system

Abstract

Provide a system. 【Solution means】 A receiving means for acquiring voice information from a user, A conversion means for converting the voice information into character information, An interpretation means for analyzing the character information and identifying the user's request, A collection means for collecting related information based on the identified request, A generation means for creating a response based on the collected information, A voice synthesis means for converting the generated response into information that can be input and output by voice, A presentation means for presenting the voice input / output-capable information to the user, A recording means for recording and storing the conversation history with the user, An adaptation means for providing information according to the user's situation in the environment, A system including the above.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a persona chatbot control method performed by at least one processor, including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance as a 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 life, users are required to obtain various information quickly and accurately on a daily basis. However, it takes time and effort to perform tasks such as information search, calculation, and management of personal past history individually, which is not efficient. There is also a problem that it is difficult to obtain consistent support when using multiple different platforms. To solve such problems, there is a need for a method to provide information according to users' needs in a unified manner and provide support in a natural dialogue form.

Means for Solving the Problems

[0005] This invention relates to a system equipped with a conversion means that acquires voice input from a user and converts the voice input into text data. Furthermore, an analysis means that analyzes the text data and identifies the user's request, enabling accurate understanding of the user's intent. Based on the identified request, an information acquisition means that collects relevant information from an external information database, enabling rapid acquisition of necessary data. In addition, a speech synthesis means that generates a response based on the collected information and converts it into data that can be output as voice is provided, enabling interactive information provision to the user in natural language. Moreover, a recording means that records and retains the history of conversations with the user makes it possible to provide personalized responses based on individual history. This enables efficient and comprehensive information support to be provided to the user.

[0006] "Input means" refers to a function or device for acquiring voice input from the user.

[0007] "Conversion means" refers to a function or device that converts voice input into digital text data.

[0008] "Analysis means" refers to a function or device that analyzes text data to identify the user's requests or intentions.

[0009] "Information acquisition means" refers to a function or device for acquiring necessary information based on an identified request, and may include access to an external information database.

[0010] "Response generation means" refers to a function or device that generates an appropriate response based on acquired information.

[0011] "Speech synthesis means" refers to a function or device that converts a generated response into speech data.

[0012] "Output means" refers to a function or device for presenting synthesized speech data to the user.

[0013] "Recording means" refers to a function or device that records and retains the history of conversations with the user.

[0014] A "conversational artificial intelligence system" is a platform that provides information and support through natural language dialogue with the user. [Brief explanation of the drawing]

[0015] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] This is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13]It is a sequence diagram showing the processing flow of the data processing system in Embodiment 2 when the emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when the emotion engine is combined.

Mode for Carrying Out the Invention

[0016] Hereinafter, an example of an embodiment of the system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0017] First, the terms used in the following description will be explained.

[0018] In the following embodiments, the numbered processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0019] In the following embodiments, the numbered RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0020] In the following embodiments, the 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, and the like.

[0021] 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).

[0022] 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."

[0023] [First Embodiment]

[0024] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0025] 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.

[0026] 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).

[0027] 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.

[0028] 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.

[0029] 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.

[0030] 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.

[0031] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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".

[0036] This invention is a "conversational artificial intelligence system" that provides information through interactive dialogue with the user and can be implemented in various forms. This system operates on a smart device and is always in standby mode. The system receives the voice input when the user speaks to it.

[0037] System Operation Overview:

[0038] 1. Acquisition of voice input

[0039] The terminal acquires voice input from the user through a microphone and processes that data as a digital signal.

[0040] 2. Converting speech to text

[0041] The server uses a speech recognition engine to convert speech data into text data. This conversion is performed in real time, and the user's speech is recognized as text.

[0042] 3. Analysis of intent

[0043] The server analyzes text data using natural language processing technology. This analysis makes it possible to understand the user's requests and intentions.

[0044] 4. Information Acquisition and Response Generation

[0045] Based on the analysis results, the server retrieves relevant information from an external information database as needed. Based on this information, the system generates an appropriate response.

[0046] 5. Generation and output of voice responses

[0047] The server converts the generated response into audio data using speech synthesis technology. The terminal plays this audio data through its speaker, providing the user with the response in voice.

[0048] 6. Recording of dialogue history

[0049] The server records the content of interactions with users as a history and maintains this history for each user. This history is used to personalize the next response.

[0050] Specific example:

[0051] If the user asks, "What time is tomorrow's meeting?", the device will capture the audio, and the server will convert the audio to text. The server will then analyze the text to understand the user's request as "to confirm the meeting time." The server will access the schedule database, retrieve the time of the relevant meeting, and generate a response such as, "Tomorrow's meeting is at 2 PM." This response will be converted back into audio and transmitted to the user by the device.

[0052] Thus, the system of the present invention can provide information immediately in response to the user's needs, enabling more efficient daily support.

[0053] The following describes the processing flow.

[0054] Step 1:

[0055] The user voice-inputs a question or instruction into the smart device. The voice input is captured by the device's microphone.

[0056] Step 2:

[0057] The terminal converts the acquired audio data into a digital signal and transmits it to the server via a secure channel. During this process, pre-processing such as noise reduction is applied to improve the quality of the audio data.

[0058] Step 3:

[0059] The server analyzes the received audio data using a speech recognition engine and converts it into text data. This process ensures that the speech is recognized as an accurate text-based command.

[0060] Step 4:

[0061] The server uses natural language processing techniques to analyze the content of text data and identify the user's intent and requests. This includes keyword extraction and semantic analysis.

[0062] Step 5:

[0063] The server retrieves the necessary information from external databases or APIs based on the identified request. For example, weather information or schedule information may be requested.

[0064] Step 6:

[0065] The server generates a response to the user based on the information it has obtained. This response is constructed in natural language so that the user can easily understand it.

[0066] Step 7:

[0067] The server converts the response text into speech data using a speech synthesis tool. This process makes the text response available in a format that can be output as speech.

[0068] Step 8:

[0069] The terminal provides a response by playing audio data received from the server to the user through its speaker.

[0070] Step 9:

[0071] The server records this interaction as a dialogue history and retains it as data to better personalize future responses.

[0072] (Example 1)

[0073] 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."

[0074] Many modern voice response systems are specialized in providing a limited range of information and struggle to generate flexible responses that meet the diverse needs of users. Furthermore, they have the challenge of not being able to fully utilize the user's past conversation history, thus failing to provide personalized, interactive dialogue.

[0075] 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.

[0076] In this invention, the server includes means for acquiring voice input through an input device, means for analyzing character data with a language processing device, and a data acquisition device for collecting information from an external database. This enables the generation of flexible responses to diverse requests and the provision of personalized information to each user.

[0077] An "input / output device" is a device that acquires, records, and provides voice input from the user to the system.

[0078] A "speech recognition device" is a device that includes technology for converting acquired speech data into text data.

[0079] A "language processing device" is a device that analyzes text data and utilizes natural language processing technology to identify the user's intentions and requests.

[0080] A "data acquisition device" is a device that collects necessary information from external or internal sources based on an identified request.

[0081] A "reaction generation device" is a device that generates an appropriate response as text based on collected information.

[0082] A "synthesis device" is a device that converts generated text responses into audio data so that they can be output as speech.

[0083] A "sound playback device" is a device such as a speaker that presents synthesized sound data to the user.

[0084] A "memory device" is a device used to record and store the history of interactions with the user.

[0085] A "personalization device" is a device that utilizes recorded dialogue history to generate subsequent responses, enabling personalized interactions.

[0086] This invention relates to a conversational artificial intelligence system that provides information through interactive dialogue with the user. The system performs a series of processes to receive voice input, analyze it, and output an appropriate voice response.

[0087] First, the terminal acquires voice input through the microphone. The acquired voice data is converted into a digital signal and sent to the server. The server uses a speech recognition engine (for example, an API provided by a common speech recognition service provider) to convert the voice data into text data. At this stage, high-precision real-time conversion of the voice input is required.

[0088] Next, the server analyzes the text data using natural language processing techniques (utilizing open-source language analysis tools such as NLTK and spaCy). This analysis recognizes the user's request and identifies their intent. Based on the analysis results, it retrieves information from an external information database (e.g., a cloud-based data management service).

[0089] Based on the acquired data, the server generates an appropriate response. This response is prepared as text data and converted into speech data using a speech synthesis engine (for example, an API provided by a common speech synthesis service provider). The terminal then plays this speech data through its speaker, delivering the answer to the user.

[0090] Furthermore, the server records the history of interactions with the user and uses this information to personalize responses in the future. This allows the system to provide increasingly efficient and personalized interactions.

[0091] For example, if a user asks, "What's the weather like today?", the device will capture this voice message. The server will convert the voice into text data and parse it as a request to provide weather information. It will then access a weather information database and generate a response such as, "Today's weather is sunny, and the temperature is 20 degrees Celsius," which it will then communicate to the user via voice. This system aims to respond quickly and accurately to a variety of everyday information requests.

[0092] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0093] Step 1:

[0094] The terminal acquires the user's voice through a microphone. This input voice is received as an analog signal and converted into a digital signal by the terminal's digital signal processor. The output of this process is digital voice data that is sent to the server.

[0095] Step 2:

[0096] The server receives digital audio data and converts it into text data using a speech recognition engine. Specifically, it calls a speech recognition API to parse the audio data and extract the corresponding strings. The input for this step is digital audio data, and the output is the parsed text data.

[0097] Step 3:

[0098] The server analyzes the parsed text data using natural language processing techniques. This procedure uses language processing libraries to analyze the sentence structure and identify the user's intent. The input is text data, and the output is a data structure representing the user's request and intent.

[0099] Step 4:

[0100] The server retrieves relevant information from the database based on the user's intent. It also calls external APIs as needed to access various cloud services and data repositories. The input in this process is a data structure representing the user's request, and the output is the requested information data.

[0101] Step 5:

[0102] The server generates a response based on the information it has acquired. This process involves selecting a response template based on conditions and embedding the actual data to create the response text. The input is informational data, and the output is the response text intended for the user.

[0103] Step 6:

[0104] The server converts the response text into speech data using a speech synthesis engine. This is done using a speech synthesis API, which obtains natural-sounding speech data from the text data. The input is the response text, and the output is the synthesized speech data.

[0105] Step 7:

[0106] The device receives the synthesized audio data and plays it back to the user through its speaker. Specifically, it uses an audio playback module for output. In this case, the input is audio data, and the output is the audio the user hears.

[0107] Step 8:

[0108] The server records the user interaction history and adds it to the training data for future interactions. This record includes utterances and system responses, and is used to provide personalized interactions. The input is the interaction history data, and the output is an updated history database.

[0109] (Application Example 1)

[0110] 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."

[0111] In the daily lives of the elderly and those requiring support, there is a need to provide necessary information quickly and accurately while appropriately responding to changes and circumstances around them. Especially when daily living support is needed, timely information and instructions are essential for improving the quality of life. However, conventional technologies do not take into account the user's surroundings and circumstances, making it difficult to provide the optimal response.

[0112] 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.

[0113] In this invention, the server includes receiving means for acquiring voice information from the user, interpreting means for analyzing text information and identifying the user's request, and adapting means for providing information according to the user's situation in the environment. This makes it possible to provide optimal information and instructions tailored to the situation to the elderly and people who need assistance.

[0114] "Receiving means" refers to a device or function that acquires voice information from the user.

[0115] "Conversion means" refers to a device or function that converts acquired audio information into text information.

[0116] "Interpretation means" refers to a device or function that analyzes textual information and identifies the user's request.

[0117] "Collection means" refers to a device or function that collects relevant information based on an identified request.

[0118] "Generating means" refers to a device or function that generates a response based on collected information.

[0119] "Speech synthesis means" refers to a device or function that converts a generated response into information that can be input or output as speech.

[0120] "Presentation means" refers to a device or function that presents information capable of audio input and output to the user.

[0121] "Recording means" refers to a device or function that records and retains the history of interactions with the user.

[0122] "Adaptive means" refers to a device or function that provides information according to the user's situation in the environment.

[0123] This system is a conversational artificial intelligence designed to support the daily lives of the elderly and people with disabilities. Specifically, it interacts with users through devices such as smart glasses and provides information via voice. The device acquires voice information transmitted by the user through a microphone. The server converts the acquired voice information into text and analyzes the text using natural language processing technology to identify the user's requests and intentions.

[0124] Once the analysis is complete, relevant information is collected from external information resources based on the request. Based on the collected information, the server generates an appropriate response and converts it into audio input / outputable information. This audio information is provided to the user through the terminal. The system also has the function to provide information appropriately, taking into account the user's situation in the environment.

[0125] The system uses the speech_recognition library for speech recognition and the pyttsx3 library for speech synthesis. Hardware-wise, the microphone and speaker built into the smart glasses play a major role.

[0126] For example, when a user asks, "Can you give me some health advice?", the system provides practical advice via voice, such as, "Make sure you stay well-hydrated." An example of a specific prompt using the generative AI model would be, "Please give me 10 pieces of health advice suitable for supporting the daily lives of elderly people."

[0127] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0128] Step 1:

[0129] The user gives voice commands to the device. The voice information is acquired through the microphone. The input is the user's voice, and the output is digital voice data. The device sends this voice data to the server.

[0130] Step 2:

[0131] The server converts received audio data into text information. The input is digital audio data, and the output is text data. The server performs speech recognition using the speech_recognition library. In this process, the user's speech is converted into text data.

[0132] Step 3:

[0133] The server analyzes text data to identify the user's intent. The input is text data, and the output is information about the user's request. Natural language processing techniques are used to interpret the user's intent and determine the specific request and expected response.

[0134] Step 4:

[0135] The server collects necessary information from external information resources based on user requests. The input is information related to the user's request, and the output is information corresponding to that request. Generative AI models and APIs are used to collect data necessary to provide appropriate information.

[0136] Step 5:

[0137] The server generates a response based on the collected information. The input is the collected information, and the output is the generated response data. Based on the collected information, it decides what kind of response to send and constructs the response in text format.

[0138] Step 6:

[0139] The server converts the generated response into audio information. The input is a text-based response, and the output is audio data. The pyttsx3 library is used to synthesize the text data into speech and generate the audio data.

[0140] Step 7:

[0141] The terminal plays audio data received from the server and delivers a response to the user. The input is audio data from the server, and the output is audio from the speaker. The user receives responses from the system in real time.

[0142] Step 8:

[0143] The server records and maintains the history of interactions with the user. The input is the entire conversation, and the output is the recorded conversation history data. This information is used to personalize responses in the future.

[0144] 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.

[0145] This invention relates to a conversational artificial intelligence system that incorporates an emotion engine to recognize the user's emotions and engage in dialogue accordingly. This system can operate through a voice interface and possesses the flexibility to provide appropriate information and responses to the user in various situations.

[0146] System configuration:

[0147] This system mainly consists of the following components: input means, conversion means, analysis means, information acquisition means including an emotion engine, response generation means, speech synthesis means, output means, and recording means.

[0148] 1. Voice input acquisition and conversion:

[0149] The device acquires the voice spoken by the user and converts it into a digital format. The voice data is then transferred to a server and converted into text using speech recognition technology.

[0150] 2. Intention Analysis and Emotion Recognition:

[0151] The server analyzes the converted text using natural language processing technology to identify the user's intent. It also uses a built-in emotion engine to analyze emotions from the audio data and text. The emotion engine estimates the user's emotional state based on factors such as voice tone, speed, and word choice.

[0152] 3. Information acquisition and response generation:

[0153] The server retrieves necessary information from an external database based on the identified intent and emotional state. Then, using a response generation mechanism, it generates an appropriate response in a natural conversational format based on the retrieved information. The generated response is then adjusted to match the user's emotions.

[0154] 4. Speech synthesis and output:

[0155] The server synthesizes the generated text response into speech and sends it to the terminal as audio data. The terminal then presents the response to the user via its speaker.

[0156] 5. Recording of dialogue history:

[0157] The server records the history of interactions, including detailed and comprehensive emotional data, and uses this information to improve future services and response quality.

[0158] Specific example:

[0159] For example, if a user asks a question about a news item they are interested in and says in a slightly anxious tone, "I'm worried about the recent economic situation," the device will pick up the voice and use its emotion engine to detect that anxiety is present. Based on this, the server will retrieve economic news and prepare a response that is tailored to provide reassurance and calmness. For example, it might include reassuring elements in the response, such as, "Many experts analyze that the recent market fluctuations are temporary." Finally, the device will output the response as audio and deliver it to the user.

[0160] In this way, dialogue that takes the user's emotions into account becomes possible, enabling more human-like communication.

[0161] The following describes the processing flow.

[0162] Step 1:

[0163] The user inputs questions and instructions by voice into the smart device. This voice is captured by the microphone built into the device.

[0164] Step 2:

[0165] The terminal converts the audio data into a digital format and transmits it to the server in real time via a secure protocol. At this time, the audio data undergoes pre-processing such as noise reduction.

[0166] Step 3:

[0167] The server analyzes the received audio data using a speech recognition engine and converts it into text data. This conversion ensures that the audio information is accurately translated into written information.

[0168] Step 4:

[0169] The server analyzes the text obtained using natural language processing technology to identify the user's intent and requests. Additionally, a built-in emotion engine references the text and original audio data to recognize the user's emotions.

[0170] Step 5:

[0171] The server retrieves the necessary information from an external database based on the request and emotional state. For example, this could include data from external resources such as news or weather information.

[0172] Step 6:

[0173] The server generates a response that is appropriate to the user's emotions based on the information it has acquired. This response generation utilizes the results of sentiment analysis to incorporate language and tone that match the user's emotions.

[0174] Step 7:

[0175] The server converts the generated response into speech data using a speech synthesis engine. This speech data is then tone-adjusted to reflect the user's emotions.

[0176] Step 8:

[0177] The terminal plays audio data received from the server using its speaker and presents an audio response to the user. This response allows the user to receive information that takes emotions into consideration.

[0178] Step 9:

[0179] The server records the content of each session's dialogue and detected emotion data as a history, and updates this as a database to make future communications more personalized and accurate.

[0180] (Example 2)

[0181] 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".

[0182] Conventional conversational artificial intelligence systems focus on identifying the user's intent from their voice, but they are insufficient at identifying emotions and generating responses accordingly. Therefore, it is difficult to have conversations that are empathetic to the user's emotions, and achieving natural, human-like communication remains a challenge.

[0183] 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.

[0184] In this invention, the server includes an input means for acquiring the user's voice information, a conversion means for converting it into text information, and an analysis means for identifying the user's intentions and emotions. This makes it possible to generate responses that take into account not only the user's intentions but also their emotions.

[0185] "User" refers to a person who interacts with the system.

[0186] "Voice information" refers to the voice signals emitted by the user, which are the subject of processing by the system.

[0187] "Input means" refers to devices or mechanisms for acquiring voice information from the user.

[0188] "Textual information" refers to text data obtained as a result of converting audio information.

[0189] "Conversion means" refers to the technology or device used to perform the process of converting audio information into text information.

[0190] "Analysis means" refers to techniques used to identify a user's intentions and emotions from textual information.

[0191] "Emotions" refers to the mental state inferred from the tone of the user's voice or text, as well as the words they choose.

[0192] A "generative AI model" is a pre-trained artificial intelligence algorithm used to generate responses from underlying data.

[0193] "Response generation means" refers to technologies and devices for generating natural-sounding, conversational responses, taking into account collected information and the user's emotions.

[0194] "Audio format" refers to the data format of the audio signal held by the generated response.

[0195] "Speech synthesis means" refers to technologies and devices for converting text information into speech format.

[0196] "Output means" refers to devices or mechanisms that provide information to the user using voice-based responses.

[0197] "Dialogue history" refers to a record of past interactions between the user and the system.

[0198] "Emotional data" refers to information obtained as a result of analyzing the user's emotions.

[0199] As an example of how to carry out the present invention, a system is provided in which an emotion engine is incorporated into a conversational artificial intelligence system to recognize the user's emotions and engage in dialogue accordingly. This system operates through a voice interface and provides appropriate information and responses to the user in various situations.

[0200] First, the user gives instructions to the system by voice. The terminal acquires this voice using an input device and converts it into a digital voice signal. Then, this signal is sent to a server and converted into text data using speech recognition technology (for example, a common speech recognition tool or service).

[0201] Next, the server processes the text data using analysis tools and identifies the user's intentions and emotions using natural language processing technology. This analysis utilizes a generative AI model to infer the user's intentions and emotions based on the input data. The emotion engine estimates the user's emotional state from factors such as tone of voice, speed, and word choice, and responds accordingly.

[0202] After analysis, the server uses information acquisition tools to collect necessary data from an external database. This database contains several information sources (e.g., weather information and news websites). The collected information is processed by a response generation tool using a generative AI model, which generates a natural conversational response that matches the user's intent and emotions. The response is appropriately adjusted according to the user's emotions.

[0203] The generated response is then converted into a format suitable for audio output by a speech synthesis system. This audio data is sent to the terminal and conveyed to the user through the speaker. Finally, the server stores the dialogue history and sentiment data using recording means, and this data is used to improve future services and response quality.

[0204] As a concrete example, suppose a user asks in an anxious tone, "What should I wear today?" The server analyzes this question using an emotion engine and recognizes the user's anxiety. Then, using a weather API, it generates a response saying, "The forecast is for sunshine today, so a light jacket would be good. Please go out with peace of mind," and delivers it to the user via voice.

[0205] An example of a prompt statement can be entered as follows:

[0206] "We have detected that the user's emotion is 'anxiety.' Please prepare a response regarding the weather forecast, including language that will alleviate the user's anxiety."

[0207] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0208] Step 1:

[0209] The user speaks to the system using their voice. This voice signal becomes the input. The terminal uses its built-in microphone to acquire this voice signal and convert it into a digital format. For example, it converts analog audio into PCM digital data and sends the output to the server.

[0210] Step 2:

[0211] The server, upon receiving the digital audio data, converts it into text using speech recognition technology. This process employs a speech recognition algorithm to analyze the input audio and output it as a string of characters. This text information is then used for further analysis.

[0212] Step 3:

[0213] The server analyzes the converted text information using natural language processing techniques. Here, text information is received as input, and contextual understanding and intent analysis are performed as data processing. A generative AI model is used to identify the user's intent and emotions, and the results are output.

[0214] Step 4:

[0215] Based on the identified intent and emotion, the server retrieves the necessary information from an external database. This step involves querying an external API and gathering relevant information. This retrieved information becomes the input for the next response generation step.

[0216] Step 5:

[0217] The server utilizes a generative AI model to generate natural, conversational responses using a response generation mechanism. This process takes into account identified emotions and makes adjustments tailored to the user. For example, this might include adjusting the language to create a sense of reassurance. The generated text response is then output.

[0218] Step 6:

[0219] The server converts the generated response text into speech data using speech synthesis technology. It receives the text response as input and outputs it as an audio file. This audio data is then sent to the terminal.

[0220] Step 7:

[0221] The device that receives the transmitted audio data plays the audio through its speaker, conveying the response to the user. This ensures that information is provided in a way that is easy for the user to understand.

[0222] Step 8:

[0223] After the conversation ends, the server records the conversation history and sentiment data and saves it to a database. This record plays an important role as it will be used for future system improvements and personalization.

[0224] (Application Example 2)

[0225] 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 device 14 will be referred to as the "terminal."

[0226] The development of conversational artificial intelligence systems that take user emotions into consideration has been a long-standing requirement. Conventional technologies can only provide simple responses to voice input, and it has been difficult to recognize user emotions and provide appropriate information and suggestions accordingly. This invention aims to realize a system that can facilitate smoother communication with senior users and provide a sense of security, particularly in caregiving settings.

[0227] 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.

[0228] In this invention, the server includes means for acquiring voice input from a user, means for converting the voice input into text data, means for analyzing the text data to identify the user's emotions, means for collecting relevant information based on the identified emotions and requests, means for generating a response tailored to the user's emotional state based on the collected information, means for converting the generated response into data that can be output as voice, means for presenting the voice-outputtable data to the user, and means for recording and maintaining the user's dialogue history and emotional data. This enables the provision of appropriate information that matches the user's emotions.

[0229] A "user" is a person who uses an interactive artificial intelligence system to acquire information or communicate through voice input.

[0230] "Voice input" refers to the words or voices that the user speaks to the system.

[0231] "Text data" refers to character information converted from voice input, and is a format used for system analysis and processing.

[0232] "Emotions" refer to an internal state that can be inferred from the user's tone of voice, word choice, and other factors.

[0233] "Relevant information" refers to necessary data that the system collects from external sources based on the user's requests and feelings.

[0234] A "response" is a verbal or written response that a system generates and outputs based on user input.

[0235] "Audio-outputtable data" refers to data that has been converted into a format that allows the generated response to be presented to the user as audio.

[0236] "Dialogue history" refers to recorded information of past conversations between the system and the user, and is used to improve response quality.

[0237] "Emotional data" refers to records of the user's emotional state, and is useful information for improving the system and generating appropriate responses.

[0238] To implement the invention, the conversational artificial intelligence system is configured using a server and a user terminal. The voice spoken by the user is collected by the terminal's voice input device. This voice data is converted into a digital signal and transmitted to the server via the network.

[0239] The server uses speech recognition software to convert audio data into text data. Google® Cloud Speech-to-Text API is used as this software. The converted text is then analyzed within the server using natural language processing technology. OpenAI® generative AI models are used in this process to identify the user's requests and intentions.

[0240] Furthermore, emotion recognition utilizes an emotion analysis engine, leveraging Amazon's Comprehend and Azure's Text Analytics API to infer user emotions from text data and voice tone. The server then retrieves necessary information from an external data store based on the user's requests and emotional state. This data retrieval is performed via database access through a REST API.

[0241] The server generates responses adapted to the user's emotional state based on the collected information. The generated text responses are converted into audio data using the Google Cloud Text-to-Speech API and sent to the device over the network. Finally, they are presented to the user as audio through the device's speaker. Dialogue history and emotional data are securely recorded and stored by the server and used to improve future responses.

[0242] As a specific example, when a caregiver wearing smart glasses interacts with a user, if the user expresses anxiety such as "I haven't been sleeping well lately," the system detects the anxiety and provides a reassuring response such as, "How about listening to relaxing music before bed to improve your sleep environment?"

[0243] An example of a prompt message for a generative AI model might be, "The user is feeling anxious. We would like to offer reassuring suggestions regarding sleep."

[0244] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0245] Step 1:

[0246] The user provides voice input to the device. The device acquires this voice data via its microphone and converts it into a digital format. This voice data is then transmitted to a server via a communication network.

[0247] Step 2:

[0248] The server converts the received audio data into text data using the Google Cloud Speech-to-Text API. This process analyzes the audio signal and generates output as text information.

[0249] Step 3:

[0250] The server analyzes the obtained text data using natural language processing techniques. Specifically, it uses OpenAI's generative AI model to identify user requests from the text and extract important information. At this stage, it identifies information such as subjects and objects from the text.

[0251] Step 4:

[0252] The server uses an emotion analysis engine to identify the user's emotional state from text data and voice tone. Specifically, it utilizes Amazon Comprehend and Azure's Text Analytics API to analyze emotion indicators in the input data and output the results.

[0253] Step 5:

[0254] Based on the identified request and sentiment state, the server retrieves relevant information from an external data store via a REST API. This process collects relevant links and article data to satisfy the user's specific request and provides them as output.

[0255] Step 6:

[0256] The server generates an appropriate response to the user based on the collected information and emotional state. In the generation process, a prompt sentence (e.g., "The user is feeling anxious. We want to offer reassuring suggestions regarding sleep") is input to the generation AI model, and a natural conversational text response is obtained.

[0257] Step 7:

[0258] The generated text response is converted to speech by the server using the Google Cloud Text-to-Speech API. In this step, the text is processed into an audio file and output as data.

[0259] Step 8:

[0260] Finally, the device receives the audio data sent from the server and presents it to the user through its built-in speaker. In this step, the user can directly hear the audio response.

[0261] Step 9:

[0262] The server securely records and stores all conversation history and sentiment data. In this step, the acquired data is stored in a database for later analysis and response improvement.

[0263] 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.

[0264] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0265] 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.

[0266] [Second Embodiment]

[0267] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0268] 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.

[0269] 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).

[0270] 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.

[0271] 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.

[0272] 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).

[0273] 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.

[0274] 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.

[0275] 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.

[0276] 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.

[0277] 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.

[0278] 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".

[0279] This invention is a "conversational artificial intelligence system" that provides information through interactive dialogue with the user and can be implemented in various forms. This system operates on a smart device and is always in standby mode. The system receives the voice input when the user speaks to it.

[0280] System Operation Overview:

[0281] 1. Acquisition of voice input

[0282] The terminal acquires voice input from the user through the microphone and processes the data as digital signals.

[0283] 2. Conversion from Voice to Text

[0284] The server uses a speech recognition engine to convert the voice data into text data. This conversion is performed in real time, and the content of the user's speech is recognized as text.

[0285] 3. Analysis of Intent

[0286] The server analyzes the text data using natural language processing techniques. Through this analysis, it is possible to grasp the user's requests and intentions.

[0287] 4. Acquisition of Information and Generation of Responses

[0288] The server obtains relevant information from an external information database as needed based on the analysis results. Based on this information, the system generates an appropriate response.

[0289] 5. Generation and Output of Voice Responses

[0290] The server converts the generated response into voice data using speech synthesis technology. The terminal plays this voice data from the speaker and provides a voice response to the user.

[0291] 6. Recording of Conversation History

[0292] The server records the content of the conversation with the user as a history and retains this for each user. This history is used to personalize the next response.

[0293] Specific Example:

[0294] If the user asks, "What time is tomorrow's meeting?", the device will capture the audio, and the server will convert the audio to text. The server will then analyze the text to understand the user's request as "to confirm the meeting time." The server will access the schedule database, retrieve the time of the relevant meeting, and generate a response such as, "Tomorrow's meeting is at 2 PM." This response will be converted back into audio and transmitted to the user by the device.

[0295] Thus, the system of the present invention can provide information immediately in response to the user's needs, enabling more efficient daily support.

[0296] The following describes the processing flow.

[0297] Step 1:

[0298] The user voice-inputs a question or instruction into the smart device. The voice input is captured by the device's microphone.

[0299] Step 2:

[0300] The terminal converts the acquired audio data into a digital signal and transmits it to the server via a secure channel. During this process, pre-processing such as noise reduction is applied to improve the quality of the audio data.

[0301] Step 3:

[0302] The server analyzes the received audio data using a speech recognition engine and converts it into text data. This process ensures that the speech is recognized as an accurate text-based command.

[0303] Step 4:

[0304] The server uses natural language processing techniques to analyze the content of text data and identify the user's intent and requests. This includes keyword extraction and semantic analysis.

[0305] Step 5:

[0306] The server obtains the necessary information from an external database or API based on the identified request. For example, weather information or schedule information may be requested.

[0307] Step 6:

[0308] Based on the obtained information, the server generates a response message for the user. This response message is assembled in natural language so that the user can easily understand it.

[0309] Step 7:

[0310] The server converts the response message into audio data using a text-to-speech tool. Through this process, the text response becomes in a format that can be output as audio.

[0311] Step 8:

[0312] The terminal provides a response by playing the audio data received from the server to the user through a speaker.

[0313] Step 9:

[0314] The server records this interaction as an interaction history and retains it as data for making subsequent responses more personalized.

[0315] (Example 1)

[0316] Next, Example 1 will be described. In the following description, the data processing device 12 is referred to as the "server", and the smart glasses 214 are referred to as the "terminal".

[0317] Many modern voice response systems are specialized in providing a limited range of information and struggle to generate flexible responses that meet the diverse needs of users. Furthermore, they have the challenge of not being able to fully utilize the user's past conversation history, thus failing to provide personalized, interactive dialogue.

[0318] 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.

[0319] In this invention, the server includes means for acquiring voice input through an input device, means for analyzing character data with a language processing device, and a data acquisition device for collecting information from an external database. This enables the generation of flexible responses to diverse requests and the provision of personalized information to each user.

[0320] An "input / output device" is a device that acquires, records, and provides voice input from the user to the system.

[0321] A "speech recognition device" is a device that includes technology for converting acquired speech data into text data.

[0322] A "language processing device" is a device that analyzes text data and utilizes natural language processing technology to identify the user's intentions and requests.

[0323] A "data acquisition device" is a device that collects necessary information from external or internal sources based on an identified request.

[0324] A "reaction generation device" is a device that generates an appropriate response as text based on collected information.

[0325] A "synthesis device" is a device that converts generated text responses into audio data so that they can be output as speech.

[0326] A "sound playback device" is a device such as a speaker that presents synthesized sound data to the user.

[0327] A "memory device" is a device used to record and store the history of interactions with the user.

[0328] A "personalization device" is a device that utilizes recorded dialogue history to generate subsequent responses, enabling personalized interactions.

[0329] This invention relates to a conversational artificial intelligence system that provides information through interactive dialogue with the user. The system performs a series of processes to receive voice input, analyze it, and output an appropriate voice response.

[0330] First, the terminal acquires voice input through the microphone. The acquired voice data is converted into a digital signal and sent to the server. The server uses a speech recognition engine (for example, an API provided by a common speech recognition service provider) to convert the voice data into text data. At this stage, high-precision real-time conversion of the voice input is required.

[0331] Next, the server analyzes the text data using natural language processing techniques (utilizing open-source language analysis tools such as NLTK and spaCy). This analysis recognizes the user's request and identifies their intent. Based on the analysis results, it retrieves information from an external information database (e.g., a cloud-based data management service).

[0332] Based on the acquired data, the server generates an appropriate response. This response is prepared as text data and converted into speech data using a speech synthesis engine (for example, an API provided by a common speech synthesis service provider). The terminal then plays this speech data through its speaker, delivering the answer to the user.

[0333] Furthermore, the server records the history of interactions with the user and uses this information to personalize responses in the future. This allows the system to provide increasingly efficient and personalized interactions.

[0334] For example, if a user asks, "What's the weather like today?", the device will capture this voice message. The server will convert the voice into text data and parse it as a request to provide weather information. It will then access a weather information database and generate a response such as, "Today's weather is sunny, and the temperature is 20 degrees Celsius," which it will then communicate to the user via voice. This system aims to respond quickly and accurately to a variety of everyday information requests.

[0335] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0336] Step 1:

[0337] The terminal acquires the user's voice through a microphone. This input voice is received as an analog signal and converted into a digital signal by the terminal's digital signal processor. The output of this process is digital voice data that is sent to the server.

[0338] Step 2:

[0339] The server receives digital audio data and converts it into text data using a speech recognition engine. Specifically, it calls a speech recognition API to parse the audio data and extract the corresponding strings. The input for this step is digital audio data, and the output is the parsed text data.

[0340] Step 3:

[0341] The server analyzes the parsed text data using natural language processing techniques. This procedure uses language processing libraries to analyze the sentence structure and identify the user's intent. The input is text data, and the output is a data structure representing the user's request and intent.

[0342] Step 4:

[0343] The server retrieves relevant information from the database based on the user's intent. It also calls external APIs as needed to access various cloud services and data repositories. The input in this process is a data structure representing the user's request, and the output is the requested information data.

[0344] Step 5:

[0345] The server generates a response based on the information it has acquired. This process involves selecting a response template based on conditions and embedding the actual data to create the response text. The input is informational data, and the output is the response text intended for the user.

[0346] Step 6:

[0347] The server converts the response text into speech data using a speech synthesis engine. This is done using a speech synthesis API, which obtains natural-sounding speech data from the text data. The input is the response text, and the output is the synthesized speech data.

[0348] Step 7:

[0349] The device receives the synthesized audio data and plays it back to the user through its speaker. Specifically, it uses an audio playback module for output. In this case, the input is audio data, and the output is the audio the user hears.

[0350] Step 8:

[0351] The server records the user interaction history and adds it to the training data for future interactions. This record includes utterances and system responses, and is used to provide personalized interactions. The input is the interaction history data, and the output is an updated history database.

[0352] (Application Example 1)

[0353] 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."

[0354] In the daily lives of the elderly and those requiring support, there is a need to provide necessary information quickly and accurately while appropriately responding to changes and circumstances around them. Especially when daily living support is needed, timely information and instructions are essential for improving the quality of life. However, conventional technologies do not take into account the user's surroundings and circumstances, making it difficult to provide the optimal response.

[0355] 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.

[0356] In this invention, the server includes receiving means for acquiring voice information from the user, interpreting means for analyzing text information and identifying the user's request, and adapting means for providing information according to the user's situation in the environment. This makes it possible to provide optimal information and instructions tailored to the situation to the elderly and people who need assistance.

[0357] "Receiving means" refers to a device or function that acquires voice information from the user.

[0358] "Conversion means" refers to a device or function that converts acquired audio information into text information.

[0359] "Interpretation means" refers to a device or function that analyzes textual information and identifies the user's request.

[0360] "Collection means" refers to a device or function that collects relevant information based on an identified request.

[0361] "Generating means" refers to a device or function that generates a response based on collected information.

[0362] "Speech synthesis means" refers to a device or function that converts a generated response into information that can be input or output as speech.

[0363] "Presentation means" refers to a device or function that presents information capable of audio input and output to the user.

[0364] "Recording means" refers to a device or function that records and retains the history of interactions with the user.

[0365] "Adaptive means" refers to a device or function that provides information according to the user's situation in the environment.

[0366] This system is a conversational artificial intelligence designed to support the daily lives of the elderly and people with disabilities. Specifically, it interacts with users through devices such as smart glasses and provides information via voice. The device acquires voice information transmitted by the user through a microphone. The server converts the acquired voice information into text and analyzes the text using natural language processing technology to identify the user's requests and intentions.

[0367] Once the analysis is complete, relevant information is collected from external information resources based on the request. Based on the collected information, the server generates an appropriate response and converts it into audio input / outputable information. This audio information is provided to the user through the terminal. The system also has the function to provide information appropriately, taking into account the user's situation in the environment.

[0368] The system uses the speech_recognition library for speech recognition and the pyttsx3 library for speech synthesis. Hardware-wise, the microphone and speaker built into the smart glasses play a major role.

[0369] For example, when a user asks, "Can you give me some health advice?", the system provides practical advice via voice, such as, "Make sure you stay well-hydrated." An example of a specific prompt using the generative AI model would be, "Please give me 10 pieces of health advice suitable for supporting the daily lives of elderly people."

[0370] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0371] Step 1:

[0372] The user gives voice commands to the device. The voice information is acquired through the microphone. The input is the user's voice, and the output is digital voice data. The device sends this voice data to the server.

[0373] Step 2:

[0374] The server converts received audio data into text information. The input is digital audio data, and the output is text data. The server performs speech recognition using the speech_recognition library. In this process, the user's speech is converted into text data.

[0375] Step 3:

[0376] The server analyzes text data to identify the user's intent. The input is text data, and the output is information about the user's request. Natural language processing techniques are used to interpret the user's intent and determine the specific request and expected response.

[0377] Step 4:

[0378] The server collects necessary information from external information resources based on user requests. The input is information related to the user's request, and the output is information corresponding to that request. Generative AI models and APIs are used to collect data necessary to provide appropriate information.

[0379] Step 5:

[0380] The server generates a response based on the collected information. The input is the collected information, and the output is the generated response data. Based on the collected information, it decides what kind of response to send and constructs the response in text format.

[0381] Step 6:

[0382] The server converts the generated response into audio information. The input is a text-based response, and the output is audio data. The pyttsx3 library is used to synthesize the text data into speech and generate the audio data.

[0383] Step 7:

[0384] The terminal plays audio data received from the server and delivers a response to the user. The input is audio data from the server, and the output is audio from the speaker. The user receives responses from the system in real time.

[0385] Step 8:

[0386] The server records and maintains the history of interactions with the user. The input is the entire conversation, and the output is the recorded conversation history data. This information is used to personalize responses in the future.

[0387] 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.

[0388] This invention relates to a conversational artificial intelligence system that incorporates an emotion engine to recognize the user's emotions and engage in dialogue accordingly. This system can operate through a voice interface and possesses the flexibility to provide appropriate information and responses to the user in various situations.

[0389] System configuration:

[0390] This system mainly consists of the following components: input means, conversion means, analysis means, information acquisition means including an emotion engine, response generation means, speech synthesis means, output means, and recording means.

[0391] 1. Voice input acquisition and conversion:

[0392] The device acquires the voice spoken by the user and converts it into a digital format. The voice data is then transferred to a server and converted into text using speech recognition technology.

[0393] 2. Intention Analysis and Emotion Recognition:

[0394] The server analyzes the converted text using natural language processing technology to identify the user's intent. It also uses a built-in emotion engine to analyze emotions from the audio data and text. The emotion engine estimates the user's emotional state based on factors such as voice tone, speed, and word choice.

[0395] 3. Information acquisition and response generation:

[0396] The server retrieves necessary information from an external database based on the identified intent and emotional state. Then, using a response generation mechanism, it generates an appropriate response in a natural conversational format based on the retrieved information. The generated response is then adjusted to match the user's emotions.

[0397] 4. Speech synthesis and output:

[0398] The server synthesizes the generated text response into speech and sends it to the terminal as audio data. The terminal then presents the response to the user via its speaker.

[0399] 5. Recording of dialogue history:

[0400] The server records the history of interactions, including detailed and comprehensive emotional data, and uses this information to improve future services and response quality.

[0401] Specific example:

[0402] For example, if a user asks a question about a news item they are interested in and says in a slightly anxious tone, "I'm worried about the recent economic situation," the device will pick up the voice and use its emotion engine to detect that anxiety is present. Based on this, the server will retrieve economic news and prepare a response that is tailored to provide reassurance and calmness. For example, it might include reassuring elements in the response, such as, "Many experts analyze that the recent market fluctuations are temporary." Finally, the device will output the response as audio and deliver it to the user.

[0403] In this way, dialogue that takes the user's emotions into account becomes possible, enabling more human-like communication.

[0404] The following describes the processing flow.

[0405] Step 1:

[0406] The user inputs questions and instructions by voice into the smart device. This voice is captured by the microphone built into the device.

[0407] Step 2:

[0408] The terminal converts the audio data into a digital format and transmits it to the server in real time via a secure protocol. At this time, the audio data undergoes pre-processing such as noise reduction.

[0409] Step 3:

[0410] The server analyzes the received audio data using a speech recognition engine and converts it into text data. This conversion ensures that the audio information is accurately translated into written information.

[0411] Step 4:

[0412] The server analyzes the text obtained using natural language processing technology to identify the user's intent and requests. Additionally, a built-in emotion engine references the text and original audio data to recognize the user's emotions.

[0413] Step 5:

[0414] The server retrieves the necessary information from an external database based on the request and emotional state. For example, this could include data from external resources such as news or weather information.

[0415] Step 6:

[0416] The server generates a response that is appropriate to the user's emotions based on the information it has acquired. This response generation utilizes the results of sentiment analysis to incorporate language and tone that match the user's emotions.

[0417] Step 7:

[0418] The server converts the generated response into speech data using a speech synthesis engine. This speech data is then tone-adjusted to reflect the user's emotions.

[0419] Step 8:

[0420] The terminal plays audio data received from the server using its speaker and presents an audio response to the user. This response allows the user to receive information that takes emotions into consideration.

[0421] Step 9:

[0422] The server records the content of each session's dialogue and detected emotion data as a history, and updates this as a database to make future communications more personalized and accurate.

[0423] (Example 2)

[0424] 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".

[0425] Conventional conversational artificial intelligence systems focus on identifying the user's intent from their voice, but they are insufficient at identifying emotions and generating responses accordingly. Therefore, it is difficult to have conversations that are empathetic to the user's emotions, and achieving natural, human-like communication remains a challenge.

[0426] 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.

[0427] In this invention, the server includes an input means for acquiring the user's voice information, a conversion means for converting it into text information, and an analysis means for identifying the user's intentions and emotions. This makes it possible to generate responses that take into account not only the user's intentions but also their emotions.

[0428] "User" refers to a person who interacts with the system.

[0429] "Voice information" refers to the voice signals emitted by the user, which are the subject of processing by the system.

[0430] "Input means" refers to devices or mechanisms for acquiring voice information from the user.

[0431] "Textual information" refers to text data obtained as a result of converting audio information.

[0432] "Conversion means" refers to the technology or device used to perform the process of converting audio information into text information.

[0433] "Analysis means" refers to techniques used to identify a user's intentions and emotions from textual information.

[0434] "Emotions" refers to the mental state inferred from the tone of the user's voice or text, as well as the words they choose.

[0435] A "generative AI model" is a pre-trained artificial intelligence algorithm used to generate responses from underlying data.

[0436] "Response generation means" refers to technologies and devices for generating natural-sounding, conversational responses, taking into account collected information and the user's emotions.

[0437] "Audio format" refers to the data format of the audio signal held by the generated response.

[0438] "Speech synthesis means" refers to technologies and devices for converting text information into speech format.

[0439] "Output means" refers to devices or mechanisms that provide information to the user using voice-based responses.

[0440] "Dialogue history" refers to a record of past interactions between the user and the system.

[0441] "Emotional data" refers to information obtained as a result of analyzing the user's emotions.

[0442] As an example of how to carry out the present invention, a system is provided in which an emotion engine is incorporated into a conversational artificial intelligence system to recognize the user's emotions and engage in dialogue accordingly. This system operates through a voice interface and provides appropriate information and responses to the user in various situations.

[0443] First, the user gives instructions to the system by voice. The terminal acquires this voice using an input device and converts it into a digital voice signal. Then, this signal is sent to a server and converted into text data using speech recognition technology (for example, a common speech recognition tool or service).

[0444] Next, the server processes the text data using analysis tools and identifies the user's intentions and emotions using natural language processing technology. This analysis utilizes a generative AI model to infer the user's intentions and emotions based on the input data. The emotion engine estimates the user's emotional state from factors such as tone of voice, speed, and word choice, and responds accordingly.

[0445] After analysis, the server uses information acquisition tools to collect necessary data from an external database. This database contains several information sources (e.g., weather information and news websites). The collected information is processed by a response generation tool using a generative AI model, which generates a natural conversational response that matches the user's intent and emotions. The response is appropriately adjusted according to the user's emotions.

[0446] The generated response is then converted into a format suitable for audio output by a speech synthesis system. This audio data is sent to the terminal and conveyed to the user through the speaker. Finally, the server stores the dialogue history and sentiment data using recording means, and this data is used to improve future services and response quality.

[0447] As a concrete example, suppose a user asks in an anxious tone, "What should I wear today?" The server analyzes this question using an emotion engine and recognizes the user's anxiety. Then, using a weather API, it generates a response saying, "The forecast is for sunshine today, so a light jacket would be good. Please go out with peace of mind," and delivers it to the user via voice.

[0448] An example of a prompt statement can be entered as follows:

[0449] "We have detected that the user's emotion is 'anxiety.' Please prepare a response regarding the weather forecast, including language that will alleviate the user's anxiety."

[0450] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0451] Step 1:

[0452] The user speaks to the system using their voice. This voice signal becomes the input. The terminal uses its built-in microphone to acquire this voice signal and convert it into a digital format. For example, it converts analog audio into PCM digital data and sends the output to the server.

[0453] Step 2:

[0454] The server, upon receiving the digital audio data, converts it into text using speech recognition technology. This process employs a speech recognition algorithm to analyze the input audio and output it as a string of characters. This text information is then used for further analysis.

[0455] Step 3:

[0456] The server analyzes the converted text information using natural language processing techniques. Here, text information is received as input, and contextual understanding and intent analysis are performed as data processing. A generative AI model is used to identify the user's intent and emotions, and the results are output.

[0457] Step 4:

[0458] Based on the identified intent and emotion, the server retrieves the necessary information from an external database. This step involves querying an external API and gathering relevant information. This retrieved information becomes the input for the next response generation step.

[0459] Step 5:

[0460] The server utilizes a generative AI model to generate natural, conversational responses using a response generation mechanism. This process takes into account identified emotions and makes adjustments tailored to the user. For example, this might include adjusting the language to create a sense of reassurance. The generated text response is then output.

[0461] Step 6:

[0462] The server converts the generated response text into speech data using speech synthesis technology. It receives the text response as input and outputs it as an audio file. This audio data is then sent to the terminal.

[0463] Step 7:

[0464] The device that receives the transmitted audio data plays the audio through its speaker, conveying the response to the user. This ensures that information is provided in a way that is easy for the user to understand.

[0465] Step 8:

[0466] After the conversation ends, the server records the conversation history and sentiment data and saves it to a database. This record plays an important role as it will be used for future system improvements and personalization.

[0467] (Application Example 2)

[0468] 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."

[0469] The development of conversational artificial intelligence systems that take user emotions into consideration has been a long-standing requirement. Conventional technologies can only provide simple responses to voice input, and it has been difficult to recognize user emotions and provide appropriate information and suggestions accordingly. This invention aims to realize a system that can facilitate smoother communication with senior users and provide a sense of security, particularly in caregiving settings.

[0470] 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.

[0471] In this invention, the server includes means for acquiring voice input from a user, means for converting the voice input into text data, means for analyzing the text data to identify the user's emotions, means for collecting relevant information based on the identified emotions and requests, means for generating a response tailored to the user's emotional state based on the collected information, means for converting the generated response into data that can be output as voice, means for presenting the voice-outputtable data to the user, and means for recording and maintaining the user's dialogue history and emotional data. This enables the provision of appropriate information that matches the user's emotions.

[0472] A "user" is a person who uses an interactive artificial intelligence system to acquire information or communicate through voice input.

[0473] "Voice input" refers to the words or voices that the user speaks to the system.

[0474] "Text data" refers to character information converted from voice input, and is a format used for system analysis and processing.

[0475] "Emotions" refer to an internal state that can be inferred from the user's tone of voice, word choice, and other factors.

[0476] "Relevant information" refers to necessary data that the system collects from external sources based on the user's requests and feelings.

[0477] A "response" is a verbal or written response that a system generates and outputs based on user input.

[0478] "Audio-outputtable data" refers to data that has been converted into a format that allows the generated response to be presented to the user as audio.

[0479] "Dialogue history" refers to recorded information of past conversations between the system and the user, and is used to improve response quality.

[0480] "Emotional data" refers to records of the user's emotional state, and is useful information for improving the system and generating appropriate responses.

[0481] To implement the invention, the conversational artificial intelligence system is configured using a server and a user terminal. The voice spoken by the user is collected by the terminal's voice input device. This voice data is converted into a digital signal and transmitted to the server via the network.

[0482] The server uses speech recognition software to convert audio data into text data. Google Cloud Speech-to-Text API is used as this software. The converted text is then analyzed within the server using natural language processing techniques. OpenAI's generative AI model is used in this process to identify the user's requests and intentions.

[0483] Furthermore, emotion recognition utilizes an emotion analysis engine, leveraging Amazon's Comprehend and Azure's Text Analytics API to infer user emotions from text data and voice tone. The server then retrieves necessary information from an external data store based on the user's requests and emotional state. This data retrieval is performed via database access through a REST API.

[0484] The server generates responses adapted to the user's emotional state based on the collected information. The generated text responses are converted into audio data using the Google Cloud Text-to-Speech API and sent to the device over the network. Finally, they are presented to the user as audio through the device's speaker. Dialogue history and emotional data are securely recorded and stored by the server and used to improve future responses.

[0485] As a specific example, when a caregiver wearing smart glasses interacts with a user, if the user expresses anxiety such as "I haven't been sleeping well lately," the system detects the anxiety and provides a reassuring response such as, "How about listening to relaxing music before bed to improve your sleep environment?"

[0486] An example of a prompt message for a generative AI model might be, "The user is feeling anxious. We would like to offer reassuring suggestions regarding sleep."

[0487] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0488] Step 1:

[0489] The user provides voice input to the device. The device acquires this voice data via its microphone and converts it into a digital format. This voice data is then transmitted to a server via a communication network.

[0490] Step 2:

[0491] The server converts the received audio data into text data using the Google Cloud Speech-to-Text API. This process analyzes the audio signal and generates output as text information.

[0492] Step 3:

[0493] The server analyzes the obtained text data using natural language processing techniques. Specifically, it uses OpenAI's generative AI model to identify user requests from the text and extract important information. At this stage, it identifies information such as subjects and objects from the text.

[0494] Step 4:

[0495] The server uses an emotion analysis engine to identify the user's emotional state from text data and voice tone. Specifically, it utilizes Amazon Comprehend and Azure's Text Analytics API to analyze emotion indicators in the input data and output the results.

[0496] Step 5:

[0497] Based on the identified request and sentiment state, the server retrieves relevant information from an external data store via a REST API. This process collects relevant links and article data to satisfy the user's specific request and provides them as output.

[0498] Step 6:

[0499] The server generates an appropriate response to the user based on the collected information and emotional state. In the generation process, a prompt sentence (e.g., "The user is feeling anxious. We want to offer reassuring suggestions regarding sleep") is input to the generation AI model, and a natural conversational text response is obtained.

[0500] Step 7:

[0501] The generated text response is converted to speech by the server using the Google Cloud Text-to-Speech API. In this step, the text is processed into an audio file and output as data.

[0502] Step 8:

[0503] Finally, the device receives the audio data sent from the server and presents it to the user through its built-in speaker. In this step, the user can directly hear the audio response.

[0504] Step 9:

[0505] The server securely records and stores all conversation history and sentiment data. In this step, the acquired data is stored in a database for later analysis and response improvement.

[0506] 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.

[0507] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0508] 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.

[0509] [Third Embodiment]

[0510] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0511] 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.

[0512] 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).

[0513] 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.

[0514] 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.

[0515] 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).

[0516] 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.

[0517] 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.

[0518] 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.

[0519] 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.

[0520] 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.

[0521] 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".

[0522] This invention is a "conversational artificial intelligence system" that provides information through interactive dialogue with the user and can be implemented in various forms. This system operates on a smart device and is always in standby mode. The system receives the voice input when the user speaks to it.

[0523] System Operation Overview:

[0524] 1. Acquisition of voice input

[0525] The terminal acquires voice input from the user through a microphone and processes that data as a digital signal.

[0526] 2. Converting speech to text

[0527] The server uses a speech recognition engine to convert speech data into text data. This conversion is performed in real time, and the user's speech is recognized as text.

[0528] 3. Analysis of intent

[0529] The server analyzes text data using natural language processing technology. This analysis makes it possible to understand the user's requests and intentions.

[0530] 4. Information Acquisition and Response Generation

[0531] Based on the analysis results, the server retrieves relevant information from an external information database as needed. Based on this information, the system generates an appropriate response.

[0532] 5. Generation and output of voice responses

[0533] The server converts the generated response into audio data using speech synthesis technology. The terminal plays this audio data through its speaker, providing the user with the response in voice.

[0534] 6. Recording of dialogue history

[0535] The server records the content of interactions with users as a history and maintains this history for each user. This history is used to personalize the next response.

[0536] Specific example:

[0537] If the user asks, "What time is tomorrow's meeting?", the device will capture the audio, and the server will convert the audio to text. The server will then analyze the text to understand the user's request as "to confirm the meeting time." The server will access the schedule database, retrieve the time of the relevant meeting, and generate a response such as, "Tomorrow's meeting is at 2 PM." This response will be converted back into audio and transmitted to the user by the device.

[0538] Thus, the system of the present invention can provide information immediately in response to the user's needs, enabling more efficient daily support.

[0539] The following describes the processing flow.

[0540] Step 1:

[0541] The user voice-inputs a question or instruction into the smart device. The voice input is captured by the device's microphone.

[0542] Step 2:

[0543] The terminal converts the acquired audio data into a digital signal and transmits it to the server via a secure channel. During this process, pre-processing such as noise reduction is applied to improve the quality of the audio data.

[0544] Step 3:

[0545] The server analyzes the received audio data using a speech recognition engine and converts it into text data. This process ensures that the speech is recognized as an accurate text-based command.

[0546] Step 4:

[0547] The server uses natural language processing techniques to analyze the content of text data and identify the user's intent and requests. This includes keyword extraction and semantic analysis.

[0548] Step 5:

[0549] The server retrieves the necessary information from external databases or APIs based on the identified request. For example, weather information or schedule information may be requested.

[0550] Step 6:

[0551] The server generates a response to the user based on the information it has obtained. This response is constructed in natural language so that the user can easily understand it.

[0552] Step 7:

[0553] The server converts the response text into speech data using a speech synthesis tool. This process makes the text response available in a format that can be output as speech.

[0554] Step 8:

[0555] The terminal provides a response by playing audio data received from the server to the user through its speaker.

[0556] Step 9:

[0557] The server records this interaction as a dialogue history and retains it as data to better personalize future responses.

[0558] (Example 1)

[0559] 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."

[0560] Many modern voice response systems are specialized in providing a limited range of information and struggle to generate flexible responses that meet the diverse needs of users. Furthermore, they have the challenge of not being able to fully utilize the user's past conversation history, thus failing to provide personalized, interactive dialogue.

[0561] 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.

[0562] In this invention, the server includes means for acquiring voice input through an input device, means for analyzing character data with a language processing device, and a data acquisition device for collecting information from an external database. This enables the generation of flexible responses to diverse requests and the provision of personalized information to each user.

[0563] An "input / output device" is a device that acquires, records, and provides voice input from the user to the system.

[0564] A "speech recognition device" is a device that includes technology for converting acquired speech data into text data.

[0565] A "language processing device" is a device that analyzes text data and utilizes natural language processing technology to identify the user's intentions and requests.

[0566] A "data acquisition device" is a device that collects necessary information from external or internal sources based on an identified request.

[0567] A "reaction generation device" is a device that generates an appropriate response as text based on collected information.

[0568] A "synthesis device" is a device that converts generated text responses into audio data so that they can be output as speech.

[0569] A "sound playback device" is a device such as a speaker that presents synthesized sound data to the user.

[0570] A "memory device" is a device used to record and store the history of interactions with the user.

[0571] A "personalization device" is a device that utilizes recorded dialogue history to generate subsequent responses, enabling personalized interactions.

[0572] This invention relates to a conversational artificial intelligence system that provides information through interactive dialogue with the user. The system performs a series of processes to receive voice input, analyze it, and output an appropriate voice response.

[0573] First, the terminal acquires voice input through the microphone. The acquired voice data is converted into a digital signal and sent to the server. The server uses a speech recognition engine (for example, an API provided by a common speech recognition service provider) to convert the voice data into text data. At this stage, high-precision real-time conversion of the voice input is required.

[0574] Next, the server analyzes the text data using natural language processing techniques (utilizing open-source language analysis tools such as NLTK and spaCy). This analysis recognizes the user's request and identifies their intent. Based on the analysis results, it retrieves information from an external information database (e.g., a cloud-based data management service).

[0575] Based on the acquired data, the server generates an appropriate response. This response is prepared as text data and converted into speech data using a speech synthesis engine (for example, an API provided by a common speech synthesis service provider). The terminal then plays this speech data through its speaker, delivering the answer to the user.

[0576] Furthermore, the server records the history of interactions with the user and uses this information to personalize responses in the future. This allows the system to provide increasingly efficient and personalized interactions.

[0577] For example, if a user asks, "What's the weather like today?", the device will capture this voice message. The server will convert the voice into text data and parse it as a request to provide weather information. It will then access a weather information database and generate a response such as, "Today's weather is sunny, and the temperature is 20 degrees Celsius," which it will then communicate to the user via voice. This system aims to respond quickly and accurately to a variety of everyday information requests.

[0578] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0579] Step 1:

[0580] The terminal acquires the user's voice through a microphone. This input voice is received as an analog signal and converted into a digital signal by the terminal's digital signal processor. The output of this process is digital voice data that is sent to the server.

[0581] Step 2:

[0582] The server receives digital audio data and converts it into text data using a speech recognition engine. Specifically, it calls a speech recognition API to parse the audio data and extract the corresponding strings. The input for this step is digital audio data, and the output is the parsed text data.

[0583] Step 3:

[0584] The server analyzes the parsed text data using natural language processing techniques. This procedure uses language processing libraries to analyze the sentence structure and identify the user's intent. The input is text data, and the output is a data structure representing the user's request and intent.

[0585] Step 4:

[0586] The server retrieves relevant information from the database based on the user's intent. It also calls external APIs as needed to access various cloud services and data repositories. The input in this process is a data structure representing the user's request, and the output is the requested information data.

[0587] Step 5:

[0588] The server generates a response based on the information it has acquired. This process involves selecting a response template based on conditions and embedding the actual data to create the response text. The input is informational data, and the output is the response text intended for the user.

[0589] Step 6:

[0590] The server converts the response text into speech data using a speech synthesis engine. This is done using a speech synthesis API, which obtains natural-sounding speech data from the text data. The input is the response text, and the output is the synthesized speech data.

[0591] Step 7:

[0592] The device receives the synthesized audio data and plays it back to the user through its speaker. Specifically, it uses an audio playback module for output. In this case, the input is audio data, and the output is the audio the user hears.

[0593] Step 8:

[0594] The server records the user interaction history and adds it to the training data for future interactions. This record includes utterances and system responses, and is used to provide personalized interactions. The input is the interaction history data, and the output is an updated history database.

[0595] (Application Example 1)

[0596] 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."

[0597] In the daily lives of the elderly and those requiring support, there is a need to provide necessary information quickly and accurately while appropriately responding to changes and circumstances around them. Especially when daily living support is needed, timely information and instructions are essential for improving the quality of life. However, conventional technologies do not take into account the user's surroundings and circumstances, making it difficult to provide the optimal response.

[0598] 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.

[0599] In this invention, the server includes receiving means for acquiring voice information from the user, interpreting means for analyzing text information and identifying the user's request, and adapting means for providing information according to the user's situation in the environment. This makes it possible to provide optimal information and instructions tailored to the situation to the elderly and people who need assistance.

[0600] "Receiving means" refers to a device or function that acquires voice information from the user.

[0601] "Conversion means" refers to a device or function that converts acquired audio information into text information.

[0602] "Interpretation means" refers to a device or function that analyzes textual information and identifies the user's request.

[0603] "Collection means" refers to a device or function that collects relevant information based on an identified request.

[0604] "Generating means" refers to a device or function that generates a response based on collected information.

[0605] "Speech synthesis means" refers to a device or function that converts a generated response into information that can be input or output as speech.

[0606] "Presentation means" refers to a device or function that presents information capable of audio input and output to the user.

[0607] "Recording means" refers to a device or function that records and retains the history of interactions with the user.

[0608] "Adaptive means" refers to a device or function that provides information according to the user's situation in the environment.

[0609] This system is a conversational artificial intelligence designed to support the daily lives of the elderly and people with disabilities. Specifically, it interacts with users through devices such as smart glasses and provides information via voice. The device acquires voice information transmitted by the user through a microphone. The server converts the acquired voice information into text and analyzes the text using natural language processing technology to identify the user's requests and intentions.

[0610] Once the analysis is complete, relevant information is collected from external information resources based on the request. Based on the collected information, the server generates an appropriate response and converts it into audio input / outputable information. This audio information is provided to the user through the terminal. The system also has the function to provide information appropriately, taking into account the user's situation in the environment.

[0611] The system uses the speech_recognition library for speech recognition and the pyttsx3 library for speech synthesis. Hardware-wise, the microphone and speaker built into the smart glasses play a major role.

[0612] For example, when a user asks, "Can you give me some health advice?", the system provides practical advice via voice, such as, "Make sure you stay well-hydrated." An example of a specific prompt using the generative AI model would be, "Please give me 10 pieces of health advice suitable for supporting the daily lives of elderly people."

[0613] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0614] Step 1:

[0615] The user gives voice commands to the device. The voice information is acquired through the microphone. The input is the user's voice, and the output is digital voice data. The device sends this voice data to the server.

[0616] Step 2:

[0617] The server converts received audio data into text information. The input is digital audio data, and the output is text data. The server performs speech recognition using the speech_recognition library. In this process, the user's speech is converted into text data.

[0618] Step 3:

[0619] The server analyzes text data to identify the user's intent. The input is text data, and the output is information about the user's request. Natural language processing techniques are used to interpret the user's intent and determine the specific request and expected response.

[0620] Step 4:

[0621] The server collects necessary information from external information resources based on user requests. The input is information related to the user's request, and the output is information corresponding to that request. Generative AI models and APIs are used to collect data necessary to provide appropriate information.

[0622] Step 5:

[0623] The server generates a response based on the collected information. The input is the collected information, and the output is the generated response data. Based on the collected information, it decides what kind of response to send and constructs the response in text format.

[0624] Step 6:

[0625] The server converts the generated response into audio information. The input is a text-based response, and the output is audio data. The pyttsx3 library is used to synthesize the text data into speech and generate the audio data.

[0626] Step 7:

[0627] The terminal plays audio data received from the server and delivers a response to the user. The input is audio data from the server, and the output is audio from the speaker. The user receives responses from the system in real time.

[0628] Step 8:

[0629] The server records and maintains the history of interactions with the user. The input is the entire conversation, and the output is the recorded conversation history data. This information is used to personalize responses in the future.

[0630] 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.

[0631] This invention relates to a conversational artificial intelligence system that incorporates an emotion engine to recognize the user's emotions and engage in dialogue accordingly. This system can operate through a voice interface and possesses the flexibility to provide appropriate information and responses to the user in various situations.

[0632] System configuration:

[0633] This system mainly consists of the following components: input means, conversion means, analysis means, information acquisition means including an emotion engine, response generation means, speech synthesis means, output means, and recording means.

[0634] 1. Voice input acquisition and conversion:

[0635] The device acquires the voice spoken by the user and converts it into a digital format. The voice data is then transferred to a server and converted into text using speech recognition technology.

[0636] 2. Intention Analysis and Emotion Recognition:

[0637] The server analyzes the converted text using natural language processing technology to identify the user's intent. It also uses a built-in emotion engine to analyze emotions from the audio data and text. The emotion engine estimates the user's emotional state based on factors such as voice tone, speed, and word choice.

[0638] 3. Information acquisition and response generation:

[0639] The server retrieves necessary information from an external database based on the identified intent and emotional state. Then, using a response generation mechanism, it generates an appropriate response in a natural conversational format based on the retrieved information. The generated response is then adjusted to match the user's emotions.

[0640] 4. Speech synthesis and output:

[0641] The server synthesizes the generated text response into speech and sends it to the terminal as audio data. The terminal then presents the response to the user via its speaker.

[0642] 5. Recording of dialogue history:

[0643] The server records the history of interactions, including detailed and comprehensive emotional data, and uses this information to improve future services and response quality.

[0644] Specific example:

[0645] For example, if a user asks a question about a news item they are interested in and says in a slightly anxious tone, "I'm worried about the recent economic situation," the device will pick up the voice and use its emotion engine to detect that anxiety is present. Based on this, the server will retrieve economic news and prepare a response that is tailored to provide reassurance and calmness. For example, it might include reassuring elements in the response, such as, "Many experts analyze that the recent market fluctuations are temporary." Finally, the device will output the response as audio and deliver it to the user.

[0646] In this way, dialogue that takes the user's emotions into account becomes possible, enabling more human-like communication.

[0647] The following describes the processing flow.

[0648] Step 1:

[0649] The user inputs questions and instructions by voice into the smart device. This voice is captured by the microphone built into the device.

[0650] Step 2:

[0651] The terminal converts the audio data into a digital format and transmits it to the server in real time via a secure protocol. At this time, the audio data undergoes pre-processing such as noise reduction.

[0652] Step 3:

[0653] The server analyzes the received audio data using a speech recognition engine and converts it into text data. This conversion ensures that the audio information is accurately translated into written information.

[0654] Step 4:

[0655] The server analyzes the text obtained using natural language processing technology to identify the user's intent and requests. Additionally, a built-in emotion engine references the text and original audio data to recognize the user's emotions.

[0656] Step 5:

[0657] The server retrieves the necessary information from an external database based on the request and emotional state. For example, this could include data from external resources such as news or weather information.

[0658] Step 6:

[0659] The server generates a response that is appropriate to the user's emotions based on the information it has acquired. This response generation utilizes the results of sentiment analysis to incorporate language and tone that match the user's emotions.

[0660] Step 7:

[0661] The server converts the generated response into speech data using a speech synthesis engine. This speech data is then tone-adjusted to reflect the user's emotions.

[0662] Step 8:

[0663] The terminal plays audio data received from the server using its speaker and presents an audio response to the user. This response allows the user to receive information that takes emotions into consideration.

[0664] Step 9:

[0665] The server records the content of each session's dialogue and detected emotion data as a history, and updates this as a database to make future communications more personalized and accurate.

[0666] (Example 2)

[0667] 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."

[0668] Conventional conversational artificial intelligence systems focus on identifying the user's intent from their voice, but they are insufficient at identifying emotions and generating responses accordingly. Therefore, it is difficult to have conversations that are empathetic to the user's emotions, and achieving natural, human-like communication remains a challenge.

[0669] 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.

[0670] In this invention, the server includes an input means for acquiring the user's voice information, a conversion means for converting it into text information, and an analysis means for identifying the user's intentions and emotions. This makes it possible to generate responses that take into account not only the user's intentions but also their emotions.

[0671] "User" refers to a person who interacts with the system.

[0672] "Voice information" refers to the voice signals emitted by the user, which are the subject of processing by the system.

[0673] "Input means" refers to devices or mechanisms for acquiring voice information from the user.

[0674] "Textual information" refers to text data obtained as a result of converting audio information.

[0675] "Conversion means" refers to the technology or device used to perform the process of converting audio information into text information.

[0676] "Analysis means" refers to techniques used to identify a user's intentions and emotions from textual information.

[0677] "Emotions" refers to the mental state inferred from the tone of the user's voice or text, as well as the words they choose.

[0678] A "generative AI model" is a pre-trained artificial intelligence algorithm used to generate responses from underlying data.

[0679] "Response generation means" refers to technologies and devices for generating natural-sounding, conversational responses, taking into account collected information and the user's emotions.

[0680] "Audio format" refers to the data format of the audio signal held by the generated response.

[0681] "Speech synthesis means" refers to technologies and devices for converting text information into speech format.

[0682] "Output means" refers to devices or mechanisms that provide information to the user using voice-based responses.

[0683] "Dialogue history" refers to a record of past interactions between the user and the system.

[0684] "Emotional data" refers to information obtained as a result of analyzing the user's emotions.

[0685] As an example of how to carry out the present invention, a system is provided in which an emotion engine is incorporated into a conversational artificial intelligence system to recognize the user's emotions and engage in dialogue accordingly. This system operates through a voice interface and provides appropriate information and responses to the user in various situations.

[0686] First, the user gives instructions to the system by voice. The terminal acquires this voice using an input device and converts it into a digital voice signal. Then, this signal is sent to a server and converted into text data using speech recognition technology (for example, a common speech recognition tool or service).

[0687] Next, the server processes the text data using analysis tools and identifies the user's intentions and emotions using natural language processing technology. This analysis utilizes a generative AI model to infer the user's intentions and emotions based on the input data. The emotion engine estimates the user's emotional state from factors such as tone of voice, speed, and word choice, and responds accordingly.

[0688] After analysis, the server uses information acquisition tools to collect necessary data from an external database. This database contains several information sources (e.g., weather information and news websites). The collected information is processed by a response generation tool using a generative AI model, which generates a natural conversational response that matches the user's intent and emotions. The response is appropriately adjusted according to the user's emotions.

[0689] The generated response is then converted into a format suitable for audio output by a speech synthesis system. This audio data is sent to the terminal and conveyed to the user through the speaker. Finally, the server stores the dialogue history and sentiment data using recording means, and this data is used to improve future services and response quality.

[0690] As a concrete example, suppose a user asks in an anxious tone, "What should I wear today?" The server analyzes this question using an emotion engine and recognizes the user's anxiety. Then, using a weather API, it generates a response saying, "The forecast is for sunshine today, so a light jacket would be good. Please go out with peace of mind," and delivers it to the user via voice.

[0691] An example of a prompt statement can be entered as follows:

[0692] "We have detected that the user's emotion is 'anxiety.' Please prepare a response regarding the weather forecast, including language that will alleviate the user's anxiety."

[0693] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0694] Step 1:

[0695] The user speaks to the system using their voice. This voice signal becomes the input. The terminal uses its built-in microphone to acquire this voice signal and convert it into a digital format. For example, it converts analog audio into PCM digital data and sends the output to the server.

[0696] Step 2:

[0697] The server, upon receiving the digital audio data, converts it into text using speech recognition technology. This process employs a speech recognition algorithm to analyze the input audio and output it as a string of characters. This text information is then used for further analysis.

[0698] Step 3:

[0699] The server analyzes the converted text information using natural language processing techniques. Here, text information is received as input, and contextual understanding and intent analysis are performed as data processing. A generative AI model is used to identify the user's intent and emotions, and the results are output.

[0700] Step 4:

[0701] Based on the identified intent and emotion, the server retrieves the necessary information from an external database. This step involves querying an external API and gathering relevant information. This retrieved information becomes the input for the next response generation step.

[0702] Step 5:

[0703] The server utilizes a generative AI model to generate natural, conversational responses using a response generation mechanism. This process takes into account identified emotions and makes adjustments tailored to the user. For example, this might include adjusting the language to create a sense of reassurance. The generated text response is then output.

[0704] Step 6:

[0705] The server converts the generated response text into speech data using speech synthesis technology. It receives the text response as input and outputs it as an audio file. This audio data is then sent to the terminal.

[0706] Step 7:

[0707] The device that receives the transmitted audio data plays the audio through its speaker, conveying the response to the user. This ensures that information is provided in a way that is easy for the user to understand.

[0708] Step 8:

[0709] After the conversation ends, the server records the conversation history and sentiment data and saves it to a database. This record plays an important role as it will be used for future system improvements and personalization.

[0710] (Application Example 2)

[0711] 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."

[0712] The development of conversational artificial intelligence systems that take user emotions into consideration has been a long-standing requirement. Conventional technologies can only provide simple responses to voice input, and it has been difficult to recognize user emotions and provide appropriate information and suggestions accordingly. This invention aims to realize a system that can facilitate smoother communication with senior users and provide a sense of security, particularly in caregiving settings.

[0713] 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.

[0714] In this invention, the server includes means for acquiring voice input from a user, means for converting the voice input into text data, means for analyzing the text data to identify the user's emotions, means for collecting relevant information based on the identified emotions and requests, means for generating a response tailored to the user's emotional state based on the collected information, means for converting the generated response into data that can be output as voice, means for presenting the voice-outputtable data to the user, and means for recording and maintaining the user's dialogue history and emotional data. This enables the provision of appropriate information that matches the user's emotions.

[0715] A "user" is a person who uses an interactive artificial intelligence system to acquire information or communicate through voice input.

[0716] "Voice input" refers to the words or voices that the user speaks to the system.

[0717] "Text data" refers to character information converted from voice input, and is a format used for system analysis and processing.

[0718] "Emotions" refer to an internal state that can be inferred from the user's tone of voice, word choice, and other factors.

[0719] "Relevant information" refers to necessary data that the system collects from external sources based on the user's requests and feelings.

[0720] A "response" is a verbal or written response that a system generates and outputs based on user input.

[0721] "Audio-outputtable data" refers to data that has been converted into a format that allows the generated response to be presented to the user as audio.

[0722] "Dialogue history" refers to recorded information of past conversations between the system and the user, and is used to improve response quality.

[0723] "Emotional data" refers to records of the user's emotional state, and is useful information for improving the system and generating appropriate responses.

[0724] To implement the invention, the conversational artificial intelligence system is configured using a server and a user terminal. The voice spoken by the user is collected by the terminal's voice input device. This voice data is converted into a digital signal and transmitted to the server via the network.

[0725] The server uses speech recognition software to convert audio data into text data. Google Cloud Speech-to-Text API is used as this software. The converted text is then analyzed within the server using natural language processing techniques. OpenAI's generative AI model is used in this process to identify the user's requests and intentions.

[0726] Furthermore, emotion recognition utilizes an emotion analysis engine, leveraging Amazon's Comprehend and Azure's Text Analytics API to infer user emotions from text data and voice tone. The server then retrieves necessary information from an external data store based on the user's requests and emotional state. This data retrieval is performed via database access through a REST API.

[0727] The server generates responses adapted to the user's emotional state based on the collected information. The generated text responses are converted into audio data using the Google Cloud Text-to-Speech API and sent to the device over the network. Finally, they are presented to the user as audio through the device's speaker. Dialogue history and emotional data are securely recorded and stored by the server and used to improve future responses.

[0728] As a specific example, when a caregiver wearing smart glasses interacts with a user, if the user expresses anxiety such as "I haven't been sleeping well lately," the system detects the anxiety and provides a reassuring response such as, "How about listening to relaxing music before bed to improve your sleep environment?"

[0729] An example of a prompt message for a generative AI model might be, "The user is feeling anxious. We would like to offer reassuring suggestions regarding sleep."

[0730] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0731] Step 1:

[0732] The user provides voice input to the device. The device acquires this voice data via its microphone and converts it into a digital format. This voice data is then transmitted to a server via a communication network.

[0733] Step 2:

[0734] The server converts the received audio data into text data using the Google Cloud Speech-to-Text API. This process analyzes the audio signal and generates output as text information.

[0735] Step 3:

[0736] The server analyzes the obtained text data using natural language processing techniques. Specifically, it uses OpenAI's generative AI model to identify user requests from the text and extract important information. At this stage, it identifies information such as subjects and objects from the text.

[0737] Step 4:

[0738] The server uses an emotion analysis engine to identify the user's emotional state from text data and voice tone. Specifically, it utilizes Amazon Comprehend and Azure's Text Analytics API to analyze emotion indicators in the input data and output the results.

[0739] Step 5:

[0740] Based on the identified request and sentiment state, the server retrieves relevant information from an external data store via a REST API. This process collects relevant links and article data to satisfy the user's specific request and provides them as output.

[0741] Step 6:

[0742] The server generates an appropriate response to the user based on the collected information and emotional state. In the generation process, a prompt sentence (e.g., "The user is feeling anxious. We want to offer reassuring suggestions regarding sleep") is input to the generation AI model, and a natural conversational text response is obtained.

[0743] Step 7:

[0744] The generated text response is converted to speech by the server using the Google Cloud Text-to-Speech API. In this step, the text is processed into an audio file and output as data.

[0745] Step 8:

[0746] Finally, the device receives the audio data sent from the server and presents it to the user through its built-in speaker. In this step, the user can directly hear the audio response.

[0747] Step 9:

[0748] The server securely records and stores all conversation history and sentiment data. In this step, the acquired data is stored in a database for later analysis and response improvement.

[0749] 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.

[0750] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0751] 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.

[0752] [Fourth Embodiment]

[0753] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0754] 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.

[0755] 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).

[0756] 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.

[0757] 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.

[0758] 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).

[0759] 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.

[0760] 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.

[0761] 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.

[0762] 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.

[0763] 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.

[0764] 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.

[0765] 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".

[0766] This invention is a "conversational artificial intelligence system" that provides information through interactive dialogue with the user and can be implemented in various forms. This system operates on a smart device and is always in standby mode. The system receives the voice input when the user speaks to it.

[0767] System Operation Overview:

[0768] 1. Acquisition of voice input

[0769] The terminal acquires voice input from the user through a microphone and processes that data as a digital signal.

[0770] 2. Converting speech to text

[0771] The server uses a speech recognition engine to convert speech data into text data. This conversion is performed in real time, and the user's speech is recognized as text.

[0772] 3. Analysis of intent

[0773] The server analyzes text data using natural language processing technology. This analysis makes it possible to understand the user's requests and intentions.

[0774] 4. Information Acquisition and Response Generation

[0775] Based on the analysis results, the server retrieves relevant information from an external information database as needed. Based on this information, the system generates an appropriate response.

[0776] 5. Generation and output of voice responses

[0777] The server converts the generated response into audio data using speech synthesis technology. The terminal plays this audio data through its speaker, providing the user with the response in voice.

[0778] 6. Recording of dialogue history

[0779] The server records the content of interactions with users as a history and maintains this history for each user. This history is used to personalize the next response.

[0780] Specific example:

[0781] If the user asks, "What time is tomorrow's meeting?", the device will capture the audio, and the server will convert the audio to text. The server will then analyze the text to understand the user's request as "to confirm the meeting time." The server will access the schedule database, retrieve the time of the relevant meeting, and generate a response such as, "Tomorrow's meeting is at 2 PM." This response will be converted back into audio and transmitted to the user by the device.

[0782] Thus, the system of the present invention can provide information immediately in response to the user's needs, enabling more efficient daily support.

[0783] The following describes the processing flow.

[0784] Step 1:

[0785] The user voice-inputs a question or instruction into the smart device. The voice input is captured by the device's microphone.

[0786] Step 2:

[0787] The terminal converts the acquired audio data into a digital signal and transmits it to the server via a secure channel. During this process, pre-processing such as noise reduction is applied to improve the quality of the audio data.

[0788] Step 3:

[0789] The server analyzes the received audio data using a speech recognition engine and converts it into text data. This process ensures that the speech is recognized as an accurate text-based command.

[0790] Step 4:

[0791] The server uses natural language processing techniques to analyze the content of text data and identify the user's intent and requests. This includes keyword extraction and semantic analysis.

[0792] Step 5:

[0793] The server retrieves the necessary information from external databases or APIs based on the identified request. For example, weather information or schedule information may be requested.

[0794] Step 6:

[0795] The server generates a response to the user based on the information it has obtained. This response is constructed in natural language so that the user can easily understand it.

[0796] Step 7:

[0797] The server converts the response text into speech data using a speech synthesis tool. This process makes the text response available in a format that can be output as speech.

[0798] Step 8:

[0799] The terminal provides a response by playing audio data received from the server to the user through its speaker.

[0800] Step 9:

[0801] The server records this interaction as a dialogue history and retains it as data to better personalize future responses.

[0802] (Example 1)

[0803] 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".

[0804] Many modern voice response systems are specialized in providing a limited range of information and struggle to generate flexible responses that meet the diverse needs of users. Furthermore, they have the challenge of not being able to fully utilize the user's past conversation history, thus failing to provide personalized, interactive dialogue.

[0805] 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.

[0806] In this invention, the server includes means for acquiring voice input through an input device, means for analyzing character data with a language processing device, and a data acquisition device for collecting information from an external database. This enables the generation of flexible responses to diverse requests and the provision of personalized information to each user.

[0807] An "input / output device" is a device that acquires, records, and provides voice input from the user to the system.

[0808] A "speech recognition device" is a device that includes technology for converting acquired speech data into text data.

[0809] A "language processing device" is a device that analyzes text data and utilizes natural language processing technology to identify the user's intentions and requests.

[0810] A "data acquisition device" is a device that collects necessary information from external or internal sources based on an identified request.

[0811] A "reaction generation device" is a device that generates an appropriate response as text based on collected information.

[0812] A "synthesis device" is a device that converts generated text responses into audio data so that they can be output as speech.

[0813] A "sound playback device" is a device such as a speaker that presents synthesized sound data to the user.

[0814] A "memory device" is a device used to record and store the history of interactions with the user.

[0815] A "personalization device" is a device that utilizes recorded dialogue history to generate subsequent responses, enabling personalized interactions.

[0816] This invention relates to a conversational artificial intelligence system that provides information through interactive dialogue with the user. The system performs a series of processes to receive voice input, analyze it, and output an appropriate voice response.

[0817] First, the terminal acquires voice input through the microphone. The acquired voice data is converted into a digital signal and sent to the server. The server uses a speech recognition engine (for example, an API provided by a common speech recognition service provider) to convert the voice data into text data. At this stage, high-precision real-time conversion of the voice input is required.

[0818] Next, the server analyzes the text data using natural language processing techniques (utilizing open-source language analysis tools such as NLTK and spaCy). This analysis recognizes the user's request and identifies their intent. Based on the analysis results, it retrieves information from an external information database (e.g., a cloud-based data management service).

[0819] Based on the acquired data, the server generates an appropriate response. This response is prepared as text data and converted into speech data using a speech synthesis engine (for example, an API provided by a common speech synthesis service provider). The terminal then plays this speech data through its speaker, delivering the answer to the user.

[0820] Furthermore, the server records the history of interactions with the user and uses this information to personalize responses in the future. This allows the system to provide increasingly efficient and personalized interactions.

[0821] For example, if a user asks, "What's the weather like today?", the device will capture this voice message. The server will convert the voice into text data and parse it as a request to provide weather information. It will then access a weather information database and generate a response such as, "Today's weather is sunny, and the temperature is 20 degrees Celsius," which it will then communicate to the user via voice. This system aims to respond quickly and accurately to a variety of everyday information requests.

[0822] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0823] Step 1:

[0824] The terminal acquires the user's voice through a microphone. This input voice is received as an analog signal and converted into a digital signal by the terminal's digital signal processor. The output of this process is digital voice data that is sent to the server.

[0825] Step 2:

[0826] The server receives digital audio data and converts it into text data using a speech recognition engine. Specifically, it calls a speech recognition API to parse the audio data and extract the corresponding strings. The input for this step is digital audio data, and the output is the parsed text data.

[0827] Step 3:

[0828] The server analyzes the parsed text data using natural language processing techniques. This procedure uses language processing libraries to analyze the sentence structure and identify the user's intent. The input is text data, and the output is a data structure representing the user's request and intent.

[0829] Step 4:

[0830] The server retrieves relevant information from the database based on the user's intent. It also calls external APIs as needed to access various cloud services and data repositories. The input in this process is a data structure representing the user's request, and the output is the requested information data.

[0831] Step 5:

[0832] The server generates a response based on the information it has acquired. This process involves selecting a response template based on conditions and embedding the actual data to create the response text. The input is informational data, and the output is the response text intended for the user.

[0833] Step 6:

[0834] The server converts the response text into speech data using a speech synthesis engine. This is done using a speech synthesis API, which obtains natural-sounding speech data from the text data. The input is the response text, and the output is the synthesized speech data.

[0835] Step 7:

[0836] The device receives the synthesized audio data and plays it back to the user through its speaker. Specifically, it uses an audio playback module for output. In this case, the input is audio data, and the output is the audio the user hears.

[0837] Step 8:

[0838] The server records the user interaction history and adds it to the training data for future interactions. This record includes utterances and system responses, and is used to provide personalized interactions. The input is the interaction history data, and the output is an updated history database.

[0839] (Application Example 1)

[0840] 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".

[0841] In the daily lives of the elderly and those requiring support, there is a need to provide necessary information quickly and accurately while appropriately responding to changes and circumstances around them. Especially when daily living support is needed, timely information and instructions are essential for improving the quality of life. However, conventional technologies do not take into account the user's surroundings and circumstances, making it difficult to provide the optimal response.

[0842] 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.

[0843] In this invention, the server includes receiving means for acquiring voice information from the user, interpreting means for analyzing text information and identifying the user's request, and adapting means for providing information according to the user's situation in the environment. This makes it possible to provide optimal information and instructions tailored to the situation to the elderly and people who need assistance.

[0844] "Receiving means" refers to a device or function that acquires voice information from the user.

[0845] "Conversion means" refers to a device or function that converts acquired audio information into text information.

[0846] "Interpretation means" refers to a device or function that analyzes textual information and identifies the user's request.

[0847] "Collection means" refers to a device or function that collects relevant information based on an identified request.

[0848] "Generating means" refers to a device or function that generates a response based on collected information.

[0849] "Speech synthesis means" refers to a device or function that converts a generated response into information that can be input or output as speech.

[0850] "Presentation means" refers to a device or function that presents information capable of audio input and output to the user.

[0851] "Recording means" refers to a device or function that records and retains the history of interactions with the user.

[0852] "Adaptive means" refers to a device or function that provides information according to the user's situation in the environment.

[0853] This system is a conversational artificial intelligence designed to support the daily lives of the elderly and people with disabilities. Specifically, it interacts with users through devices such as smart glasses and provides information via voice. The device acquires voice information transmitted by the user through a microphone. The server converts the acquired voice information into text and analyzes the text using natural language processing technology to identify the user's requests and intentions.

[0854] Once the analysis is complete, relevant information is collected from external information resources based on the request. Based on the collected information, the server generates an appropriate response and converts it into audio input / outputable information. This audio information is provided to the user through the terminal. The system also has the function to provide information appropriately, taking into account the user's situation in the environment.

[0855] The system uses the speech_recognition library for speech recognition and the pyttsx3 library for speech synthesis. Hardware-wise, the microphone and speaker built into the smart glasses play a major role.

[0856] For example, when a user asks, "Can you give me some health advice?", the system provides practical advice via voice, such as, "Make sure you stay well-hydrated." An example of a specific prompt using the generative AI model would be, "Please give me 10 pieces of health advice suitable for supporting the daily lives of elderly people."

[0857] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0858] Step 1:

[0859] The user gives voice commands to the device. The voice information is acquired through the microphone. The input is the user's voice, and the output is digital voice data. The device sends this voice data to the server.

[0860] Step 2:

[0861] The server converts received audio data into text information. The input is digital audio data, and the output is text data. The server performs speech recognition using the speech_recognition library. In this process, the user's speech is converted into text data.

[0862] Step 3:

[0863] The server analyzes text data to identify the user's intent. The input is text data, and the output is information about the user's request. Natural language processing techniques are used to interpret the user's intent and determine the specific request and expected response.

[0864] Step 4:

[0865] The server collects necessary information from external information resources based on user requests. The input is information related to the user's request, and the output is information corresponding to that request. Generative AI models and APIs are used to collect data necessary to provide appropriate information.

[0866] Step 5:

[0867] The server generates a response based on the collected information. The input is the collected information, and the output is the generated response data. Based on the collected information, it decides what kind of response to send and constructs the response in text format.

[0868] Step 6:

[0869] The server converts the generated response into audio information. The input is a text-based response, and the output is audio data. The pyttsx3 library is used to synthesize the text data into speech and generate the audio data.

[0870] Step 7:

[0871] The terminal plays audio data received from the server and delivers a response to the user. The input is audio data from the server, and the output is audio from the speaker. The user receives responses from the system in real time.

[0872] Step 8:

[0873] The server records and maintains the history of interactions with the user. The input is the entire conversation, and the output is the recorded conversation history data. This information is used to personalize responses in the future.

[0874] 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.

[0875] This invention relates to a conversational artificial intelligence system that incorporates an emotion engine to recognize the user's emotions and engage in dialogue accordingly. This system can operate through a voice interface and possesses the flexibility to provide appropriate information and responses to the user in various situations.

[0876] System configuration:

[0877] This system mainly consists of the following components: input means, conversion means, analysis means, information acquisition means including an emotion engine, response generation means, speech synthesis means, output means, and recording means.

[0878] 1. Voice input acquisition and conversion:

[0879] The device acquires the voice spoken by the user and converts it into a digital format. The voice data is then transferred to a server and converted into text using speech recognition technology.

[0880] 2. Intention Analysis and Emotion Recognition:

[0881] The server analyzes the converted text using natural language processing technology to identify the user's intent. It also uses a built-in emotion engine to analyze emotions from the audio data and text. The emotion engine estimates the user's emotional state based on factors such as voice tone, speed, and word choice.

[0882] 3. Information acquisition and response generation:

[0883] The server retrieves necessary information from an external database based on the identified intent and emotional state. Then, using a response generation mechanism, it generates an appropriate response in a natural conversational format based on the retrieved information. The generated response is then adjusted to match the user's emotions.

[0884] 4. Speech synthesis and output:

[0885] The server synthesizes the generated text response into speech and sends it to the terminal as audio data. The terminal then presents the response to the user via its speaker.

[0886] 5. Recording of dialogue history:

[0887] The server records the history of interactions, including detailed and comprehensive emotional data, and uses this information to improve future services and response quality.

[0888] Specific example:

[0889] For example, if a user asks a question about a news item they are interested in and says in a slightly anxious tone, "I'm worried about the recent economic situation," the device will pick up the voice and use its emotion engine to detect that anxiety is present. Based on this, the server will retrieve economic news and prepare a response that is tailored to provide reassurance and calmness. For example, it might include reassuring elements in the response, such as, "Many experts analyze that the recent market fluctuations are temporary." Finally, the device will output the response as audio and deliver it to the user.

[0890] In this way, dialogue that takes the user's emotions into account becomes possible, enabling more human-like communication.

[0891] The following describes the processing flow.

[0892] Step 1:

[0893] The user inputs questions and instructions by voice into the smart device. This voice is captured by the microphone built into the device.

[0894] Step 2:

[0895] The terminal converts the audio data into a digital format and transmits it to the server in real time via a secure protocol. At this time, the audio data undergoes pre-processing such as noise reduction.

[0896] Step 3:

[0897] The server analyzes the received audio data using a speech recognition engine and converts it into text data. This conversion ensures that the audio information is accurately translated into written information.

[0898] Step 4:

[0899] The server analyzes the text obtained using natural language processing technology to identify the user's intent and requests. Additionally, a built-in emotion engine references the text and original audio data to recognize the user's emotions.

[0900] Step 5:

[0901] The server retrieves the necessary information from an external database based on the request and emotional state. For example, this could include data from external resources such as news or weather information.

[0902] Step 6:

[0903] The server generates a response that is appropriate to the user's emotions based on the information it has acquired. This response generation utilizes the results of sentiment analysis to incorporate language and tone that match the user's emotions.

[0904] Step 7:

[0905] The server converts the generated response into speech data using a speech synthesis engine. This speech data is then tone-adjusted to reflect the user's emotions.

[0906] Step 8:

[0907] The terminal plays audio data received from the server using its speaker and presents an audio response to the user. This response allows the user to receive information that takes emotions into consideration.

[0908] Step 9:

[0909] The server records the content of each session's dialogue and detected emotion data as a history, and updates this as a database to make future communications more personalized and accurate.

[0910] (Example 2)

[0911] 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".

[0912] Conventional conversational artificial intelligence systems focus on identifying the user's intent from their voice, but they are insufficient at identifying emotions and generating responses accordingly. Therefore, it is difficult to have conversations that are empathetic to the user's emotions, and achieving natural, human-like communication remains a challenge.

[0913] 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.

[0914] In this invention, the server includes an input means for acquiring the user's voice information, a conversion means for converting it into text information, and an analysis means for identifying the user's intentions and emotions. This makes it possible to generate responses that take into account not only the user's intentions but also their emotions.

[0915] "User" refers to a person who interacts with the system.

[0916] "Voice information" refers to the voice signals emitted by the user, which are the subject of processing by the system.

[0917] "Input means" refers to devices or mechanisms for acquiring voice information from the user.

[0918] "Textual information" refers to text data obtained as a result of converting audio information.

[0919] "Conversion means" refers to the technology or device used to perform the process of converting audio information into text information.

[0920] "Analysis means" refers to techniques used to identify a user's intentions and emotions from textual information.

[0921] "Emotions" refers to the mental state inferred from the tone of the user's voice or text, as well as the words they choose.

[0922] A "generative AI model" is a pre-trained artificial intelligence algorithm used to generate responses from underlying data.

[0923] "Response generation means" refers to technologies and devices for generating natural-sounding, conversational responses, taking into account collected information and the user's emotions.

[0924] "Audio format" refers to the data format of the audio signal held by the generated response.

[0925] "Speech synthesis means" refers to technologies and devices for converting text information into speech format.

[0926] "Output means" refers to devices or mechanisms that provide information to the user using voice-based responses.

[0927] "Dialogue history" refers to a record of past interactions between the user and the system.

[0928] "Emotional data" refers to information obtained as a result of analyzing the user's emotions.

[0929] As an example of how to carry out the present invention, a system is provided in which an emotion engine is incorporated into a conversational artificial intelligence system to recognize the user's emotions and engage in dialogue accordingly. This system operates through a voice interface and provides appropriate information and responses to the user in various situations.

[0930] First, the user gives instructions to the system by voice. The terminal acquires this voice using an input device and converts it into a digital voice signal. Then, this signal is sent to a server and converted into text data using speech recognition technology (for example, a common speech recognition tool or service).

[0931] Next, the server processes the text data using analysis tools and identifies the user's intentions and emotions using natural language processing technology. This analysis utilizes a generative AI model to infer the user's intentions and emotions based on the input data. The emotion engine estimates the user's emotional state from factors such as tone of voice, speed, and word choice, and responds accordingly.

[0932] After analysis, the server uses information acquisition tools to collect necessary data from an external database. This database contains several information sources (e.g., weather information and news websites). The collected information is processed by a response generation tool using a generative AI model, which generates a natural conversational response that matches the user's intent and emotions. The response is appropriately adjusted according to the user's emotions.

[0933] The generated response is then converted into a format suitable for audio output by a speech synthesis system. This audio data is sent to the terminal and conveyed to the user through the speaker. Finally, the server stores the dialogue history and sentiment data using recording means, and this data is used to improve future services and response quality.

[0934] As a concrete example, suppose a user asks in an anxious tone, "What should I wear today?" The server analyzes this question using an emotion engine and recognizes the user's anxiety. Then, using a weather API, it generates a response saying, "The forecast is for sunshine today, so a light jacket would be good. Please go out with peace of mind," and delivers it to the user via voice.

[0935] An example of a prompt statement can be entered as follows:

[0936] "We have detected that the user's emotion is 'anxiety.' Please prepare a response regarding the weather forecast, including language that will alleviate the user's anxiety."

[0937] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0938] Step 1:

[0939] The user speaks to the system using their voice. This voice signal becomes the input. The terminal uses its built-in microphone to acquire this voice signal and convert it into a digital format. For example, it converts analog audio into PCM digital data and sends the output to the server.

[0940] Step 2:

[0941] The server, upon receiving the digital audio data, converts it into text using speech recognition technology. This process employs a speech recognition algorithm to analyze the input audio and output it as a string of characters. This text information is then used for further analysis.

[0942] Step 3:

[0943] The server analyzes the converted text information using natural language processing techniques. Here, text information is received as input, and contextual understanding and intent analysis are performed as data processing. A generative AI model is used to identify the user's intent and emotions, and the results are output.

[0944] Step 4:

[0945] Based on the identified intent and emotion, the server retrieves the necessary information from an external database. This step involves querying an external API and gathering relevant information. This retrieved information becomes the input for the next response generation step.

[0946] Step 5:

[0947] The server utilizes a generative AI model to generate natural, conversational responses using a response generation mechanism. This process takes into account identified emotions and makes adjustments tailored to the user. For example, this might include adjusting the language to create a sense of reassurance. The generated text response is then output.

[0948] Step 6:

[0949] The server converts the generated response text into speech data using speech synthesis technology. It receives the text response as input and outputs it as an audio file. This audio data is then sent to the terminal.

[0950] Step 7:

[0951] The device that receives the transmitted audio data plays the audio through its speaker, conveying the response to the user. This ensures that information is provided in a way that is easy for the user to understand.

[0952] Step 8:

[0953] After the conversation ends, the server records the conversation history and sentiment data and saves it to a database. This record plays an important role as it will be used for future system improvements and personalization.

[0954] (Application Example 2)

[0955] 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".

[0956] The development of conversational artificial intelligence systems that take user emotions into consideration has been a long-standing requirement. Conventional technologies can only provide simple responses to voice input, and it has been difficult to recognize user emotions and provide appropriate information and suggestions accordingly. This invention aims to realize a system that can facilitate smoother communication with senior users and provide a sense of security, particularly in caregiving settings.

[0957] 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.

[0958] In this invention, the server includes means for acquiring voice input from a user, means for converting the voice input into text data, means for analyzing the text data to identify the user's emotions, means for collecting relevant information based on the identified emotions and requests, means for generating a response tailored to the user's emotional state based on the collected information, means for converting the generated response into data that can be output as voice, means for presenting the voice-outputtable data to the user, and means for recording and maintaining the user's dialogue history and emotional data. This enables the provision of appropriate information that matches the user's emotions.

[0959] A "user" is a person who uses an interactive artificial intelligence system to acquire information or communicate through voice input.

[0960] "Voice input" refers to the words or voices that the user speaks to the system.

[0961] "Text data" refers to character information converted from voice input, and is a format used for system analysis and processing.

[0962] "Emotions" refer to an internal state that can be inferred from the user's tone of voice, word choice, and other factors.

[0963] "Relevant information" refers to necessary data that the system collects from external sources based on the user's requests and feelings.

[0964] A "response" is a verbal or written response that a system generates and outputs based on user input.

[0965] "Audio-outputtable data" refers to data that has been converted into a format that allows the generated response to be presented to the user as audio.

[0966] "Dialogue history" refers to recorded information of past conversations between the system and the user, and is used to improve response quality.

[0967] "Emotional data" refers to records of the user's emotional state, and is useful information for improving the system and generating appropriate responses.

[0968] To implement the invention, the conversational artificial intelligence system is configured using a server and a user terminal. The voice spoken by the user is collected by the terminal's voice input device. This voice data is converted into a digital signal and transmitted to the server via the network.

[0969] The server uses speech recognition software to convert audio data into text data. Google Cloud Speech-to-Text API is used as this software. The converted text is then analyzed within the server using natural language processing techniques. OpenAI's generative AI model is used in this process to identify the user's requests and intentions.

[0970] Furthermore, emotion recognition utilizes an emotion analysis engine, leveraging Amazon's Comprehend and Azure's Text Analytics API to infer user emotions from text data and voice tone. The server then retrieves necessary information from an external data store based on the user's requests and emotional state. This data retrieval is performed via database access through a REST API.

[0971] The server generates responses adapted to the user's emotional state based on the collected information. The generated text responses are converted into audio data using the Google Cloud Text-to-Speech API and sent to the device over the network. Finally, they are presented to the user as audio through the device's speaker. Dialogue history and emotional data are securely recorded and stored by the server and used to improve future responses.

[0972] As a specific example, when a caregiver wearing smart glasses interacts with a user, if the user expresses anxiety such as "I haven't been sleeping well lately," the system detects the anxiety and provides a reassuring response such as, "How about listening to relaxing music before bed to improve your sleep environment?"

[0973] An example of a prompt message for a generative AI model might be, "The user is feeling anxious. We would like to offer reassuring suggestions regarding sleep."

[0974] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0975] Step 1:

[0976] The user provides voice input to the device. The device acquires this voice data via its microphone and converts it into a digital format. This voice data is then transmitted to a server via a communication network.

[0977] Step 2:

[0978] The server converts the received audio data into text data using the Google Cloud Speech-to-Text API. This process analyzes the audio signal and generates output as text information.

[0979] Step 3:

[0980] The server analyzes the obtained text data using natural language processing techniques. Specifically, it uses OpenAI's generative AI model to identify user requests from the text and extract important information. At this stage, it identifies information such as subjects and objects from the text.

[0981] Step 4:

[0982] The server uses an emotion analysis engine to identify the user's emotional state from text data and voice tone. Specifically, it utilizes Amazon Comprehend and Azure's Text Analytics API to analyze emotion indicators in the input data and output the results.

[0983] Step 5:

[0984] Based on the identified request and sentiment state, the server retrieves relevant information from an external data store via a REST API. This process collects relevant links and article data to satisfy the user's specific request and provides them as output.

[0985] Step 6:

[0986] The server generates an appropriate response to the user based on the collected information and emotional state. In the generation process, a prompt sentence (e.g., "The user is feeling anxious. We want to offer reassuring suggestions regarding sleep") is input to the generation AI model, and a natural conversational text response is obtained.

[0987] Step 7:

[0988] The generated text response is converted to speech by the server using the Google Cloud Text-to-Speech API. In this step, the text is processed into an audio file and output as data.

[0989] Step 8:

[0990] Finally, the device receives the audio data sent from the server and presents it to the user through its built-in speaker. In this step, the user can directly hear the audio response.

[0991] Step 9:

[0992] The server securely records and stores all conversation history and sentiment data. In this step, the acquired data is stored in a database for later analysis and response improvement.

[0993] 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.

[0994] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0995] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0996] 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.

[0997] 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.

[0998] 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.

[0999] 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.

[1000] 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.

[1001] 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."

[1002] 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.

[1003] 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.

[1004] 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.

[1005] 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.

[1006] 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.

[1007] 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.

[1008] 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.

[1009] 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.

[1010] 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.

[1011] 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.

[1012] 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.

[1013] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[1014] The following is further disclosed regarding the embodiments described above.

[1015] (Claim 1)

[1016] An input means for acquiring voice input from the user,

[1017] A conversion means for converting the voice input into text data,

[1018] An analysis means for analyzing the aforementioned text data and identifying the user's request,

[1019] Information acquisition means for collecting relevant information based on identified requests,

[1020] A response generation means that generates a response based on the collected information,

[1021] A speech synthesis means that converts the generated response into data that can be output as speech,

[1022] An output means for presenting the aforementioned audio-outputtable data to the user,

[1023] A recording means for recording and retaining the history of conversations with the user,

[1024] A conversational artificial intelligence system that includes this.

[1025] (Claim 2)

[1026] The conversational artificial intelligence system according to claim 1, characterized in that the analysis means identifies the intent of the voice input using natural language processing technology.

[1027] (Claim 3)

[1028] The conversational artificial intelligence system according to claim 1, characterized in that the information acquisition means acquires data from an external information database.

[1029] "Example 1"

[1030] (Claim 1)

[1031] An input / output device that acquires voice input from the user,

[1032] A speech recognition device that converts the voice input into text data,

[1033] A language processing device that analyzes the aforementioned character data and identifies the user's request,

[1034] A data acquisition device that collects relevant information based on identified requests,

[1035] A reaction generation device that generates a response based on the collected information,

[1036] A synthesis device that converts the generated response into data that can be output as speech,

[1037] A sound playback device that presents the aforementioned audio-outputtable data to the user,

[1038] A storage device that records and retains the history of interactions with the user,

[1039] A personalization device that uses the recorded dialogue history for generating the next response,

[1040] A system that includes this.

[1041] (Claim 2)

[1042] The system according to claim 1, characterized in that the language processing device identifies the intent of the voice input using natural language processing technology.

[1043] (Claim 3)

[1044] The system according to claim 1, characterized in that the data acquisition device collects information from an external information database.

[1045] "Application Example 1"

[1046] (Claim 1)

[1047] A means for receiving voice information from the user,

[1048] A conversion means for converting audio information into text information,

[1049] An interpretation means for analyzing the aforementioned textual information and identifying the user's request,

[1050] A collection means for collecting relevant information based on identified requests,

[1051] A generation means for creating a response based on the collected information,

[1052] A speech synthesis means that converts the generated response into information that can be input or output as speech,

[1053] A presentation means for presenting the aforementioned audio input / output-capable information to the user,

[1054] A recording means for recording and retaining the history of conversations with the user,

[1055] Adaptive means that provide information according to the user's situation in the environment,

[1056] A system that includes this.

[1057] (Claim 2)

[1058] The system according to claim 1, characterized in that the interpretation means identifies the intent of the speech information using natural language processing technology.

[1059] (Claim 3)

[1060] The system according to claim 1, characterized in that the collection means acquires information from an external information resource.

[1061] "Example 2 of combining an emotion engine"

[1062] (Claim 1)

[1063] An input means for acquiring voice information from the user,

[1064] A conversion means for converting audio information into text information,

[1065] An analysis means for analyzing the aforementioned textual information and identifying the user's intentions and emotions,

[1066] Information acquisition means for collecting relevant information based on identified intentions and emotions,

[1067] A response generation means that generates a response using a generation AI model according to the collected information and the user's emotions,

[1068] A speech synthesis means that converts the generated response into a speech format,

[1069] An output means for outputting the response in the aforementioned voice format to the user,

[1070] Means for recording and retaining emotional data, including the history of conversations with the user,

[1071] A system that includes this.

[1072] (Claim 2)

[1073] The system according to claim 1, characterized in that the analysis means identifies the intent and emotion of the speech information using natural language processing technology.

[1074] (Claim 3)

[1075] The system according to claim 1, characterized in that the information acquisition means collects information from an external database.

[1076] "Application example 2 when combining with an emotional engine"

[1077] (Claim 1)

[1078] A means of obtaining voice input from the user,

[1079] A means of converting that voice input into text data,

[1080] A means for analyzing the aforementioned text data and identifying the user's emotions,

[1081] Means for collecting relevant information based on identified emotions and needs,

[1082] A means of generating a response tailored to the user's emotional state based on the collected information,

[1083] A means for converting the generated response into data that can be output as audio,

[1084] A means for presenting the aforementioned audio-outputtable data to the user,

[1085] Means for recording and retaining user interaction history and emotional data,

[1086] A system that includes this.

[1087] (Claim 2)

[1088] The system according to claim 1, characterized in that the analysis means identifies intentions and emotions from voice input using natural language processing technology.

[1089] (Claim 3)

[1090] The system according to claim 1, characterized in that the information gathering means acquires data from an external information data store. [Explanation of Symbols]

[1091] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for receiving voice information from the user, A conversion means for converting audio information into text information, An interpretation means for analyzing the aforementioned textual information and identifying the user's request, A collection means for collecting relevant information based on identified requests, A generation means for creating a response based on the collected information, A speech synthesis means that converts the generated response into information that can be input or output as speech, A presentation means for presenting the aforementioned audio input / output-capable information to the user, A recording means for recording and retaining the history of conversations with the user, Adaptive means that provide information according to the user's situation in the environment, A system that includes this.

2. The system according to claim 1, characterized in that the interpretation means identifies the intent of the speech information using natural language processing technology.

3. The system according to claim 1, characterized in that the collection means acquires information from an external information resource.

Images