: User's question data.

[0549] The server uses a generative AI model to analyze the questions and extract important keywords.

[0550] Output: Extracted keyword list.

[0551] Step 6:

[0552] The server searches the database based on the extracted keywords to find the most relevant answer.

[0553] Input: Keyword list.

[0554] The server searches the database for relevant information using SQL queries and selects the appropriate answer.

[0555] Output: Selected answer.

[0556] Step 7:

[0557] The server sends the selected answer to the user.

[0558] Input: Selected answer.

[0559] The server sends the response to the terminal, allowing the user to view it on the screen.

[0560] Output: The answer displayed on the user's terminal.

[0561] Step 8:

[0562] Users provide feedback on the answers.

[0563] Input: Feedback on the answer (e.g., "It was helpful").

[0564] The user sends feedback information to the server.

[0565] Output: Feedback data sent to the server.

[0566] Step 9:

[0567] The server updates the AI ​​model using feedback.

[0568] Input: Feedback data.

[0569] The server analyzes the feedback and uses it as further training data for the AI ​​model.

[0570] Output: An improved AI model.

[0571] (Application Example 1)

[0572] 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."

[0573] In recent years, content distribution services have faced the challenge of users finding it difficult to select appropriate content from a vast amount of information. Furthermore, establishing a system for promptly responding to user inquiries is crucial for improving service quality. However, current systems have difficulty fully utilizing past viewing history and inquiry information to provide personalized recommendations to individual users.

[0574] 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.

[0575] In this invention, the server includes means for collecting past inquiry information, preprocessing the information, and providing it as training data; means for training an artificial intelligence model, analyzing questions, and searching for the most relevant answers; means for receiving questions from users and providing answers to those questions; means for obtaining feedback on the answers and updating the artificial intelligence model; and means for collecting past viewing history information, analyzing the user's viewing trends, and recommending relevant content. This enables personalized content recommendations to users and quick and appropriate responses to inquiries.

[0576] "Past inquiry information" refers to a collection of data related to questions and requests previously submitted by users.

[0577] "Preprocessing" refers to the initial stages of processing to convert collected data into a format suitable for analysis and learning.

[0578] "Training data" refers to the dataset used to train an artificial intelligence model.

[0579] An "artificial intelligence model" is a part of a program that automates specific tasks based on large amounts of data.

[0580] "Natural language processing technology" refers to the technology that allows computers to analyze and understand human speech.

[0581] "Analyzing a question" means understanding the content of the question submitted by the user and extracting important information.

[0582] A "highly relevant answer" is a response that contains the most accurate and useful information in response to a user's question.

[0583] "Feedback" refers to user feedback, such as impressions of using the product or suggestions for improvement.

[0584] "Viewing history information" refers to a record of content that a user has viewed in the past.

[0585] "Analyzing viewing trends" means analyzing a user's past viewing history to identify patterns in content they prefer.

[0586] "Personalized recommendations" refer to presenting the most suitable content based on the user's individual preferences and history.

[0587] The system that realizes this invention consists of a server, a terminal, and a user. The server first obtains past inquiry information from an external system using communication means, preprocesses this data, and provides it as training data. Next, it uses an artificial intelligence model and natural language processing technology to analyze questions from the user and searches the database for the most relevant answers to those questions. Furthermore, it also collects past viewing history information and analyzes the user's viewing trends to recommend relevant content in a personalized manner.

[0588] The server applies a generative AI model to analyze user preferences and past behavioral patterns based on collected inquiry and viewing history information. This enables the rapid delivery of the information and content users are looking for. User feedback is then fed back into the AI ​​model by the server and used for further learning to improve accuracy. In this way, a system is created that provides users with more appropriate answers and recommendations.

[0589] For example, if a user has a viewing history indicating they "like science fiction and action movies," the server will use that information to recommend new movies and other relevant content that match their preferences. Furthermore, based on a prompt like, "I'm looking for a romantic comedy to relax in on my day off, do you have any recommendations?", the system will generate answers tailored to the user's new questions. By using this prompt to power the AI ​​model, the system can quickly suggest the most suitable content for the user.

[0590] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0591] Step 1:

[0592] The server uses communication methods to retrieve past query information from an external system. The input is query data from the external system, and the output is pre-processable query data stored on the server. The server converts this data into an initial format so that it can be supplied to the artificial intelligence model.

[0593] Step 2:

[0594] The server preprocesses the collected query information and prepares it as training data for the generating AI model. The input is the query data obtained in the previous step, and the output is a dataset optimized for training. The server cleans up the data, extracts important keywords, and constructs the dataset.

[0595] Step 3:

[0596] The server receives new questions from users and analyzes them using natural language processing techniques. The input is the user's question, and the output is information identifying important keywords and their semantic relationships. The server uses a generative AI model to analyze the question content and understand its structure and context.

[0597] Step 4:

[0598] The server searches past answers in the database to identify the most relevant answers. The input is the analysis results obtained in step 3, and the output is a list of candidate answers to the user's question. The server queries the existing database and selects answers based on their relevance scores.

[0599] Step 5:

[0600] The server collects user viewing history information, analyzes viewing trends, and recommends relevant content. The input is user viewing history data, and the output is a personalized list of content. The server uses a generative AI model to analyze viewing history and extract content that matches the user's preferences.

[0601] Step 6:

[0602] The server provides the user with answers and recommended content, and receives feedback from the user. The input is the information obtained in steps 4 and 5, and the output is the answers and related content displayed to the user. The server collects user feedback and records it to improve the system.

[0603] Step 7:

[0604] The server updates its artificial intelligence model using the feedback it receives, improving the system's accuracy. The input is user feedback, and the output is an updated version of the generative AI model that reflects that feedback. The server retrains the model to improve the accuracy of its responses based on prompts.

[0605] 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.

[0606] The system of the present invention consists of a server, a terminal, a user, and an emotion engine. The server collects past inquiry information via the API of a communication tool, preprocesses it, and provides it as training data for an artificial intelligence model. The preprocessed data is supplied to the artificial intelligence model through the server, enhancing its ability to analyze questions using natural language processing techniques.

[0607] The artificial intelligence model receives new questions from users, analyzes them, and then searches for highly relevant answers. The server uses a sentiment engine in conjunction with the question analysis to recognize emotions from the user's questions and feedback. The sentiment engine analyzes the user's emotions and adjusts the tone and content of the answers based on the results.

[0608] As a concrete example, consider a scenario where a user sends a question via their device saying, "The system has become slow since the recent update; is there anything that can be done?" The server receives this question, analyzes it using an artificial intelligence model, and recognizes the user's emotions using an emotion engine. In this case, the emotion engine can determine that the user is dissatisfied.

[0609] As a result, the server selects a response that includes a more apologetic tone and suggested solutions to alleviate dissatisfaction, and presents it to the user through the terminal. For example, it might adjust its response to say, "We are very sorry for the inconvenience. As a solution to the slowdown that occurred after the update, please try clearing your cache."

[0610] The server then acquires further feedback and uses that feedback to retrain the artificial intelligence model to improve its accuracy. In this way, the system of the present invention enables responses that take user emotions into account, resulting in the provision of more consistent support quality.

[0611] The following describes the processing flow.

[0612] Step 1:

[0613] The server collects past query information through the communication tool's API and performs preprocessing. This preprocessing cleans the data and prepares it for training.

[0614] Step 2:

[0615] The server feeds pre-processed data to an artificial intelligence model for training. Through this process, the model improves its ability to find the optimal answer to a question.

[0616] Step 3:

[0617] The user uses a device to enter a new question and submit it. The device then forwards the question to the server.

[0618] Step 4:

[0619] The server receives the user's question and passes it to an artificial intelligence model for analysis. Simultaneously, an emotion engine is used to recognize the user's emotions from the question.

[0620] Step 5:

[0621] The emotion engine analyzes the user's emotions, and the server adjusts the tone and content of the response based on the results. At this stage, the most appropriate answer to the question is selected.

[0622] Step 6:

[0623] The server sends a prepared response to the terminal and presents it to the user. This allows the user to obtain information corresponding to the question.

[0624] Step 7:

[0625] The user sends feedback on their response to the server via their device.

[0626] Step 8:

[0627] The server retrains its artificial intelligence model based on user feedback to improve accuracy.

[0628] (Example 2)

[0629] 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."

[0630] Conventional AI-powered question answering systems often fail to consider user emotions, potentially leading to a degraded user experience. Furthermore, this can result in inconsistent support quality. There is a growing need for flexible responses that take user emotions into account, thereby providing higher-quality support.

[0631] 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.

[0632] In this invention, the server includes means for collecting past inquiry information, preprocessing the information, and providing it as input data for machine learning; means for training a generative AI model, analyzing questions using natural language processing technology, and searching for the most relevant response; and means for receiving questions from users, recognizing emotions, and adjusting the tone and content of the response before providing it. This makes it possible to generate optimal responses tailored to the user's emotions and consistently provide high-quality user support.

[0633] "Past inquiry information" refers to data related to questions and comments submitted by users to date.

[0634] "Preprocessing" refers to the process of preparing collected data into a format suitable for analysis and learning, and includes processing such as correcting typos and normalizing text.

[0635] "Machine learning input data" refers to the basic data that generative AI models use to learn, and the information used to make predictions and decisions.

[0636] A "generative AI model" is a type of artificial intelligence equipped with an algorithm that generates new information and creates responses based on input data.

[0637] "Natural language processing technology" refers to the technology that enables computers to understand, interpret, and generate human language, and is used when dealing with text and audio data.

[0638] "Emotion recognition" is a technology that identifies the emotions contained in a user's text and adjusts the tone and content of the response accordingly.

[0639] "Feedback" refers to the opinions and evaluations received from users regarding the provided answers, and is information that can be used to improve the system.

[0640] The embodiments for carrying out the present invention will be described in detail below.

[0641] This system primarily consists of a server, terminals, users, and a generative AI model equipped with emotion recognition capabilities. First, the server collects past inquiry information from external information processing systems via multiple communication methods. This information is preprocessed to enable accurate analysis and prepared as input data for machine learning. As a result, the generative AI model learns based on the collected data and utilizes natural language processing techniques to respond to new questions from users.

[0642] Specifically, the server receives questions sent by users through their devices and analyzes their content. The generative AI model understands the context and intent of the text during this analysis process and generates highly relevant responses. The server also has the ability to determine the user's emotions using sentiment recognition capabilities and provide answers in an appropriate tone.

[0643] A typical example of this behavior is when a user asks, "My system has become slow since the recent update; is there anything that can be done?" The server uses a generative AI model to generate a response that includes an apology. For example, it might say, "We are very sorry for the inconvenience. As a solution to the slowdown that occurred after the update, please try clearing your cache."

[0644] An example of a prompt message to be input into the generation AI model would be, "Generate an appropriate apology message for when a user is dissatisfied with the system's slowdown."

[0645] Through feedback obtained throughout this process, the server can continuously update the generated AI model, enabling it to provide even more advanced user support. This allows the entire system to be continuously improved through user interaction, ensuring consistent high-quality service.

[0646] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0647] Step 1:

[0648] The server collects past query information from external information processing systems using communication methods. This input data is in text format, and since the server may not be able to analyze it directly, a preprocessing step is necessary. Specifically, this involves normalization, removing spelling errors and redundant information to maintain data consistency, and formatting the text data into a unified format. The output is data formatted in a format that can be used by machine learning models.

[0649] Step 2:

[0650] The server supplies pre-processed data to a generative AI model and runs the learning process. This input data consists of past queries and their responses. The server feeds this to the generative AI model, and the model's algorithm uses natural language processing techniques to analyze it and improve its ability to generate more relevant responses. The output is a computational model that produces more accurate and relevant responses.

[0651] Step 3:

[0652] The user submits a question via their device. This input data, the question, is in natural language text format. The server receives this data and sends it to the AI ​​model for response generation. Specifically, the server analyzes the context and intent of the question, classifies the question content appropriately, and prepares for response generation. The output is the analyzed question data.

[0653] Step 4:

[0654] The server generates appropriate responses to questions analyzed using a generative AI model. During this process, emotion recognition is employed to determine the user's emotions, and therefore emotional information is included in the input. For example, if emotion recognition determines that the user is feeling dissatisfied, a response with appropriate tone will be generated. The output will be a response text with an appropriate tone corresponding to the emotion.

[0655] Step 5:

[0656] The server sends the generated response to the user via the terminal. The input here is a response text corresponding to the emotion. The server formats it and presents it to the user in an appropriate format. The output is the final response presented to the user.

[0657] Step 6:

[0658] Users send feedback on their answers to the server via their device. This feedback arrives at the server as input data and is then used for the continuous improvement of the generated AI model. Specifically, the server systematically analyzes this feedback and uses it as data for retraining the model. As output, feedback data for improvement is obtained.

[0659] (Application Example 2)

[0660] 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."

[0661] Modern e-commerce sites are required to respond to user inquiries quickly and appropriately. However, it is difficult to respond while considering the user's feelings, and in some cases, this can lead to dissatisfaction. Furthermore, if the answers provided do not match the user's feelings, it can result in a decline in the quality of customer support. Solving these challenges is essential.

[0662] 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.

[0663] In this invention, the server includes means for collecting past query data and preprocessing the data to provide it as learning information; means for analyzing questions using natural language processing techniques with an artificial intelligence model to search for the most appropriate response; and means for receiving questions from users and providing responses to those questions in a manner that recognizes the user's emotions and adjusts the tone accordingly. This enables a quick and appropriate response while taking the user's emotions into consideration.

[0664] "Past inquiry data" refers to records of questions and requests previously made by users.

[0665] "Preprocessing" refers to the process of converting raw query data into a format suitable for analysis and learning.

[0666] "Learning information" refers to information used as training data to improve artificial intelligence models.

[0667] An "artificial intelligence model" is a computer program designed to understand and analyze language like a human.

[0668] "Natural language processing technology" is a technology that enables computers to understand and generate human language.

[0669] A "response" is an answer or reaction to a user's question or request.

[0670] "Evaluation" refers to data used to measure user satisfaction and reactions to the responses provided.

[0671] An "emotion recognition engine" is an algorithm or application that identifies emotions from a user's words and actions.

[0672] "Adjusting the tone" means changing the expression and attitude of the response according to the user's emotions.

[0673] To implement this invention, the following hardware and software are used. The server uses a dedicated data processing module to collect past query data and preprocess it. This data is analyzed by an artificial intelligence model (generative AI model).

[0674] The server uses natural language processing technology to analyze user questions. For this analysis, it utilizes the Google Cloud Natural Language API, a commercial natural language processing API.

[0675] The server also features an emotion recognition engine that identifies the user's emotions. This uses an emotion analysis API to analyze the sentiment of messages sent by the user.

[0676] Based on the analysis results, the tone of the response to the user is adjusted. The adjusted response is then delivered to the user via the device. This enables customer support that is sensitive to the user's feelings.

[0677] As a concrete example, if a user sends a message expressing concern such as, "My ordered item hasn't arrived yet," the server can recognize this concern and quickly generate a response such as, "We apologize for your concern. We are checking the delivery status, so please wait a moment."

[0678] In situations where a generative AI model is used, an example of a prompt message would be: "For the user's inquiry, 'User's Question,' we request an answer that combines appropriate emotions and relevant information."

[0679] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0680] Step 1:

[0681] The server collects past query data. This data collection is performed via an API from an external system. The collected data is then preprocessed to remove unnecessary information and extract necessary information. The input is raw query data, and the output is preprocessed training data.

[0682] Step 2:

[0683] The server trains a generative AI model using pre-processed training data. Natural language processing techniques are used to improve its ability to find the most relevant responses to queries. The input is pre-processed query data, and the output is the optimized generative AI model.

[0684] Step 3:

[0685] The user submits a query using a terminal. The input from the terminal is the user's question, and this information is transmitted to the server. The output is data containing the query details.

[0686] Step 4:

[0687] The server uses a generative AI model to analyze questions received from users. It extracts relevant information related to the questions and searches for the most relevant responses. The input is the user's question, and the output is the relevant response information.

[0688] Step 5:

[0689] The server uses an emotion recognition engine to analyze the emotions expressed in the user's questions. Based on the analysis results, it adjusts the tone of its responses. The input is the user's question data, and the output is a set of adjusted response options.

[0690] Step 6:

[0691] The server determines the final response and delivers it to the user via the terminal. Here, it selects the most appropriate response from several generated options and sends it to the terminal. The input consists of response options adjusted based on emotion, and the output is the selected response message.

[0692] Step 7:

[0693] The user reviews the provided response and may send feedback to the server. Based on this feedback, the generative AI model continuously learns and improves response quality. The input is the feedback data, and the output is the adjustment and improvement of the model.

[0694] 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.

[0695] 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.

[0696] 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.

[0697] [Fourth Embodiment]

[0698] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0699] 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.

[0700] 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).

[0701] 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.

[0702] 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.

[0703] 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).

[0704] 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.

[0705] 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.

[0706] 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.

[0707] 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.

[0708] 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.

[0709] 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.

[0710] 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".

[0711] The system of this invention consists of a server, a terminal, and a user. The server connects to the API of a communication tool and collects past inquiry information. This data is preprocessed within the server and supplied as training data for an artificial intelligence model.

[0712] The artificial intelligence model is built using natural language processing technology and, by learning from collected data, acquires the ability to analyze user questions and identify the most relevant answers. The server uses this model to analyze new questions received from users and searches past answers in the database.

[0713] For example, if a user sends a question via their device such as "What should I do if my VPN connection fails?", the server receives this question. The artificial intelligence model analyzes this question and extracts important keywords such as "VPN," "connection," and "failure." Then, the server selects the most relevant answer from its past database and prepares the response.

[0714] The server sends answers to the user via the terminal, helping the user resolve the problem based on that information. If the user provides feedback on the answers, the server feeds this feedback back into the artificial intelligence model for additional learning to improve the model's accuracy. In this way, the system enables more efficient support operations and faster user response.

[0715] The following describes the processing flow.

[0716] Step 1:

[0717] The server collects past inquiry information through the communication tool's API. This includes past questions and their answers.

[0718] Step 2:

[0719] The server preprocesses the collected query information. This preprocessing involves removing noise from the data, standardizing the format, correcting typos, and converting it into a format suitable for the next learning stage.

[0720] Step 3:

[0721] The server supplies pre-processed data to the artificial intelligence model, allowing the model to train. Here, the model learns question-and-answer pairs and recognizes their relationships.

[0722] Step 4:

[0723] Users use their devices to post new questions to the system. These questions are sent to the server via a communication tool.

[0724] Step 5:

[0725] The server receives user questions and passes them to an artificial intelligence model for analysis. This model uses natural language processing techniques to analyze the questions and identify important keywords and context.

[0726] Step 6:

[0727] The server uses an artificial intelligence model to search a historical database and select the most relevant answer. This answer is then appropriately mapped based on the analysis of the question.

[0728] Step 7:

[0729] The server transfers the selected answer to the terminal and presents it to the user. The user then attempts to solve the problem based on that answer.

[0730] Step 8:

[0731] When a user provides feedback on the answers they have provided, the server receives that feedback and feeds it back into the artificial intelligence model as training data. This improves the system's accuracy.

[0732] (Example 1)

[0733] 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".

[0734] In today's information and communication landscape, where rapid and accurate information delivery is essential, efficiently providing optimal answers to user inquiries is a challenge. Traditional systems often struggle with rapid response times and provide insufficient accuracy. As a result, user satisfaction declines, and the efficiency of support operations decreases.

[0735] 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.

[0736] In this invention, the server includes means for collecting past inquiry information, preprocessing the information, and providing it as training data; means for analyzing questions and searching for the most relevant answers based on machine learning techniques using natural language processing; and means for receiving questions from users and providing answers to those questions. This enables the provision of answers quickly and accurately.

[0737] "Past inquiry information" refers to data about questions and problems previously submitted by users, which can be referenced in future inquiry processing.

[0738] "Preprocessing" refers to data processing operations such as cleaning and formatting performed on collected data, preparing it into a format that can be used by artificial intelligence models for learning.

[0739] "Training data" refers to a dataset used by artificial intelligence models to learn patterns and rules, and contributes to improving the accuracy of the model.

[0740] "Natural language processing" refers to the technology that enables computers to understand, analyze, and generate human language, and includes the analysis of grammatical structure and the understanding of meaning.

[0741] "Machine learning technology" is a technique that automatically learns rules and patterns from data and uses that learned information to make predictions and perform analyses on new data.

[0742] A "database" is an information management system built to efficiently manage, search, and update organized information.

[0743] A "generated prompt sentence" is a guiding sentence that an artificial intelligence model generates to analyze a user's question and guide them to an appropriate answer.

[0744] "Information and communication technology" refers to all technologies used to process and transmit information, and includes hardware, software, and network technologies.

[0745] To implement this invention, the involvement of a server, a terminal, and a user is required. The server first collects past inquiry information from external systems (e.g., email systems and chat platforms) via communication means. Since this information may contain noise in its raw state, the server performs data cleaning using natural language processing libraries (e.g., NLTK or spaCy) and preprocessing such as tokenization.

[0746] The server feeds pre-processed data to a machine learning model, which then uses a generative AI model (such as BERT or GPT) to learn how to generate the best possible answers to queries. This enables the server to analyze user questions.

[0747] The user enters a question through a terminal, and the server receives this input. For example, if the user enters "What should I do if the VPN connection fails?", the server uses an artificial intelligence model to analyze this input. The model extracts keywords such as "VPN," "connection," and "failure," and searches a database based on these keywords. This database stores past inquiry response data, enabling efficient searching.

[0748] The server generates relevant answers from search results and provides them to the user. For example, it might provide an answer such as, "Please try resetting your VPN settings and then reconfiguring them," to help the user solve their problem.

[0749] Furthermore, if a user provides feedback on the response they have received, the server feeds that feedback back into the artificial intelligence model, thereby improving the model's accuracy. This gradually improves the overall response accuracy and speed of the system.

[0750] As a concrete example of a prompt, it is envisioned that the user will directly input a question in the form of, "What should I do if I forget my password?". This type of input allows the server to respond immediately.

[0751] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0752] Step 1:

[0753] The server uses communication methods to collect past query information from external systems.

[0754] Input: Original data from emails or chat messages.

[0755] The server retrieves message data via the API and stores it in a local database.

[0756] Output: Saved raw data.

[0757] Step 2:

[0758] The server performs preprocessing on the collected raw data.

[0759] Input: Raw data from past inquiry information.

[0760] The server performs data cleaning, removing unnecessary characters and special symbols. Furthermore, it tokenizes the data using a natural language processing library.

[0761] Output: Pre-processed query information.

[0762] Step 3:

[0763] The server supplies pre-processed data as training data for the generating AI model.

[0764] Input: Pre-processed query information.

[0765] The server inputs this data into models such as BERT and GPT, allowing them to learn patterns and rules.

[0766] Output: A pre-trained AI model.

[0767] Step 4:

[0768] The user enters the question through the terminal.

[0769] Input: User's question text (e.g., "What should I do if my VPN connection fails?").

[0770] The user sends a question from their device to the server.

[0771] Output: Question data sent to the server.

[0772] Step 5:

[0773] The server analyzes the questions received from the user.

[0774] Input: User's question data.

[0775] The server uses a generative AI model to analyze the questions and extract important keywords.

[0776] Output: Extracted keyword list.

[0777] Step 6:

[0778] The server searches the database based on the extracted keywords to find the most relevant answer.

[0779] Input: Keyword list.

[0780] The server searches the database for relevant information using SQL queries and selects the appropriate answer.

[0781] Output: Selected answer.

[0782] Step 7:

[0783] The server sends the selected answer to the user.

[0784] Input: Selected answer.

[0785] The server sends the response to the terminal, allowing the user to view it on the screen.

[0786] Output: The answer displayed on the user's terminal.

[0787] Step 8:

[0788] Users provide feedback on the answers.

[0789] Input: Feedback on the answer (e.g., "It was helpful").

[0790] The user sends feedback information to the server.

[0791] Output: Feedback data sent to the server.

[0792] Step 9:

[0793] The server updates the AI ​​model using feedback.

[0794] Input: Feedback data.

[0795] The server analyzes the feedback and uses it as further training data for the AI ​​model.

[0796] Output: An improved AI model.

[0797] (Application Example 1)

[0798] 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".

[0799] In recent years, content distribution services have faced the challenge of users finding it difficult to select appropriate content from a vast amount of information. Furthermore, establishing a system for promptly responding to user inquiries is crucial for improving service quality. However, current systems have difficulty fully utilizing past viewing history and inquiry information to provide personalized recommendations to individual users.

[0800] 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.

[0801] In this invention, the server includes means for collecting past inquiry information, preprocessing the information, and providing it as training data; means for training an artificial intelligence model, analyzing questions, and searching for the most relevant answers; means for receiving questions from users and providing answers to those questions; means for obtaining feedback on the answers and updating the artificial intelligence model; and means for collecting past viewing history information, analyzing the user's viewing trends, and recommending relevant content. This enables personalized content recommendations to users and quick and appropriate responses to inquiries.

[0802] "Past inquiry information" refers to a collection of data related to questions and requests previously submitted by users.

[0803] "Preprocessing" refers to the initial stages of processing to convert collected data into a format suitable for analysis and learning.

[0804] "Training data" refers to the dataset used to train an artificial intelligence model.

[0805] An "artificial intelligence model" is a part of a program that automates specific tasks based on large amounts of data.

[0806] "Natural language processing technology" refers to the technology that allows computers to analyze and understand human speech.

[0807] "Analyzing a question" means understanding the content of the question submitted by the user and extracting important information.

[0808] A "highly relevant answer" is a response that contains the most accurate and useful information in response to a user's question.

[0809] "Feedback" refers to user feedback, such as impressions of using the product or suggestions for improvement.

[0810] "Viewing history information" refers to a record of content that a user has viewed in the past.

[0811] "Analyzing viewing trends" means analyzing a user's past viewing history to identify patterns in content they prefer.

[0812] "Personalized recommendations" refer to presenting the most suitable content based on the user's individual preferences and history.

[0813] The system that realizes this invention consists of a server, a terminal, and a user. The server first obtains past inquiry information from an external system using communication means, preprocesses this data, and provides it as training data. Next, it uses an artificial intelligence model and natural language processing technology to analyze questions from the user and searches the database for the most relevant answers to those questions. Furthermore, it also collects past viewing history information and analyzes the user's viewing trends to recommend relevant content in a personalized manner.

[0814] The server applies a generative AI model to analyze user preferences and past behavioral patterns based on collected inquiry and viewing history information. This enables the rapid delivery of the information and content users are looking for. User feedback is then fed back into the AI ​​model by the server and used for further learning to improve accuracy. In this way, a system is created that provides users with more appropriate answers and recommendations.

[0815] For example, if a user has a viewing history indicating they "like science fiction and action movies," the server will use that information to recommend new movies and other relevant content that match their preferences. Furthermore, based on a prompt like, "I'm looking for a romantic comedy to relax in on my day off, do you have any recommendations?", the system will generate answers tailored to the user's new questions. By using this prompt to power the AI ​​model, the system can quickly suggest the most suitable content for the user.

[0816] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0817] Step 1:

[0818] The server uses communication methods to retrieve past query information from an external system. The input is query data from the external system, and the output is pre-processable query data stored on the server. The server converts this data into an initial format so that it can be supplied to the artificial intelligence model.

[0819] Step 2:

[0820] The server preprocesses the collected query information and prepares it as training data for the generating AI model. The input is the query data obtained in the previous step, and the output is a dataset optimized for training. The server cleans up the data, extracts important keywords, and constructs the dataset.

[0821] Step 3:

[0822] The server receives new questions from users and analyzes them using natural language processing techniques. The input is the user's question, and the output is information identifying important keywords and their semantic relationships. The server uses a generative AI model to analyze the question content and understand its structure and context.

[0823] Step 4:

[0824] The server searches past answers in the database to identify the most relevant answers. The input is the analysis results obtained in step 3, and the output is a list of candidate answers to the user's question. The server queries the existing database and selects answers based on their relevance scores.

[0825] Step 5:

[0826] The server collects user viewing history information, analyzes viewing trends, and recommends relevant content. The input is user viewing history data, and the output is a personalized list of content. The server uses a generative AI model to analyze viewing history and extract content that matches the user's preferences.

[0827] Step 6:

[0828] The server provides the user with answers and recommended content, and receives feedback from the user. The input is the information obtained in steps 4 and 5, and the output is the answers and related content displayed to the user. The server collects user feedback and records it to improve the system.

[0829] Step 7:

[0830] The server updates its artificial intelligence model using the feedback it receives, improving the system's accuracy. The input is user feedback, and the output is an updated version of the generative AI model that reflects that feedback. The server retrains the model to improve the accuracy of its responses based on prompts.

[0831] 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.

[0832] The system of the present invention consists of a server, a terminal, a user, and an emotion engine. The server collects past inquiry information via the API of a communication tool, preprocesses it, and provides it as training data for an artificial intelligence model. The preprocessed data is supplied to the artificial intelligence model through the server, enhancing its ability to analyze questions using natural language processing techniques.

[0833] The artificial intelligence model receives new questions from users, analyzes them, and then searches for highly relevant answers. The server uses a sentiment engine in conjunction with the question analysis to recognize emotions from the user's questions and feedback. The sentiment engine analyzes the user's emotions and adjusts the tone and content of the answers based on the results.

[0834] As a concrete example, consider a scenario where a user sends a question via their device saying, "The system has become slow since the recent update; is there anything that can be done?" The server receives this question, analyzes it using an artificial intelligence model, and recognizes the user's emotions using an emotion engine. In this case, the emotion engine can determine that the user is dissatisfied.

[0835] As a result, the server selects a response that includes a more apologetic tone and suggested solutions to alleviate dissatisfaction, and presents it to the user through the terminal. For example, it might adjust its response to say, "We are very sorry for the inconvenience. As a solution to the slowdown that occurred after the update, please try clearing your cache."

[0836] The server then acquires further feedback and uses that feedback to retrain the artificial intelligence model to improve its accuracy. In this way, the system of the present invention enables responses that take user emotions into account, resulting in the provision of more consistent support quality.

[0837] The following describes the processing flow.

[0838] Step 1:

[0839] The server collects past query information through the communication tool's API and performs preprocessing. This preprocessing cleans the data and prepares it for training.

[0840] Step 2:

[0841] The server feeds pre-processed data to an artificial intelligence model for training. Through this process, the model improves its ability to find the optimal answer to a question.

[0842] Step 3:

[0843] The user uses a device to enter a new question and submit it. The device then forwards the question to the server.

[0844] Step 4:

[0845] The server receives the user's question and passes it to an artificial intelligence model for analysis. Simultaneously, an emotion engine is used to recognize the user's emotions from the question.

[0846] Step 5:

[0847] The emotion engine analyzes the user's emotions, and the server adjusts the tone and content of the response based on the results. At this stage, the most appropriate answer to the question is selected.

[0848] Step 6:

[0849] The server sends a prepared response to the terminal and presents it to the user. This allows the user to obtain information corresponding to the question.

[0850] Step 7:

[0851] The user sends feedback on their response to the server via their device.

[0852] Step 8:

[0853] The server retrains its artificial intelligence model based on user feedback to improve accuracy.

[0854] (Example 2)

[0855] 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".

[0856] Conventional AI-powered question answering systems often fail to consider user emotions, potentially leading to a degraded user experience. Furthermore, this can result in inconsistent support quality. There is a growing need for flexible responses that take user emotions into account, thereby providing higher-quality support.

[0857] 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.

[0858] In this invention, the server includes means for collecting past inquiry information, preprocessing the information, and providing it as input data for machine learning; means for training a generative AI model, analyzing questions using natural language processing technology, and searching for the most relevant response; and means for receiving questions from users, recognizing emotions, and adjusting the tone and content of the response before providing it. This makes it possible to generate optimal responses tailored to the user's emotions and consistently provide high-quality user support.

[0859] "Past inquiry information" refers to data related to questions and comments submitted by users to date.

[0860] "Preprocessing" refers to the process of preparing collected data into a format suitable for analysis and learning, and includes processing such as correcting typos and normalizing text.

[0861] "Machine learning input data" refers to the basic data that generative AI models use to learn, and the information used to make predictions and decisions.

[0862] A "generative AI model" is a type of artificial intelligence equipped with an algorithm that generates new information and creates responses based on input data.

[0863] "Natural language processing technology" refers to the technology that enables computers to understand, interpret, and generate human language, and is used when dealing with text and audio data.

[0864] "Emotion recognition" is a technology that identifies the emotions contained in a user's text and adjusts the tone and content of the response accordingly.

[0865] "Feedback" refers to the opinions and evaluations received from users regarding the provided answers, and is information that can be used to improve the system.

[0866] The embodiments for carrying out the present invention will be described in detail below.

[0867] This system primarily consists of a server, terminals, users, and a generative AI model equipped with emotion recognition capabilities. First, the server collects past inquiry information from external information processing systems via multiple communication methods. This information is preprocessed to enable accurate analysis and prepared as input data for machine learning. As a result, the generative AI model learns based on the collected data and utilizes natural language processing techniques to respond to new questions from users.

[0868] Specifically, the server receives questions sent by users through their devices and analyzes their content. The generative AI model understands the context and intent of the text during this analysis process and generates highly relevant responses. The server also has the ability to determine the user's emotions using sentiment recognition capabilities and provide answers in an appropriate tone.

[0869] A typical example of this behavior is when a user asks, "My system has become slow since the recent update; is there anything that can be done?" The server uses a generative AI model to generate a response that includes an apology. For example, it might say, "We are very sorry for the inconvenience. As a solution to the slowdown that occurred after the update, please try clearing your cache."

[0870] An example of a prompt message to be input into the generation AI model would be, "Generate an appropriate apology message for when a user is dissatisfied with the system's slowdown."

[0871] Through feedback obtained throughout this process, the server can continuously update the generated AI model, enabling it to provide even more advanced user support. This allows the entire system to be continuously improved through user interaction, ensuring consistent high-quality service.

[0872] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0873] Step 1:

[0874] The server collects past query information from external information processing systems using communication methods. This input data is in text format, and since the server may not be able to analyze it directly, a preprocessing step is necessary. Specifically, this involves normalization, removing spelling errors and redundant information to maintain data consistency, and formatting the text data into a unified format. The output is data formatted in a format that can be used by machine learning models.

[0875] Step 2:

[0876] The server supplies pre-processed data to a generative AI model and runs the learning process. This input data consists of past queries and their responses. The server feeds this to the generative AI model, and the model's algorithm uses natural language processing techniques to analyze it and improve its ability to generate more relevant responses. The output is a computational model that produces more accurate and relevant responses.

[0877] Step 3:

[0878] The user submits a question via their device. This input data, the question, is in natural language text format. The server receives this data and sends it to the AI ​​model for response generation. Specifically, the server analyzes the context and intent of the question, classifies the question content appropriately, and prepares for response generation. The output is the analyzed question data.

[0879] Step 4:

[0880] The server generates appropriate responses to questions analyzed using a generative AI model. During this process, emotion recognition is employed to determine the user's emotions, and therefore emotional information is included in the input. For example, if emotion recognition determines that the user is feeling dissatisfied, a response with appropriate tone will be generated. The output will be a response text with an appropriate tone corresponding to the emotion.

[0881] Step 5:

[0882] The server sends the generated response to the user via the terminal. The input here is a response text corresponding to the emotion. The server formats it and presents it to the user in an appropriate format. The output is the final response presented to the user.

[0883] Step 6:

[0884] Users send feedback on their answers to the server via their device. This feedback arrives at the server as input data and is then used for the continuous improvement of the generated AI model. Specifically, the server systematically analyzes this feedback and uses it as data for retraining the model. As output, feedback data for improvement is obtained.

[0885] (Application Example 2)

[0886] 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".

[0887] Modern e-commerce sites are required to respond to user inquiries quickly and appropriately. However, it is difficult to respond while considering the user's feelings, and in some cases, this can lead to dissatisfaction. Furthermore, if the answers provided do not match the user's feelings, it can result in a decline in the quality of customer support. Solving these challenges is essential.

[0888] 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.

[0889] In this invention, the server includes means for collecting past query data and preprocessing the data to provide it as learning information; means for analyzing questions using natural language processing techniques with an artificial intelligence model to search for the most appropriate response; and means for receiving questions from users and providing responses to those questions in a manner that recognizes the user's emotions and adjusts the tone accordingly. This enables a quick and appropriate response while taking the user's emotions into consideration.

[0890] "Past inquiry data" refers to records of questions and requests previously made by users.

[0891] "Preprocessing" refers to the process of converting raw query data into a format suitable for analysis and learning.

[0892] "Learning information" refers to information used as training data to improve artificial intelligence models.

[0893] An "artificial intelligence model" is a computer program designed to understand and analyze language like a human.

[0894] "Natural language processing technology" is a technology that enables computers to understand and generate human language.

[0895] A "response" is an answer or reaction to a user's question or request.

[0896] "Evaluation" refers to data used to measure user satisfaction and reactions to the responses provided.

[0897] An "emotion recognition engine" is an algorithm or application that identifies emotions from a user's words and actions.

[0898] "Adjusting the tone" means changing the expression and attitude of the response according to the user's emotions.

[0899] To implement this invention, the following hardware and software are used. The server uses a dedicated data processing module to collect past query data and preprocess it. This data is analyzed by an artificial intelligence model (generative AI model).

[0900] The server uses natural language processing technology to analyze user questions. For this analysis, it utilizes the Google Cloud Natural Language API, a commercial natural language processing API.

[0901] The server also features an emotion recognition engine that identifies the user's emotions. This uses an emotion analysis API to analyze the sentiment of messages sent by the user.

[0902] Based on the analysis results, the tone of the response to the user is adjusted. The adjusted response is then delivered to the user via the device. This enables customer support that is sensitive to the user's feelings.

[0903] As a concrete example, if a user sends a message expressing concern such as, "My ordered item hasn't arrived yet," the server can recognize this concern and quickly generate a response such as, "We apologize for your concern. We are checking the delivery status, so please wait a moment."

[0904] In situations where a generative AI model is used, an example of a prompt message would be: "For the user's inquiry, 'User's Question,' we request an answer that combines appropriate emotions and relevant information."

[0905] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0906] Step 1:

[0907] The server collects past query data. This data collection is performed via an API from an external system. The collected data is then preprocessed to remove unnecessary information and extract necessary information. The input is raw query data, and the output is preprocessed training data.

[0908] Step 2:

[0909] The server trains a generative AI model using pre-processed training data. Natural language processing techniques are used to improve its ability to find the most relevant responses to queries. The input is pre-processed query data, and the output is the optimized generative AI model.

[0910] Step 3:

[0911] The user submits a query using a terminal. The input from the terminal is the user's question, and this information is transmitted to the server. The output is data containing the query details.

[0912] Step 4:

[0913] The server uses a generative AI model to analyze questions received from users. It extracts relevant information related to the questions and searches for the most relevant responses. The input is the user's question, and the output is the relevant response information.

[0914] Step 5:

[0915] The server uses an emotion recognition engine to analyze the emotions expressed in the user's questions. Based on the analysis results, it adjusts the tone of its responses. The input is the user's question data, and the output is a set of adjusted response options.

[0916] Step 6:

[0917] The server determines the final response and delivers it to the user via the terminal. Here, it selects the most appropriate response from several generated options and sends it to the terminal. The input consists of response options adjusted based on emotion, and the output is the selected response message.

[0918] Step 7:

[0919] The user reviews the provided response and may send feedback to the server. Based on this feedback, the generative AI model continuously learns and improves response quality. The input is the feedback data, and the output is the adjustment and improvement of the model.

[0920] 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.

[0921] 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.

[0922] 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.

[0923] 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.

[0924] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. In the upper and lower directions of the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. Also, the upper side of the concentric circles is where "pleasant" emotions are located, and the lower side is where "unpleasant" emotions are located. In this way, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0925] 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.

[0926] 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.

[0927] 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.

[0928] 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."

[0929] 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.

[0930] 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.

[0931] 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.

[0932] 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.

[0933] 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.

[0934] 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.

[0935] 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.

[0936] 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.

[0937] 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.

[0938] 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.

[0939] 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.

[0940] 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 to be incorporated by reference.

[0941] The following is further disclosed regarding the embodiments described above.

[0942] (Claim 1)

[0943] A means for collecting past inquiry information, preprocessing the information, and providing it as training data,

[0944] A method for training an artificial intelligence model, analyzing questions, and searching for the most relevant answers,

[0945] A means for receiving questions from users and providing answers to those questions,

[0946] A means for obtaining feedback on the answer and updating the artificial intelligence model,

[0947] A system that includes this.

[0948] (Claim 2)

[0949] The system according to claim 1, which obtains past inquiry information from an external system using communication means.

[0950] (Claim 3)

[0951] The system according to claim 1, wherein an artificial intelligence model analyzes a question using natural language processing technology.

[0952] "Example 1"

[0953] (Claim 1)

[0954] A means for collecting past inquiry information, preprocessing the information, and providing it as training data,

[0955] A means of analyzing a question and searching for the most relevant answer based on machine learning techniques using natural language processing,

[0956] A means for receiving questions from users and providing answers to those questions,

[0957] A means for obtaining feedback on the answer and updating the artificial intelligence model,

[0958] A means of searching for information from a database and generating and providing highly accurate answers,

[0959] A means to help users solve problems based on the generated answers,

[0960] A system that includes this.

[0961] (Claim 2)

[0962] The system according to claim 1, which obtains past inquiry information from an external information source using information and communication technology.

[0963] (Claim 3)

[0964] The system according to claim 1, which analyzes a question using prompt sentences generated by a machine learning model.

[0965] "Application Example 1"

[0966] (Claim 1)

[0967] A means for collecting past inquiry information, preprocessing the information, and providing it as training data,

[0968] A method for training an artificial intelligence model, analyzing questions, and searching for the most relevant answers,

[0969] A means for receiving questions from users and providing answers to those questions,

[0970] A means for obtaining feedback on the answer and updating the artificial intelligence model,

[0971] A means of collecting past viewing history information, analyzing the user's viewing trends, and recommending relevant content,

[0972] A system that includes this.

[0973] (Claim 2)

[0974] The system according to claim 1, which obtains past inquiry information from an external system using communication means and obtains past viewing history information from a user terminal.

[0975] (Claim 3)

[0976] The system according to claim 1, wherein an artificial intelligence model analyzes questions and viewing trends using natural language processing technology.

[0977] "Example 2 of combining an emotion engine"

[0978] (Claim 1)

[0979] A means for collecting past inquiry information, preprocessing the information, and providing it as input data for machine learning,

[0980] A method for training a generative AI model, analyzing a question using natural language processing technology, and searching for the most relevant response,

[0981] A means of receiving questions from users, and adjusting the tone and content of the answers by recognizing their emotions,

[0982] A means for obtaining feedback on the response and updating the generated AI model,

[0983] A computer system including a computer system.

[0984] (Claim 2)

[0985] The computer system according to claim 1, which obtains past inquiry information from an external information processing system using communication means.

[0986] (Claim 3)

[0987] The computer system according to claim 1, which uses emotion recognition functionality to modify the style of response based on the user's emotions.

[0988] "Application example 2 when combining with an emotional engine"

[0989] (Claim 1)

[0990] A means for collecting past inquiry data, preprocessing the data, and providing it as training information,

[0991] A means of analyzing a question using an artificial intelligence model and natural language processing technology to search for the most appropriate response,

[0992] A means for receiving questions from users and providing responses to those questions in a manner that recognizes the user's emotions and adjusts the tone accordingly,

[0993] A means for obtaining an evaluation of the response and for enhancing the artificial intelligence model,

[0994] A system that includes this.

[0995] (Claim 2)

[0996] The system according to claim 1, which obtains past inquiry data from an external device using a communication device.

[0997] (Claim 3)

[0998] The system according to claim 1, which uses an emotion recognition engine to analyze the user's emotions and adjust the tone of the response. [Explanation of Symbols]

[0999] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for collecting past inquiry information, preprocessing the information, and providing it as training data, A method for training an artificial intelligence model, analyzing questions, and searching for the most relevant answers, A means for receiving questions from users and providing answers to those questions, A means for obtaining feedback on the answer and updating the artificial intelligence model, A means of collecting past viewing history information, analyzing the user's viewing trends, and recommending relevant content, A system that includes this.

2. The system according to claim 1, which obtains past inquiry information from an external system using communication means and obtains past viewing history information from a user terminal.

3. The system according to claim 1, wherein an artificial intelligence model analyzes questions and viewing trends using natural language processing technology.

Images