system

Abstract

Provide a system. 【Solution means】 Means for obtaining the user's voice input, Means for converting the voice input into text data, Means for analyzing the intention of the user's question, Means for searching for relevant information based on the analysis result, Means for generating the searched information as text data, Means for converting the text data into voice data, Means for providing the voice data to the user, A system including the above.

Description

Technical Field

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

Background Art

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

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In modern audiobooks, listeners need to listen to a large amount of information and it is difficult to quickly find specific information. Also, in order to understand complex stories or specialized content, often re-listening or long searches are required, which causes stress and reduced efficiency. Therefore, there is a need for a means for a user to quickly and accurately obtain necessary information while listening to an audiobook.

Means for Solving the Problems

[0005] The present invention solves the above problems by providing a system that includes means for acquiring user voice input, means for converting voice input into text data, means for analyzing the intent of the user's question, means for searching for relevant information based on the analysis results, means for generating the searched information as text data, means for converting the text data into voice data, and means for providing the voice data to the user. With this system, the user can acquire information from audiobooks in real time by asking questions by voice and efficiently understand the content.

[0006]

[0007] "Means for acquiring user voice input" refers to a device or system for capturing the voice spoken by a user in digital format.

[0008] "Means for converting voice input into text data" refers to a speech recognition engine or algorithm for converting captured voice data into text data.

[0009] "Means for analyzing the intent behind a user's question" refers to natural language processing algorithms that analyze text data to clarify the information the user wants to know and the intent behind their question.

[0010] "Means for searching for relevant information based on analysis results" refers to a search engine or algorithm that efficiently retrieves relevant information from databases or indexes based on the analyzed intent.

[0011] "Means for generating searched information as text data" refers to a procedure or algorithm for formatting search results into an appropriate text format for providing to the user.

[0012] "Means for converting text data to audio data" refers to a speech synthesis engine or algorithm for converting the generated text data into audio data.

[0013] "Means of providing audio data to the user" refers to a speaker or playback device that plays back synthesized audio data so that the user can hear it. [Brief explanation of the drawing]

[0014] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] This is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] This is a sequence diagram showing the processing flow of the data processing system in Example 2, which incorporates an emotion engine. [Figure 14]It is a sequence diagram showing the processing flow of a data processing system in Application Example 2 when a sentiment engine is combined.

Embodiments for Carrying Out the Invention

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

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

[0017] In the following embodiments, a 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, a 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, a numbered storage is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, etc.

[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 present invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to an audiobook. The system consists of the following main components: voice input acquisition means, voice recognition means, question analysis means, information retrieval means, answer generation means, voice synthesis means, and voice output means.

[0036] Overall processing flow

[0037] 1. Acquisition of voice input

[0038] The user can ask questions about what they want to know using voice commands.

[0039] The device captures the user's voice through the microphone.

[0040] 2. Speech Recognition

[0041] The device converts the captured audio into text data and sends it to the speech recognition engine.

[0042] 3. Question analysis

[0043] The server receives text data sent from the speech recognition engine and uses natural language processing to analyze the intent of the question.

[0044] 4. Information Retrieval

[0045] Based on the intent of the analyzed question, the server searches for relevant information within the audiobook content.

[0046] 5. Answer generation

[0047] The server generates an answer based on the search results.

[0048] 6. Speech synthesis

[0049] The server uses a speech synthesis engine to convert the generated responses from text to speech.

[0050] 7. Audio Output

[0051] The server sends the synthesized voice data to the terminal.

[0052] The device plays back the received audio data and provides the user with an answer.

[0053] Detailed explanations for each step are provided below.

[0054] Acquiring voice input

[0055] The user says to the device, "What is the name of the main character in this book?"

[0056] The device captures this audio through its built-in microphone and saves it as audio data.

[0057] Speech recognition

[0058] The device sends the captured audio data to a speech recognition engine in the cloud.

[0059] The speech recognition engine converts the audio data into text data such as "What is the name of the main character in this book?".

[0060] The terminal receives the converted text data and sends it to the server.

[0061] question analysis

[0062] The server analyzes the received text data using a natural language processing algorithm.

[0063] Understand the intent of the question and identify the specific information the user wants to know (in this case, "the protagonist's name").

[0064] Information Retrieval

[0065] Based on the identified intent, the server issues a search query to the audiobook content database.

[0066] The server retrieves relevant information, such as the protagonist's name "John Smith," as a search result.

[0067] Answer generation

[0068] Based on the retrieved search results, the server generates the answer, "The name of the main character in this book is John Smith."

[0069] Speech synthesis

[0070] The server sends the generated text-based response to the speech synthesis engine.

[0071] The speech synthesis engine converts this text into speech data.

[0072] Audio output

[0073] The server sends the audio data to the terminal.

[0074] The device plays audio data and tells the user, "The name of the main character in this book is John Smith."

[0075] Specific example

[0076] For example, if a user asks, "What is the most important point in this chapter?", the same process would apply.

[0077] The server generates the response, "The key point of this chapter is the scene where the protagonist makes a firm decision," and plays it back to the user as audio.

[0078] This invention allows users to instantly obtain the information they want while listening to audiobooks, deepening their understanding of the content. In addition, it reduces the wasted time caused by conventional methods of re-searching and re-listening, enabling more efficient access to information.

[0079] The following describes the processing flow.

[0080] Step 1:

[0081] The user speaks a question into the device. Example: "What is the name of the main character in this book?"

[0082] Step 2:

[0083] The device uses its microphone to capture the user's voice and saves it as digital audio data.

[0084] Step 3:

[0085] The device sends the captured audio data to the speech recognition engine.

[0086] Step 4:

[0087] The speech recognition engine analyzes the audio data and converts it into corresponding text data. For example, it might generate text data such as, "What is the name of the main character in this book?"

[0088] Step 5:

[0089] The speech recognition engine sends the converted text data back to the device.

[0090] Step 6:

[0091] The terminal sends text data to the server.

[0092] Step 7:

[0093] The server analyzes the received text data using natural language processing algorithms. Specifically, it grasps the intent of the user's question and extracts important keywords (e.g., "the protagonist's name").

[0094] Step 8:

[0095] Based on the analysis results, the server issues a query to the audiobook content database to retrieve relevant information.

[0096] Step 9:

[0097] The server retrieves relevant information from the database. For example, the data "John Smith" is searched for.

[0098] Step 10:

[0099] The server generates the answer, "The name of the main character in this book is John Smith," based on the search results.

[0100] Step 11:

[0101] The server sends the generated text-based response to the speech synthesis engine.

[0102] Step 12:

[0103] The speech synthesis engine analyzes text data and converts it into speech data.

[0104] Step 13:

[0105] The speech synthesis engine sends the generated audio data back to the server.

[0106] Step 14:

[0107] The server sends the audio data to the terminal.

[0108] Step 15:

[0109] The device plays the received audio data and tells the user, "The name of the main character in this book is John Smith."

[0110] (Example 1)

[0111] 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."

[0112] Conventional audio content listening systems were inefficient because users had to manually re-search or re-listen to obtain the information they needed. Furthermore, systems that converted voice input to text and provided appropriate answers to user questions lacked sufficient technology to accurately analyze the intent of the questions and provide timely, appropriate responses. As a result, it was difficult for users to quickly and accurately obtain the information they needed while listening to audio content.

[0113] 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.

[0114] In this invention, the server includes means for enabling the user to ask questions in real time while listening to audio content, means for generating answers to the questions using a natural language processing algorithm and a generative AI model, and means for converting the generated answers into audio data using a speech synthesizer. This enables the user to quickly and accurately obtain the necessary information without interrupting the audio content.

[0115] A "voice input device" is a device that captures a user's voice and converts it into digital voice data.

[0116] A "speech recognition engine" is a program or service that converts speech data into text data.

[0117] A "natural language processing algorithm" is a computational method for analyzing text data and understanding its meaning, context, and intent.

[0118] A "generative AI model" is an artificial intelligence model designed to generate natural language responses based on text data.

[0119] A "speech synthesis device" is a device or service used to convert text data into speech data.

[0120] "Audio content" refers to information and entertainment provided in audio format.

[0121] A "prompt sentence" is an input sentence given to a generative AI model, and it is an instruction sentence used to generate a specific answer.

[0122] "Real-time" refers to the fact that user requests and operations are processed instantly with virtually no delay.

[0123] A "database" is a system designed to organize and manage large amounts of data, enabling efficient searching and updating.

[0124] A "user" is a person who attempts to obtain information by operating a system.

[0125] A "terminal" is an electronic device used by a user to access a system.

[0126] A "server" is a computer that acts as the central hub of a system, processing, storing, and providing data.

[0127] This invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to audio content. This system acquires the user's voice input and performs a series of processes to provide an appropriate response using natural language processing and a generative AI model, and its specific implementation method is described below.

[0128] System Overview

[0129] This system consists of the following main parts:

[0130] 1. Voice input device

[0131] 2. Speech Recognition Engine

[0132] 3. Natural Language Processing Algorithms

[0133] 4. Generative AI Models

[0134] 5. Speech synthesis device

[0135] 6. Server

[0136] 7. Terminal

[0137] Required hardware and software

[0138] Voice input device: Used to capture the user's voice. Example: Microphone.

[0139] Speech recognition engine: Used to convert speech data into text data. Examples: Google® Cloud Speech-to-Text API, IBM Watson® Speech to Text, etc.

[0140] Natural Language Processing (NLP) algorithms: Used to analyze the intent behind a user's question. Examples: Natural Language Processing libraries such as NLTK and spaCy.

[0141] Generative AI models: Used to generate appropriate answers to questions. Example: GPT-3 (registered trademark) from OpenAI (registered trademark).

[0142] Text-to-speech (Speech synthesis) devices are used to convert text data into speech data. Examples include Amazon Polly and Google Cloud Text-to-Speech.

[0143] Server: Manages the processing of the entire system.

[0144] Terminal: A device used by a user to access a system, input questions, and receive answers. Examples: smartphones, tablets, personal computers, etc.

[0145] Specific example

[0146] When a user is listening to audio content, for example, they might be asked, "What is the name of the main character in this book?" In this case, the system operates as follows:

[0147] 1. Acquisition of voice input

[0148] The user says to the device, "What is the name of the main character in this book?"

[0149] The device captures this audio through its built-in microphone and saves it as audio data.

[0150] 2. Speech Recognition

[0151] The device sends the captured audio data to a speech recognition engine in the cloud.

[0152] The speech recognition engine converts the audio data into text data such as "What is the name of the main character in this book?".

[0153] The terminal receives the converted text data and sends it to the server.

[0154] 3. Question analysis

[0155] The server analyzes the received text data using a natural language processing algorithm.

[0156] Understand the intent of the question and identify the specific information the user wants to know (in this case, "the protagonist's name").

[0157] 4. Information Retrieval

[0158] Based on the intent of the analyzed question, the server searches for relevant information from the audio content.

[0159] The server retrieves relevant information as search results. For example, the protagonist's name is "John Smith".

[0160] 5. Answer generation

[0161] Based on the retrieved search results, the server generates the answer, "The name of the main character in this book is John Smith."

[0162] 6. Speech synthesis

[0163] The server sends the generated text-based response to the speech synthesis engine.

[0164] The speech synthesis engine converts this text into speech data.

[0165] 7. Audio Output

[0166] The server sends the audio data to the terminal.

[0167] The device plays audio data and tells the user, "The name of the main character in this book is John Smith."

[0168] Example of a prompt

[0169] The following is an example of a prompt message used to generate an appropriate answer to a user's question using a generative AI model.

[0170] Prompt example (English):

[0171] "User asks in English: 'What is the name of the main character in this book?'. The audiobooks database has the following relevant information: 'The main character of the book is John Smith.'. Respond to the user's question with this information."

[0172] Prompt example (Japanese):

[0173] The user asks, "What is the name of the main character in this book?" The audio content database contains the related information, "The main character in this book is John Smith." Use this information to answer the user's question.

[0174] This invention allows users to instantly obtain the information they want while listening to audio content, leading to a deeper understanding of the content. This reduces the wasted time caused by traditional methods of re-searching or re-listening, enabling more efficient access to information.

[0175] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0176] Step 1:

[0177] The user asks a question aloud, "What is the name of the main character in this book?" The input is the user's voice, which is captured by the device's built-in microphone. Specifically, the device runs a voice capture program and saves it as audio data.

[0178] Step 2:

[0179] The device sends the captured audio data to a speech recognition engine in the cloud. The input is audio data, and the output is text data. Specifically, the device uses a cloud communication module to send audio data, which the speech recognition engine converts into text, such as "What is the name of the main character in this book?"

[0180] Step 3:

[0181] The device receives converted text data from the cloud and sends it to the server. The input is text data from the speech recognition engine, and the output is the transmission of text data to the server. Specifically, the device uses a cloud communication module to transfer the text data to the server.

[0182] Step 4:

[0183] The server analyzes the received text data using a natural language processing algorithm. The input is text data, and the output is the result of analyzing the intent of the user's question. Specifically, the server uses natural language processing libraries such as NLTK or spaCy to identify the intent of the question as "I want to know the name of the main character."

[0184] Step 5:

[0185] The server searches for relevant information based on the analysis results. The input is the analysis result (the intent of the question), and the output is the relevant information (search results). Specifically, the server performs a search on the database using SQL queries or ElasticSearch®, and as a result retrieves the name "John Smith".

[0186] Step 6:

[0187] The server generates an answer based on the search results. The input is the information from the search results, and the output is the text data of the answer. Specifically, the server uses a generative AI model (for example, GPT-3) to generate the text "The name of the main character in this book is John Smith."

[0188] Step 7:

[0189] The server sends the generated text-based response to a text-to-speech engine, where it is converted into audio data. The input is the text data of the response, and the output is audio data. Specifically, the server uses Amazon Polly or Google Text-to-Speech to generate the audio data.

[0190] Step 8:

[0191] The server sends the generated audio data to the terminal. The input is audio data, and the output is the transmission of audio data to the terminal. Specifically, the server transfers the audio data to the terminal using a cloud communication module.

[0192] Step 9:

[0193] The terminal plays the received audio data and provides it to the user. The input is audio data, and the output is audio playback. Specifically, the terminal uses an audio playback program to play the sentence, "The name of the main character in this book is John Smith."

[0194] (Application Example 1)

[0195] 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."

[0196] Currently, in physical stores, customers need to either ask store staff directly or search for information themselves to get instant product information, which can be time-consuming. Furthermore, some stores may not have sufficient staff, making it difficult for customers to obtain the information they need immediately. Therefore, there is a need for an efficient system that allows customers to get immediate answers to their product inquiries within physical stores.

[0197] 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.

[0198] In this invention, the server includes means for acquiring user voice input, means for converting voice input into text data, means for analyzing the intent of the user's question, means for searching for relevant information based on the analysis results, means for generating the retrieved information as text data, means for converting the text data into voice data, means for providing the voice data to the user, and software that operates on a smart device, including applications for in-store applications where customers ask questions about product features and specifications. As a result, even when a customer asks a question about a product in a physical store using voice input, relevant information is provided immediately, making it possible to obtain the necessary information quickly and accurately.

[0199] definition statement

[0200] "Means for acquiring user voice input" refers to devices or systems that capture the voice spoken by a user and save it as digital data.

[0201] "Means of converting voice input into text data" refers to the process or system of converting acquired voice data into text data.

[0202] "Means for analyzing the intent behind user questions" refers to methods and algorithms for analyzing questions that have been converted into text data and understanding their content and intent.

[0203] "Means for searching for relevant information based on analysis results" refers to a function that retrieves relevant data from appropriate sources in accordance with the analyzed intent.

[0204] "Means for generating retrieved information as text data" refers to a mechanism that processes and generates acquired information as text data in an appropriate format.

[0205] "Means of converting text data into audio data" refers to technologies and tools that convert generated text into synthesized speech.

[0206] "Means of providing audio data to users" refers to devices or systems that play synthesized audio and allow users to listen to it.

[0207] "Application to in-store applications where customers ask questions about product features and specifications" refers to an application technology that allows customers in physical stores to ask questions about products using voice and receive immediate answers to those questions.

[0208] "Software that runs on smart devices" refers to a series of programs and applications that run on electronic devices such as smartphones and smart glasses.

[0209] Modes for carrying out the invention

[0210] System Program Overview

[0211] The system of this invention acquires voice input from a user, analyzes it, and immediately provides information about products. The entire system includes a series of processes in which the user inputs voice via a smart device (such as a smartphone or smart glasses), the voice is converted into text, analyzed, and an appropriate response is generated, synthesized into speech, and then provided back to the user.

[0212] Hardware and software usage

[0213] Hardware:

[0214] 1. Smart devices: Examples include smartphones and smart glasses. These devices have voice input and voice output capabilities.

[0215] 2. Microphone: The device's built-in microphone captures voice input.

[0216] software:

[0217] 1. Speech recognition engine: Use the speech_recognition library to convert speech data into text data.

[0218] 2. Natural Language Processing Algorithm: The intent of the question will be analyzed using OpenAI's GPT-3 model, and an appropriate answer will be generated. Specifically, the OpenAI API will be used.

[0219] 3. Text-to-Speech Engine: The GTTS (Google Text-to-Speech) library and the playsound library are used to convert text-based responses into audio data and then play it back.

[0220] Data processing and data calculation

[0221] 1. Acquisition of voice input:

[0222] The user voice-inputs questions about the product into the device.

[0223] The microphone captures the sound and saves it as digital data.

[0224] 2. Speech recognition:

[0225] The acquired audio data is converted into text data using the speech_recognition library.

[0226] 3. Question analysis and answer generation:

[0227] Text data is sent as a prompt to the OpenAI API, analyzed using a GPT-3 model, and an appropriate response is generated.

[0228] Examples of prompt statements to use:

[0229] A customer asked: 'What are the features of this shampoo?' Please provide information about the product.

[0230] 4. Speech synthesis and output:

[0231] The generated text data is converted into audio data using the gtts library.

[0232] The playsound library is used to play the generated audio data and provide it to the user.

[0233] Specific example

[0234] For example, if a user asks "What are the features of this shampoo?" while in a physical store, the system will operate as follows:

[0235] 1. The microphone captures the audio, and the speech recognition engine generates text data that says, "What are the features of this shampoo?"

[0236] 2. This text data is sent to the OpenAI API, and the GPT-3 model generates the response, "This shampoo is rich in natural ingredients and makes hair soft."

[0237] 3. The generated text response is converted into audio data by a speech synthesis engine and provided to the user as audio using the playsound library.

[0238] This system allows customers to quickly obtain product information within physical stores, reducing the burden on store staff. Furthermore, it allows customers to deepen their understanding of the products.

[0239] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0240] Program processing steps

[0241] Step 1:

[0242] The user voice-inputs a question about the product into a smart device. For example, the user might say, "What are the features of this shampoo?" The input is voice data, which is then captured as digital data using a microphone.

[0243] Step 2:

[0244] The device converts the captured audio data into text data using the speech_recognition library. In this step, the audio data is the input, and the converted text data is the output. Specifically, the speech recognition engine converts it into the text "What are the features of this shampoo?".

[0245] Step 3:

[0246] The terminal sends the converted text data to the server. The server then sends it to the OpenAI API in the form of a prompt sentence to analyze the user's intent. The input is text data, which is converted into data in the form of a prompt sentence. For example, the prompt sentence "The customer asked: 'What are the features of this shampoo?' Please provide information about the product." is generated.

[0247] Step 4:

[0248] The server sends the generated prompt to the OpenAI GPT-3 model, which analyzes the intent of the question and generates an appropriate answer. The input is the prompt, and the output is the text data of the answer. Specifically, GPT-3 generates the answer "This shampoo is rich in natural ingredients and makes hair soft."

[0249] Step 5:

[0250] The server returns the generated response as text data to the terminal. The input is the generated text data, and the output and transmitted data are also text data.

[0251] Step 6:

[0252] The device converts received text data into audio data using GTTS. The input is text data, and the output is audio data. Specifically, the sentence "This shampoo is rich in natural ingredients and makes hair soft" is converted into audio data.

[0253] Step 7:

[0254] The device plays the converted audio data using the playsound library and provides it to the user. The input is audio data, and the output is the played audio. The user can receive a response via voice from their smart device.

[0255] 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.

[0256] The present invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to audiobooks, and more particularly to a system that incorporates an emotion engine to recognize the user's emotions. The system consists of the following main parts: voice input acquisition means, voice recognition means, question analysis means, information retrieval means, answer generation means, voice synthesis means, voice output means, and emotion recognition means.

[0257] Overall processing flow

[0258] 1. Acquisition of voice input

[0259] The user can ask questions about what they want to know using voice commands.

[0260] The device captures the user's voice through the microphone.

[0261] 2. Speech Recognition

[0262] The device converts the captured audio into text data and sends it to the speech recognition engine.

[0263] 3. Question analysis

[0264] The server receives text data sent from the speech recognition engine and uses natural language processing to analyze the intent of the question.

[0265] 4. Information Retrieval

[0266] Based on the intent of the analyzed question, the server searches for relevant information within the audiobook content.

[0267] 5. Answer generation

[0268] The server generates an answer based on the search results.

[0269] 6. Speech synthesis

[0270] The server uses a text-to-speech engine to convert the generated answer from text to speech.

[0271] 7. Emotion Recognition

[0272] The server analyzes the user's emotion using the emotion engine with the voice input and text data.

[0273] 8. Answer Adjustment According to Emotion

[0274] The server adjusts the content and tone of the answer based on the recognized emotion of the user.

[0275] 9. Voice Output <000087\4> The server sends the synthesized voice data to the terminal.

[0277] The terminal plays the received voice data and provides an answer to the user.

[0278] Detailed explanations for each step are shown below.

[0279] Obtaining Voice Input

[0280] \4>The user speaks to the terminal, for example, "What is the name of the protagonist in this book?"

[0281] The terminal captures the voice through the built-in microphone and saves it as voice data.

[0282] Voice Recognition

[0283] The terminal sends the captured voice data to the voice recognition engine on the cloud. <00\40897>

[0284] The voice recognition engine converts the voice data into text data "What is the name of the protagonist in this book?"

[0285] The terminal receives the converted text data and sends it to the server.

[0286] Question analysis

[0287] The server analyzes the received text data using natural language processing algorithms.

[0288] Understand the intent of the question and specifically identify the information the user wants to know (in this case, "the name of the protagonist").

[0289] Information search

[0290] Based on the identified intent, the server issues a search query to the audio book content database.

[0291] The server obtains relevant information, such as the name of the protagonist "John Smith", as the search result.

[0292] Answer generation

[0293] Based on the obtained search results, the server generates an answer such as "The name of the protagonist in this book is John Smith".

[0294] Voice synthesis

[0295] The server sends the generated text-form answer to the voice synthesis engine.

[0296] The voice synthesis engine converts this text into voice data.

[0297] Emotion recognition

[0298] The server uses the user's voice input and text data for analysis by the emotion engine. For example, recognize whether the user is feeling stressed when asking questions.

[0299] Answer adjustment according to emotion

[0300] The server adjusts the content and tone of the response based on the recognized emotions of the user. For example, if the user is irritated, the server responds in a gentle tone.

[0301] Voice output

[0302] The server sends the adjusted voice data to the terminal.

[0303] The terminal plays the voice data and provides it to the user, saying, "The name of the protagonist in this book is John Smith."

[0304] Specific example

[0305] [[ID=I9]] For example, when the user is tired and asks, "What is the name of the protagonist in this book?" and the emotion engine recognizes the user's fatigue. The server generates a response in a gentle tone, saying, "The name of the protagonist in this book is John Smith. Please enjoy the rest at your own pace." and plays it through voice synthesis.

[0306] The present invention can adjust responses according to the emotions of the user and is an advanced adaptation system for improving the user experience.

[0307] The following describes the processing flow.

[0308] Step 1:

[0309] The user verbally asks a question towards the terminal. For example, "What is the name of the protagonist in this book?"

[0310] Step 2:

[0311] The terminal captures the user's voice using the terminal's microphone and saves it as digital voice data. [[ID=:46]]

[0312] I Step 3:

[0313] The terminal sends the voice data captured to the voice recognition engine.

[0314] Step 4:

[0315] The speech recognition engine analyzes the audio data and converts it into corresponding text data. For example, it might generate text data such as, "What is the name of the main character in this book?"

[0316] Step 5:

[0317] The speech recognition engine sends the converted text data back to the device.

[0318] Step 6:

[0319] The terminal sends text data to the server.

[0320] Step 7:

[0321] The server analyzes the received text data using natural language processing algorithms. Specifically, it grasps the intent of the user's question and extracts key keywords (e.g., "the protagonist's name").

[0322] Step 8:

[0323] Based on the analysis results, the server issues a query to the audiobook content database to retrieve relevant information.

[0324] Step 9:

[0325] The server retrieves relevant information from the database. For example, the data "John Smith" is searched for.

[0326] Step 10:

[0327] Based on the search results obtained by the server, the system generates the answer, "The name of the main character in this book is John Smith."

[0328] Step 11:

[0329] The server sends the generated text-formatted response to the sentiment engine.

[0330] Step 12:

[0331] The server uses an emotion engine to recognize the user's emotions from their voice input. For example, it can determine if the user is stressed or has questions.

[0332] Step 13:

[0333] The server adjusts the response content to match the user's emotions based on the emotions it recognizes. For example, if the user is tired, the response will be made more kind and include elements of encouragement.

[0334] Step 14:

[0335] The server sends text that has been adjusted according to the emotion to the speech synthesis engine.

[0336] Step 15:

[0337] The speech synthesis engine analyzes text data and converts it into speech data.

[0338] Step 16:

[0339] The speech synthesis engine sends the generated audio data back to the server.

[0340] Step 17:

[0341] The server sends the audio data to the terminal.

[0342] Step 18:

[0343] The device plays the received audio data and tells the user, "The main character of this book is named John Smith. Please enjoy the rest of the story."

[0344] In this way, users can ask questions in their own voice, and the system can not only respond appropriately to those questions but also adjust its answers according to the user's emotions. This process allows users to enjoy a more personalized experience.

[0345] (Example 2)

[0346] 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".

[0347] Conventional voice dialogue systems have the problem of not improving the user experience because they provide mechanical responses without considering the user's emotions. Furthermore, the accuracy of answers to user questions is low, making it difficult to provide accurate information quickly. In addition, there was a lack of efficient ways to obtain necessary information when listening to audio content such as audiobooks.

[0348] 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.

[0349] In this invention, the server includes means for acquiring user voice input, means for converting voice input into text data, means for analyzing the intent of the user's question, means for searching for relevant information based on the analysis results, means for generating the retrieved information as text data, means for converting the text data into voice data, means for analyzing the user's emotions, means for adjusting the tone of the response based on the analyzed emotions, and means for providing the adjusted voice data to the user. This enables responses that are in line with the user's emotions, improving the user experience and enabling the rapid provision of accurate information.

[0350] "Means for acquiring voice input" refers to a device or method for capturing the voice spoken by a user and converting it into digital data.

[0351] "Means for converting voice input into text data" refers to technologies and algorithms for analyzing captured voice data and converting it into a corresponding text format.

[0352] "Means for analyzing the intent behind a user's question" refers to natural language processing techniques that analyze the content of text data to understand the information the user wants to know and the purpose of their question.

[0353] "Means for retrieving relevant information based on analysis results" refers to techniques for retrieving relevant information from databases and other sources based on the intent of the analyzed question.

[0354] "Means for generating searched information as text data" refers to technologies for generating information in text format based on search results and providing it to users.

[0355] "Methods for converting text data into audio data" refers to technologies that convert generated text data into audio data using speech synthesis technology.

[0356] "Methods for analyzing user emotions" refer to technologies that recognize and analyze the emotions a user is feeling based on their voice input and text data.

[0357] "Means for adjusting the tone of responses based on analyzed emotions" refers to technologies that appropriately modify the content and tone of generated responses according to the user's emotions.

[0358] "Means for providing adjusted audio data to a user" refers to devices or methods for playing or providing adjusted audio data to a user.

[0359] Modes for carrying out the invention

[0360] This invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to audiobooks, and simultaneously recognize and respond to the user's emotions. This system includes the following hardware and software.

[0361] Main components and devices

[0362] 1. Voice input acquisition method

[0363] The device uses its built-in microphone to acquire voice input from the user.

[0364] An external microphone or a microphone built into the device will be used.

[0365] 2. Speech recognition means

[0366] The device sends the acquired voice data to a speech recognition engine in the cloud.

[0367] The specific software used is the Google Cloud Speech-to-Text API.

[0368] The speech recognition engine converts speech data into text data.

[0369] 3. Question analysis means

[0370] The server analyzes the received text data using a natural language processing algorithm.

[0371] Specifically, the Python libraries NLTK and spaCy will be used.

[0372] The server analyzes the intent behind the user's question and identifies specific information.

[0373] 4. Information Retrieval Methods

[0374] Based on the analysis results, the server issues a search query to the audiobook content database.

[0375] SQL queries are used for the search.

[0376] The server retrieves relevant information as search results.

[0377] 5. Answer generation means

[0378] Based on the information it retrieves, the server generates an appropriate response using a template.

[0379] The response will be generated as text data.

[0380] 6. Speech synthesis means

[0381] The server uses a speech synthesis engine to convert the generated text data into speech data.

[0382] The Google Text-to-Speech API is used.

[0383] 7. Emotion recognition means

[0384] The server analyzes the user's emotions based on their voice input and text data.

[0385] OpenAI's sentiment analysis model will be used.

[0386] 8. Response adjustment mechanisms based on emotions

[0387] The server adjusts the tone of its responses based on the analyzed user's emotions.

[0388] By adding gentle tones and encouraging words, we can respond to the user's emotions.

[0389] 9. Audio output means

[0390] The server sends the adjusted audio data to the terminal.

[0391] The device plays audio data and provides a response to the user.

[0392] Specific example

[0393] For example, if a user asks, "What is the name of the main character in this book?", the device captures this audio and converts it to text using the Google Cloud Speech-to-Text API. The server then analyzes this text using natural language processing algorithms (such as NLTK or spaCy) and searches for relevant information in a database. Based on the search results, the answer "The main character in this book is named John Smith." is generated and converted to speech using the Google Text-to-Speech API. The user's emotions are then analyzed using an OpenAI sentiment analysis model, and the tone is adjusted as needed. Finally, the adjusted audio data is sent to the device and provided to the user.

[0394] Example of a prompt

[0395] "Convert the user's spoken questions into text and generate corresponding answers."

[0396] "Analyze the user's emotions and adjust the tone of your responses accordingly."

[0397] "Please search for the main character's name in the audiobook and tell it to the user."

[0398] "Adjust the content and tone of your responses according to the user's emotions."

[0399] This system aims to significantly improve the user experience by allowing users to quickly and accurately obtain necessary information while enjoying audiobooks, and by providing responses that take the user's emotions into consideration.

[0400] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0401] System program processing steps

[0402] Step 1: Obtaining voice input

[0403] The user asks a question to the device.

[0404] Input: User's voice (e.g., "What is the name of the main character in this book?")

[0405] Operation: The device's built-in microphone captures the user's voice and saves it as audio data (in WAV format, etc.).

[0406] Output: Captured audio data

[0407] Step 2: Speech Recognition

[0408] The device sends the captured audio data to a speech recognition engine in the cloud.

[0409] Input: Audio data

[0410] How it works: The Google Cloud Speech-to-Text API converts speech data into text data.

[0411] Output: Converted text data (e.g., "What is the name of the main character in this book?")

[0412] Step 3: Questionnaire Analysis

[0413] The server analyzes the received text data using a natural language processing algorithm.

[0414] Input: Text data

[0415] Operation: Uses Python's NLTK and spaCy libraries to parse text data and identify the intent behind user questions. For example, it parses the subject, predicate, and object of a question.

[0416] Output: Analysis results (Example: "I want to know the protagonist's name")

[0417] Step 4: Information Retrieval

[0418] Based on the analysis results, the server issues a search query to the audiobook content database.

[0419] Input: Analysis results

[0420] Operation: Generates an SQL query and performs a search on the audiobook database. For example, it searches for information related to "the main character's name".

[0421] Output: Search results (e.g., "John Smith")

[0422] Step 5: Generate Response

[0423] The server generates an answer based on the retrieved search results.

[0424] Input: Search Results

[0425] Operation: Generates an answer using a template. Converts it to a format such as "The name of the main character in this book is John Smith."

[0426] Output: Text-formatted answer (Example: "The name of the main character in this book is John Smith.")

[0427] Step 6: Speech Synthesis

[0428] The server sends the generated text-formatted response to the speech synthesis engine.

[0429] Input: Text data response

[0430] How it works: The Google Text-to-Speech API converts text into speech data.

[0431] Output: Audio data (e.g., audio file)

[0432] Step 7: Emotion Recognition

[0433] The server analyzes the user's emotions based on their voice input and text data.

[0434] Input: Audio data and text data

[0435] Operation: Uses OpenAI's emotion analysis model to analyze the user's emotions. For example, it identifies stress and fatigue from voice tone and text content.

[0436] Output: User's emotional state (e.g., "tired")

[0437] Step 8: Adjusting responses based on emotions

[0438] The server adjusts the tone of its responses based on the analyzed user's emotions.

[0439] Input: Emotional state, text-formatted response

[0440] Action: If the user is tired, add a gentle tone and encouraging words to the response. For example, adjust it to a format like, "The main character of this book is named John Smith. Please take a break and enjoy the rest of the story."

[0441] Output: Answer in adjusted text format

[0442] Step 9: Audio Output

[0443] The server sends the adjusted audio data to the terminal.

[0444] Input: Adjusted audio data

[0445] Operation: The device plays the received audio data.

[0446] Output: The user listens to pre-recorded audio data (e.g., "The main character of this book is named John Smith. Please relax and enjoy the rest of the story.")

[0447] In this way, by clearly defining the specific actions, inputs, and outputs at each step, we have created a system that allows users to quickly and accurately obtain the information they need, and furthermore, can respond to the user's emotions during the process.

[0448] (Application Example 2)

[0449] 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".

[0450] Conventional speech recognition systems have made it difficult for users to quickly and accurately obtain necessary information while listening to audiobooks. Furthermore, conventional systems cannot recognize the user's emotions and adjust the content and tone of the responses accordingly, resulting in a limited user experience. The present invention aims to provide a system that not only provides information based on the user's voice input but also analyzes the user's emotions in real time and provides responses with the optimal tone and content.

[0451] 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.

[0452] In this invention, the server includes means for converting voice input into text data, means for analyzing the intent of the user's question, and means for retrieving relevant information. This enables a personalized audiobook experience that responds to the user's emotions.

[0453] "Means for acquiring voice input" refers to devices or mechanisms for capturing the voice spoken by a user and recording it as digitized voice data.

[0454] "Means for converting voice input into text data" refers to engines or software that analyze captured voice data and convert it into text information.

[0455] "Means for analyzing the intent behind user questions" refer to natural language processing algorithms and systems that understand the intent behind user questions obtained as text data and identify relevant information.

[0456] "Means for searching for related information" refers to mechanisms or engines for detecting necessary information from databases and related resources based on the analyzed intent.

[0457] "Means for generating searched information as text data" refers to software or algorithms that organize searched information into text in a format that is easy for users to understand.

[0458] "Means for converting text data into audio data" refers to speech synthesis engines or systems that convert the generated text data into an audio format.

[0459] "Means of providing audio data to the user" refers to devices such as speakers and headsets that play back converted audio data and convey information to the user by voice.

[0460] "Means of recognizing user emotions" refer to emotion recognition engines and software that analyze user voice input, facial expression data, etc., to identify their emotional state.

[0461] "Means of adjusting responses according to emotions" refers to algorithms and control systems that appropriately change the content and tone of the responses provided based on the recognized emotional state of the user.

[0462] This invention provides a system that allows users to quickly and accurately obtain necessary information while listening to audiobooks, and in particular, incorporates an emotion engine that recognizes the user's emotions. The system includes a voice input acquisition means, a voice recognition means, a question analysis means, an information retrieval means, an answer generation means, a voice synthesis means, a voice output means, an emotion recognition means, and an answer adjustment means according to emotion.

[0463] Specific examples of hardware and software

[0464] hardware

[0465] Smart glasses: Equipped with a built-in microphone and speaker to capture the user's voice and provide the results to the user.

[0466] Server: Processes and manages data.

[0467] software

[0468] Speech recognition engine: Uses Google Cloud Speech-to-Text to convert speech input into text.

[0469] The natural language processing algorithm, TENSORFLOW®, is used to analyze the user's question intent.

[0470] Speech synthesis engine: Uses Amazon Polly to convert text data into speech data.

[0471] Emotion recognition engine: Uses Affectiva or Microsoft® Azure® Emotion API to analyze user emotions.

[0472] Content distribution service APIs: Utilizing APIs from services like Audible and Spotify.

[0473] Program processing details

[0474] 1. Acquisition of voice input:

[0475] The smart glasses capture the user's voice commands using a microphone.

[0476] 2. Speech recognition:

[0477] The captured audio data is sent to a cloud-based speech recognition engine (Google Cloud Speech-to-Text) and converted into text data.

[0478] 3. Question analysis:

[0479] The converted text data is sent to a server and analyzed using a natural language processing algorithm (TensorFlow model) to identify the intent of the question.

[0480] 4. Information Retrieval:

[0481] The server searches for relevant information from the audiobook content database based on the intent of the question.

[0482] 5. Answer generation:

[0483] Based on the search results, it generates specific answers such as, "The name of the main character in this book is XX."

[0484] 6. Speech synthesis:

[0485] The generated responses are converted into audio data using Amazon Polly.

[0486] 7. Emotion recognition:

[0487] Affectiva is used to analyze user voice input and text data to identify the user's emotional state.

[0488] 8. Answer adjustment:

[0489] Based on the perceived emotions, the tone and content of the response will be adjusted. For example, if the user is feeling tired, the response will encourage them to relax.

[0490] 9. Audio output:

[0491] The adjusted audio data is sent to the smart glasses, and the information is provided audibly through the speaker.

[0492] Specific example

[0493] If the user asks a question: For example, if the user says, "What is the name of the main character in this book?"

[0494] A concrete example of emotion recognition: When the emotion recognition engine determines that the user is tired.

[0495] Adapted response: The smart glasses' speaker responds in a gentle tone, "The main character of this book is named XX. Please relax and enjoy the rest of the story."

[0496] Example of a prompt

[0497] "We want to develop an application that provides a personalized experience by offering appropriate responses based on voice input instructions, tailored to the user's emotional state. It would quickly search for the information the user is looking for within the audiobook content, provide the answer via voice, and adjust the tone of response based on the user's emotional state. For example, if the user is tired, the response would include a message encouraging relaxation."

[0498] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0499] Step 1:

[0500] The user listens to an audiobook and makes a voice input into the microphone. The device (smart glasses) captures the user's voice through its built-in microphone and records it as digital audio data.

[0501] Input: User voice input (for example, "What is the name of the main character in this book?")

[0502] Output: Digital audio data

[0503] Operation: The microphone records the user's speech as audio data.

[0504] Step 2:

[0505] The device sends the captured digital audio data to a cloud-based speech recognition engine (Google Cloud Speech-to-Text), which converts the audio data into text data.

[0506] Input: Digital audio data

[0507] Output: Text data (for example, "What is the name of the main character in this book?")

[0508] Operation: The device uploads voice data to the cloud engine, and the speech recognition engine converts it into text.

[0509] Step 3:

[0510] The server receives the converted text data and uses a natural language processing algorithm (TensorFlow model) to analyze the user's intent behind the question.

[0511] Input: Text data

[0512] Output: Analyzed question intent (e.g., "I want to know the protagonist's name")

[0513] Operation: The server inputs text data into a natural language processing algorithm to analyze the intent of the question.

[0514] Step 4:

[0515] The server searches for relevant information from the audiobook content database based on the analyzed question intent.

[0516] Input: Analyzed question intent

[0517] Output: Related information (e.g., "Protagonist's name: John Smith")

[0518] Operation: The server issues a query to the database and retrieves the corresponding information.

[0519] Step 5:

[0520] The server generates a text-based response based on the relevant information it has retrieved.

[0521] Input: Related Information

[0522] Output: Text-formatted answer (e.g., "The protagonist's name is John Smith")

[0523] Operation: The server combines relevant information and generates a response in text format to provide to the user.

[0524] Step 6:

[0525] The server sends the generated text-based response to a speech synthesis engine (Amazon Polly) to convert it into audio data.

[0526] Input: Text format answer

[0527] Output: Audio data

[0528] Operation: The server sends text data to the speech synthesis engine and retrieves the data output as speech.

[0529] Step 7:

[0530] The server analyzes voice input and text data using an emotion recognition engine (Affectiva) to recognize the user's emotions.

[0531] Input: Voice input, text data

[0532] Output: User's emotional state (e.g., "tired")

[0533] Operation: The server sends voice input and text data to the emotion recognition engine, which then analyzes the user's emotions.

[0534] Step 8:

[0535] The server adjusts the tone and content of its responses based on the emotions it perceives.

[0536] Input: User's emotional state, text-formatted response.

[0537] Output: Adjusted audio data

[0538] Operation: The server uses the results of emotion recognition to change the response to a gentler tone or content that promotes relaxation.

[0539] Step 9:

[0540] The device transmits the adjusted audio data to the smart glasses' speaker, providing information to the user.

[0541] Input: Adjusted audio data

[0542] Output: Provides an answer via voice (for example, "The main character's name is John Smith. Relax and enjoy the rest.")

[0543] Operation: The device retrieves audio data from the server, plays it through the speaker, and provides it to the user.

[0544] This processing flow allows users to ask questions in real time while listening to audiobooks and receive personalized answers that are tailored to their emotions.

[0545] 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.

[0546] 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 those described above. 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 shown 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.

[0547] 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.

[0548] [Second Embodiment]

[0549] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0550] 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.

[0551] 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).

[0552] 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.

[0553] 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.

[0554] 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).

[0555] 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.

[0556] 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.

[0557] 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.

[0558] 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.

[0559] 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.

[0560] 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".

[0561] The present invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to an audiobook. The system consists of the following main components: voice input acquisition means, voice recognition means, question analysis means, information retrieval means, answer generation means, voice synthesis means, and voice output means.

[0562] Overall processing flow

[0563] 1. Acquisition of voice input

[0564] The user can ask questions about what they want to know using voice commands.

[0565] The device captures the user's voice through the microphone.

[0566] 2. Speech Recognition

[0567] The device converts the captured audio into text data and sends it to the speech recognition engine.

[0568] 3. Question analysis

[0569] The server receives text data sent from the speech recognition engine and uses natural language processing to analyze the intent of the question.

[0570] 4. Information Retrieval

[0571] Based on the intent of the analyzed question, the server searches for relevant information within the audiobook content.

[0572] 5. Answer generation

[0573] The server generates an answer based on the search results.

[0574] 6. Speech synthesis

[0575] The server uses a speech synthesis engine to convert the generated responses from text to speech.

[0576] 7. Audio Output

[0577] The server sends the synthesized voice data to the terminal.

[0578] The device plays back the received audio data and provides the user with an answer.

[0579] Detailed explanations for each step are provided below.

[0580] Acquiring voice input

[0581] The user says to the device, "What is the name of the main character in this book?"

[0582] The device captures this audio through its built-in microphone and saves it as audio data.

[0583] Speech recognition

[0584] The device sends the captured audio data to a speech recognition engine in the cloud.

[0585] The speech recognition engine converts the audio data into text data such as "What is the name of the main character in this book?".

[0586] The terminal receives the converted text data and sends it to the server.

[0587] question analysis

[0588] The server analyzes the received text data using a natural language processing algorithm.

[0589] Understand the intent of the question and identify the specific information the user wants to know (in this case, "the protagonist's name").

[0590] Information Retrieval

[0591] Based on the identified intent, the server issues a search query to the audiobook content database.

[0592] The server retrieves relevant information, such as the protagonist's name "John Smith," as a search result.

[0593] Answer generation

[0594] Based on the retrieved search results, the server generates the answer, "The name of the main character in this book is John Smith."

[0595] Speech synthesis

[0596] The server sends the generated text-based response to the speech synthesis engine.

[0597] The speech synthesis engine converts this text into speech data.

[0598] Audio output

[0599] The server sends the audio data to the terminal.

[0600] The device plays audio data and tells the user, "The name of the main character in this book is John Smith."

[0601] Specific example

[0602] For example, if a user asks, "What is the most important point in this chapter?", the same process would apply.

[0603] The server generates the response, "The key point of this chapter is the scene where the protagonist makes a firm decision," and plays it back to the user as audio.

[0604] This invention allows users to instantly obtain the information they want while listening to audiobooks, deepening their understanding of the content. In addition, it reduces the wasted time caused by conventional methods of re-searching and re-listening, enabling more efficient access to information.

[0605] The following describes the processing flow.

[0606] Step 1:

[0607] The user speaks a question into the device. Example: "What is the name of the main character in this book?"

[0608] Step 2:

[0609] The device uses its microphone to capture the user's voice and saves it as digital audio data.

[0610] Step 3:

[0611] The device sends the captured audio data to the speech recognition engine.

[0612] Step 4:

[0613] The speech recognition engine analyzes the audio data and converts it into corresponding text data. For example, it might generate text data such as, "What is the name of the main character in this book?"

[0614] Step 5:

[0615] The speech recognition engine sends the converted text data back to the device.

[0616] Step 6:

[0617] The terminal sends text data to the server.

[0618] Step 7:

[0619] The server analyzes the received text data using natural language processing algorithms. Specifically, it grasps the intent of the user's question and extracts important keywords (e.g., "the protagonist's name").

[0620] Step 8:

[0621] Based on the analysis results, the server issues a query to the audiobook content database to retrieve relevant information.

[0622] Step 9:

[0623] The server retrieves relevant information from the database. For example, the data "John Smith" is searched for.

[0624] Step 10:

[0625] The server generates the answer, "The name of the main character in this book is John Smith," based on the search results.

[0626] Step 11:

[0627] The server sends the generated text-based response to the speech synthesis engine.

[0628] Step 12:

[0629] The speech synthesis engine analyzes text data and converts it into speech data.

[0630] Step 13:

[0631] The speech synthesis engine sends the generated audio data back to the server.

[0632] Step 14:

[0633] The server sends the audio data to the terminal.

[0634] Step 15:

[0635] The device plays the received audio data and tells the user, "The name of the main character in this book is John Smith."

[0636] (Example 1)

[0637] 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".

[0638] Conventional audio content listening systems were inefficient because users had to manually re-search or re-listen to obtain the information they needed. Furthermore, systems that converted voice input to text and provided appropriate answers to user questions lacked sufficient technology to accurately analyze the intent of the questions and provide timely, appropriate responses. As a result, it was difficult for users to quickly and accurately obtain the information they needed while listening to audio content.

[0639] 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.

[0640] In this invention, the server includes means for enabling the user to ask questions in real time while listening to audio content, means for generating answers to the questions using a natural language processing algorithm and a generative AI model, and means for converting the generated answers into audio data using a speech synthesizer. This enables the user to quickly and accurately obtain the necessary information without interrupting the audio content.

[0641] A "voice input device" is a device that captures a user's voice and converts it into digital voice data.

[0642] A "speech recognition engine" is a program or service that converts speech data into text data.

[0643] A "natural language processing algorithm" is a computational method for analyzing text data and understanding its meaning, context, and intent.

[0644] A "generative AI model" is an artificial intelligence model designed to generate natural language responses based on text data.

[0645] A "speech synthesis device" is a device or service used to convert text data into speech data.

[0646] "Audio content" refers to information and entertainment provided in audio format.

[0647] A "prompt sentence" is an input sentence given to a generative AI model, and it is an instruction sentence used to generate a specific answer.

[0648] "Real-time" refers to the fact that user requests and operations are processed instantly with virtually no delay.

[0649] A "database" is a system designed to organize and manage large amounts of data, enabling efficient searching and updating.

[0650] A "user" is a person who attempts to obtain information by operating a system.

[0651] A "terminal" is an electronic device used by a user to access a system.

[0652] A "server" is a computer that acts as the central hub of a system, processing, storing, and providing data.

[0653] This invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to audio content. This system acquires the user's voice input and performs a series of processes to provide an appropriate response using natural language processing and a generative AI model, and its specific implementation method is described below.

[0654] System Overview

[0655] This system consists of the following main parts:

[0656] 1. Voice input device

[0657] 2. Speech Recognition Engine

[0658] 3. Natural Language Processing Algorithms

[0659] 4. Generative AI Models

[0660] 5. Speech synthesis device

[0661] 6. Server

[0662] 7. Terminal

[0663] Required hardware and software

[0664] Voice input device: Used to capture the user's voice. Example: Microphone.

[0665] Speech recognition engine: Used to convert speech data into text data. Examples: Google Cloud Speech-to-Text API, IBM Watson Speech to Text, etc.

[0666] Natural Language Processing (NLP) algorithms: Used to analyze the intent behind a user's question. Examples: Natural Language Processing libraries such as NLTK and spaCy.

[0667] Generative AI models: Used to generate appropriate answers to questions. Example: OpenAI's GPT-3.

[0668] Text-to-speech (Speech synthesis) devices are used to convert text data into speech data. Examples include Amazon Polly and Google Cloud Text-to-Speech.

[0669] Server: Manages the processing of the entire system.

[0670] Terminal: A device used by a user to access a system, input questions, and receive answers. Examples: smartphones, tablets, personal computers, etc.

[0671] Specific example

[0672] When a user is listening to audio content, for example, they might be asked, "What is the name of the main character in this book?" In this case, the system operates as follows:

[0673] 1. Acquisition of voice input

[0674] The user says to the device, "What is the name of the main character in this book?"

[0675] The device captures this audio through its built-in microphone and saves it as audio data.

[0676] 2. Speech Recognition

[0677] The device sends the captured audio data to a speech recognition engine in the cloud.

[0678] The speech recognition engine converts the audio data into text data such as "What is the name of the main character in this book?".

[0679] The terminal receives the converted text data and sends it to the server.

[0680] 3. Question analysis

[0681] The server analyzes the received text data using a natural language processing algorithm.

[0682] Understand the intent of the question and identify the specific information the user wants to know (in this case, "the protagonist's name").

[0683] 4. Information Retrieval

[0684] Based on the intent of the analyzed question, the server searches for relevant information from the audio content.

[0685] The server retrieves relevant information as search results. For example, the protagonist's name is "John Smith".

[0686] 5. Answer generation

[0687] Based on the retrieved search results, the server generates the answer, "The name of the main character in this book is John Smith."

[0688] 6. Speech synthesis

[0689] The server sends the generated text-based response to the speech synthesis engine.

[0690] The speech synthesis engine converts this text into speech data.

[0691] 7. Audio Output

[0692] The server sends the audio data to the terminal.

[0693] The device plays audio data and tells the user, "The name of the main character in this book is John Smith."

[0694] Example of a prompt

[0695] The following is an example of a prompt message used to generate an appropriate answer to a user's question using a generative AI model.

[0696] Prompt example (English):

[0697] "User asks in English: 'What is the name of the main character in this book?'. The audiobooks database has the following relevant information: 'The main character of the book is John Smith.'. Respond to the user's question with this information."

[0698] Prompt example (Japanese):

[0699] The user asks, "What is the name of the main character in this book?" The audio content database contains the related information, "The main character in this book is John Smith." Use this information to answer the user's question.

[0700] This invention allows users to instantly obtain the information they want while listening to audio content, leading to a deeper understanding of the content. This reduces the wasted time caused by traditional methods of re-searching or re-listening, enabling more efficient access to information.

[0701] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0702] Step 1:

[0703] The user asks a question aloud, "What is the name of the main character in this book?" The input is the user's voice, which is captured by the device's built-in microphone. Specifically, the device runs a voice capture program and saves it as audio data.

[0704] Step 2:

[0705] The device sends the captured audio data to a speech recognition engine in the cloud. The input is audio data, and the output is text data. Specifically, the device uses a cloud communication module to send audio data, which the speech recognition engine converts into text, such as "What is the name of the main character in this book?"

[0706] Step 3:

[0707] The device receives converted text data from the cloud and sends it to the server. The input is text data from the speech recognition engine, and the output is the transmission of text data to the server. Specifically, the device uses a cloud communication module to transfer the text data to the server.

[0708] Step 4:

[0709] The server analyzes the received text data using a natural language processing algorithm. The input is text data, and the output is the result of analyzing the intent of the user's question. Specifically, the server uses natural language processing libraries such as NLTK or spaCy to identify the intent of the question as "I want to know the name of the main character."

[0710] Step 5:

[0711] The server searches for relevant information based on the analysis results. The input is the analysis result (the intent of the question), and the output is the relevant information (search results). Specifically, the server performs a search on the database using SQL queries or Elasticsearch, and as a result retrieves the name "John Smith".

[0712] Step 6:

[0713] The server generates an answer based on the search results. The input is the information from the search results, and the output is the text data of the answer. Specifically, the server uses a generative AI model (for example, GPT-3) to generate the text "The name of the main character in this book is John Smith."

[0714] Step 7:

[0715] The server sends the generated text-based response to a text-to-speech engine, where it is converted into audio data. The input is the text data of the response, and the output is audio data. Specifically, the server uses Amazon Polly or Google Text-to-Speech to generate the audio data.

[0716] Step 8:

[0717] The server sends the generated audio data to the terminal. The input is audio data, and the output is the transmission of audio data to the terminal. Specifically, the server transfers the audio data to the terminal using a cloud communication module.

[0718] Step 9:

[0719] The terminal plays the received audio data and provides it to the user. The input is audio data, and the output is audio playback. Specifically, the terminal uses an audio playback program to play the sentence, "The name of the main character in this book is John Smith."

[0720] (Application Example 1)

[0721] 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."

[0722] Currently, in physical stores, customers need to either ask store staff directly or search for information themselves to get instant product information, which can be time-consuming. Furthermore, some stores may not have sufficient staff, making it difficult for customers to obtain the information they need immediately. Therefore, there is a need for an efficient system that allows customers to get immediate answers to their product inquiries within physical stores.

[0723] 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.

[0724] In this invention, the server includes means for acquiring user voice input, means for converting voice input into text data, means for analyzing the intent of the user's question, means for searching for relevant information based on the analysis results, means for generating the retrieved information as text data, means for converting the text data into voice data, means for providing the voice data to the user, and software that operates on a smart device, including applications for in-store applications where customers ask questions about product features and specifications. As a result, even when a customer asks a question about a product in a physical store using voice input, relevant information is provided immediately, making it possible to obtain the necessary information quickly and accurately.

[0725] definition statement

[0726] "Means for acquiring user voice input" refers to devices or systems that capture the voice spoken by a user and save it as digital data.

[0727] "Means of converting voice input into text data" refers to the process or system of converting acquired voice data into text data.

[0728] "Means for analyzing the intent behind user questions" refers to methods and algorithms for analyzing questions that have been converted into text data and understanding their content and intent.

[0729] "Means for searching for relevant information based on analysis results" refers to a function that retrieves relevant data from appropriate sources in accordance with the analyzed intent.

[0730] "Means for generating retrieved information as text data" refers to a mechanism that processes and generates acquired information as text data in an appropriate format.

[0731] "Means of converting text data into audio data" refers to technologies and tools that convert generated text into synthesized speech.

[0732] "Means of providing audio data to users" refers to devices or systems that play synthesized audio and allow users to listen to it.

[0733] "Application to in-store applications where customers ask questions about product features and specifications" refers to an application technology that allows customers in physical stores to ask questions about products using voice and receive immediate answers to those questions.

[0734] "Software that runs on smart devices" refers to a series of programs and applications that run on electronic devices such as smartphones and smart glasses.

[0735] Modes for carrying out the invention

[0736] System Program Overview

[0737] The system of this invention acquires voice input from a user, analyzes it, and immediately provides information about products. The entire system includes a series of processes in which the user inputs voice via a smart device (such as a smartphone or smart glasses), the voice is converted into text, analyzed, and an appropriate response is generated, synthesized into speech, and then provided back to the user.

[0738] Hardware and software usage

[0739] Hardware:

[0740] 1. Smart devices: Examples include smartphones and smart glasses. These devices have voice input and voice output capabilities.

[0741] 2. Microphone: The device's built-in microphone captures voice input.

[0742] software:

[0743] 1. Speech recognition engine: Use the speech_recognition library to convert speech data into text data.

[0744] 2. Natural Language Processing Algorithm: The intent of the question will be analyzed using OpenAI's GPT-3 model, and an appropriate answer will be generated. Specifically, the OpenAI API will be used.

[0745] 3. Text-to-Speech Engine: The GTTS (Google Text-to-Speech) library and the playsound library are used to convert text-based responses into audio data and then play it back.

[0746] Data processing and data calculation

[0747] 1. Acquisition of voice input:

[0748] The user voice-inputs questions about the product into the device.

[0749] The microphone captures the sound and saves it as digital data.

[0750] 2. Speech recognition:

[0751] The acquired audio data is converted into text data using the speech_recognition library.

[0752] 3. Question analysis and answer generation:

[0753] Text data is sent as a prompt to the OpenAI API, analyzed using a GPT-3 model, and an appropriate response is generated.

[0754] Examples of prompt statements to use:

[0755] A customer asked: 'What are the features of this shampoo?' Please provide information about the product.

[0756] 4. Speech synthesis and output:

[0757] The generated text data is converted into audio data using the gtts library.

[0758] The playsound library is used to play the generated audio data and provide it to the user.

[0759] Specific example

[0760] For example, if a user asks "What are the features of this shampoo?" while in a physical store, the system will operate as follows:

[0761] 1. The microphone captures the audio, and the speech recognition engine generates text data that says, "What are the features of this shampoo?"

[0762] 2. This text data is sent to the OpenAI API, and the GPT-3 model generates the response, "This shampoo is rich in natural ingredients and makes hair soft."

[0763] 3. The generated text response is converted into audio data by a speech synthesis engine and provided to the user as audio using the playsound library.

[0764] This system allows customers to quickly obtain product information within physical stores, reducing the burden on store staff. Furthermore, it allows customers to deepen their understanding of the products.

[0765] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0766] Program processing steps

[0767] Step 1:

[0768] The user voice-inputs a question about the product into a smart device. For example, the user might say, "What are the features of this shampoo?" The input is voice data, which is then captured as digital data using a microphone.

[0769] Step 2:

[0770] The device converts the captured audio data into text data using the speech_recognition library. In this step, the audio data is the input, and the converted text data is the output. Specifically, the speech recognition engine converts it into the text "What are the features of this shampoo?".

[0771] Step 3:

[0772] The terminal sends the converted text data to the server. The server then sends it to the OpenAI API in the form of a prompt sentence to analyze the user's intent. The input is text data, which is converted into data in the form of a prompt sentence. For example, the prompt sentence "The customer asked: 'What are the features of this shampoo?' Please provide information about the product." is generated.

[0773] Step 4:

[0774] The server sends the generated prompt to the OpenAI GPT-3 model, which analyzes the intent of the question and generates an appropriate answer. The input is the prompt, and the output is the text data of the answer. Specifically, GPT-3 generates the answer "This shampoo is rich in natural ingredients and makes hair soft."

[0775] Step 5:

[0776] The server returns the generated response as text data to the terminal. The input is the generated text data, and the output and transmitted data are also text data.

[0777] Step 6:

[0778] The device converts received text data into audio data using GTTS. The input is text data, and the output is audio data. Specifically, the sentence "This shampoo is rich in natural ingredients and makes hair soft" is converted into audio data.

[0779] Step 7:

[0780] The device plays the converted audio data using the playsound library and provides it to the user. The input is audio data, and the output is the played audio. The user can receive a response via voice from their smart device.

[0781] 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.

[0782] The present invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to audiobooks, and more particularly to a system that incorporates an emotion engine to recognize the user's emotions. The system consists of the following main parts: voice input acquisition means, voice recognition means, question analysis means, information retrieval means, answer generation means, voice synthesis means, voice output means, and emotion recognition means.

[0783] Overall processing flow

[0784] 1. Acquisition of voice input

[0785] The user can ask questions about what they want to know using voice commands.

[0786] The device captures the user's voice through the microphone.

[0787] 2. Speech Recognition

[0788] The device converts the captured audio into text data and sends it to the speech recognition engine.

[0789] 3. Question analysis

[0790] The server receives text data sent from the speech recognition engine and uses natural language processing to analyze the intent of the question.

[0791] 4. Information Retrieval

[0792] Based on the intent of the analyzed question, the server searches for relevant information within the audiobook content.

[0793] 5. Answer generation

[0794] The server generates an answer based on the search results.

[0795] 6. Speech synthesis

[0796] The server uses a speech synthesis engine to convert the generated responses from text to speech.

[0797] 7. Emotion recognition

[0798] The server uses voice input and text data to analyze the user's emotions using an emotion engine.

[0799] 8. Adjusting responses based on emotions.

[0800] The server adjusts the content and tone of its responses based on the recognized emotions of the user.

[0801] 9. Audio Output

[0802] The server sends the synthesized voice data to the terminal.

[0803] The device plays back the received audio data and provides the user with an answer.

[0804] Detailed explanations for each step are provided below.

[0805] Acquiring voice input

[0806] For example, the user might say to the device, "What is the name of the main character in this book?"

[0807] The device captures audio through its built-in microphone and saves it as audio data.

[0808] Speech recognition

[0809] The device sends the captured audio data to a speech recognition engine in the cloud.

[0810] The speech recognition engine converts the audio data into text data such as "What is the name of the main character in this book?".

[0811] The terminal receives the converted text data and sends it to the server.

[0812] question analysis

[0813] The server analyzes the received text data using a natural language processing algorithm.

[0814] Understand the intent of the question and identify the specific information the user wants to know (in this case, "the protagonist's name").

[0815] Information Retrieval

[0816] Based on the identified intent, the server issues a search query to the audiobook content database.

[0817] The server retrieves relevant information, such as the protagonist's name "John Smith," as a search result.

[0818] Answer generation

[0819] Based on the retrieved search results, the server generates the answer, "The name of the main character in this book is John Smith."

[0820] Speech synthesis

[0821] The server sends the generated text-based response to the speech synthesis engine.

[0822] The speech synthesis engine converts this text into speech data.

[0823] emotion recognition

[0824] The server uses user voice input and text data and analyzes it with an emotion engine. For example, it can recognize whether the user is feeling stressed when asking a question.

[0825] Emotion-based response adjustments

[0826] The server adjusts the content and tone of its responses based on the perceived emotions of the user. For example, if the user is irritated, it will respond in a gentle tone.

[0827] Audio output

[0828] The server sends the adjusted audio data to the terminal.

[0829] The device plays audio data and tells the user, "The name of the main character in this book is John Smith."

[0830] Specific example

[0831] For example, if the user is tired, and the emotion engine recognizes the user's fatigue, it might ask, "What is the name of the main character in this book?" The server will then generate a gentle response, such as, "The main character in this book is named John Smith. Please take a break and enjoy the rest of the book," and play it back through speech synthesis.

[0832] This invention is an advanced adaptive system that can adjust responses according to the user's emotions, thereby improving the user experience.

[0833] The following describes the processing flow.

[0834] Step 1:

[0835] The user speaks a question into the device. For example, "What is the name of the main character in this book?"

[0836] Step 2:

[0837] The device uses its microphone to capture the user's voice and saves it as digital audio data.

[0838] Step 3:

[0839] The device sends the captured audio data to the speech recognition engine.

[0840] Step 4:

[0841] The speech recognition engine analyzes the audio data and converts it into corresponding text data. For example, it might generate text data such as, "What is the name of the main character in this book?"

[0842] Step 5:

[0843] The speech recognition engine sends the converted text data back to the device.

[0844] Step 6:

[0845] The terminal sends text data to the server.

[0846] Step 7:

[0847] The server analyzes the received text data using natural language processing algorithms. Specifically, it grasps the intent of the user's question and extracts key keywords (e.g., "the protagonist's name").

[0848] Step 8:

[0849] Based on the analysis results, the server issues a query to the audiobook content database to retrieve relevant information.

[0850] Step 9:

[0851] The server retrieves relevant information from the database. For example, the data "John Smith" is searched for.

[0852] Step 10:

[0853] Based on the search results obtained by the server, the system generates the answer, "The name of the main character in this book is John Smith."

[0854] Step 11:

[0855] The server sends the generated text-formatted response to the sentiment engine.

[0856] Step 12:

[0857] The server uses an emotion engine to recognize the user's emotions from their voice input. For example, it can determine if the user is stressed or has questions.

[0858] Step 13:

[0859] The server adjusts the response content to match the user's emotions based on the emotions it recognizes. For example, if the user is tired, the response will be made more kind and include elements of encouragement.

[0860] Step 14:

[0861] The server sends text that has been adjusted according to the emotion to the speech synthesis engine.

[0862] Step 15:

[0863] The speech synthesis engine analyzes text data and converts it into speech data.

[0864] Step 16:

[0865] The speech synthesis engine sends the generated audio data back to the server.

[0866] Step 17:

[0867] The server sends the audio data to the terminal.

[0868] Step 18:

[0869] The device plays the received audio data and tells the user, "The main character of this book is named John Smith. Please enjoy the rest of the story."

[0870] In this way, users can ask questions in their own voice, and the system can not only respond appropriately to those questions but also adjust its answers according to the user's emotions. This process allows users to enjoy a more personalized experience.

[0871] (Example 2)

[0872] 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".

[0873] Conventional voice dialogue systems have the problem of not improving the user experience because they provide mechanical responses without considering the user's emotions. Furthermore, the accuracy of answers to user questions is low, making it difficult to provide accurate information quickly. In addition, there was a lack of efficient ways to obtain necessary information when listening to audio content such as audiobooks.

[0874] 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.

[0875] In this invention, the server includes means for acquiring user voice input, means for converting voice input into text data, means for analyzing the intent of the user's question, means for searching for relevant information based on the analysis results, means for generating the retrieved information as text data, means for converting the text data into voice data, means for analyzing the user's emotions, means for adjusting the tone of the response based on the analyzed emotions, and means for providing the adjusted voice data to the user. This enables responses that are in line with the user's emotions, improving the user experience and enabling the rapid provision of accurate information.

[0876] "Means for acquiring voice input" refers to a device or method for capturing the voice spoken by a user and converting it into digital data.

[0877] "Means for converting voice input into text data" refers to technologies and algorithms for analyzing captured voice data and converting it into a corresponding text format.

[0878] "Means for analyzing the intent behind a user's question" refers to natural language processing techniques that analyze the content of text data to understand the information the user wants to know and the purpose of their question.

[0879] "Means for retrieving relevant information based on analysis results" refers to techniques for retrieving relevant information from databases and other sources based on the intent of the analyzed question.

[0880] "Means for generating searched information as text data" refers to technologies for generating information in text format based on search results and providing it to users.

[0881] "Methods for converting text data into audio data" refers to technologies that convert generated text data into audio data using speech synthesis technology.

[0882] "Methods for analyzing user emotions" refer to technologies that recognize and analyze the emotions a user is feeling based on their voice input and text data.

[0883] "Means for adjusting the tone of responses based on analyzed emotions" refers to technologies that appropriately modify the content and tone of generated responses according to the user's emotions.

[0884] "Means for providing adjusted audio data to a user" refers to devices or methods for playing or providing adjusted audio data to a user.

[0885] Modes for carrying out the invention

[0886] This invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to audiobooks, and simultaneously recognize and respond to the user's emotions. This system includes the following hardware and software.

[0887] Main components and devices

[0888] 1. Voice input acquisition method

[0889] The device uses its built-in microphone to acquire voice input from the user.

[0890] An external microphone or a microphone built into the device will be used.

[0891] 2. Speech recognition means

[0892] The device sends the acquired voice data to a speech recognition engine in the cloud.

[0893] The specific software used is the Google Cloud Speech-to-Text API.

[0894] The speech recognition engine converts speech data into text data.

[0895] 3. Question analysis means

[0896] The server analyzes the received text data using a natural language processing algorithm.

[0897] Specifically, the Python libraries NLTK and spaCy will be used.

[0898] The server analyzes the intent behind the user's question and identifies specific information.

[0899] 4. Information Retrieval Methods

[0900] Based on the analysis results, the server issues a search query to the audiobook content database.

[0901] SQL queries are used for the search.

[0902] The server retrieves relevant information as search results.

[0903] 5. Answer generation means

[0904] Based on the information it retrieves, the server generates an appropriate response using a template.

[0905] The response will be generated as text data.

[0906] 6. Speech synthesis means

[0907] The server uses a speech synthesis engine to convert the generated text data into speech data.

[0908] The Google Text-to-Speech API is used.

[0909] 7. Emotion recognition means

[0910] The server analyzes the user's emotions based on their voice input and text data.

[0911] OpenAI's sentiment analysis model will be used.

[0912] 8. Response adjustment mechanisms based on emotions

[0913] The server adjusts the tone of its responses based on the analyzed user's emotions.

[0914] By adding gentle tones and encouraging words, we can respond to the user's emotions.

[0915] 9. Audio output means

[0916] The server sends the adjusted audio data to the terminal.

[0917] The device plays audio data and provides a response to the user.

[0918] Specific example

[0919] For example, if a user asks, "What is the name of the main character in this book?", the device captures this audio and converts it to text using the Google Cloud Speech-to-Text API. The server then analyzes this text using natural language processing algorithms (such as NLTK or spaCy) and searches for relevant information in a database. Based on the search results, the answer "The main character in this book is named John Smith." is generated and converted to speech using the Google Text-to-Speech API. The user's emotions are then analyzed using an OpenAI sentiment analysis model, and the tone is adjusted as needed. Finally, the adjusted audio data is sent to the device and provided to the user.

[0920] Example of a prompt

[0921] "Convert the user's spoken questions into text and generate corresponding answers."

[0922] "Analyze the user's emotions and adjust the tone of your responses accordingly."

[0923] "Please search for the main character's name in the audiobook and tell it to the user."

[0924] "Adjust the content and tone of your responses according to the user's emotions."

[0925] This system aims to significantly improve the user experience by allowing users to quickly and accurately obtain necessary information while enjoying audiobooks, and by providing responses that take the user's emotions into consideration.

[0926] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0927] System program processing steps

[0928] Step 1: Obtaining voice input

[0929] The user asks a question to the device.

[0930] Input: User's voice (e.g., "What is the name of the main character in this book?")

[0931] Operation: The device's built-in microphone captures the user's voice and saves it as audio data (in WAV format, etc.).

[0932] Output: Captured audio data

[0933] Step 2: Speech Recognition

[0934] The device sends the captured audio data to a speech recognition engine in the cloud.

[0935] Input: Audio data

[0936] How it works: The Google Cloud Speech-to-Text API converts speech data into text data.

[0937] Output: Converted text data (e.g., "What is the name of the main character in this book?")

[0938] Step 3: Questionnaire Analysis

[0939] The server analyzes the received text data using a natural language processing algorithm.

[0940] Input: Text data

[0941] Operation: Uses Python's NLTK and spaCy libraries to parse text data and identify the intent behind user questions. For example, it parses the subject, predicate, and object of a question.

[0942] Output: Analysis results (Example: "I want to know the protagonist's name")

[0943] Step 4: Information Retrieval

[0944] Based on the analysis results, the server issues a search query to the audiobook content database.

[0945] Input: Analysis results

[0946] Operation: Generates an SQL query and performs a search on the audiobook database. For example, it searches for information related to "the main character's name".

[0947] Output: Search results (e.g., "John Smith")

[0948] Step 5: Generate Response

[0949] The server generates an answer based on the retrieved search results.

[0950] Input: Search Results

[0951] Operation: Generates an answer using a template. Converts it to a format such as "The name of the main character in this book is John Smith."

[0952] Output: Text-formatted answer (Example: "The name of the main character in this book is John Smith.")

[0953] Step 6: Speech Synthesis

[0954] The server sends the generated text-formatted response to the speech synthesis engine.

[0955] Input: Text data response

[0956] How it works: The Google Text-to-Speech API converts text into speech data.

[0957] Output: Audio data (e.g., audio file)

[0958] Step 7: Emotion Recognition

[0959] The server analyzes the user's emotions based on their voice input and text data.

[0960] Input: Audio data and text data

[0961] Operation: Uses OpenAI's emotion analysis model to analyze the user's emotions. For example, it identifies stress and fatigue from voice tone and text content.

[0962] Output: User's emotional state (e.g., "tired")

[0963] Step 8: Adjusting responses based on emotions

[0964] The server adjusts the tone of its responses based on the analyzed user's emotions.

[0965] Input: Emotional state, text-formatted response

[0966] Action: If the user is tired, add a gentle tone and encouraging words to the response. For example, adjust it to a format like, "The main character of this book is named John Smith. Please take a break and enjoy the rest of the story."

[0967] Output: Answer in adjusted text format

[0968] Step 9: Audio Output

[0969] The server sends the adjusted audio data to the terminal.

[0970] Input: Adjusted audio data

[0971] Operation: The device plays the received audio data.

[0972] Output: The user listens to pre-recorded audio data (e.g., "The main character of this book is named John Smith. Please relax and enjoy the rest of the story.")

[0973] In this way, by clearly defining the specific actions, inputs, and outputs at each step, we have created a system that allows users to quickly and accurately obtain the information they need, and furthermore, can respond to the user's emotions during the process.

[0974] (Application Example 2)

[0975] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0976] Conventional speech recognition systems have made it difficult for users to quickly and accurately obtain necessary information while listening to audiobooks. Furthermore, conventional systems cannot recognize the user's emotions and adjust the content and tone of the responses accordingly, resulting in a limited user experience. The present invention aims to provide a system that not only provides information based on the user's voice input but also analyzes the user's emotions in real time and provides responses with the optimal tone and content.

[0977] 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.

[0978] In this invention, the server includes means for converting voice input into text data, means for analyzing the intent of the user's question, and means for retrieving relevant information. This enables a personalized audiobook experience that responds to the user's emotions.

[0979] "Means for acquiring voice input" refers to devices or mechanisms for capturing the voice spoken by a user and recording it as digitized voice data.

[0980] "Means for converting voice input into text data" refers to engines or software that analyze captured voice data and convert it into text information.

[0981] "Means for analyzing the intent behind user questions" refer to natural language processing algorithms and systems that understand the intent behind user questions obtained as text data and identify relevant information.

[0982] "Means for searching for related information" refers to mechanisms or engines for detecting necessary information from databases and related resources based on the analyzed intent.

[0983] "Means for generating searched information as text data" refers to software or algorithms that organize searched information into text in a format that is easy for users to understand.

[0984] "Means for converting text data into audio data" refers to speech synthesis engines or systems that convert the generated text data into an audio format.

[0985] "Means of providing audio data to the user" refers to devices such as speakers and headsets that play back converted audio data and convey information to the user by voice.

[0986] "Means of recognizing user emotions" refer to emotion recognition engines and software that analyze user voice input, facial expression data, etc., to identify their emotional state.

[0987] "Means of adjusting responses according to emotions" refers to algorithms and control systems that appropriately change the content and tone of the responses provided based on the recognized emotional state of the user.

[0988] This invention provides a system that allows users to quickly and accurately obtain necessary information while listening to audiobooks, and in particular, incorporates an emotion engine that recognizes the user's emotions. The system includes a voice input acquisition means, a voice recognition means, a question analysis means, an information retrieval means, an answer generation means, a voice synthesis means, a voice output means, an emotion recognition means, and an answer adjustment means according to emotion.

[0989] Specific examples of hardware and software

[0990] hardware

[0991] Smart glasses: Equipped with a built-in microphone and speaker to capture the user's voice and provide the results to the user.

[0992] Server: Processes and manages data.

[0993] software

[0994] Speech recognition engine: Uses Google Cloud Speech-to-Text to convert speech input into text.

[0995] Natural language processing algorithm: TensorFlow is used to analyze the user's intent behind the question.

[0996] Speech synthesis engine: Uses Amazon Polly to convert text data into speech data.

[0997] Emotion recognition engine: Uses Affectiva or Microsoft Azure Emotion API to analyze user emotions.

[0998] Content distribution service APIs: Utilizing APIs from services like Audible and Spotify.

[0999] Program processing details

[1000] 1. Acquisition of voice input:

[1001] The smart glasses capture the user's voice commands using a microphone.

[1002] 2. Speech recognition:

[1003] The captured audio data is sent to a cloud-based speech recognition engine (Google Cloud Speech-to-Text) and converted into text data.

[1004] 3. Question analysis:

[1005] The converted text data is sent to a server and analyzed using a natural language processing algorithm (TensorFlow model) to identify the intent of the question.

[1006] 4. Information Retrieval:

[1007] The server searches for relevant information from the audiobook content database based on the intent of the question.

[1008] 5. Answer generation:

[1009] Based on the search results, it generates specific answers such as, "The name of the main character in this book is XX."

[1010] 6. Speech synthesis:

[1011] The generated responses are converted into audio data using Amazon Polly.

[1012] 7. Emotion recognition:

[1013] Affectiva is used to analyze user voice input and text data to identify the user's emotional state.

[1014] 8. Answer adjustment:

[1015] Based on the perceived emotions, the tone and content of the response will be adjusted. For example, if the user is feeling tired, the response will encourage them to relax.

[1016] 9. Audio output:

[1017] The adjusted audio data is sent to the smart glasses, and the information is provided audibly through the speaker.

[1018] Specific example

[1019] If the user asks a question: For example, if the user says, "What is the name of the main character in this book?"

[1020] A concrete example of emotion recognition: When the emotion recognition engine determines that the user is tired.

[1021] Adapted response: The smart glasses' speaker responds in a gentle tone, "The main character of this book is named XX. Please relax and enjoy the rest of the story."

[1022] Example of a prompt

[1023] "We want to develop an application that provides a personalized experience by offering appropriate responses based on voice input instructions, tailored to the user's emotional state. It would quickly search for the information the user is looking for within the audiobook content, provide the answer via voice, and adjust the tone of response based on the user's emotional state. For example, if the user is tired, the response would include a message encouraging relaxation."

[1024] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1025] Step 1:

[1026] The user listens to an audiobook and makes a voice input into the microphone. The device (smart glasses) captures the user's voice through its built-in microphone and records it as digital audio data.

[1027] Input: User voice input (for example, "What is the name of the main character in this book?")

[1028] Output: Digital audio data

[1029] Operation: The microphone records the user's speech as audio data.

[1030] Step 2:

[1031] The device sends the captured digital audio data to a cloud-based speech recognition engine (Google Cloud Speech-to-Text), which converts the audio data into text data.

[1032] Input: Digital audio data

[1033] Output: Text data (for example, "What is the name of the main character in this book?")

[1034] Operation: The device uploads voice data to the cloud engine, and the speech recognition engine converts it into text.

[1035] Step 3:

[1036] The server receives the converted text data and uses a natural language processing algorithm (TensorFlow model) to analyze the user's intent behind the question.

[1037] Input: Text data

[1038] Output: Analyzed question intent (e.g., "I want to know the protagonist's name")

[1039] Operation: The server inputs text data into a natural language processing algorithm to analyze the intent of the question.

[1040] Step 4:

[1041] The server searches for relevant information from the audiobook content database based on the analyzed question intent.

[1042] Input: Analyzed question intent

[1043] Output: Related information (e.g., "Protagonist's name: John Smith")

[1044] Operation: The server issues a query to the database and retrieves the corresponding information.

[1045] Step 5:

[1046] The server generates a text-based response based on the relevant information it has retrieved.

[1047] Input: Related Information

[1048] Output: Text-formatted answer (e.g., "The protagonist's name is John Smith")

[1049] Operation: The server combines relevant information and generates a response in text format to provide to the user.

[1050] Step 6:

[1051] The server sends the generated text-based response to a speech synthesis engine (Amazon Polly) to convert it into audio data.

[1052] Input: Text format answer

[1053] Output: Audio data

[1054] Operation: The server sends text data to the speech synthesis engine and retrieves the data output as speech.

[1055] Step 7:

[1056] The server analyzes voice input and text data using an emotion recognition engine (Affectiva) to recognize the user's emotions.

[1057] Input: Voice input, text data

[1058] Output: User's emotional state (e.g., "tired")

[1059] Operation: The server sends voice input and text data to the emotion recognition engine, which then analyzes the user's emotions.

[1060] Step 8:

[1061] The server adjusts the tone and content of its responses based on the emotions it perceives.

[1062] Input: User's emotional state, text-formatted response.

[1063] Output: Adjusted audio data

[1064] Operation: The server uses the results of emotion recognition to change the response to a gentler tone or content that promotes relaxation.

[1065] Step 9:

[1066] The device transmits the adjusted audio data to the smart glasses' speaker, providing information to the user.

[1067] Input: Adjusted audio data

[1068] Output: Provides an answer via voice (for example, "The main character's name is John Smith. Relax and enjoy the rest.")

[1069] Operation: The device retrieves audio data from the server, plays it through the speaker, and provides it to the user.

[1070] This processing flow allows users to ask questions in real time while listening to audiobooks and receive personalized answers that are tailored to their emotions.

[1071] 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.

[1072] 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 those described above. 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 shown 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.

[1073] 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.

[1074] [Third Embodiment]

[1075] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[1076] 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.

[1077] 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).

[1078] 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.

[1079] 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.

[1080] 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).

[1081] 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.

[1082] 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.

[1083] 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.

[1084] 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.

[1085] 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.

[1086] 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".

[1087] The present invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to an audiobook. The system consists of the following main components: voice input acquisition means, voice recognition means, question analysis means, information retrieval means, answer generation means, voice synthesis means, and voice output means.

[1088] Overall processing flow

[1089] 1. Acquisition of voice input

[1090] The user can ask questions about what they want to know using voice commands.

[1091] The device captures the user's voice through the microphone.

[1092] 2. Speech Recognition

[1093] The device converts the captured audio into text data and sends it to the speech recognition engine.

[1094] 3. Question analysis

[1095] The server receives text data sent from the speech recognition engine and uses natural language processing to analyze the intent of the question.

[1096] 4. Information Retrieval

[1097] Based on the intent of the analyzed question, the server searches for relevant information within the audiobook content.

[1098] 5. Answer generation

[1099] The server generates an answer based on the search results.

[1100] 6. Speech synthesis

[1101] The server uses a speech synthesis engine to convert the generated responses from text to speech.

[1102] 7. Audio Output

[1103] The server sends the synthesized voice data to the terminal.

[1104] The device plays back the received audio data and provides the user with an answer.

[1105] Detailed explanations for each step are provided below.

[1106] Acquiring voice input

[1107] The user says to the device, "What is the name of the main character in this book?"

[1108] The device captures this audio through its built-in microphone and saves it as audio data.

[1109] Speech recognition

[1110] The device sends the captured audio data to a speech recognition engine in the cloud.

[1111] The speech recognition engine converts the audio data into text data such as "What is the name of the main character in this book?".

[1112] The terminal receives the converted text data and sends it to the server.

[1113] question analysis

[1114] The server analyzes the received text data using a natural language processing algorithm.

[1115] Understand the intent of the question and identify the specific information the user wants to know (in this case, "the protagonist's name").

[1116] Information Retrieval

[1117] Based on the identified intent, the server issues a search query to the audiobook content database.

[1118] The server retrieves relevant information, such as the protagonist's name "John Smith," as a search result.

[1119] Answer generation

[1120] Based on the retrieved search results, the server generates the answer, "The name of the main character in this book is John Smith."

[1121] Speech synthesis

[1122] The server sends the generated text-based response to the speech synthesis engine.

[1123] The speech synthesis engine converts this text into speech data.

[1124] Audio output

[1125] The server sends the audio data to the terminal.

[1126] The device plays audio data and tells the user, "The name of the main character in this book is John Smith."

[1127] Specific example

[1128] For example, if a user asks, "What is the most important point in this chapter?", the same process would apply.

[1129] The server generates the response, "The key point of this chapter is the scene where the protagonist makes a firm decision," and plays it back to the user as audio.

[1130] This invention allows users to instantly obtain the information they want while listening to audiobooks, deepening their understanding of the content. In addition, it reduces the wasted time caused by conventional methods of re-searching and re-listening, enabling more efficient access to information.

[1131] The following describes the processing flow.

[1132] Step 1:

[1133] The user speaks a question into the device. Example: "What is the name of the main character in this book?"

[1134] Step 2:

[1135] The device uses its microphone to capture the user's voice and saves it as digital audio data.

[1136] Step 3:

[1137] The device sends the captured audio data to the speech recognition engine.

[1138] Step 4:

[1139] The speech recognition engine analyzes the audio data and converts it into corresponding text data. For example, it might generate text data such as, "What is the name of the main character in this book?"

[1140] Step 5:

[1141] The speech recognition engine sends the converted text data back to the device.

[1142] Step 6:

[1143] The terminal sends text data to the server.

[1144] Step 7:

[1145] The server analyzes the received text data using natural language processing algorithms. Specifically, it grasps the intent of the user's question and extracts important keywords (e.g., "the protagonist's name").

[1146] Step 8:

[1147] Based on the analysis results, the server issues a query to the audiobook content database to retrieve relevant information.

[1148] Step 9:

[1149] The server retrieves relevant information from the database. For example, the data "John Smith" is searched for.

[1150] Step 10:

[1151] The server generates the answer, "The name of the main character in this book is John Smith," based on the search results.

[1152] Step 11:

[1153] The server sends the generated text-based response to the speech synthesis engine.

[1154] Step 12:

[1155] The speech synthesis engine analyzes text data and converts it into speech data.

[1156] Step 13:

[1157] The speech synthesis engine sends the generated audio data back to the server.

[1158] Step 14:

[1159] The server sends the audio data to the terminal.

[1160] Step 15:

[1161] The device plays the received audio data and tells the user, "The name of the main character in this book is John Smith."

[1162] (Example 1)

[1163] 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."

[1164] Conventional audio content listening systems were inefficient because users had to manually re-search or re-listen to obtain the information they needed. Furthermore, systems that converted voice input to text and provided appropriate answers to user questions lacked sufficient technology to accurately analyze the intent of the questions and provide timely, appropriate responses. As a result, it was difficult for users to quickly and accurately obtain the information they needed while listening to audio content.

[1165] 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.

[1166] In this invention, the server includes means for enabling the user to ask questions in real time while listening to audio content, means for generating answers to the questions using a natural language processing algorithm and a generative AI model, and means for converting the generated answers into audio data using a speech synthesizer. This enables the user to quickly and accurately obtain the necessary information without interrupting the audio content.

[1167] A "voice input device" is a device that captures a user's voice and converts it into digital voice data.

[1168] A "speech recognition engine" is a program or service that converts speech data into text data.

[1169] A "natural language processing algorithm" is a computational method for analyzing text data and understanding its meaning, context, and intent.

[1170] A "generative AI model" is an artificial intelligence model designed to generate natural language responses based on text data.

[1171] A "speech synthesis device" is a device or service used to convert text data into speech data.

[1172] "Audio content" refers to information and entertainment provided in audio format.

[1173] A "prompt sentence" is an input sentence given to a generative AI model, and it is an instruction sentence used to generate a specific answer.

[1174] "Real-time" refers to the fact that user requests and operations are processed instantly with virtually no delay.

[1175] A "database" is a system designed to organize and manage large amounts of data, enabling efficient searching and updating.

[1176] A "user" is a person who attempts to obtain information by operating a system.

[1177] A "terminal" is an electronic device used by a user to access a system.

[1178] A "server" is a computer that acts as the central hub of a system, processing, storing, and providing data.

[1179] This invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to audio content. This system acquires the user's voice input and performs a series of processes to provide an appropriate response using natural language processing and a generative AI model, and its specific implementation method is described below.

[1180] System Overview

[1181] This system consists of the following main parts:

[1182] 1. Voice input device

[1183] 2. Speech Recognition Engine

[1184] 3. Natural Language Processing Algorithms

[1185] 4. Generative AI Models

[1186] 5. Speech synthesis device

[1187] 6. Server

[1188] 7. Terminal

[1189] Required hardware and software

[1190] Voice input device: Used to capture the user's voice. Example: Microphone.

[1191] Speech recognition engine: Used to convert speech data into text data. Examples: Google Cloud Speech-to-Text API, IBM Watson Speech to Text, etc.

[1192] Natural Language Processing (NLP) algorithms: Used to analyze the intent behind a user's question. Examples: Natural Language Processing libraries such as NLTK and spaCy.

[1193] Generative AI models: Used to generate appropriate answers to questions. Example: OpenAI's GPT-3.

[1194] Text-to-speech (Speech synthesis) devices are used to convert text data into speech data. Examples include Amazon Polly and Google Cloud Text-to-Speech.

[1195] Server: Manages the processing of the entire system.

[1196] Terminal: A device used by a user to access a system, input questions, and receive answers. Examples: smartphones, tablets, personal computers, etc.

[1197] Specific example

[1198] When a user is listening to audio content, for example, they might be asked, "What is the name of the main character in this book?" In this case, the system operates as follows:

[1199] 1. Acquisition of voice input

[1200] The user says to the device, "What is the name of the main character in this book?"

[1201] The device captures this audio through its built-in microphone and saves it as audio data.

[1202] 2. Speech Recognition

[1203] The device sends the captured audio data to a speech recognition engine in the cloud.

[1204] The speech recognition engine converts the audio data into text data such as "What is the name of the main character in this book?".

[1205] The terminal receives the converted text data and sends it to the server.

[1206] 3. Question analysis

[1207] The server analyzes the received text data using a natural language processing algorithm.

[1208] Understand the intent of the question and identify the specific information the user wants to know (in this case, "the protagonist's name").

[1209] 4. Information Retrieval

[1210] Based on the intent of the analyzed question, the server searches for relevant information from the audio content.

[1211] The server retrieves relevant information as search results. For example, the protagonist's name is "John Smith".

[1212] 5. Answer generation

[1213] Based on the retrieved search results, the server generates the answer, "The name of the main character in this book is John Smith."

[1214] 6. Speech synthesis

[1215] The server sends the generated text-based response to the speech synthesis engine.

[1216] The speech synthesis engine converts this text into speech data.

[1217] 7. Audio Output

[1218] The server sends the audio data to the terminal.

[1219] The device plays audio data and tells the user, "The name of the main character in this book is John Smith."

[1220] Example of a prompt

[1221] The following is an example of a prompt message used to generate an appropriate answer to a user's question using a generative AI model.

[1222] Prompt example (English):

[1223] "User asks in English: 'What is the name of the main character in this book?'. The audiobooks database has the following relevant information: 'The main character of the book is John Smith.'. Respond to the user's question with this information."

[1224] Prompt example (Japanese):

[1225] The user asks, "What is the name of the main character in this book?" The audio content database contains the related information, "The main character in this book is John Smith." Use this information to answer the user's question.

[1226] This invention allows users to instantly obtain the information they want while listening to audio content, leading to a deeper understanding of the content. This reduces the wasted time caused by traditional methods of re-searching or re-listening, enabling more efficient access to information.

[1227] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1228] Step 1:

[1229] The user asks a question aloud, "What is the name of the main character in this book?" The input is the user's voice, which is captured by the device's built-in microphone. Specifically, the device runs a voice capture program and saves it as audio data.

[1230] Step 2:

[1231] The device sends the captured audio data to a speech recognition engine in the cloud. The input is audio data, and the output is text data. Specifically, the device uses a cloud communication module to send audio data, which the speech recognition engine converts into text, such as "What is the name of the main character in this book?"

[1232] Step 3:

[1233] The device receives converted text data from the cloud and sends it to the server. The input is text data from the speech recognition engine, and the output is the transmission of text data to the server. Specifically, the device uses a cloud communication module to transfer the text data to the server.

[1234] Step 4:

[1235] The server analyzes the received text data using a natural language processing algorithm. The input is text data, and the output is the result of analyzing the intent of the user's question. Specifically, the server uses natural language processing libraries such as NLTK or spaCy to identify the intent of the question as "I want to know the name of the main character."

[1236] Step 5:

[1237] The server searches for relevant information based on the analysis results. The input is the analysis result (the intent of the question), and the output is the relevant information (search results). Specifically, the server performs a search on the database using SQL queries or Elasticsearch, and as a result retrieves the name "John Smith".

[1238] Step 6:

[1239] The server generates an answer based on the search results. The input is the information from the search results, and the output is the text data of the answer. Specifically, the server uses a generative AI model (for example, GPT-3) to generate the text "The name of the main character in this book is John Smith."

[1240] Step 7:

[1241] The server sends the generated text-based response to a text-to-speech engine, where it is converted into audio data. The input is the text data of the response, and the output is audio data. Specifically, the server uses Amazon Polly or Google Text-to-Speech to generate the audio data.

[1242] Step 8:

[1243] The server sends the generated audio data to the terminal. The input is audio data, and the output is the transmission of audio data to the terminal. Specifically, the server transfers the audio data to the terminal using a cloud communication module.

[1244] Step 9:

[1245] The terminal plays the received audio data and provides it to the user. The input is audio data, and the output is audio playback. Specifically, the terminal uses an audio playback program to play the sentence, "The name of the main character in this book is John Smith."

[1246] (Application Example 1)

[1247] 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."

[1248] Currently, in physical stores, customers need to either ask store staff directly or search for information themselves to get instant product information, which can be time-consuming. Furthermore, some stores may not have sufficient staff, making it difficult for customers to obtain the information they need immediately. Therefore, there is a need for an efficient system that allows customers to get immediate answers to their product inquiries within physical stores.

[1249] 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.

[1250] In this invention, the server includes means for acquiring user voice input, means for converting voice input into text data, means for analyzing the intent of the user's question, means for searching for relevant information based on the analysis results, means for generating the retrieved information as text data, means for converting the text data into voice data, means for providing the voice data to the user, and software that operates on a smart device, including applications for in-store applications where customers ask questions about product features and specifications. As a result, even when a customer asks a question about a product in a physical store using voice input, relevant information is provided immediately, making it possible to obtain the necessary information quickly and accurately.

[1251] definition statement

[1252] "Means for acquiring user voice input" refers to devices or systems that capture the voice spoken by a user and save it as digital data.

[1253] "Means of converting voice input into text data" refers to the process or system of converting acquired voice data into text data.

[1254] "Means for analyzing the intent behind user questions" refers to methods and algorithms for analyzing questions that have been converted into text data and understanding their content and intent.

[1255] "Means for searching for relevant information based on analysis results" refers to a function that retrieves relevant data from appropriate sources in accordance with the analyzed intent.

[1256] "Means for generating retrieved information as text data" refers to a mechanism that processes and generates acquired information as text data in an appropriate format.

[1257] "Means of converting text data into audio data" refers to technologies and tools that convert generated text into synthesized speech.

[1258] "Means of providing audio data to users" refers to devices or systems that play synthesized audio and allow users to listen to it.

[1259] "Application to in-store applications where customers ask questions about product features and specifications" refers to an application technology that allows customers in physical stores to ask questions about products using voice and receive immediate answers to those questions.

[1260] "Software that runs on smart devices" refers to a series of programs and applications that run on electronic devices such as smartphones and smart glasses.

[1261] Modes for carrying out the invention

[1262] System Program Overview

[1263] The system of this invention acquires voice input from a user, analyzes it, and immediately provides information about products. The entire system includes a series of processes in which the user inputs voice via a smart device (such as a smartphone or smart glasses), the voice is converted into text, analyzed, and an appropriate response is generated, synthesized into speech, and then provided back to the user.

[1264] Hardware and software usage

[1265] Hardware:

[1266] 1. Smart devices: Examples include smartphones and smart glasses. These devices have voice input and voice output capabilities.

[1267] 2. Microphone: The device's built-in microphone captures voice input.

[1268] software:

[1269] 1. Speech recognition engine: Use the speech_recognition library to convert speech data into text data.

[1270] 2. Natural Language Processing Algorithm: The intent of the question will be analyzed using OpenAI's GPT-3 model, and an appropriate answer will be generated. Specifically, the OpenAI API will be used.

[1271] 3. Text-to-Speech Engine: The GTTS (Google Text-to-Speech) library and the playsound library are used to convert text-based responses into audio data and then play it back.

[1272] Data processing and data calculation

[1273] 1. Acquisition of voice input:

[1274] The user voice-inputs questions about the product into the device.

[1275] The microphone captures the sound and saves it as digital data.

[1276] 2. Speech recognition:

[1277] The acquired audio data is converted into text data using the speech_recognition library.

[1278] 3. Question analysis and answer generation:

[1279] Text data is sent as a prompt to the OpenAI API, analyzed using a GPT-3 model, and an appropriate response is generated.

[1280] Examples of prompt statements to use:

[1281] A customer asked: 'What are the features of this shampoo?' Please provide information about the product.

[1282] 4. Speech synthesis and output:

[1283] The generated text data is converted into audio data using the gtts library.

[1284] The playsound library is used to play the generated audio data and provide it to the user.

[1285] Specific example

[1286] For example, if a user asks "What are the features of this shampoo?" while in a physical store, the system will operate as follows:

[1287] 1. The microphone captures the audio, and the speech recognition engine generates text data that says, "What are the features of this shampoo?"

[1288] 2. This text data is sent to the OpenAI API, and the GPT-3 model generates the response, "This shampoo is rich in natural ingredients and makes hair soft."

[1289] 3. The generated text response is converted into audio data by a speech synthesis engine and provided to the user as audio using the playsound library.

[1290] This system allows customers to quickly obtain product information within physical stores, reducing the burden on store staff. Furthermore, it allows customers to deepen their understanding of the products.

[1291] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1292] Program processing steps

[1293] Step 1:

[1294] The user voice-inputs a question about the product into a smart device. For example, the user might say, "What are the features of this shampoo?" The input is voice data, which is then captured as digital data using a microphone.

[1295] Step 2:

[1296] The device converts the captured audio data into text data using the speech_recognition library. In this step, the audio data is the input, and the converted text data is the output. Specifically, the speech recognition engine converts it into the text "What are the features of this shampoo?".

[1297] Step 3:

[1298] The terminal sends the converted text data to the server. The server then sends it to the OpenAI API in the form of a prompt sentence to analyze the user's intent. The input is text data, which is converted into data in the form of a prompt sentence. For example, the prompt sentence "The customer asked: 'What are the features of this shampoo?' Please provide information about the product." is generated.

[1299] Step 4:

[1300] The server sends the generated prompt to the OpenAI GPT-3 model, which analyzes the intent of the question and generates an appropriate answer. The input is the prompt, and the output is the text data of the answer. Specifically, GPT-3 generates the answer "This shampoo is rich in natural ingredients and makes hair soft."

[1301] Step 5:

[1302] The server returns the generated response as text data to the terminal. The input is the generated text data, and the output and transmitted data are also text data.

[1303] Step 6:

[1304] The device converts received text data into audio data using GTTS. The input is text data, and the output is audio data. Specifically, the sentence "This shampoo is rich in natural ingredients and makes hair soft" is converted into audio data.

[1305] Step 7:

[1306] The device plays the converted audio data using the playsound library and provides it to the user. The input is audio data, and the output is the played audio. The user can receive a response via voice from their smart device.

[1307] 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.

[1308] The present invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to audiobooks, and more particularly to a system that incorporates an emotion engine to recognize the user's emotions. The system consists of the following main parts: voice input acquisition means, voice recognition means, question analysis means, information retrieval means, answer generation means, voice synthesis means, voice output means, and emotion recognition means.

[1309] Overall processing flow

[1310] 1. Acquisition of voice input

[1311] The user can ask questions about what they want to know using voice commands.

[1312] The device captures the user's voice through the microphone.

[1313] 2. Speech Recognition

[1314] The device converts the captured audio into text data and sends it to the speech recognition engine.

[1315] 3. Question analysis

[1316] The server receives text data sent from the speech recognition engine and uses natural language processing to analyze the intent of the question.

[1317] 4. Information Retrieval

[1318] Based on the intent of the analyzed question, the server searches for relevant information within the audiobook content.

[1319] 5. Answer generation

[1320] The server generates an answer based on the search results.

[1321] 6. Speech synthesis

[1322] The server uses a speech synthesis engine to convert the generated responses from text to speech.

[1323] 7. Emotion recognition

[1324] The server uses voice input and text data to analyze the user's emotions using an emotion engine.

[1325] 8. Adjusting responses based on emotions.

[1326] The server adjusts the content and tone of its responses based on the recognized emotions of the user.

[1327] 9. Audio Output

[1328] The server sends the synthesized voice data to the terminal.

[1329] The device plays back the received audio data and provides the user with an answer.

[1330] Detailed explanations for each step are provided below.

[1331] Acquiring voice input

[1332] For example, the user might say to the device, "What is the name of the main character in this book?"

[1333] The device captures audio through its built-in microphone and saves it as audio data.

[1334] Speech recognition

[1335] The device sends the captured audio data to a speech recognition engine in the cloud.

[1336] The speech recognition engine converts the audio data into text data such as "What is the name of the main character in this book?".

[1337] The terminal receives the converted text data and sends it to the server.

[1338] question analysis

[1339] The server analyzes the received text data using a natural language processing algorithm.

[1340] Understand the intent of the question and identify the specific information the user wants to know (in this case, "the protagonist's name").

[1341] Information Retrieval

[1342] Based on the identified intent, the server issues a search query to the audiobook content database.

[1343] The server retrieves relevant information, such as the protagonist's name "John Smith," as a search result.

[1344] Answer generation

[1345] Based on the retrieved search results, the server generates the answer, "The name of the main character in this book is John Smith."

[1346] Speech synthesis

[1347] The server sends the generated text-based response to the speech synthesis engine.

[1348] The speech synthesis engine converts this text into speech data.

[1349] emotion recognition

[1350] The server uses user voice input and text data and analyzes it with an emotion engine. For example, it can recognize whether the user is feeling stressed when asking a question.

[1351] Emotion-based response adjustments

[1352] The server adjusts the content and tone of its responses based on the perceived emotions of the user. For example, if the user is irritated, it will respond in a gentle tone.

[1353] Audio output

[1354] The server sends the adjusted audio data to the terminal.

[1355] The device plays audio data and tells the user, "The name of the main character in this book is John Smith."

[1356] Specific example

[1357] For example, if the user is tired, and the emotion engine recognizes the user's fatigue, it might ask, "What is the name of the main character in this book?" The server will then generate a gentle response, such as, "The main character in this book is named John Smith. Please take a break and enjoy the rest of the book," and play it back through speech synthesis.

[1358] This invention is an advanced adaptive system that can adjust responses according to the user's emotions, thereby improving the user experience.

[1359] The following describes the processing flow.

[1360] Step 1:

[1361] The user speaks a question into the device. For example, "What is the name of the main character in this book?"

[1362] Step 2:

[1363] The device uses its microphone to capture the user's voice and saves it as digital audio data.

[1364] Step 3:

[1365] The device sends the captured audio data to the speech recognition engine.

[1366] Step 4:

[1367] The speech recognition engine analyzes the audio data and converts it into corresponding text data. For example, it might generate text data such as, "What is the name of the main character in this book?"

[1368] Step 5:

[1369] The speech recognition engine sends the converted text data back to the device.

[1370] Step 6:

[1371] The terminal sends text data to the server.

[1372] Step 7:

[1373] The server analyzes the received text data using natural language processing algorithms. Specifically, it grasps the intent of the user's question and extracts key keywords (e.g., "the protagonist's name").

[1374] Step 8:

[1375] Based on the analysis results, the server issues a query to the audiobook content database to retrieve relevant information.

[1376] Step 9:

[1377] The server retrieves relevant information from the database. For example, the data "John Smith" is searched for.

[1378] Step 10:

[1379] Based on the search results obtained by the server, the system generates the answer, "The name of the main character in this book is John Smith."

[1380] Step 11:

[1381] The server sends the generated text-formatted response to the sentiment engine.

[1382] Step 12:

[1383] The server uses an emotion engine to recognize the user's emotions from their voice input. For example, it can determine if the user is stressed or has questions.

[1384] Step 13:

[1385] The server adjusts the response content to match the user's emotions based on the emotions it recognizes. For example, if the user is tired, the response will be made more kind and include elements of encouragement.

[1386] Step 14:

[1387] The server sends text that has been adjusted according to the emotion to the speech synthesis engine.

[1388] Step 15:

[1389] The speech synthesis engine analyzes text data and converts it into speech data.

[1390] Step 16:

[1391] The speech synthesis engine sends the generated audio data back to the server.

[1392] Step 17:

[1393] The server sends the audio data to the terminal.

[1394] Step 18:

[1395] The device plays the received audio data and tells the user, "The main character of this book is named John Smith. Please enjoy the rest of the story."

[1396] In this way, users can ask questions in their own voice, and the system can not only respond appropriately to those questions but also adjust its answers according to the user's emotions. This process allows users to enjoy a more personalized experience.

[1397] (Example 2)

[1398] 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."

[1399] Conventional voice dialogue systems have the problem of not improving the user experience because they provide mechanical responses without considering the user's emotions. Furthermore, the accuracy of answers to user questions is low, making it difficult to provide accurate information quickly. In addition, there was a lack of efficient ways to obtain necessary information when listening to audio content such as audiobooks.

[1400] 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.

[1401] In this invention, the server includes means for acquiring user voice input, means for converting voice input into text data, means for analyzing the intent of the user's question, means for searching for relevant information based on the analysis results, means for generating the retrieved information as text data, means for converting the text data into voice data, means for analyzing the user's emotions, means for adjusting the tone of the response based on the analyzed emotions, and means for providing the adjusted voice data to the user. This enables responses that are in line with the user's emotions, improving the user experience and enabling the rapid provision of accurate information.

[1402] "Means for acquiring voice input" refers to a device or method for capturing the voice spoken by a user and converting it into digital data.

[1403] "Means for converting voice input into text data" refers to technologies and algorithms for analyzing captured voice data and converting it into a corresponding text format.

[1404] "Means for analyzing the intent behind a user's question" refers to natural language processing techniques that analyze the content of text data to understand the information the user wants to know and the purpose of their question.

[1405] "Means for retrieving relevant information based on analysis results" refers to techniques for retrieving relevant information from databases and other sources based on the intent of the analyzed question.

[1406] "Means for generating searched information as text data" refers to technologies for generating information in text format based on search results and providing it to users.

[1407] "Methods for converting text data into audio data" refers to technologies that convert generated text data into audio data using speech synthesis technology.

[1408] "Methods for analyzing user emotions" refer to technologies that recognize and analyze the emotions a user is feeling based on their voice input and text data.

[1409] "Means for adjusting the tone of responses based on analyzed emotions" refers to technologies that appropriately modify the content and tone of generated responses according to the user's emotions.

[1410] "Means for providing adjusted audio data to a user" refers to devices or methods for playing or providing adjusted audio data to a user.

[1411] Modes for carrying out the invention

[1412] This invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to audiobooks, and simultaneously recognize and respond to the user's emotions. This system includes the following hardware and software.

[1413] Main components and devices

[1414] 1. Voice input acquisition method

[1415] The device uses its built-in microphone to acquire voice input from the user.

[1416] An external microphone or a microphone built into the device will be used.

[1417] 2. Speech recognition means

[1418] The device sends the acquired voice data to a speech recognition engine in the cloud.

[1419] The specific software used is the Google Cloud Speech-to-Text API.

[1420] The speech recognition engine converts speech data into text data.

[1421] 3. Question analysis means

[1422] The server analyzes the received text data using a natural language processing algorithm.

[1423] Specifically, the Python libraries NLTK and spaCy will be used.

[1424] The server analyzes the intent behind the user's question and identifies specific information.

[1425] 4. Information Retrieval Methods

[1426] Based on the analysis results, the server issues a search query to the audiobook content database.

[1427] SQL queries are used for the search.

[1428] The server retrieves relevant information as search results.

[1429] 5. Answer generation means

[1430] Based on the information it retrieves, the server generates an appropriate response using a template.

[1431] The response will be generated as text data.

[1432] 6. Speech synthesis means

[1433] The server uses a speech synthesis engine to convert the generated text data into speech data.

[1434] The Google Text-to-Speech API is used.

[1435] 7. Emotion recognition means

[1436] The server analyzes the user's emotions based on their voice input and text data.

[1437] OpenAI's sentiment analysis model will be used.

[1438] 8. Response adjustment mechanisms based on emotions

[1439] The server adjusts the tone of its responses based on the analyzed user's emotions.

[1440] By adding gentle tones and encouraging words, we can respond to the user's emotions.

[1441] 9. Audio output means

[1442] The server sends the adjusted audio data to the terminal.

[1443] The device plays audio data and provides a response to the user.

[1444] Specific example

[1445] For example, if a user asks, "What is the name of the main character in this book?", the device captures this audio and converts it to text using the Google Cloud Speech-to-Text API. The server then analyzes this text using natural language processing algorithms (such as NLTK or spaCy) and searches for relevant information in a database. Based on the search results, the answer "The main character in this book is named John Smith." is generated and converted to speech using the Google Text-to-Speech API. The user's emotions are then analyzed using an OpenAI sentiment analysis model, and the tone is adjusted as needed. Finally, the adjusted audio data is sent to the device and provided to the user.

[1446] Example of a prompt

[1447] "Convert the user's spoken questions into text and generate corresponding answers."

[1448] "Analyze the user's emotions and adjust the tone of your responses accordingly."

[1449] "Please search for the main character's name in the audiobook and tell it to the user."

[1450] "Adjust the content and tone of your responses according to the user's emotions."

[1451] This system aims to significantly improve the user experience by allowing users to quickly and accurately obtain necessary information while enjoying audiobooks, and by providing responses that take the user's emotions into consideration.

[1452] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1453] System program processing steps

[1454] Step 1: Obtaining voice input

[1455] The user asks a question to the device.

[1456] Input: User's voice (e.g., "What is the name of the main character in this book?")

[1457] Operation: The device's built-in microphone captures the user's voice and saves it as audio data (in WAV format, etc.).

[1458] Output: Captured audio data

[1459] Step 2: Speech Recognition

[1460] The device sends the captured audio data to a speech recognition engine in the cloud.

[1461] Input: Audio data

[1462] How it works: The Google Cloud Speech-to-Text API converts speech data into text data.

[1463] Output: Converted text data (e.g., "What is the name of the main character in this book?")

[1464] Step 3: Questionnaire Analysis

[1465] The server analyzes the received text data using a natural language processing algorithm.

[1466] Input: Text data

[1467] Operation: Uses Python's NLTK and spaCy libraries to parse text data and identify the intent behind user questions. For example, it parses the subject, predicate, and object of a question.

[1468] Output: Analysis results (Example: "I want to know the protagonist's name")

[1469] Step 4: Information Retrieval

[1470] Based on the analysis results, the server issues a search query to the audiobook content database.

[1471] Input: Analysis results

[1472] Operation: Generates an SQL query and performs a search on the audiobook database. For example, it searches for information related to "the main character's name".

[1473] Output: Search results (e.g., "John Smith")

[1474] Step 5: Generate Response

[1475] The server generates an answer based on the retrieved search results.

[1476] Input: Search Results

[1477] Operation: Generates an answer using a template. Converts it to a format such as "The name of the main character in this book is John Smith."

[1478] Output: Text-formatted answer (Example: "The name of the main character in this book is John Smith.")

[1479] Step 6: Speech Synthesis

[1480] The server sends the generated text-formatted response to the speech synthesis engine.

[1481] Input: Text data response

[1482] How it works: The Google Text-to-Speech API converts text into speech data.

[1483] Output: Audio data (e.g., audio file)

[1484] Step 7: Emotion Recognition

[1485] The server analyzes the user's emotions based on their voice input and text data.

[1486] Input: Audio data and text data

[1487] Operation: Uses OpenAI's emotion analysis model to analyze the user's emotions. For example, it identifies stress and fatigue from voice tone and text content.

[1488] Output: User's emotional state (e.g., "tired")

[1489] Step 8: Adjusting responses based on emotions

[1490] The server adjusts the tone of its responses based on the analyzed user's emotions.

[1491] Input: Emotional state, text-formatted response

[1492] Action: If the user is tired, add a gentle tone and encouraging words to the response. For example, adjust it to a format like, "The main character of this book is named John Smith. Please take a break and enjoy the rest of the story."

[1493] Output: Answer in adjusted text format

[1494] Step 9: Audio Output

[1495] The server sends the adjusted audio data to the terminal.

[1496] Input: Adjusted audio data

[1497] Operation: The device plays the received audio data.

[1498] Output: The user listens to pre-recorded audio data (e.g., "The main character of this book is named John Smith. Please relax and enjoy the rest of the story.")

[1499] In this way, by clearly defining the specific actions, inputs, and outputs at each step, we have created a system that allows users to quickly and accurately obtain the information they need, and furthermore, can respond to the user's emotions during the process.

[1500] (Application Example 2)

[1501] 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."

[1502] Conventional speech recognition systems have made it difficult for users to quickly and accurately obtain necessary information while listening to audiobooks. Furthermore, conventional systems cannot recognize the user's emotions and adjust the content and tone of the responses accordingly, resulting in a limited user experience. The present invention aims to provide a system that not only provides information based on the user's voice input but also analyzes the user's emotions in real time and provides responses with the optimal tone and content.

[1503] 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.

[1504] In this invention, the server includes means for converting voice input into text data, means for analyzing the intent of the user's question, and means for retrieving relevant information. This enables a personalized audiobook experience that responds to the user's emotions.

[1505] "Means for acquiring voice input" refers to devices or mechanisms for capturing the voice spoken by a user and recording it as digitized voice data.

[1506] "Means for converting voice input into text data" refers to engines or software that analyze captured voice data and convert it into text information.

[1507] "Means for analyzing the intent behind user questions" refer to natural language processing algorithms and systems that understand the intent behind user questions obtained as text data and identify relevant information.

[1508] "Means for searching for related information" refers to mechanisms or engines for detecting necessary information from databases and related resources based on the analyzed intent.

[1509] "Means for generating searched information as text data" refers to software or algorithms that organize searched information into text in a format that is easy for users to understand.

[1510] "Means for converting text data into audio data" refers to speech synthesis engines or systems that convert the generated text data into an audio format.

[1511] "Means of providing audio data to the user" refers to devices such as speakers and headsets that play back converted audio data and convey information to the user by voice.

[1512] "Means of recognizing user emotions" refer to emotion recognition engines and software that analyze user voice input, facial expression data, etc., to identify their emotional state.

[1513] "Means of adjusting responses according to emotions" refers to algorithms and control systems that appropriately change the content and tone of the responses provided based on the recognized emotional state of the user.

[1514] This invention provides a system that allows users to quickly and accurately obtain necessary information while listening to audiobooks, and in particular, incorporates an emotion engine that recognizes the user's emotions. The system includes a voice input acquisition means, a voice recognition means, a question analysis means, an information retrieval means, an answer generation means, a voice synthesis means, a voice output means, an emotion recognition means, and an answer adjustment means according to emotion.

[1515] Specific examples of hardware and software

[1516] hardware

[1517] Smart glasses: Equipped with a built-in microphone and speaker to capture the user's voice and provide the results to the user.

[1518] Server: Processes and manages data.

[1519] software

[1520] Speech recognition engine: Uses Google Cloud Speech-to-Text to convert speech input into text.

[1521] Natural language processing algorithm: TensorFlow is used to analyze the user's intent behind the question.

[1522] Speech synthesis engine: Uses Amazon Polly to convert text data into speech data.

[1523] Emotion recognition engine: Uses Affectiva or Microsoft Azure Emotion API to analyze user emotions.

[1524] Content distribution service APIs: Utilizing APIs from services like Audible and Spotify.

[1525] Program processing details

[1526] 1. Acquisition of voice input:

[1527] The smart glasses capture the user's voice commands using a microphone.

[1528] 2. Speech recognition:

[1529] The captured audio data is sent to a cloud-based speech recognition engine (Google Cloud Speech-to-Text) and converted into text data.

[1530] 3. Question analysis:

[1531] The converted text data is sent to a server and analyzed using a natural language processing algorithm (TensorFlow model) to identify the intent of the question.

[1532] 4. Information Retrieval:

[1533] The server searches for relevant information from the audiobook content database based on the intent of the question.

[1534] 5. Answer generation:

[1535] Based on the search results, it generates specific answers such as, "The name of the main character in this book is XX."

[1536] 6. Speech synthesis:

[1537] The generated responses are converted into audio data using Amazon Polly.

[1538] 7. Emotion recognition:

[1539] Affectiva is used to analyze user voice input and text data to identify the user's emotional state.

[1540] 8. Answer adjustment:

[1541] Based on the perceived emotions, the tone and content of the response will be adjusted. For example, if the user is feeling tired, the response will encourage them to relax.

[1542] 9. Audio output:

[1543] The adjusted audio data is sent to the smart glasses, and the information is provided audibly through the speaker.

[1544] Specific example

[1545] If the user asks a question: For example, if the user says, "What is the name of the main character in this book?"

[1546] A concrete example of emotion recognition: When the emotion recognition engine determines that the user is tired.

[1547] Adapted response: The smart glasses' speaker responds in a gentle tone, "The main character of this book is named XX. Please relax and enjoy the rest of the story."

[1548] Example of a prompt

[1549] "We want to develop an application that provides a personalized experience by offering appropriate responses based on voice input instructions, tailored to the user's emotional state. It would quickly search for the information the user is looking for within the audiobook content, provide the answer via voice, and adjust the tone of response based on the user's emotional state. For example, if the user is tired, the response would include a message encouraging relaxation."

[1550] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1551] Step 1:

[1552] The user listens to an audiobook and makes a voice input into the microphone. The device (smart glasses) captures the user's voice through its built-in microphone and records it as digital audio data.

[1553] Input: User voice input (for example, "What is the name of the main character in this book?")

[1554] Output: Digital audio data

[1555] Operation: The microphone records the user's speech as audio data.

[1556] Step 2:

[1557] The device sends the captured digital audio data to a cloud-based speech recognition engine (Google Cloud Speech-to-Text), which converts the audio data into text data.

[1558] Input: Digital audio data

[1559] Output: Text data (for example, "What is the name of the main character in this book?")

[1560] Operation: The device uploads voice data to the cloud engine, and the speech recognition engine converts it into text.

[1561] Step 3:

[1562] The server receives the converted text data and uses a natural language processing algorithm (TensorFlow model) to analyze the user's intent behind the question.

[1563] Input: Text data

[1564] Output: Analyzed question intent (e.g., "I want to know the protagonist's name")

[1565] Operation: The server inputs text data into a natural language processing algorithm to analyze the intent of the question.

[1566] Step 4:

[1567] The server searches for relevant information from the audiobook content database based on the analyzed question intent.

[1568] Input: Analyzed question intent

[1569] Output: Related information (e.g., "Protagonist's name: John Smith")

[1570] Operation: The server issues a query to the database and retrieves the corresponding information.

[1571] Step 5:

[1572] The server generates a text-based response based on the relevant information it has retrieved.

[1573] Input: Related Information

[1574] Output: Text-formatted answer (e.g., "The protagonist's name is John Smith")

[1575] Operation: The server combines relevant information and generates a response in text format to provide to the user.

[1576] Step 6:

[1577] The server sends the generated text-based response to a speech synthesis engine (Amazon Polly) to convert it into audio data.

[1578] Input: Text format answer

[1579] Output: Audio data

[1580] Operation: The server sends text data to the speech synthesis engine and retrieves the data output as speech.

[1581] Step 7:

[1582] The server analyzes voice input and text data using an emotion recognition engine (Affectiva) to recognize the user's emotions.

[1583] Input: Voice input, text data

[1584] Output: User's emotional state (e.g., "tired")

[1585] Operation: The server sends voice input and text data to the emotion recognition engine, which then analyzes the user's emotions.

[1586] Step 8:

[1587] The server adjusts the tone and content of its responses based on the emotions it perceives.

[1588] Input: User's emotional state, text-formatted response.

[1589] Output: Adjusted audio data

[1590] Operation: The server uses the results of emotion recognition to change the response to a gentler tone or content that promotes relaxation.

[1591] Step 9:

[1592] The device transmits the adjusted audio data to the smart glasses' speaker, providing information to the user.

[1593] Input: Adjusted audio data

[1594] Output: Provides an answer via voice (for example, "The main character's name is John Smith. Relax and enjoy the rest.")

[1595] Operation: The device retrieves audio data from the server, plays it through the speaker, and provides it to the user.

[1596] This processing flow allows users to ask questions in real time while listening to audiobooks and receive personalized answers that are tailored to their emotions.

[1597] 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.

[1598] 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 those described above. 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 shown 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.

[1599] 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.

[1600] [Fourth Embodiment]

[1601] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1602] 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.

[1603] 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).

[1604] 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.

[1605] 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.

[1606] 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).

[1607] 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.

[1608] 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.

[1609] 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.

[1610] 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.

[1611] 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.

[1612] 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.

[1613] 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".

[1614] The present invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to an audiobook. The system consists of the following main components: voice input acquisition means, voice recognition means, question analysis means, information retrieval means, answer generation means, voice synthesis means, and voice output means.

[1615] Overall processing flow

[1616] 1. Acquisition of voice input

[1617] The user can ask questions about what they want to know using voice commands.

[1618] The device captures the user's voice through the microphone.

[1619] 2. Speech Recognition

[1620] The device converts the captured audio into text data and sends it to the speech recognition engine.

[1621] 3. Question analysis

[1622] The server receives text data sent from the speech recognition engine and uses natural language processing to analyze the intent of the question.

[1623] 4. Information Retrieval

[1624] Based on the intent of the analyzed question, the server searches for relevant information within the audiobook content.

[1625] 5. Answer generation

[1626] The server generates an answer based on the search results.

[1627] 6. Speech synthesis

[1628] The server uses a speech synthesis engine to convert the generated responses from text to speech.

[1629] 7. Audio Output

[1630] The server sends the synthesized voice data to the terminal.

[1631] The device plays back the received audio data and provides the user with an answer.

[1632] Detailed explanations for each step are provided below.

[1633] Acquiring voice input

[1634] The user says to the device, "What is the name of the main character in this book?"

[1635] The device captures this audio through its built-in microphone and saves it as audio data.

[1636] Speech recognition

[1637] The device sends the captured audio data to a speech recognition engine in the cloud.

[1638] The speech recognition engine converts the audio data into text data such as "What is the name of the main character in this book?".

[1639] The terminal receives the converted text data and sends it to the server.

[1640] question analysis

[1641] The server analyzes the received text data using a natural language processing algorithm.

[1642] Understand the intent of the question and identify the specific information the user wants to know (in this case, "the protagonist's name").

[1643] Information Retrieval

[1644] Based on the identified intent, the server issues a search query to the audiobook content database.

[1645] The server retrieves relevant information, such as the protagonist's name "John Smith," as a search result.

[1646] Answer generation

[1647] Based on the retrieved search results, the server generates the answer, "The name of the main character in this book is John Smith."

[1648] Speech synthesis

[1649] The server sends the generated text-based response to the speech synthesis engine.

[1650] The speech synthesis engine converts this text into speech data.

[1651] Audio output

[1652] The server sends the audio data to the terminal.

[1653] The device plays audio data and tells the user, "The name of the main character in this book is John Smith."

[1654] Specific example

[1655] For example, if a user asks, "What is the most important point in this chapter?", the same process would apply.

[1656] The server generates the response, "The key point of this chapter is the scene where the protagonist makes a firm decision," and plays it back to the user as audio.

[1657] This invention allows users to instantly obtain the information they want while listening to audiobooks, deepening their understanding of the content. In addition, it reduces the wasted time caused by conventional methods of re-searching and re-listening, enabling more efficient access to information.

[1658] The following describes the processing flow.

[1659] Step 1:

[1660] The user speaks a question into the device. Example: "What is the name of the main character in this book?"

[1661] Step 2:

[1662] The device uses its microphone to capture the user's voice and saves it as digital audio data.

[1663] Step 3:

[1664] The device sends the captured audio data to the speech recognition engine.

[1665] Step 4:

[1666] The speech recognition engine analyzes the audio data and converts it into corresponding text data. For example, it might generate text data such as, "What is the name of the main character in this book?"

[1667] Step 5:

[1668] The speech recognition engine sends the converted text data back to the device.

[1669] Step 6:

[1670] The terminal sends text data to the server.

[1671] Step 7:

[1672] The server analyzes the received text data using natural language processing algorithms. Specifically, it grasps the intent of the user's question and extracts important keywords (e.g., "the protagonist's name").

[1673] Step 8:

[1674] Based on the analysis results, the server issues a query to the audiobook content database to retrieve relevant information.

[1675] Step 9:

[1676] The server retrieves relevant information from the database. For example, the data "John Smith" is searched for.

[1677] Step 10:

[1678] The server generates the answer, "The name of the main character in this book is John Smith," based on the search results.

[1679] Step 11:

[1680] The server sends the generated text-based response to the speech synthesis engine.

[1681] Step 12:

[1682] The speech synthesis engine analyzes text data and converts it into speech data.

[1683] Step 13:

[1684] The speech synthesis engine sends the generated audio data back to the server.

[1685] Step 14:

[1686] The server sends the audio data to the terminal.

[1687] Step 15:

[1688] The device plays the received audio data and tells the user, "The name of the main character in this book is John Smith."

[1689] (Example 1)

[1690] 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".

[1691] Conventional audio content listening systems were inefficient because users had to manually re-search or re-listen to obtain the information they needed. Furthermore, systems that converted voice input to text and provided appropriate answers to user questions lacked sufficient technology to accurately analyze the intent of the questions and provide timely, appropriate responses. As a result, it was difficult for users to quickly and accurately obtain the information they needed while listening to audio content.

[1692] 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.

[1693] In this invention, the server includes means for enabling the user to ask questions in real time while listening to audio content, means for generating answers to the questions using a natural language processing algorithm and a generative AI model, and means for converting the generated answers into audio data using a speech synthesizer. This enables the user to quickly and accurately obtain the necessary information without interrupting the audio content.

[1694] A "voice input device" is a device that captures a user's voice and converts it into digital voice data.

[1695] A "speech recognition engine" is a program or service that converts speech data into text data.

[1696] A "natural language processing algorithm" is a computational method for analyzing text data and understanding its meaning, context, and intent.

[1697] A "generative AI model" is an artificial intelligence model designed to generate natural language responses based on text data.

[1698] A "speech synthesis device" is a device or service used to convert text data into speech data.

[1699] "Audio content" refers to information and entertainment provided in audio format.

[1700] A "prompt sentence" is an input sentence given to a generative AI model, and it is an instruction sentence used to generate a specific answer.

[1701] "Real-time" refers to the fact that user requests and operations are processed instantly with virtually no delay.

[1702] A "database" is a system designed to organize and manage large amounts of data, enabling efficient searching and updating.

[1703] A "user" is a person who attempts to obtain information by operating a system.

[1704] A "terminal" is an electronic device used by a user to access a system.

[1705] A "server" is a computer that acts as the central hub of a system, processing, storing, and providing data.

[1706] This invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to audio content. This system acquires the user's voice input and performs a series of processes to provide an appropriate response using natural language processing and a generative AI model, and its specific implementation method is described below.

[1707] System Overview

[1708] This system consists of the following main parts:

[1709] 1. Voice input device

[1710] 2. Speech Recognition Engine

[1711] 3. Natural Language Processing Algorithms

[1712] 4. Generative AI Models

[1713] 5. Speech synthesis device

[1714] 6. Server

[1715] 7. Terminal

[1716] Required hardware and software

[1717] Voice input device: Used to capture the user's voice. Example: Microphone.

[1718] Speech recognition engine: Used to convert speech data into text data. Examples: Google Cloud Speech-to-Text API, IBM Watson Speech to Text, etc.

[1719] Natural Language Processing (NLP) algorithms: Used to analyze the intent behind a user's question. Examples: Natural Language Processing libraries such as NLTK and spaCy.

[1720] Generative AI models: Used to generate appropriate answers to questions. Example: OpenAI's GPT-3.

[1721] Text-to-speech (Speech synthesis) devices are used to convert text data into speech data. Examples include Amazon Polly and Google Cloud Text-to-Speech.

[1722] Server: Manages the processing of the entire system.

[1723] Terminal: A device used by a user to access a system, input questions, and receive answers. Examples: smartphones, tablets, personal computers, etc.

[1724] Specific example

[1725] When a user is listening to audio content, for example, they might be asked, "What is the name of the main character in this book?" In this case, the system operates as follows:

[1726] 1. Acquisition of voice input

[1727] The user says to the device, "What is the name of the main character in this book?"

[1728] The device captures this audio through its built-in microphone and saves it as audio data.

[1729] 2. Speech Recognition

[1730] The device sends the captured audio data to a speech recognition engine in the cloud.

[1731] The speech recognition engine converts the audio data into text data such as "What is the name of the main character in this book?".

[1732] The terminal receives the converted text data and sends it to the server.

[1733] 3. Question analysis

[1734] The server analyzes the received text data using a natural language processing algorithm.

[1735] Understand the intent of the question and identify the specific information the user wants to know (in this case, "the protagonist's name").

[1736] 4. Information Retrieval

[1737] Based on the intent of the analyzed question, the server searches for relevant information from the audio content.

[1738] The server retrieves relevant information as search results. For example, the protagonist's name is "John Smith".

[1739] 5. Answer generation

[1740] Based on the retrieved search results, the server generates the answer, "The name of the main character in this book is John Smith."

[1741] 6. Speech synthesis

[1742] The server sends the generated text-based response to the speech synthesis engine.

[1743] The speech synthesis engine converts this text into speech data.

[1744] 7. Audio Output

[1745] The server sends the audio data to the terminal.

[1746] The device plays audio data and tells the user, "The name of the main character in this book is John Smith."

[1747] Example of a prompt

[1748] The following is an example of a prompt message used to generate an appropriate answer to a user's question using a generative AI model.

[1749] Prompt example (English):

[1750] "User asks in English: 'What is the name of the main character in this book?'. The audiobooks database has the following relevant information: 'The main character of the book is John Smith.'. Respond to the user's question with this information."

[1751] Prompt example (Japanese):

[1752] The user asks, "What is the name of the main character in this book?" The audio content database contains the related information, "The main character in this book is John Smith." Use this information to answer the user's question.

[1753] This invention allows users to instantly obtain the information they want while listening to audio content, leading to a deeper understanding of the content. This reduces the wasted time caused by traditional methods of re-searching or re-listening, enabling more efficient access to information.

[1754] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1755] Step 1:

[1756] The user asks a question aloud, "What is the name of the main character in this book?" The input is the user's voice, which is captured by the device's built-in microphone. Specifically, the device runs a voice capture program and saves it as audio data.

[1757] Step 2:

[1758] The device sends the captured audio data to a speech recognition engine in the cloud. The input is audio data, and the output is text data. Specifically, the device uses a cloud communication module to send audio data, which the speech recognition engine converts into text, such as "What is the name of the main character in this book?"

[1759] Step 3:

[1760] The device receives converted text data from the cloud and sends it to the server. The input is text data from the speech recognition engine, and the output is the transmission of text data to the server. Specifically, the device uses a cloud communication module to transfer the text data to the server.

[1761] Step 4:

[1762] The server analyzes the received text data using a natural language processing algorithm. The input is text data, and the output is the result of analyzing the intent of the user's question. Specifically, the server uses natural language processing libraries such as NLTK or spaCy to identify the intent of the question as "I want to know the name of the main character."

[1763] Step 5:

[1764] The server searches for relevant information based on the analysis results. The input is the analysis result (the intent of the question), and the output is the relevant information (search results). Specifically, the server performs a search on the database using SQL queries or Elasticsearch, and as a result retrieves the name "John Smith".

[1765] Step 6:

[1766] The server generates an answer based on the search results. The input is the information from the search results, and the output is the text data of the answer. Specifically, the server uses a generative AI model (for example, GPT-3) to generate the text "The name of the main character in this book is John Smith."

[1767] Step 7:

[1768] The server sends the generated text-based response to a text-to-speech engine, where it is converted into audio data. The input is the text data of the response, and the output is audio data. Specifically, the server uses Amazon Polly or Google Text-to-Speech to generate the audio data.

[1769] Step 8:

[1770] The server sends the generated audio data to the terminal. The input is audio data, and the output is the transmission of audio data to the terminal. Specifically, the server transfers the audio data to the terminal using a cloud communication module.

[1771] Step 9:

[1772] The terminal plays the received audio data and provides it to the user. The input is audio data, and the output is audio playback. Specifically, the terminal uses an audio playback program to play the sentence, "The name of the main character in this book is John Smith."

[1773] (Application Example 1)

[1774] 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".

[1775] Currently, in physical stores, customers need to either ask store staff directly or search for information themselves to get instant product information, which can be time-consuming. Furthermore, some stores may not have sufficient staff, making it difficult for customers to obtain the information they need immediately. Therefore, there is a need for an efficient system that allows customers to get immediate answers to their product inquiries within physical stores.

[1776] 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.

[1777] In this invention, the server includes means for acquiring user voice input, means for converting voice input into text data, means for analyzing the intent of the user's question, means for searching for relevant information based on the analysis results, means for generating the retrieved information as text data, means for converting the text data into voice data, means for providing the voice data to the user, and software that operates on a smart device, including applications for in-store applications where customers ask questions about product features and specifications. As a result, even when a customer asks a question about a product in a physical store using voice input, relevant information is provided immediately, making it possible to obtain the necessary information quickly and accurately.

[1778] definition statement

[1779] "Means for acquiring user voice input" refers to devices or systems that capture the voice spoken by a user and save it as digital data.

[1780] "Means of converting voice input into text data" refers to the process or system of converting acquired voice data into text data.

[1781] "Means for analyzing the intent behind user questions" refers to methods and algorithms for analyzing questions that have been converted into text data and understanding their content and intent.

[1782] "Means for searching for relevant information based on analysis results" refers to a function that retrieves relevant data from appropriate sources in accordance with the analyzed intent.

[1783] "Means for generating retrieved information as text data" refers to a mechanism that processes and generates acquired information as text data in an appropriate format.

[1784] "Means of converting text data into audio data" refers to technologies and tools that convert generated text into synthesized speech.

[1785] "Means of providing audio data to users" refers to devices or systems that play synthesized audio and allow users to listen to it.

[1786] "Application to in-store applications where customers ask questions about product features and specifications" refers to an application technology that allows customers in physical stores to ask questions about products using voice and receive immediate answers to those questions.

[1787] "Software that runs on smart devices" refers to a series of programs and applications that run on electronic devices such as smartphones and smart glasses.

[1788] Modes for carrying out the invention

[1789] System Program Overview

[1790] The system of this invention acquires voice input from a user, analyzes it, and immediately provides information about products. The entire system includes a series of processes in which the user inputs voice via a smart device (such as a smartphone or smart glasses), the voice is converted into text, analyzed, and an appropriate response is generated, synthesized into speech, and then provided back to the user.

[1791] Hardware and software usage

[1792] Hardware:

[1793] 1. Smart devices: Examples include smartphones and smart glasses. These devices have voice input and voice output capabilities.

[1794] 2. Microphone: The device's built-in microphone captures voice input.

[1795] software:

[1796] 1. Speech recognition engine: Use the speech_recognition library to convert speech data into text data.

[1797] 2. Natural Language Processing Algorithm: The intent of the question will be analyzed using OpenAI's GPT-3 model, and an appropriate answer will be generated. Specifically, the OpenAI API will be used.

[1798] 3. Text-to-Speech Engine: The GTTS (Google Text-to-Speech) library and the playsound library are used to convert text-based responses into audio data and then play it back.

[1799] Data processing and data calculation

[1800] 1. Acquisition of voice input:

[1801] The user voice-inputs questions about the product into the device.

[1802] The microphone captures the sound and saves it as digital data.

[1803] 2. Speech recognition:

[1804] The acquired audio data is converted into text data using the speech_recognition library.

[1805] 3. Question analysis and answer generation:

[1806] Text data is sent as a prompt to the OpenAI API, analyzed using a GPT-3 model, and an appropriate response is generated.

[1807] Examples of prompt statements to use:

[1808] A customer asked: 'What are the features of this shampoo?' Please provide information about the product.

[1809] 4. Speech synthesis and output:

[1810] The generated text data is converted into audio data using the gtts library.

[1811] The playsound library is used to play the generated audio data and provide it to the user.

[1812] Specific example

[1813] For example, if a user asks "What are the features of this shampoo?" while in a physical store, the system will operate as follows:

[1814] 1. The microphone captures the audio, and the speech recognition engine generates text data that says, "What are the features of this shampoo?"

[1815] 2. This text data is sent to the OpenAI API, and the GPT-3 model generates the response, "This shampoo is rich in natural ingredients and makes hair soft."

[1816] 3. The generated text response is converted into audio data by a speech synthesis engine and provided to the user as audio using the playsound library.

[1817] This system allows customers to quickly obtain product information within physical stores, reducing the burden on store staff. Furthermore, it allows customers to deepen their understanding of the products.

[1818] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1819] Program processing steps

[1820] Step 1:

[1821] The user voice-inputs a question about the product into a smart device. For example, the user might say, "What are the features of this shampoo?" The input is voice data, which is then captured as digital data using a microphone.

[1822] Step 2:

[1823] The device converts the captured audio data into text data using the speech_recognition library. In this step, the audio data is the input, and the converted text data is the output. Specifically, the speech recognition engine converts it into the text "What are the features of this shampoo?".

[1824] Step 3:

[1825] The terminal sends the converted text data to the server. The server then sends it to the OpenAI API in the form of a prompt sentence to analyze the user's intent. The input is text data, which is converted into data in the form of a prompt sentence. For example, the prompt sentence "The customer asked: 'What are the features of this shampoo?' Please provide information about the product." is generated.

[1826] Step 4:

[1827] The server sends the generated prompt to the OpenAI GPT-3 model, which analyzes the intent of the question and generates an appropriate answer. The input is the prompt, and the output is the text data of the answer. Specifically, GPT-3 generates the answer "This shampoo is rich in natural ingredients and makes hair soft."

[1828] Step 5:

[1829] The server returns the generated response as text data to the terminal. The input is the generated text data, and the output and transmitted data are also text data.

[1830] Step 6:

[1831] The device converts received text data into audio data using GTTS. The input is text data, and the output is audio data. Specifically, the sentence "This shampoo is rich in natural ingredients and makes hair soft" is converted into audio data.

[1832] Step 7:

[1833] The device plays the converted audio data using the playsound library and provides it to the user. The input is audio data, and the output is the played audio. The user can receive a response via voice from their smart device.

[1834] 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.

[1835] The present invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to audiobooks, and more particularly to a system that incorporates an emotion engine to recognize the user's emotions. The system consists of the following main parts: voice input acquisition means, voice recognition means, question analysis means, information retrieval means, answer generation means, voice synthesis means, voice output means, and emotion recognition means.

[1836] Overall processing flow

[1837] 1. Acquisition of voice input

[1838] The user can ask questions about what they want to know using voice commands.

[1839] The device captures the user's voice through the microphone.

[1840] 2. Speech Recognition

[1841] The device converts the captured audio into text data and sends it to the speech recognition engine.

[1842] 3. Question analysis

[1843] The server receives text data sent from the speech recognition engine and uses natural language processing to analyze the intent of the question.

[1844] 4. Information Retrieval

[1845] Based on the intent of the analyzed question, the server searches for relevant information within the audiobook content.

[1846] 5. Answer generation

[1847] The server generates an answer based on the search results.

[1848] 6. Speech synthesis

[1849] The server uses a speech synthesis engine to convert the generated responses from text to speech.

[1850] 7. Emotion recognition

[1851] The server uses voice input and text data to analyze the user's emotions using an emotion engine.

[1852] 8. Adjusting responses based on emotions.

[1853] The server adjusts the content and tone of its responses based on the recognized emotions of the user.

[1854] 9. Audio Output

[1855] The server sends the synthesized voice data to the terminal.

[1856] The device plays back the received audio data and provides the user with an answer.

[1857] Detailed explanations for each step are provided below.

[1858] Acquiring voice input

[1859] For example, the user might say to the device, "What is the name of the main character in this book?"

[1860] The device captures audio through its built-in microphone and saves it as audio data.

[1861] Speech recognition

[1862] The device sends the captured audio data to a speech recognition engine in the cloud.

[1863] The speech recognition engine converts the audio data into text data such as "What is the name of the main character in this book?".

[1864] The terminal receives the converted text data and sends it to the server.

[1865] question analysis

[1866] The server analyzes the received text data using a natural language processing algorithm.

[1867] Understand the intent of the question and identify the specific information the user wants to know (in this case, "the protagonist's name").

[1868] Information Retrieval

[1869] Based on the identified intent, the server issues a search query to the audiobook content database.

[1870] The server retrieves relevant information, such as the protagonist's name "John Smith," as a search result.

[1871] Answer generation

[1872] Based on the retrieved search results, the server generates the answer, "The name of the main character in this book is John Smith."

[1873] Speech synthesis

[1874] The server sends the generated text-based response to the speech synthesis engine.

[1875] The speech synthesis engine converts this text into speech data.

[1876] emotion recognition

[1877] The server uses user voice input and text data and analyzes it with an emotion engine. For example, it can recognize whether the user is feeling stressed when asking a question.

[1878] Emotion-based response adjustments

[1879] The server adjusts the content and tone of its responses based on the perceived emotions of the user. For example, if the user is irritated, it will respond in a gentle tone.

[1880] Audio output

[1881] The server sends the adjusted audio data to the terminal.

[1882] The device plays audio data and tells the user, "The name of the main character in this book is John Smith."

[1883] Specific example

[1884] For example, if the user is tired, and the emotion engine recognizes the user's fatigue, it might ask, "What is the name of the main character in this book?" The server will then generate a gentle response, such as, "The main character in this book is named John Smith. Please take a break and enjoy the rest of the book," and play it back through speech synthesis.

[1885] This invention is an advanced adaptive system that can adjust responses according to the user's emotions, thereby improving the user experience.

[1886] The following describes the processing flow.

[1887] Step 1:

[1888] The user speaks a question into the device. For example, "What is the name of the main character in this book?"

[1889] Step 2:

[1890] The device uses its microphone to capture the user's voice and saves it as digital audio data.

[1891] Step 3:

[1892] The device sends the captured audio data to the speech recognition engine.

[1893] Step 4:

[1894] The speech recognition engine analyzes the audio data and converts it into corresponding text data. For example, it might generate text data such as, "What is the name of the main character in this book?"

[1895] Step 5:

[1896] The speech recognition engine sends the converted text data back to the device.

[1897] Step 6:

[1898] The terminal sends text data to the server.

[1899] Step 7:

[1900] The server analyzes the received text data using natural language processing algorithms. Specifically, it grasps the intent of the user's question and extracts key keywords (e.g., "the protagonist's name").

[1901] Step 8:

[1902] Based on the analysis results, the server issues a query to the audiobook content database to retrieve relevant information.

[1903] Step 9:

[1904] The server retrieves relevant information from the database. For example, the data "John Smith" is searched for.

[1905] Step 10:

[1906] Based on the search results obtained by the server, the system generates the answer, "The name of the main character in this book is John Smith."

[1907] Step 11:

[1908] The server sends the generated text-formatted response to the sentiment engine.

[1909] Step 12:

[1910] The server uses an emotion engine to recognize the user's emotions from their voice input. For example, it can determine if the user is stressed or has questions.

[1911] Step 13:

[1912] The server adjusts the response content to match the user's emotions based on the emotions it recognizes. For example, if the user is tired, the response will be made more kind and include elements of encouragement.

[1913] Step 14:

[1914] The server sends text that has been adjusted according to the emotion to the speech synthesis engine.

[1915] Step 15:

[1916] The speech synthesis engine analyzes text data and converts it into speech data.

[1917] Step 16:

[1918] The speech synthesis engine sends the generated audio data back to the server.

[1919] Step 17:

[1920] The server sends the audio data to the terminal.

[1921] Step 18:

[1922] The device plays the received audio data and tells the user, "The main character of this book is named John Smith. Please enjoy the rest of the story."

[1923] In this way, users can ask questions in their own voice, and the system can not only respond appropriately to those questions but also adjust its answers according to the user's emotions. This process allows users to enjoy a more personalized experience.

[1924] (Example 2)

[1925] 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".

[1926] Conventional voice dialogue systems have the problem of not improving the user experience because they provide mechanical responses without considering the user's emotions. Furthermore, the accuracy of answers to user questions is low, making it difficult to provide accurate information quickly. In addition, there was a lack of efficient ways to obtain necessary information when listening to audio content such as audiobooks.

[1927] 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.

[1928] In this invention, the server includes means for acquiring user voice input, means for converting voice input into text data, means for analyzing the intent of the user's question, means for searching for relevant information based on the analysis results, means for generating the retrieved information as text data, means for converting the text data into voice data, means for analyzing the user's emotions, means for adjusting the tone of the response based on the analyzed emotions, and means for providing the adjusted voice data to the user. This enables responses that are in line with the user's emotions, improving the user experience and enabling the rapid provision of accurate information.

[1929] "Means for acquiring voice input" refers to a device or method for capturing the voice spoken by a user and converting it into digital data.

[1930] "Means for converting voice input into text data" refers to technologies and algorithms for analyzing captured voice data and converting it into a corresponding text format.

[1931] "Means for analyzing the intent behind a user's question" refers to natural language processing techniques that analyze the content of text data to understand the information the user wants to know and the purpose of their question.

[1932] "Means for retrieving relevant information based on analysis results" refers to techniques for retrieving relevant information from databases and other sources based on the intent of the analyzed question.

[1933] "Means for generating searched information as text data" refers to technologies for generating information in text format based on search results and providing it to users.

[1934] "Methods for converting text data into audio data" refers to technologies that convert generated text data into audio data using speech synthesis technology.

[1935] "Methods for analyzing user emotions" refer to technologies that recognize and analyze the emotions a user is feeling based on their voice input and text data.

[1936] "Means for adjusting the tone of responses based on analyzed emotions" refers to technologies that appropriately modify the content and tone of generated responses according to the user's emotions.

[1937] "Means for providing adjusted audio data to a user" refers to devices or methods for playing or providing adjusted audio data to a user.

[1938] Modes for carrying out the invention

[1939] This invention relates to a system that allows users to quickly and accurately obtain necessary information while listening to audiobooks, and simultaneously recognize and respond to the user's emotions. This system includes the following hardware and software.

[1940] Main components and devices

[1941] 1. Voice input acquisition method

[1942] The device uses its built-in microphone to acquire voice input from the user.

[1943] An external microphone or a microphone built into the device will be used.

[1944] 2. Speech recognition means

[1945] The device sends the acquired voice data to a speech recognition engine in the cloud.

[1946] The specific software used is the Google Cloud Speech-to-Text API.

[1947] The speech recognition engine converts speech data into text data.

[1948] 3. Question analysis means

[1949] The server analyzes the received text data using a natural language processing algorithm.

[1950] Specifically, the Python libraries NLTK and spaCy will be used.

[1951] The server analyzes the intent behind the user's question and identifies specific information.

[1952] 4. Information Retrieval Methods

[1953] Based on the analysis results, the server issues a search query to the audiobook content database.

[1954] SQL queries are used for the search.

[1955] The server retrieves relevant information as search results.

[1956] 5. Answer generation means

[1957] Based on the information it retrieves, the server generates an appropriate response using a template.

[1958] The response will be generated as text data.

[1959] 6. Speech synthesis means

[1960] The server uses a speech synthesis engine to convert the generated text data into speech data.

[1961] The Google Text-to-Speech API is used.

[1962] 7. Emotion recognition means

[1963] The server analyzes the user's emotions based on their voice input and text data.

[1964] OpenAI's sentiment analysis model will be used.

[1965] 8. Response adjustment mechanisms based on emotions

[1966] The server adjusts the tone of its responses based on the analyzed user's emotions.

[1967] By adding gentle tones and encouraging words, we can respond to the user's emotions.

[1968] 9. Audio output means

[1969] The server sends the adjusted audio data to the terminal.

[1970] The device plays audio data and provides a response to the user.

[1971] Specific example

[1972] For example, if a user asks, "What is the name of the main character in this book?", the device captures this audio and converts it to text using the Google Cloud Speech-to-Text API. The server then analyzes this text using natural language processing algorithms (such as NLTK or spaCy) and searches for relevant information in a database. Based on the search results, the answer "The main character in this book is named John Smith." is generated and converted to speech using the Google Text-to-Speech API. The user's emotions are then analyzed using an OpenAI sentiment analysis model, and the tone is adjusted as needed. Finally, the adjusted audio data is sent to the device and provided to the user.

[1973] Example of a prompt

[1974] "Convert the user's spoken questions into text and generate corresponding answers."

[1975] "Analyze the user's emotions and adjust the tone of your responses accordingly."

[1976] "Please search for the main character's name in the audiobook and tell it to the user."

[1977] "Adjust the content and tone of your responses according to the user's emotions."

[1978] This system aims to significantly improve the user experience by allowing users to quickly and accurately obtain necessary information while enjoying audiobooks, and by providing responses that take the user's emotions into consideration.

[1979] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1980] System program processing steps

[1981] Step 1: Obtaining voice input

[1982] The user asks a question to the device.

[1983] Input: User's voice (e.g., "What is the name of the main character in this book?")

[1984] Operation: The device's built-in microphone captures the user's voice and saves it as audio data (in WAV format, etc.).

[1985] Output: Captured audio data

[1986] Step 2: Speech Recognition

[1987] The device sends the captured audio data to a speech recognition engine in the cloud.

[1988] Input: Audio data

[1989] How it works: The Google Cloud Speech-to-Text API converts speech data into text data.

[1990] Output: Converted text data (e.g., "What is the name of the main character in this book?")

[1991] Step 3: Questionnaire Analysis

[1992] The server analyzes the received text data using a natural language processing algorithm.

[1993] Input: Text data

[1994] Operation: Uses Python's NLTK and spaCy libraries to parse text data and identify the intent behind user questions. For example, it parses the subject, predicate, and object of a question.

[1995] Output: Analysis results (Example: "I want to know the protagonist's name")

[1996] Step 4: Information Retrieval

[1997] Based on the analysis results, the server issues a search query to the audiobook content database.

[1998] Input: Analysis results

[1999] Operation: Generates an SQL query and performs a search on the audiobook database. For example, it searches for information related to "the main character's name".

[2000] Output: Search results (e.g., "John Smith")

[2001] Step 5: Generate Response

[2002] The server generates an answer based on the retrieved search results.

[2003] Input: Search Results

[2004] Operation: Generates an answer using a template. Converts it to a format such as "The name of the main character in this book is John Smith."

[2005] Output: Text-formatted answer (Example: "The name of the main character in this book is John Smith.")

[2006] Step 6: Speech Synthesis

[2007] The server sends the generated text-formatted response to the speech synthesis engine.

[2008] Input: Text data response

[2009] How it works: The Google Text-to-Speech API converts text into speech data.

[2010] Output: Audio data (e.g., audio file)

[2011] Step 7: Emotion Recognition

[2012] The server analyzes the user's emotions based on their voice input and text data.

[2013] Input: Audio data and text data

[2014] Operation: Uses OpenAI's emotion analysis model to analyze the user's emotions. For example, it identifies stress and fatigue from voice tone and text content.

[2015] Output: User's emotional state (e.g., "tired")

[2016] Step 8: Adjusting responses based on emotions

[2017] The server adjusts the tone of its responses based on the analyzed user's emotions.

[2018] Input: Emotional state, text-formatted response

[2019] Action: If the user is tired, add a gentle tone and encouraging words to the response. For example, adjust it to a format like, "The main character of this book is named John Smith. Please take a break and enjoy the rest of the story."

[2020] Output: Answer in adjusted text format

[2021] Step 9: Audio Output

[2022] The server sends the adjusted audio data to the terminal.

[2023] Input: Adjusted audio data

[2024] Operation: The device plays the received audio data.

[2025] Output: The user listens to pre-recorded audio data (e.g., "The main character of this book is named John Smith. Please relax and enjoy the rest of the story.")

[2026] In this way, by clearly defining the specific actions, inputs, and outputs at each step, we have created a system that allows users to quickly and accurately obtain the information they need, and furthermore, can respond to the user's emotions during the process.

[2027] (Application Example 2)

[2028] 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".

[2029] Conventional speech recognition systems have made it difficult for users to quickly and accurately obtain necessary information while listening to audiobooks. Furthermore, conventional systems cannot recognize the user's emotions and adjust the content and tone of the responses accordingly, resulting in a limited user experience. The present invention aims to provide a system that not only provides information based on the user's voice input but also analyzes the user's emotions in real time and provides responses with the optimal tone and content.

[2030] 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.

[2031] In this invention, the server includes means for converting voice input into text data, means for analyzing the intent of the user's question, and means for retrieving relevant information. This enables a personalized audiobook experience that responds to the user's emotions.

[2032] "Means for acquiring voice input" refers to devices or mechanisms for capturing the voice spoken by a user and recording it as digitized voice data.

[2033] "Means for converting voice input into text data" refers to engines or software that analyze captured voice data and convert it into text information.

[2034] "Means for analyzing the intent behind user questions" refer to natural language processing algorithms and systems that understand the intent behind user questions obtained as text data and identify relevant information.

[2035] "Means for searching for related information" refers to mechanisms or engines for detecting necessary information from databases and related resources based on the analyzed intent.

[2036] "Means for generating searched information as text data" refers to software or algorithms that organize searched information into text in a format that is easy for users to understand.

[2037] "Means for converting text data into audio data" refers to speech synthesis engines or systems that convert the generated text data into an audio format.

[2038] "Means of providing audio data to the user" refers to devices such as speakers and headsets that play back converted audio data and convey information to the user by voice.

[2039] "Means of recognizing user emotions" refer to emotion recognition engines and software that analyze user voice input, facial expression data, etc., to identify their emotional state.

[2040] "Means of adjusting responses according to emotions" refers to algorithms and control systems that appropriately change the content and tone of the responses provided based on the recognized emotional state of the user.

[2041] This invention provides a system that allows users to quickly and accurately obtain necessary information while listening to audiobooks, and in particular, incorporates an emotion engine that recognizes the user's emotions. The system includes a voice input acquisition means, a voice recognition means, a question analysis means, an information retrieval means, an answer generation means, a voice synthesis means, a voice output means, an emotion recognition means, and an answer adjustment means according to emotion.

[2042] Specific examples of hardware and software

[2043] hardware

[2044] Smart glasses: Equipped with a built-in microphone and speaker to capture the user's voice and provide the results to the user.

[2045] Server: Processes and manages data.

[2046] software

[2047] Speech recognition engine: Uses Google Cloud Speech-to-Text to convert speech input into text.

[2048] Natural language processing algorithm: TensorFlow is used to analyze the user's intent behind the question.

[2049] Speech synthesis engine: Uses Amazon Polly to convert text data into speech data.

[2050] Emotion recognition engine: Uses Affectiva or Microsoft Azure Emotion API to analyze user emotions.

[2051] Content distribution service APIs: Utilizing APIs from services like Audible and Spotify.

[2052] Program processing details

[2053] 1. Acquisition of voice input:

[2054] The smart glasses capture the user's voice commands using a microphone.

[2055] 2. Speech recognition:

[2056] The captured audio data is sent to a cloud-based speech recognition engine (Google Cloud Speech-to-Text) and converted into text data.

[2057] 3. Question analysis:

[2058] The converted text data is sent to a server and analyzed using a natural language processing algorithm (TensorFlow model) to identify the intent of the question.

[2059] 4. Information Retrieval:

[2060] The server searches for relevant information from the audiobook content database based on the intent of the question.

[2061] 5. Answer generation:

[2062] Based on the search results, it generates specific answers such as, "The name of the main character in this book is XX."

[2063] 6. Speech synthesis:

[2064] The generated responses are converted into audio data using Amazon Polly.

[2065] 7. Emotion recognition:

[2066] Affectiva is used to analyze user voice input and text data to identify the user's emotional state.

[2067] 8. Answer adjustment:

[2068] Based on the perceived emotions, the tone and content of the response will be adjusted. For example, if the user is feeling tired, the response will encourage them to relax.

[2069] 9. Audio output:

[2070] The adjusted audio data is sent to the smart glasses, and the information is provided audibly through the speaker.

[2071] Specific example

[2072] If the user asks a question: For example, if the user says, "What is the name of the main character in this book?"

[2073] A concrete example of emotion recognition: When the emotion recognition engine determines that the user is tired.

[2074] Adapted response: The smart glasses' speaker responds in a gentle tone, "The main character of this book is named XX. Please relax and enjoy the rest of the story."

[2075] Example of a prompt

[2076] "We want to develop an application that provides a personalized experience by offering appropriate responses based on voice input instructions, tailored to the user's emotional state. It would quickly search for the information the user is looking for within the audiobook content, provide the answer via voice, and adjust the tone of response based on the user's emotional state. For example, if the user is tired, the response would include a message encouraging relaxation."

[2077] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[2078] Step 1:

[2079] The user listens to an audiobook and makes a voice input into the microphone. The device (smart glasses) captures the user's voice through its built-in microphone and records it as digital audio data.

[2080] Input: User voice input (for example, "What is the name of the main character in this book?")

[2081] Output: Digital audio data

[2082] Operation: The microphone records the user's speech as audio data.

[2083] Step 2:

[2084] The device sends the captured digital audio data to a cloud-based speech recognition engine (Google Cloud Speech-to-Text), which converts the audio data into text data.

[2085] Input: Digital audio data

[2086] Output: Text data (for example, "What is the name of the main character in this book?")

[2087] Operation: The device uploads voice data to the cloud engine, and the speech recognition engine converts it into text.

[2088] Step 3:

[2089] The server receives the converted text data and uses a natural language processing algorithm (TensorFlow model) to analyze the user's intent behind the question.

[2090] Input: Text data

[2091] Output: Analyzed question intent (e.g., "I want to know the protagonist's name")

[2092] Operation: The server inputs text data into a natural language processing algorithm to analyze the intent of the question.

[2093] Step 4:

[2094] The server searches for relevant information from the audiobook content database based on the analyzed question intent.

[2095] Input: Analyzed question intent

[2096] Output: Related information (e.g., "Protagonist's name: John Smith")

[2097] Operation: The server issues a query to the database and retrieves the corresponding information.

[2098] Step 5:

[2099] The server generates a text-based response based on the relevant information it has retrieved.

[2100] Input: Related Information

[2101] Output: Text-formatted answer (e.g., "The protagonist's name is John Smith")

[2102] Operation: The server combines relevant information and generates a response in text format to provide to the user.

[2103] Step 6:

[2104] The server sends the generated text-based response to a speech synthesis engine (Amazon Polly) to convert it into audio data.

[2105] Input: Text format answer

[2106] Output: Audio data

[2107] Operation: The server sends text data to the speech synthesis engine and retrieves the data output as speech.

[2108] Step 7:

[2109] The server analyzes voice input and text data using an emotion recognition engine (Affectiva) to recognize the user's emotions.

[2110] Input: Voice input, text data

[2111] Output: User's emotional state (e.g., "tired")

[2112] Operation: The server sends voice input and text data to the emotion recognition engine, which then analyzes the user's emotions.

[2113] Step 8:

[2114] The server adjusts the tone and content of its responses based on the emotions it perceives.

[2115] Input: User's emotional state, text-formatted response.

[2116] Output: Adjusted audio data

[2117] Operation: The server uses the results of emotion recognition to change the response to a gentler tone or content that promotes relaxation.

[2118] Step 9:

[2119] The device transmits the adjusted audio data to the smart glasses' speaker, providing information to the user.

[2120] Input: Adjusted audio data

[2121] Output: Provides an answer via voice (for example, "The main character's name is John Smith. Relax and enjoy the rest.")

[2122] Operation: The device retrieves audio data from the server, plays it through the speaker, and provides it to the user.

[2123] This processing flow allows users to ask questions in real time while listening to audiobooks and receive personalized answers that are tailored to their emotions.

[2124] 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.

[2125] 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 those described above. 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 shown 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.

[2126] 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 robot 414.

[2127] 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.

[2128] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[2129] 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.

[2130] 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.

[2131] 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.

[2132] 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."

[2133] 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.

[2134] 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.

[2135] 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.

[2136] 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.

[2137] 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.

[2138] 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.

[2139] 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.

[2140] 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.

[2141] 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.

[2142] 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.

[2143] 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.

[2144] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[2145] The following is further disclosed regarding the embodiments described above.

[2146] (Claim 1)

[2147] A means of obtaining user voice input,

[2148] A means of converting voice input into text data,

[2149] A means of analyzing the intent behind a user's question,

[2150] A means of searching for relevant information based on the analysis results,

[2151] A means of generating the searched information as text data,

[2152] A means of converting text data into audio data,

[2153] Means of providing audio data to users,

[2154] A system that includes this.

[2155] (Claim 2)

[2156] The system according to claim 1, characterized by having a microphone for capturing user voice input.

[2157] (Claim 3)

[2158] The system according to claim 1, characterized in that it analyzes the intent of the user's question using a natural language processing algorithm.

[2159]

[2160] "Example 1"

[2161] (Claim 1)

[2162] A means of obtaining user voice input,

[2163] A means of converting voice input into text data,

[2164] A means of analyzing the intent behind a user's question,

[2165] A means of searching for relevant information based on the analysis results,

[2166] A means of generating the searched information as text data,

[2167] A means of converting text data into audio data,

[2168] Means of providing audio data to users,

[2169] A means by which users can ask questions in real time while listening to audio content,

[2170] A means for generating answers to questions using natural language processing algorithms and generative AI models,

[2171] A system that includes this.

[2172] (Claim 2)

[2173] The system according to claim 1, characterized by comprising a voice input device for capturing the user's voice input.

[2174] (Claim 3)

[2175] The system according to claim 1, characterized in that it analyzes the intent of a user's question using a natural language processing algorithm and a generative AI model, and generates an appropriate prompt sentence in response to the question.

[2176] "Application Example 1"

[2177] (Claim 1)

[2178] A means of obtaining user voice input,

[2179] A means of converting voice input into text data,

[2180] A means of analyzing the intent behind a user's question,

[2181] A means of searching for relevant information based on the analysis results,

[2182] A means of generating the searched information as text data,

[2183] A means of converting text data into audio data,

[2184] Means of providing audio data to users,

[2185] This includes applications for in-store applications where customers ask questions about product features and specifications.

[2186] A system including software that runs on a smart device.

[2187] (Claim 2)

[2188] The system according to claim 1, characterized by having a microphone for capturing user voice input.

[2189] (Claim 3)

[2190] The system according to claim 1, characterized in that it analyzes the intent of the user's question using a natural language processing algorithm.

[2191] "Example 2 of combining an emotion engine"

[2192] (Claim 1)

[2193] A means of obtaining user voice input,

[2194] A means of converting voice input into text data,

[2195] A means of analyzing the intent behind a user's question,

[2196] A means of searching for relevant information based on the analysis results,

[2197] A means of generating the searched information as text data,

[2198] A means of converting text data into audio data,

[2199] A means of analyzing user emotions,

[2200] A means of adjusting the tone of responses based on analyzed emotions,

[2201] A means of providing the user with adjusted audio data,

[2202] A system that includes this.

[2203] (Claim 2)

[2204] The system according to claim 1, characterized by comprising a voice input device for capturing the user's voice input.

[2205] (Claim 3)

[2206] The system according to claim 1, characterized in that it analyzes the intent of the user's question using natural language processing technology.

[2207] "Application example 2 when combining with an emotional engine"

[2208] (Claim 1)

[2209] A means of obtaining user voice input,

[2210] A means of converting voice input into text data,

[2211] A means of analyzing the intent behind a user's question,

[2212] A means of searching for relevant information based on the analysis results,

[2213] A means of generating the searched information as text data,

[2214] A means of converting text data into audio data,

[2215] Means of providing audio data to users,

[2216] Means of recognizing user emotions,

[2217] A means of adjusting responses according to emotions,

[2218] A system that includes this.

[2219] (Claim 2)

[2220] The system according to claim 1, characterized by having a microphone for capturing user voice input.

[2221] (Claim 3)

[2222] The system according to claim 1, characterized in that it analyzes the intent of the user's question using a natural language processing algorithm. [Explanation of Symbols]

[2223] 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 of obtaining user voice input, A means of converting voice input into text data, A means of analyzing the intent behind a user's question, A means of searching for relevant information based on the analysis results, A means of generating the searched information as text data, A means of converting text data into audio data, Means of providing audio data to users, A system that includes this.

2. The system according to claim 1, characterized by comprising a microphone for capturing user voice input.

3. The system according to claim 1, characterized in that it analyzes the intent of the user's question using a natural language processing algorithm.

Images