system

Abstract

A system is provided.SOLUTION: Means for receiving request data in a form of at least one of voice, text, image, and video from a user; means for transmitting the received request data to a server; means for preprocessing the request data and converting voice data into text; A system comprising: means for analyzing image data; means for analyzing a user's intention using pre-processed data; means for referring to an internal database or an external API to obtain necessary information; means for generating an appropriate response based on the obtained information; means for transmitting the generated response to a terminal; and means for the terminal to display the response to the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] Conventional customer service systems were unable to process a variety of data formats, including text, voice, images, and video, making it difficult to respond quickly and appropriately. Delays in data processing and inaccuracies in analysis also led to problems with low user satisfaction. Furthermore, while systems are required to flexibly adapt to advances in digital technology and respond quickly to market changes, their slow response time has also been an issue. [Means for solving the problem]

[0005] The present invention solves the aforementioned problems by the following means. Specifically, it includes a means for receiving request data from a user in at least one of the following formats: voice, text, image, and video; and a means for transmitting the received request data to a server. The server also includes means for preprocessing the request data, converting voice data into text, and analyzing image data. The server further includes means for analyzing the user's intent using the preprocessed data and means for obtaining necessary information by referencing an internal database or an external API. The system also includes means for generating an appropriate response based on the obtained information, means for transmitting the generated response to a terminal, and means for the terminal to display the response to the user. This series of means enables efficient processing of multiple data formats through integration, enabling prompt and accurate responses to the user.

[0006] "Request data" is data in the form of text, audio, images, video, etc. that a user sends to the system.

[0007] A "server" is a computer system that provides services to other computers over the Internet or a local network.

[0008] A "terminal" refers to a device, such as a smartphone or computer, through which a user inputs request data and receives a response result.

[0009] A "speech recognition system" is a system that takes speech as input and converts it into digital text.

[0010] An "image analysis system" is a system that receives image data as input and extracts useful information from it.

[0011] An "internal database" is a data storage device that stores data held internally by the system and allows searching and referencing as needed.

[0012] An "external API" is an interface provided to enable communication and data exchange with other systems.

[0013] "Preprocessing" refers to the process of converting request data into a format suitable for analysis and response generation.

[0014] "User intent" refers to the purpose or request that the user has when sending request data.

[0015] "Response generation" is the process of creating appropriate answers or instructions as a result of analyzing the user's intent.

[0016] "Display" refers to the act of providing response data on a terminal in a form that can be visually recognized by the user. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9]1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0018] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0019] First, the terms used in the following description will be explained.

[0020] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0021] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0022] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0023] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0024] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0025] [First embodiment]

[0026] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0027] 1, a 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.

[0028] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0029] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0030] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the 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.

[0031] 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 of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0032] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0033] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0034] 2, in the data processing device 12, a specific process 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" according to the technology of the present 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 process 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.

[0035] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0036] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0037] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0038] The present invention relates to a customer service system that can process requests in various modalities. It supports each modality (text, voice, image, video) and provides appropriate responses to user requests in real time. Specific embodiments of this system are described below.

[0039] Program Overview

[0040] This system consists of three main components: a server, a terminal, and a user. The server is responsible for processing the AI ​​model and managing data, while the terminal provides an interface for users to send requests and receive responses, allowing users to ask questions or make requests. The cooperation of these components creates a mechanism that can appropriately respond to a variety of user requests.

[0041] Program Details

[0042] Receiving a user request

[0043] The user makes a request through the device, which can be in the form of text, voice, image, or video.

[0044] Example: A user types the text "I want to see my electricity bill details" into the terminal.

[0045] Submitting a Request

[0046] The terminal transmits the received request data to the server.

[0047] Example: The terminal sends text data to the server saying, "I want to see my electricity bill details."

[0048] Data Preprocessing

[0049] The server preprocesses the received data: voice data is converted into text using a voice recognition system, and image data is extracted using an image analysis system.

[0050] Example: When the server converts voice data to text, it uses speech recognition to generate the text "I would like to see my electricity bill details."

[0051] Intent Analysis

[0052] The server inputs the preprocessed data into the AI ​​model to understand the user's intent, and then proceeds with the process based on the user's intent.

[0053] Example: From the text "I want to see my electricity bill details," the user's intent is interpreted as "I want to check the details of my current electricity bill."

[0054] Database Reference

[0055] The server references an internal database or external API to obtain the necessary information.

[0056] Example: The server looks up the user's account information in a database and retrieves the latest electricity bill data.

[0057] Generating a response

[0058] Based on the information obtained, the server generates an appropriate response to the user.

[0059] Example: Generate a response in text format saying "Your latest electricity bill is 5000 yen."

[0060] Sending a Response

[0061] The server generates a response and sends it to the terminal.

[0062] Example: The server sends text data to the terminal saying, "The latest electricity bill is 5,000 yen."

[0063] Viewing the response

[0064] The terminal displays the response received from the server to the user.

[0065] Example: The device displays a text message to the user saying, "Your latest electricity bill is 5000 yen."

[0066] Specific examples

[0067] The following is an example of the process when a user verbally asks the terminal, "I'm having trouble reading the electricity meter."

[0068] 1. Receiving a user request

[0069] The user verbally asks the terminal, "I'm having trouble reading the electric meter."

[0070] The terminal receives the audio data.

[0071] 2. Submitting a Request

[0072] The terminal transmits the voice data to the server.

[0073] 3. Data Preprocessing

[0074] The server converts the audio data into text.

[0075] The speech recognition system generates the text "I'm having trouble reading my electric meter."

[0076] 4. Intent Analysis

[0077] The server inputs the text data into an AI model and analyzes the user's intent.

[0078] The user's intent is interpreted as "I want to know how to read my electric meter."

[0079] 5. Referencing the database

[0080] The server retrieves guide information on how to read the electricity meter from an internal database.

[0081] 6. Generating the Response

[0082] The server generates a response saying, "Take a photo of your electric meter and tap the camera icon in the chat window."

[0083] 7. Sending the Response

[0084] The server generates a response and sends it to the terminal.

[0085] 8. Displaying the Response

[0086] The device displays a text message to the user saying, "Take a photo of your electricity meter and tap the camera icon on the chat screen."

[0087] This system can handle a variety of data formats and provide users with prompt and appropriate responses in real time.

[0088] The processing flow will be explained below.

[0089] Step 1:

[0090] The user makes a request through the device, which can be input in the form of text, voice, image, or video.

[0091] Example: A user texts a device asking, "What's my electricity bill this month?"

[0092] Step 2:

[0093] The terminal transmits the received request data to the server.

[0094] Example: The device sends text data such as "What is the electricity bill this month?" to the server.

[0095] Step 3:

[0096] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[0097] Example: The server receives voice data and converts it into text using a speech recognition system: "What is the electricity bill this month?"

[0098] Step 4:

[0099] The server inputs the preprocessed data into an AI model to analyze the user's intent.

[0100] Example: An AI model interprets the text "What is the electricity bill this month?" as the user's intent to "I want to know the current bill amount."

[0101] Step 5:

[0102] The server references an internal database or external API to obtain the necessary information.

[0103] Example: The server retrieves a user's account information from a database and retrieves the latest electricity bill information.

[0104] Step 6:

[0105] The server generates an appropriate response based on the information it has obtained.

[0106] Example: The server generates a text response saying "This month's electricity bill is 5000 yen."

[0107] Step 7:

[0108] The server generates a response and sends it to the terminal.

[0109] Example: The server sends the text "This month's electricity bill is 5,000 yen" to the terminal.

[0110] Step 8:

[0111] The terminal displays the response received from the server to the user.

[0112] Example: The device displays a text message to the user saying, "This month's electricity bill is 5000 yen."

[0113] The above is the flow of processing for a user request.

[0114] Example 1

[0115] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0116] Conventional customer service systems have had difficulty providing appropriate responses to user requests in real time. Furthermore, few systems support multiple modalities (voice, text, images, and video), creating a need for improved service quality. Furthermore, accurately analyzing user intent and quickly providing appropriate information is a challenge.

[0117] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0118] In this invention, the server includes means for receiving request data from a user in at least one form of voice, text, image, or video, means for preprocessing the received request data, converting voice data into text, and analyzing image data, means for using a generative model to analyze the user's intention using the preprocessed data, means for obtaining necessary information by referring to an internal database or an external interface, means for generating an appropriate response based on the obtained information, means for transmitting the generated response to a terminal, and means for the terminal to display the response to the user. This makes it possible to support various modalities, accurately analyze the user's intention, and provide appropriate information in real time.

[0119] A "user" is a person or entity that submits requests to a system to obtain information.

[0120] A "terminal" is a device through which a user inputs requests and receives responses.

[0121] A "server" is a computing system that processes requests from users and generates and sends appropriate responses to terminals.

[0122] "Request data" refers to voice, text, image, or video data that a user sends to the system through a terminal.

[0123] "Preprocessing" is the process of converting received request data into a format that is easy to analyze.

[0124] A "voice recognition system" is a technology that converts voice data into text data.

[0125] An "image analysis system" is a technology that analyzes image data and extracts necessary information.

[0126] A "generative model" is an AI technology that analyzes user intent based on preprocessed data.

[0127] An "internal database" is a system that manages data within a server.

[0128] An "external interface" is a means by which a server interacts with external systems and services to obtain data.

[0129] An "appropriate response" is data that contains information or instructions that are most appropriate to the user's request.

[0130] The present invention relates to a multi-modal customer service system that receives voice, text, image, and video request data from users and generates appropriate responses in real time.

[0131] System configuration

[0132] The system consists of three main components: a server, a terminal, and a user. The server is responsible for processing the AI ​​model and managing data, while the terminal provides an interface where users can send requests and receive responses. Users can ask questions or make requests and use the system's services.

[0133] Hardware and Software

[0134] Server: A high-performance computer system that preprocesses data, runs generative models, manages databases, and generates responses. Specific software includes speech recognition engines (e.g., Google Cloud Speech-to-Text) and generative AI models (e.g., OpenAI's GPT-3).

[0135] Terminal: A device that provides a user interface, such as a smartphone, tablet, or PC. The terminal uses a communication protocol (e.g., HTTP / HTTPS) to enable sending and receiving requests.

[0136] Initial Setup and Operation

[0137] 1. User request input: The user uses the device to send a request via voice, text, image, or video.

[0138] Example: A user makes a voice request to their smartphone saying, "I want to see my electricity bill details."

[0139] 2. Sending request data: The terminal sends the request data to the server via a communication protocol.

[0140] Example: A smartphone sends voice data saying "I want to see my electricity bill details" to a server via an HTTP request.

[0141] 3. Data preprocessing: The server preprocesses the received data. Voice data is converted into text using a voice recognition system, and image data is analyzed using an image analysis system.

[0142] Example: The server uses a speech recognition engine to convert the voice data into text, generating the text "I would like to see my electricity bill details."

[0143] 4. Intent analysis: The server inputs the preprocessed data into the generative AI model to analyze the user's intent.

[0144] Example: The server inputs the text "I want to see my electricity bill details" into the generated AI model, and analyzes the user's intent as "I want to know the details of my current electricity bill."

[0145] 5. Data retrieval: The server retrieves the required information using an internal database or an external interface.

[0146] Example: A server queries an internal database to get the latest electricity bill (e.g., 5000 yen).

[0147] 6. Generate a response: The server uses the information it has obtained to generate an appropriate response for the user.

[0148] Example: The server generates a text response saying "The latest electricity bill is 5000 yen."

[0149] 7. Send and display the response: Send the generated response to the terminal, which displays it to the user.

[0150] Example: The server sends generated text data to the terminal, and the terminal displays the message "The latest electricity bill is 5,000 yen."

[0151] Example prompt

[0152] An example prompt for using a generative AI model is:

[0153] "Please tell me the details of your latest electricity bill."

[0154] How do I read an electric meter?

[0155] As a result, the system can handle requests in a variety of data formats and provide quick and appropriate responses.

[0156] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0157] Step 1:

[0158] A user uses a terminal to input a request in the form of voice, text, image, or video. For example, a user makes a voice request to a smartphone saying, "I want to see my electricity bill details." In this case, the input is voice data, and the voice data is saved in the terminal as output.

[0159] Step 2:

[0160] The device sends the request data received from the user to the server. At this time, the device sends the voice data to the server as an HTTP request. The input is the voice data stored on the device, and the voice data is transferred to the server as the output.

[0161] Step 3:

[0162] The server preprocesses the received voice data. The server uses a speech recognition engine to convert the voice data into text data. Specifically, it uses the Google Cloud Speech-to-Text API to convert the voice data, such as "I want to see my electricity bill details," into text. The input is voice data, and text data is generated as the output.

[0163] Step 4:

[0164] The server inputs the preprocessed text data into a generative AI model to analyze the user's intent. Using OpenAI's GPT-3, the server inputs the text "I want to see my electricity bill details" and analyzes the user's intent as "I want to know the details of my current electricity bill." The input is text data, and the analyzed user's intent is output.

[0165] Step 5:

[0166] Based on the user's intent, the server retrieves the necessary information by referencing an internal database or an external interface. The server queries the internal database and retrieves the latest electricity bill data (e.g., 5,000 yen). The input is the parsed user's intent, and the electricity bill data is obtained as the output.

[0167] Step 6:

[0168] Based on the information acquired, the server generates an appropriate response for the user. The server uses a generative AI model to generate a text response such as "The latest electricity bill is 5,000 yen." The input is the electricity bill data, and the generated text response is obtained as the output.

[0169] Step 7:

[0170] The server sends the generated response to the terminal. The server sends the generated text data to the terminal as an HTTP response. The input is the generated text response, and the text data is sent to the terminal as output.

[0171] Step 8:

[0172] The terminal displays the response received from the server to the user. The terminal displays a text message on the screen saying "The latest electricity bill is 5000 yen." The input is the text data sent from the server, and the output is the text message displayed on the user's screen.

[0173] The above steps show in detail the process from when a user sends a request to the system to when an appropriate response is received. This system realizes quick and appropriate responses to a variety of data formats.

[0174] (Application example 1)

[0175] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0176] Modern content delivery services require systems that allow users to make requests using a variety of modalities (voice, text, images, and videos) and provide prompt and appropriate responses. However, existing systems often only support specific modalities, and there are few general-purpose systems that support all modalities. Furthermore, there are also limited systems that utilize generative AI models to provide optimal recommendations for user requests and improve the quality of information provided. Therefore, a system that can respond to requests in different modalities and generate optimal responses using generative AI models is needed.

[0177] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0178] In this invention, the server includes: means for receiving request data from a user in at least one of the following formats: voice, text, image, and video; means for transmitting the received request data to the server; means for preprocessing the request data, converting voice data to text, and analyzing image data; means for analyzing the user's intent using the preprocessed data; means for obtaining necessary information by referencing an internal database or an external API; means for generating an appropriate response based on the obtained information; means for generating a prompt sentence in response to the user's request using a generative AI model; means for generating an optimal response using the generated prompt sentence; means for transmitting the generated response to a terminal; and means for the terminal to display the response to the user. This enables prompt and appropriate responses to be provided even when the user makes requests in various formats. Furthermore, the use of a generative AI model enables higher quality recommendations and information provision.

[0179] A "user" is a person using a terminal that makes requests and receives responses in a variety of forms.

[0180] "Request data" is data sent by a user in one or more formats: voice, text, image, video.

[0181] "Preprocessing" refers to the process in which the server analyzes the request data received, converts voice data into text, and extracts necessary information from image data.

[0182] A "voice recognition system" is a system that converts voice data into text data.

[0183] An "image analysis system" is a system that analyzes image data and extracts necessary information.

[0184] "Intent analysis" is the process of using preprocessed data to understand the purpose and intent of a user request.

[0185] An "internal database" is a database stored within the system that is referenced to obtain required information.

[0186] An "external API" is an interface for communicating with services or databases outside the system.

[0187] A "generative AI model" is an artificial intelligence model that generates appropriate responses to user requests.

[0188] A "prompt sentence" is an input sentence to a generative AI model, which is generated based on a user request.

[0189] A "terminal" is a device through which a user sends requests and receives responses from a server.

[0190] A "server" is a computer system that processes request data, analyzes intent, retrieves information, and generates a response.

[0191] The present invention relates to a customer service system capable of processing requests of various modalities, and a specific embodiment thereof will be described below. The system is composed of three main components: a server, a terminal, and a user.

[0192] Receiving a user request

[0193] 1. The user sends a request to the device. The request can be in the form of voice, text, image, video, etc. For example, a user can make a voice request such as, "Tell me the next action movie I should watch."

[0194] 2. The terminal sends the received request data to the server.

[0195] Data Preprocessing

[0196] 1. The server preprocesses the data it receives: if it is voice data, it converts it into text using a voice recognition system, and if it is image data, it uses an image analysis system to extract the necessary information.

[0197] 2. As a speech recognition system, for example, use the speech_recognition library.

[0198] 3. PIL (Python Imaging Library) and pytesseract are used as image analysis systems.

[0199] Intent Analysis

[0200] 1. The server analyzes the user's intent based on the preprocessed data. For example, it understands the intent of "what action movie should I watch next" from the speech data.

[0201] Obtaining information

[0202] 1. The server retrieves the required information from an internal database or an external API. For example, it retrieves a list of recommended action movies from an internal database.

[0203] Generating a response

[0204] 1. Based on the information obtained by the server, an appropriate prompt sentence is generated using a generative AI model.

[0205] 2. The generated prompt sentence will be something like, "What's the next action movie I should watch?"

[0206] 3. Using this prompt, the generative AI model generates an appropriate response, such as "The recommended action movie is 'Recommended Movie Title'."

[0207] Sending and Displaying Responses

[0208] 1. The server generates a response and sends it to the device.

[0209] 2. The device displays a response to the user, for example, "The recommended action movie is 'Recommended Movie Title'."

[0210] Specific examples

[0211] For example, if a user speaks to their device, "What's the next action movie I should watch?", the process goes like this:

[0212] 1. The user's voice data is sent by the terminal to the server.

[0213] 2. The server uses a voice recognition system to convert the voice data into text.

[0214] 3. The server analyzes the user's intent based on the converted text data and generates an appropriate prompt.

[0215] 4. The generative AI model generates a response based on the prompt, "What's the next action movie I should watch?"

[0216] 5. The generated response will be "The recommended action movie is 'Recommended Movie Title'."

[0217] 6. The terminal displays this response to the user.

[0218] This embodiment allows users to easily make different types of requests and receive a quick and appropriate response, an example of a prompt sentence could be "What's the next action movie I should watch?"

[0219] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0220] Step 1:

[0221] The user sends request data. The request can be in the form of voice, text, image, or video. For example, the user might request by voice, "Tell me the next action movie I should watch." This voice data is input into the terminal.

[0222] Step 2:

[0223] The terminal transmits the received voice data to the server. The terminal uses the network to transmit the voice data and sends this data to the server. This inputs the voice data to the server.

[0224] Step 3:

[0225] The server preprocesses the received voice data. Specifically, it converts the voice data into text using the speech_recognition library. This process converts the voice data into text data.

[0226] Step 4:

[0227] The server analyzes the user's intent based on the converted text data. For example, from the text "Tell me the next action movie I should watch," it extracts the intent that the user is looking for movie recommendations.

[0228] Step 5:

[0229] The server retrieves the required information from an internal database or an external API. For example, it retrieves a list of action movies from an internal database. This process retrieves the data corresponding to the user's request.

[0230] Step 6:

[0231] Based on the information acquired by the server, a prompt sentence is generated using a generative AI model. For example, the prompt sentence might be, "Tell me the next action movie I should watch." This generated prompt sentence is input into the generative AI model.

[0232] Step 7:

[0233] The server uses the generative AI model to generate the optimal response based on the prompt. For example, the response generated is "The recommended action movie is 'Recommended Movie Title'." This response becomes the output of the generative AI model.

[0234] Step 8:

[0235] The server sends the generated response to the terminal, and this response message is sent to the terminal via the network and is input to the terminal.

[0236] Step 9:

[0237] The device displays the response received from the server to the user. For example, a message such as "The recommended action movie is 'Recommended Movie Title'" is displayed on the device screen.

[0238] This series of processing steps allows users to make requests in a variety of formats and receive appropriate responses in real time.

[0239] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0240] The present invention provides a customer service system capable of processing requests from multiple modalities, and in particular adds a function for recognizing a user's emotions and adjusting the content of responses accordingly. This makes it possible to provide a more sophisticated and personalized user experience. Specific embodiments of this system are described below.

[0241] Program Overview

[0242] The system's main components are a server, a terminal, and a user, combined with an emotion engine. The emotion engine analyzes emotions from user request data and reflects them in the response. The server is responsible for processing the AI ​​model and managing data, while the terminal provides the user interface, i.e., inputting requests and displaying responses.

[0243] Program Details

[0244] Receiving a user request

[0245] The user makes a request through the device, which can be input in the form of text, voice, image, or video.

[0246] Example: A user texts a device asking, "What's my electricity bill this month?"

[0247] Submitting a Request

[0248] The terminal transmits the received request data to the server.

[0249] Example: The device sends text data such as "What is the electricity bill this month?" to the server.

[0250] Data Preprocessing

[0251] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[0252] Example: The server receives voice data and converts it into text using a speech recognition system: "What is the electricity bill this month?"

[0253] Intention analysis and emotion recognition

[0254] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions and also analyzes the user's emotions using an emotion engine.

[0255] Example: From the text "What's the electricity bill this month?", an AI model interprets the user's intent as "I want to know the current bill amount," and an emotion engine recognizes the user's tension or irritation.

[0256] Database Reference

[0257] The server references an internal database or external API to obtain the necessary information.

[0258] Example: The server retrieves a user's account information from a database and retrieves the latest electricity bill information.

[0259] Response generation and emotional reflection

[0260] Based on the information acquired by the server, an appropriate response is generated taking into account the user's emotions.

[0261] Example: The server generates "This month's electricity bill is 5000 yen," but recognizing the user's frustration, adds an additional "Do you need help with payment methods?"

[0262] Sending a Response

[0263] The server generates a response and sends it to the terminal.

[0264] Example: The server sends a text to the device saying, "This month's electricity bill is 5000 yen. Do you need help with payment?"

[0265] Viewing the response

[0266] The terminal displays the response received from the server to the user.

[0267] Example: The device displays a text message to the user saying, "Your electricity bill this month is 5000 yen. Do you need help paying?"

[0268] Specific examples

[0269] As an example, the process will be described when a user asks a question by voice, "The electric meter reading is not working properly," and feels uneasy.

[0270] 1. Receiving a user request

[0271] The user asks the device a question by voice, "The electric meter reading is not working properly." The voice contains a sense of anxiety.

[0272] The terminal receives the audio data.

[0273] 2. Submitting a Request

[0274] The terminal transmits the voice data to the server.

[0275] 3. Data Preprocessing

[0276] The server converts the audio data into text.

[0277] The speech recognition system generates the text "I'm having trouble reading my electric meter."

[0278] 4. Intention Analysis and Emotion Recognition

[0279] The server inputs the text data into an AI model and analyzes the user's intent.

[0280] The emotion engine analyzes the user's anxiety.

[0281] 5. Referencing the database

[0282] The server retrieves guide information on how to read the electricity meter from an internal database.

[0283] 6. Response generation and emotional reflection

[0284] The server generates a response saying, "Take a photo of your electric meter and tap the camera icon in the chat window," and includes an additional message to address the user's concerns: "If you have any concerns, please contact support."

[0285] 7. Sending the Response

[0286] The server generates a response and sends it to the terminal.

[0287] 8. Displaying the Response

[0288] The device will display a text message to the user saying, "Take a photo of your electric meter and tap the camera icon in the chat screen. If you have any concerns, please contact support."

[0289] This system enables us to provide detailed responses that take into account the user's emotions, improving the quality of the user experience.

[0290] The processing flow will be explained below.

[0291] Step 1:

[0292] The user makes a request through the device, which can be input in the form of text, voice, image, or video.

[0293] Example: A user texts a device asking, "What's my electricity bill this month?"

[0294] Step 2:

[0295] The terminal transmits the received request data to the server.

[0296] Example: The device sends text data such as "What is the electricity bill this month?" to the server.

[0297] Step 3:

[0298] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[0299] Example: The server receives voice data and converts it into text using a speech recognition system: "What is the electricity bill this month?"

[0300] Step 4:

[0301] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions, and then analyzes the user's emotions using an emotion engine.

[0302] Example: From the text "What's the electricity bill this month?", an AI model interprets the user's intent as "I want to know the current bill amount," and the emotion engine recognizes the user's tension or irritation.

[0303] Step 5:

[0304] The server references an internal database or external API to obtain the necessary information.

[0305] Example: The server retrieves a user's account information from a database and retrieves the latest electricity bill information.

[0306] Step 6:

[0307] The server generates an appropriate response based on the acquired information, taking into account the user's emotions based on the emotion engine.

[0308] Example: The server generates the text "This month's electricity bill is 5000 yen", but recognizes the user's frustration and adds "This month's electricity bill is 5000 yen. Do you need help paying?"

[0309] Step 7:

[0310] The server generates a response and sends it to the terminal.

[0311] Example: The server sends a text to the device saying, "This month's electricity bill is 5000 yen. Do you need help with payment?"

[0312] Step 8:

[0313] The terminal displays the response received from the server to the user.

[0314] Example: The device displays a text message to the user saying, "Your electricity bill this month is 5000 yen. Do you need help paying?"

[0315] The above is the specific flow of processing for a user request.

[0316] Example 2

[0317] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0318] Conventional customer service systems simply respond to user requests and do not take the user's emotions into consideration when responding. This does not alleviate user dissatisfaction or frustration, resulting in a poor user experience. The present invention aims to provide a more sophisticated and personalized user experience by analyzing the user's emotions and generating responses accordingly.

[0319] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0320] In this invention, the server includes means for analyzing the user's intention, means for analyzing the user's intention using preprocessed data, means for obtaining necessary information by referencing an internal database or an external API, means for analyzing the user's emotions and adjusting the response content to reflect the emotions, and means for transmitting the generated response to the terminal, thereby enabling the server to generate an appropriate response that takes the user's emotions into consideration and improve the quality of the user experience.

[0321] A "user" is a user who inputs a request to the system.

[0322] "Audio, text, images, video" are different forms of data that users use to make requests to the system.

[0323] "Request data" is data containing information sent by a user, and represents a question or request.

[0324] A "terminal" is a device used by a user to enter requests and receive responses.

[0325] A "server" is a central computer system that processes request data sent by users and generates and sends responses.

[0326] "Preprocessing" refers to initial processing for converting request data into a format that is easy to analyze, and includes, for example, converting voice data into text and analyzing image data.

[0327] A "voice recognition system" is a technology for converting voice data into text data.

[0328] An "image analysis system" is a technology for extracting necessary information from image data.

[0329] "Intent analysis" is the process of understanding the intent of a user request.

[0330] "Sentiment analysis" is the process of identifying a user's emotional state from their request data.

[0331] An "internal database" is an information storage location installed within the system, which contains information about users and data necessary for providing services.

[0332] An "external API" is an interface for accessing external services and databases and obtaining information.

[0333] "Response generation" is the process of creating an appropriate reply to the user based on the analyzed intent and sentiment.

[0334] "Adjusting the content of a response" refers to changing or complementing the content of a generated response to match the emotional state of the user.

[0335] The present invention is a customer service system that can process requests from multiple modalities, and in particular has the function of recognizing the user's emotions and adjusting the response content. This system is composed of major components: a server, a terminal, a user, and an emotion engine, and operates as follows.

[0336] The user makes a request through the device. The request can be input in the form of voice, text, image, or video. For example, the user might ask the device in text, "What is my electricity bill this month?" The device then sends the received request data to the server. If the data is voice, the server uses a voice recognition system to convert it into text, and if it is image data, it uses an image analysis system to extract the required information. Specifically, this could involve using a voice recognition service such as Google Cloud Speech-to-Text API or Amazon Transcribe.

[0337] The preprocessed data is input into an AI model on the server, where the user's intent is analyzed. The emotion engine also analyzes the user's emotions. For example, the server uses OpenAI's GPT model to analyze the intent from the text "What is my electricity bill this month?", and the emotion engine recognizes the user's tension or irritation. This allows the system to interpret the user's intent as "I want to know the current bill amount" and also take the user's emotions into account.

[0338] The server then references internal databases and external APIs to retrieve the necessary information. Specifically, it retrieves the user's account information and latest electricity bill information from databases such as AWS DynamoDB and Google Firestore. Based on this information, the server generates an appropriate response that takes the user's emotions into account. For example, the server uses a generative AI model to generate a response such as "This month's electricity bill is 5000 yen," but recognizes the user's frustration and includes an additional message such as "Do you need help with your payment method?"

[0339] The generated response is sent from the server to the terminal, which then displays this response to the user. For example, the terminal might display the following text message to the user: "This month's electricity bill is 5000 yen. Do you need help with payment?"

[0340] In this way, the present invention enables a finely tuned response that takes into account the user's emotions, improving the quality of the user experience.

[0341] Specific examples

[0342] Consider a case where a user voices a question saying, "I'm having trouble reading my electric meter," and feels anxious.

[0343] 1. The user asks the device a question by voice, "The electric meter reading is not working properly." The user expresses anxiety.

[0344] 2. The device sends the audio data to the server.

[0345] 3. The server converts the voice data into text, typically using the Google Cloud Speech-to-Text API.

[0346] 4. Use AI models to analyze intent and emotion engines to analyze anxiety.

[0347] 5. The server retrieves guidance information on how to read the electricity meter from its internal database.

[0348] 6. The server generates a response that says "Take a photo of your electric meter and tap the camera icon in the chat window," with an additional message that says "If you have any concerns, please contact support."

[0349] 7. The response is sent to the terminal, which displays it to the user.

[0350] This allows the user to feel at ease and solve the problem.

[0351] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0352] Step 1:

[0353] The user inputs a request into the device, which can be in the form of voice, text, image, or video.

[0354] Example: A user texts a device asking, "What's my electricity bill this month?"

[0355] Input: User request (e.g., text format)

[0356] Output: Request data in the terminal

[0357] Step 2:

[0358] The terminal transmits the received request data to the server.

[0359] Specific operation: The terminal sends the request data to the server via the network.

[0360] Input: Request data in the terminal

[0361] Output: Request data to the server

[0362] Step 3:

[0363] The server preprocesses the request data received: if it is voice data, it converts it into text using a voice recognition system, and if it is image data, it uses an image analysis system to extract the necessary information.

[0364] What happens: The server uses a speech recognition API (e.g., Google Cloud Speech-to-Text) to convert the speech to text.

[0365] Input: Request data to the server (e.g., audio data)

[0366] Output: Preprocessed text data

[0367] Step 4:

[0368] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions, and also analyzes the user's emotions using an emotion engine.

[0369] How it works: The server uses a generative AI model (e.g., OpenAI GPT) to analyze the intent and evaluates the user's emotions with an emotion engine.

[0370] Input: Preprocessed text data

[0371] Output: Parsed intent and emotion data

[0372] Step 5:

[0373] The server retrieves the required information by referencing an internal database or an external API.

[0374] Specific operation: The server retrieves the user's account information from a database (e.g., AWS DynamoDB) and collects related information.

[0375] Input: Parsed intent and emotion data

[0376] Output: Required information (e.g. electricity bill information)

[0377] Step 6:

[0378] Based on the information obtained by the server, the user's emotions are taken into consideration and an appropriate response is generated.

[0379] What it does: The server uses a generative AI model to generate a response like "Your electricity bill this month is 5000 yen" and includes an additional message based on sentiment data: "Do you need help with payment methods?"

[0380] Input: Required information (e.g., electricity bill information) and emotion data

[0381] Output: The generated response

[0382] Step 7:

[0383] The server generates a response and sends it to the terminal.

[0384] Specific operation: The server sends response data to the terminal using the network.

[0385] Input: The generated response

[0386] Output: Response data to the terminal

[0387] Step 8:

[0388] The terminal displays the response received from the server to the user.

[0389] Specific operation: The terminal displays a response message on the display.

[0390] Input: Response data to the terminal

[0391] Output: The response message that is displayed to the user

[0392] (Application example 2)

[0393] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0394] Current customer service systems can only provide standard responses to user requests, making it difficult to provide a personalized user experience. Furthermore, they lack the ability to recognize user emotions and respond appropriately, resulting in insufficient responses when users feel dissatisfied or anxious. Therefore, improving the quality of the user experience is a challenge in areas such as food delivery, where fast and personalized responses are required.

[0395] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0396] In this invention, the server includes a means for analyzing user emotions, a means for analyzing user intentions using preprocessed data, and a means for obtaining necessary information by referencing an internal database or an external API, thereby enabling the server to generate an appropriate response according to the user's emotions.

[0397] "User request data" refers to information provided by a user to the system in the form of voice, text, images, or video.

[0398] A "server" is a computing system that receives, pre-processes, and generates a response from a user's request data.

[0399] A "voice recognition system" is a technology that converts voice data into text.

[0400] An "image analysis system" is a technology that extracts specific information from image data.

[0401] "Means for analyzing user intent" refers to a process for understanding the user's purpose and requirements from the user's request data.

[0402] "Means for analyzing user emotions" refers to technology that recognizes and classifies the user's emotional state based on request data.

[0403] An "internal database" is a database that stores internal information managed by the system.

[0404] An "External API" is an application programming interface for accessing services or data outside the system.

[0405] The "means for generating an appropriate response" is a process for generating an optimal response to the user based on the acquired information and the analyzed user's emotions.

[0406] A "terminal" is a device through which a user inputs requests and receives responses.

[0407] The present invention provides a customer service system capable of processing requests from multiple modalities, and in particular adds a function for recognizing a user's emotions and adjusting the content of responses, thereby enabling a more sophisticated and personalized user experience. An embodiment of this system will be described below.

[0408] System Configuration

[0409] The system consists of a server, a terminal, a user, and an emotion engine as its main components. The server is responsible for processing the AI ​​model and managing data, while the terminal provides the user interface, i.e., inputting requests and displaying responses.

[0410] Hardware and Software Specifications

[0411] Server: A high-performance computer system is required.

[0412] Device: Using a user interface device such as a smartphone, smart glasses, a head-mounted display, or a robot.

[0413] Software used:

[0414] AI models: Use Natural Language Processing (NLP) libraries and models for sentiment analysis.

[0415] Speech recognition system: Used to convert voice data into text.

[0416] Image analysis system: Used to analyze image data and extract necessary information.

[0417] Emotion Engine: Used to analyze emotions from user request data.

[0418] Program processing flow

[0419] Receiving a user request

[0420] The user makes a request through the device. The request can be input in the form of voice, text, image, or video. For example, the user can type "My order hasn't arrived" into their smartphone.

[0421] Submitting a Request

[0422] The terminal transmits the received request data to the server.

[0423] Data Preprocessing

[0424] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[0425] Intention analysis and emotion recognition

[0426] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions and also analyzes the user's emotions using an emotion engine.

[0427] Database Reference

[0428] The server references an internal database or external API to obtain the necessary information.

[0429] Response generation and emotional reflection

[0430] Based on the information acquired by the server, the server generates an appropriate response taking into account the user's emotions. For example, it generates a response that takes into account emotions, such as "We apologize for the delay in your order. We are currently checking the current status. Would you mind waiting?"

[0431] Sending a Response

[0432] The server generates a response and sends it to the terminal.

[0433] Viewing the response

[0434] The device displays the response it receives from the server to the user. For example, a message like "We apologize for the delay in your order. We are currently checking the current status. Would you mind waiting?" appears on the smartphone screen.

[0435] Specific examples

[0436] Let's take a concrete example: when a user types "My order hasn't arrived" into their smartphone, the request is sent to the server, which preprocesses the data using a speech recognition system and image analysis system. The emotion engine then analyzes the user's frustration and generates an appropriate response using a generative AI model.

[0437] Prompt Sentence Examples

[0438] Analyze the sentiment when a user inquires about the status of an order in a food delivery app. Analyze the following user input text and generate a response based on the sentiment:

[0439] User Input: "Order not received"

[0440] Expected output:

[0441] User Sentiment: Annoyance

[0442] Response: "Sorry for the delay in your order. We're currently checking the status. Would you mind waiting?"

[0443] By using prompt sentences in this way, the generative AI model can recognize emotions and generate responses appropriately.

[0444] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0445] Step 1:

[0446] The user enters a request

[0447] Input: Text or speak "My order hasn't arrived" into your smartphone or other device.

[0448] Action: The user interacts with the device and types a request, which may include expressions of frustration or annoyance.

[0449] Output: The device receives the user's request data.

[0450] Step 2:

[0451] Sends request data to the server

[0452] Input: User request data obtained by the device.

[0453] What happens: Your device sends data to a server over Wi-Fi or your mobile network.

[0454] Output: The server receives the request data.

[0455] Step 3:

[0456] Data Preprocessing

[0457] Input: The request data received by the server.

[0458] How it works: The server converts voice data into text using a voice recognition system, and analyzes image data using an image analysis system.

[0459] Output: Text data and analysis results are obtained.

[0460] Step 4:

[0461] Intention analysis and emotion recognition

[0462] Input: Text data preprocessed by the server and analysis results.

[0463] How it works: The server inputs text data into the AI ​​model, analyzes the user's intention, and recognizes the user's emotions using the emotion engine.

[0464] Output: Data is obtained that analyzes the user's intent and emotions.

[0465] Step 5:

[0466] Referencing a database or external API

[0467] Input: User intent and sentiment parsed by the server.

[0468] What happens: The server looks up an internal database or an external API to get the information it needs. For example, it looks up a database to check the status of an order.

[0469] Output: The required information (e.g., current order status) is obtained.

[0470] Step 6:

[0471] Response generation and emotional reflection

[0472] Input: Captured information and analyzed user sentiment.

[0473] How it works: Based on the information the server obtains, it generates a response that takes into account the user's feelings. For example, it generates a message like, "We're sorry your order is delayed. We're currently checking the current status. Would you mind waiting?"

[0474] Output: A response message is generated that takes the user's feelings into consideration.

[0475] Step 7:

[0476] Sending a Response

[0477] Input: The server-generated response message.

[0478] Action: The server sends a response message to the terminal.

[0479] Output: The terminal receives the response message.

[0480] Step 8:

[0481] Viewing the response

[0482] Input: Response message received by the terminal.

[0483] What it does: The device displays a response message to the user. For example, a message on a smartphone screen might say, "We apologize for the delay in your order. We're currently checking the current status. Would you mind waiting?"

[0484] Output: The user confirms the response message and takes further action if necessary.

[0485] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the 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.

[0486] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0487] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0488] [Second embodiment]

[0489] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0490] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0491] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0492] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0493] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0494] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0495] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0496] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0497] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[0498] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0499] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0500] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. 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."

[0501] The present invention relates to a customer service system that can process requests in various modalities. It supports each modality (text, voice, image, video) and provides appropriate responses to user requests in real time. Specific embodiments of this system are described below.

[0502] Program Overview

[0503] This system consists of three main components: a server, a terminal, and a user. The server is responsible for processing the AI ​​model and managing data, while the terminal provides an interface for users to send requests and receive responses, allowing users to ask questions or make requests. The cooperation of these components creates a mechanism that can appropriately respond to a variety of user requests.

[0504] Program Details

[0505] Receiving a user request

[0506] The user makes a request through the device, which can be in the form of text, voice, image, or video.

[0507] Example: A user types the text "I want to see my electricity bill details" into the terminal.

[0508] Submitting a Request

[0509] The terminal transmits the received request data to the server.

[0510] Example: The terminal sends text data to the server saying, "I want to see my electricity bill details."

[0511] Data Preprocessing

[0512] The server preprocesses the received data: voice data is converted into text using a voice recognition system, and image data is extracted using an image analysis system.

[0513] Example: When the server converts voice data to text, it uses speech recognition to generate the text "I would like to see my electricity bill details."

[0514] Intent Analysis

[0515] The server inputs the preprocessed data into the AI ​​model to understand the user's intent, and then proceeds with the process based on the user's intent.

[0516] Example: From the text "I want to see my electricity bill details," the user's intent is interpreted as "I want to check the details of my current electricity bill."

[0517] Database Reference

[0518] The server references an internal database or external API to obtain the necessary information.

[0519] Example: The server looks up the user's account information in a database and retrieves the latest electricity bill data.

[0520] Generating a response

[0521] Based on the information obtained, the server generates an appropriate response to the user.

[0522] Example: Generate a response in text format saying "Your latest electricity bill is 5000 yen."

[0523] Sending a Response

[0524] The server generates a response and sends it to the terminal.

[0525] Example: The server sends text data to the terminal saying, "The latest electricity bill is 5,000 yen."

[0526] Viewing the response

[0527] The terminal displays the response received from the server to the user.

[0528] Example: The device displays a text message to the user saying, "Your latest electricity bill is 5000 yen."

[0529] Specific examples

[0530] The following is an example of the process when a user verbally asks the terminal, "I'm having trouble reading the electricity meter."

[0531] 1. Receiving a user request

[0532] The user verbally asks the terminal, "I'm having trouble reading the electric meter."

[0533] The terminal receives the audio data.

[0534] 2. Submitting a Request

[0535] The terminal transmits the voice data to the server.

[0536] 3. Data Preprocessing

[0537] The server converts the audio data into text.

[0538] The speech recognition system generates the text "I'm having trouble reading my electric meter."

[0539] 4. Intent Analysis

[0540] The server inputs the text data into an AI model and analyzes the user's intent.

[0541] The user's intent is interpreted as "I want to know how to read my electric meter."

[0542] 5. Referencing the database

[0543] The server retrieves guide information on how to read the electricity meter from an internal database.

[0544] 6. Generating the Response

[0545] The server generates a response saying, "Take a photo of your electric meter and tap the camera icon in the chat window."

[0546] 7. Sending the Response

[0547] The server generates a response and sends it to the terminal.

[0548] 8. Displaying the Response

[0549] The device displays a text message to the user saying, "Take a photo of your electricity meter and tap the camera icon on the chat screen."

[0550] This system can handle a variety of data formats and provide users with prompt and appropriate responses in real time.

[0551] The processing flow will be explained below.

[0552] Step 1:

[0553] The user makes a request through the device, which can be input in the form of text, voice, image, or video.

[0554] Example: A user texts a device asking, "What's my electricity bill this month?"

[0555] Step 2:

[0556] The terminal transmits the received request data to the server.

[0557] Example: The device sends text data such as "What is the electricity bill this month?" to the server.

[0558] Step 3:

[0559] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[0560] Example: The server receives voice data and converts it into text using a speech recognition system: "What is the electricity bill this month?"

[0561] Step 4:

[0562] The server inputs the preprocessed data into an AI model to analyze the user's intent.

[0563] Example: An AI model interprets the text "What is the electricity bill this month?" as the user's intent to "I want to know the current bill amount."

[0564] Step 5:

[0565] The server references an internal database or external API to obtain the necessary information.

[0566] Example: The server retrieves a user's account information from a database and retrieves the latest electricity bill information.

[0567] Step 6:

[0568] The server generates an appropriate response based on the information it has obtained.

[0569] Example: The server generates a text response saying "This month's electricity bill is 5000 yen."

[0570] Step 7:

[0571] The server generates a response and sends it to the terminal.

[0572] Example: The server sends the text "This month's electricity bill is 5,000 yen" to the terminal.

[0573] Step 8:

[0574] The terminal displays the response received from the server to the user.

[0575] Example: The device displays a text message to the user saying, "This month's electricity bill is 5000 yen."

[0576] The above is the flow of processing for a user request.

[0577] Example 1

[0578] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0579] Conventional customer service systems have had difficulty providing appropriate responses to user requests in real time. Furthermore, few systems support multiple modalities (voice, text, images, and video), creating a need for improved service quality. Furthermore, accurately analyzing user intent and quickly providing appropriate information is a challenge.

[0580] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0581] In this invention, the server includes means for receiving request data from a user in at least one form of voice, text, image, or video, means for preprocessing the received request data, converting voice data into text, and analyzing image data, means for using a generative model to analyze the user's intention using the preprocessed data, means for obtaining necessary information by referring to an internal database or an external interface, means for generating an appropriate response based on the obtained information, means for transmitting the generated response to a terminal, and means for the terminal to display the response to the user. This makes it possible to support various modalities, accurately analyze the user's intention, and provide appropriate information in real time.

[0582] A "user" is a person or entity that submits requests to a system to obtain information.

[0583] A "terminal" is a device through which a user inputs requests and receives responses.

[0584] A "server" is a computing system that processes requests from users and generates and sends appropriate responses to terminals.

[0585] "Request data" refers to voice, text, image, or video data that a user sends to the system through a terminal.

[0586] "Preprocessing" is the process of converting received request data into a format that is easy to analyze.

[0587] A "voice recognition system" is a technology that converts voice data into text data.

[0588] An "image analysis system" is a technology that analyzes image data and extracts necessary information.

[0589] A "generative model" is an AI technology that analyzes user intent based on preprocessed data.

[0590] An "internal database" is a system that manages data within a server.

[0591] An "external interface" is a means by which a server interacts with external systems and services to obtain data.

[0592] An "appropriate response" is data that contains information or instructions that are most appropriate to the user's request.

[0593] The present invention relates to a multi-modal customer service system that receives voice, text, image, and video request data from users and generates appropriate responses in real time.

[0594] System configuration

[0595] The system consists of three main components: a server, a terminal, and a user. The server is responsible for processing the AI ​​model and managing data, while the terminal provides an interface where users can send requests and receive responses. Users can ask questions or make requests and use the system's services.

[0596] Hardware and Software

[0597] Server: A high-performance computer system that preprocesses data, runs generative models, manages databases, and generates responses. Specific software includes speech recognition engines (e.g., Google Cloud Speech-to-Text) and generative AI models (e.g., OpenAI's GPT-3).

[0598] Terminal: A device that provides a user interface, such as a smartphone, tablet, or PC. The terminal uses a communication protocol (e.g., HTTP / HTTPS) to enable sending and receiving requests.

[0599] Initial Setup and Operation

[0600] 1. User request input: The user uses the device to send a request via voice, text, image, or video.

[0601] Example: A user makes a voice request to their smartphone saying, "I want to see my electricity bill details."

[0602] 2. Sending request data: The terminal sends the request data to the server via a communication protocol.

[0603] Example: A smartphone sends voice data saying "I want to see my electricity bill details" to a server via an HTTP request.

[0604] 3. Data preprocessing: The server preprocesses the received data. Voice data is converted into text using a voice recognition system, and image data is analyzed using an image analysis system.

[0605] Example: The server uses a speech recognition engine to convert the voice data into text, generating the text "I would like to see my electricity bill details."

[0606] 4. Intent analysis: The server inputs the preprocessed data into the generative AI model to analyze the user's intent.

[0607] Example: The server inputs the text "I want to see my electricity bill details" into the generated AI model, and analyzes the user's intent as "I want to know the details of my current electricity bill."

[0608] 5. Data retrieval: The server retrieves the required information using an internal database or an external interface.

[0609] Example: A server queries an internal database to get the latest electricity bill (e.g., 5000 yen).

[0610] 6. Generate a response: The server uses the information it has obtained to generate an appropriate response for the user.

[0611] Example: The server generates a text response saying "The latest electricity bill is 5000 yen."

[0612] 7. Send and display the response: Send the generated response to the terminal, which displays it to the user.

[0613] Example: The server sends generated text data to the terminal, and the terminal displays the message "The latest electricity bill is 5,000 yen."

[0614] Example prompt

[0615] An example prompt for using a generative AI model is:

[0616] "Please tell me the details of your latest electricity bill."

[0617] How do I read an electric meter?

[0618] As a result, the system can handle requests in a variety of data formats and provide quick and appropriate responses.

[0619] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0620] Step 1:

[0621] A user uses a terminal to input a request in the form of voice, text, image, or video. For example, a user makes a voice request to a smartphone saying, "I want to see my electricity bill details." In this case, the input is voice data, and the voice data is saved in the terminal as output.

[0622] Step 2:

[0623] The device sends the request data received from the user to the server. At this time, the device sends the voice data to the server as an HTTP request. The input is the voice data stored on the device, and the voice data is transferred to the server as the output.

[0624] Step 3:

[0625] The server preprocesses the received voice data. The server uses a speech recognition engine to convert the voice data into text data. Specifically, it uses the Google Cloud Speech-to-Text API to convert the voice data, such as "I want to see my electricity bill details," into text. The input is voice data, and text data is generated as the output.

[0626] Step 4:

[0627] The server inputs the preprocessed text data into a generative AI model to analyze the user's intent. Using OpenAI's GPT-3, the server inputs the text "I want to see my electricity bill details" and analyzes the user's intent as "I want to know the details of my current electricity bill." The input is text data, and the analyzed user's intent is output.

[0628] Step 5:

[0629] Based on the user's intent, the server retrieves the necessary information by referencing an internal database or an external interface. The server queries the internal database and retrieves the latest electricity bill data (e.g., 5,000 yen). The input is the parsed user's intent, and the electricity bill data is obtained as the output.

[0630] Step 6:

[0631] Based on the information acquired, the server generates an appropriate response for the user. The server uses a generative AI model to generate a text response such as "The latest electricity bill is 5,000 yen." The input is the electricity bill data, and the generated text response is obtained as the output.

[0632] Step 7:

[0633] The server sends the generated response to the terminal. The server sends the generated text data to the terminal as an HTTP response. The input is the generated text response, and the text data is sent to the terminal as output.

[0634] Step 8:

[0635] The terminal displays the response received from the server to the user. The terminal displays a text message on the screen saying "The latest electricity bill is 5000 yen." The input is the text data sent from the server, and the output is the text message displayed on the user's screen.

[0636] The above steps show in detail the process from when a user sends a request to the system to when an appropriate response is received. This system realizes quick and appropriate responses to a variety of data formats.

[0637] (Application example 1)

[0638] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0639] Modern content delivery services require systems that allow users to make requests using a variety of modalities (voice, text, images, and videos) and provide prompt and appropriate responses. However, existing systems often only support specific modalities, and there are few general-purpose systems that support all modalities. Furthermore, there are also limited systems that utilize generative AI models to provide optimal recommendations for user requests and improve the quality of information provided. Therefore, a system that can respond to requests in different modalities and generate optimal responses using generative AI models is needed.

[0640] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0641] In this invention, the server includes: means for receiving request data from a user in at least one of the following formats: voice, text, image, and video; means for transmitting the received request data to the server; means for preprocessing the request data, converting voice data to text, and analyzing image data; means for analyzing the user's intent using the preprocessed data; means for obtaining necessary information by referencing an internal database or an external API; means for generating an appropriate response based on the obtained information; means for generating a prompt sentence in response to the user's request using a generative AI model; means for generating an optimal response using the generated prompt sentence; means for transmitting the generated response to a terminal; and means for the terminal to display the response to the user. This enables prompt and appropriate responses to be provided even when the user makes requests in various formats. Furthermore, the use of a generative AI model enables higher quality recommendations and information provision.

[0642] A "user" is a person using a terminal that makes requests and receives responses in a variety of forms.

[0643] "Request data" is data sent by a user in one or more formats: voice, text, image, video.

[0644] "Preprocessing" refers to the process in which the server analyzes the request data received, converts voice data into text, and extracts necessary information from image data.

[0645] A "voice recognition system" is a system that converts voice data into text data.

[0646] An "image analysis system" is a system that analyzes image data and extracts necessary information.

[0647] "Intent analysis" is the process of using preprocessed data to understand the purpose and intent of a user request.

[0648] An "internal database" is a database stored within the system that is referenced to obtain required information.

[0649] An "external API" is an interface for communicating with services or databases outside the system.

[0650] A "generative AI model" is an artificial intelligence model that generates appropriate responses to user requests.

[0651] A "prompt sentence" is an input sentence to a generative AI model, which is generated based on a user request.

[0652] A "terminal" is a device through which a user sends requests and receives responses from a server.

[0653] A "server" is a computer system that processes request data, analyzes intent, retrieves information, and generates a response.

[0654] The present invention relates to a customer service system capable of processing requests of various modalities, and a specific embodiment thereof will be described below. The system is composed of three main components: a server, a terminal, and a user.

[0655] Receiving a user request

[0656] 1. The user sends a request to the device. The request can be in the form of voice, text, image, video, etc. For example, a user can make a voice request such as, "Tell me the next action movie I should watch."

[0657] 2. The terminal sends the received request data to the server.

[0658] Data Preprocessing

[0659] 1. The server preprocesses the data it receives: if it is voice data, it converts it into text using a voice recognition system, and if it is image data, it uses an image analysis system to extract the necessary information.

[0660] 2. As a speech recognition system, for example, use the speech_recognition library.

[0661] 3. PIL (Python Imaging Library) and pytesseract are used as image analysis systems.

[0662] Intent Analysis

[0663] 1. The server analyzes the user's intent based on the preprocessed data. For example, it understands the intent of "what action movie should I watch next" from the speech data.

[0664] Obtaining information

[0665] 1. The server retrieves the required information from an internal database or an external API. For example, it retrieves a list of recommended action movies from an internal database.

[0666] Generating a response

[0667] 1. Based on the information obtained by the server, an appropriate prompt sentence is generated using a generative AI model.

[0668] 2. The generated prompt sentence will be something like, "What's the next action movie I should watch?"

[0669] 3. Using this prompt, the generative AI model generates an appropriate response, such as "The recommended action movie is 'Recommended Movie Title'."

[0670] Sending and Displaying Responses

[0671] 1. The server generates a response and sends it to the device.

[0672] 2. The device displays a response to the user, for example, "The recommended action movie is 'Recommended Movie Title'."

[0673] Specific examples

[0674] For example, if a user speaks to their device, "What's the next action movie I should watch?", the process goes like this:

[0675] 1. The user's voice data is sent by the terminal to the server.

[0676] 2. The server uses a voice recognition system to convert the voice data into text.

[0677] 3. The server analyzes the user's intent based on the converted text data and generates an appropriate prompt.

[0678] 4. The generative AI model generates a response based on the prompt, "What's the next action movie I should watch?"

[0679] 5. The generated response will be "The recommended action movie is 'Recommended Movie Title'."

[0680] 6. The terminal displays this response to the user.

[0681] This embodiment allows users to easily make different types of requests and receive a quick and appropriate response, an example of a prompt sentence could be "What's the next action movie I should watch?"

[0682] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0683] Step 1:

[0684] The user sends request data. The request can be in the form of voice, text, image, or video. For example, the user might request by voice, "Tell me the next action movie I should watch." This voice data is input into the terminal.

[0685] Step 2:

[0686] The terminal transmits the received voice data to the server. The terminal uses the network to transmit the voice data and sends this data to the server. This inputs the voice data to the server.

[0687] Step 3:

[0688] The server preprocesses the received voice data. Specifically, it converts the voice data into text using the speech_recognition library. This process converts the voice data into text data.

[0689] Step 4:

[0690] The server analyzes the user's intent based on the converted text data. For example, from the text "Tell me the next action movie I should watch," it extracts the intent that the user is looking for movie recommendations.

[0691] Step 5:

[0692] The server retrieves the required information from an internal database or an external API. For example, it retrieves a list of action movies from an internal database. This process retrieves the data corresponding to the user's request.

[0693] Step 6:

[0694] Based on the information acquired by the server, a prompt sentence is generated using a generative AI model. For example, the prompt sentence might be, "Tell me the next action movie I should watch." This generated prompt sentence is input into the generative AI model.

[0695] Step 7:

[0696] The server uses the generative AI model to generate the optimal response based on the prompt. For example, the response generated is "The recommended action movie is 'Recommended Movie Title'." This response becomes the output of the generative AI model.

[0697] Step 8:

[0698] The server sends the generated response to the terminal, and this response message is sent to the terminal via the network and is input to the terminal.

[0699] Step 9:

[0700] The device displays the response received from the server to the user. For example, a message such as "The recommended action movie is 'Recommended Movie Title'" is displayed on the device screen.

[0701] This series of processing steps allows users to make requests in a variety of formats and receive appropriate responses in real time.

[0702] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0703] The present invention provides a customer service system capable of processing requests from multiple modalities, and in particular adds a function for recognizing a user's emotions and adjusting the content of responses accordingly. This makes it possible to provide a more sophisticated and personalized user experience. Specific embodiments of this system are described below.

[0704] Program Overview

[0705] The system's main components are a server, a terminal, and a user, combined with an emotion engine. The emotion engine analyzes emotions from user request data and reflects them in the response. The server is responsible for processing the AI ​​model and managing data, while the terminal provides the user interface, i.e., inputting requests and displaying responses.

[0706] Program Details

[0707] Receiving a user request

[0708] The user makes a request through the device, which can be input in the form of text, voice, image, or video.

[0709] Example: A user texts a device asking, "What's my electricity bill this month?"

[0710] Submitting a Request

[0711] The terminal transmits the received request data to the server.

[0712] Example: The device sends text data such as "What is the electricity bill this month?" to the server.

[0713] Data Preprocessing

[0714] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[0715] Example: The server receives voice data and converts it into text using a speech recognition system: "What is the electricity bill this month?"

[0716] Intention analysis and emotion recognition

[0717] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions and also analyzes the user's emotions using an emotion engine.

[0718] Example: From the text "What's the electricity bill this month?", an AI model interprets the user's intent as "I want to know the current bill amount," and an emotion engine recognizes the user's tension or irritation.

[0719] Database Reference

[0720] The server references an internal database or external API to obtain the necessary information.

[0721] Example: The server retrieves a user's account information from a database and retrieves the latest electricity bill information.

[0722] Response generation and emotional reflection

[0723] Based on the information acquired by the server, an appropriate response is generated taking into account the user's emotions.

[0724] Example: The server generates "This month's electricity bill is 5000 yen," but recognizing the user's frustration, adds an additional "Do you need help with payment methods?"

[0725] Sending a Response

[0726] The server generates a response and sends it to the terminal.

[0727] Example: The server sends a text to the device saying, "This month's electricity bill is 5000 yen. Do you need help with payment?"

[0728] Viewing the response

[0729] The terminal displays the response received from the server to the user.

[0730] Example: The device displays a text message to the user saying, "Your electricity bill this month is 5000 yen. Do you need help paying?"

[0731] Specific examples

[0732] As an example, the process will be described when a user asks a question by voice, "The electric meter reading is not working properly," and feels uneasy.

[0733] 1. Receiving a user request

[0734] The user asks the device a question by voice, "The electric meter reading is not working properly." The voice contains a sense of anxiety.

[0735] The terminal receives the audio data.

[0736] 2. Submitting a Request

[0737] The terminal transmits the voice data to the server.

[0738] 3. Data Preprocessing

[0739] The server converts the audio data into text.

[0740] The speech recognition system generates the text "I'm having trouble reading my electric meter."

[0741] 4. Intention Analysis and Emotion Recognition

[0742] The server inputs the text data into an AI model and analyzes the user's intent.

[0743] The emotion engine analyzes the user's anxiety.

[0744] 5. Referencing the database

[0745] The server retrieves guide information on how to read the electricity meter from an internal database.

[0746] 6. Response generation and emotional reflection

[0747] The server generates a response saying, "Take a photo of your electric meter and tap the camera icon in the chat window," and includes an additional message to address the user's concerns: "If you have any concerns, please contact support."

[0748] 7. Sending the Response

[0749] The server generates a response and sends it to the terminal.

[0750] 8. Displaying the Response

[0751] The device will display a text message to the user saying, "Take a photo of your electric meter and tap the camera icon in the chat screen. If you have any concerns, please contact support."

[0752] This system enables us to provide detailed responses that take into account the user's emotions, improving the quality of the user experience.

[0753] The processing flow will be explained below.

[0754] Step 1:

[0755] The user makes a request through the device, which can be input in the form of text, voice, image, or video.

[0756] Example: A user texts a device asking, "What's my electricity bill this month?"

[0757] Step 2:

[0758] The terminal transmits the received request data to the server.

[0759] Example: The device sends text data such as "What is the electricity bill this month?" to the server.

[0760] Step 3:

[0761] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[0762] Example: The server receives voice data and converts it into text using a speech recognition system: "What is the electricity bill this month?"

[0763] Step 4:

[0764] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions, and then analyzes the user's emotions using an emotion engine.

[0765] Example: From the text "What's the electricity bill this month?", an AI model interprets the user's intent as "I want to know the current bill amount," and the emotion engine recognizes the user's tension or irritation.

[0766] Step 5:

[0767] The server references an internal database or external API to obtain the necessary information.

[0768] Example: The server retrieves a user's account information from a database and retrieves the latest electricity bill information.

[0769] Step 6:

[0770] The server generates an appropriate response based on the acquired information, taking into account the user's emotions based on the emotion engine.

[0771] Example: The server generates the text "This month's electricity bill is 5000 yen", but recognizes the user's frustration and adds "This month's electricity bill is 5000 yen. Do you need help paying?"

[0772] Step 7:

[0773] The server generates a response and sends it to the terminal.

[0774] Example: The server sends a text to the device saying, "This month's electricity bill is 5000 yen. Do you need help with payment?"

[0775] Step 8:

[0776] The terminal displays the response received from the server to the user.

[0777] Example: The device displays a text message to the user saying, "Your electricity bill this month is 5000 yen. Do you need help paying?"

[0778] The above is the specific flow of processing for a user request.

[0779] Example 2

[0780] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0781] Conventional customer service systems simply respond to user requests and do not take the user's emotions into consideration when responding. This does not alleviate user dissatisfaction or frustration, resulting in a poor user experience. The present invention aims to provide a more sophisticated and personalized user experience by analyzing the user's emotions and generating responses accordingly.

[0782] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0783] In this invention, the server includes means for analyzing the user's intention, means for analyzing the user's intention using preprocessed data, means for obtaining necessary information by referencing an internal database or an external API, means for analyzing the user's emotions and adjusting the response content to reflect the emotions, and means for transmitting the generated response to the terminal, thereby enabling the server to generate an appropriate response that takes the user's emotions into consideration and improve the quality of the user experience.

[0784] A "user" is a user who inputs a request to the system.

[0785] "Audio, text, images, video" are different forms of data that users use to make requests to the system.

[0786] "Request data" is data containing information sent by a user, and represents a question or request.

[0787] A "terminal" is a device used by a user to enter requests and receive responses.

[0788] A "server" is a central computer system that processes request data sent by users and generates and sends responses.

[0789] "Preprocessing" refers to initial processing for converting request data into a format that is easy to analyze, and includes, for example, converting voice data into text and analyzing image data.

[0790] A "voice recognition system" is a technology for converting voice data into text data.

[0791] An "image analysis system" is a technology for extracting necessary information from image data.

[0792] "Intent analysis" is the process of understanding the intent of a user request.

[0793] "Sentiment analysis" is the process of identifying a user's emotional state from their request data.

[0794] An "internal database" is an information storage location installed within the system, which contains information about users and data necessary for providing services.

[0795] An "external API" is an interface for accessing external services and databases and obtaining information.

[0796] "Response generation" is the process of creating an appropriate reply to the user based on the analyzed intent and sentiment.

[0797] "Adjusting the content of a response" refers to changing or complementing the content of a generated response to match the emotional state of the user.

[0798] The present invention is a customer service system that can process requests from multiple modalities, and in particular has the function of recognizing the user's emotions and adjusting the response content. This system is composed of major components: a server, a terminal, a user, and an emotion engine, and operates as follows.

[0799] The user makes a request through the device. The request can be input in the form of voice, text, image, or video. For example, the user might ask the device in text, "What is my electricity bill this month?" The device then sends the received request data to the server. If the data is voice, the server uses a voice recognition system to convert it into text, and if it is image data, it uses an image analysis system to extract the required information. Specifically, this could involve using a voice recognition service such as Google Cloud Speech-to-Text API or Amazon Transcribe.

[0800] The preprocessed data is input into an AI model on the server, where the user's intent is analyzed. The emotion engine also analyzes the user's emotions. For example, the server uses OpenAI's GPT model to analyze the intent from the text "What is my electricity bill this month?", and the emotion engine recognizes the user's tension or irritation. This allows the system to interpret the user's intent as "I want to know the current bill amount" and also take the user's emotions into account.

[0801] The server then references internal databases and external APIs to retrieve the necessary information. Specifically, it retrieves the user's account information and latest electricity bill information from databases such as AWS DynamoDB and Google Firestore. Based on this information, the server generates an appropriate response that takes the user's emotions into account. For example, the server uses a generative AI model to generate a response such as "This month's electricity bill is 5000 yen," but recognizes the user's frustration and includes an additional message such as "Do you need help with your payment method?"

[0802] The generated response is sent from the server to the terminal, which then displays this response to the user. For example, the terminal might display the following text message to the user: "This month's electricity bill is 5000 yen. Do you need help with payment?"

[0803] In this way, the present invention enables a finely tuned response that takes into account the user's emotions, improving the quality of the user experience.

[0804] Specific examples

[0805] Consider a case where a user voices a question saying, "I'm having trouble reading my electric meter," and feels anxious.

[0806] 1. The user asks the device a question by voice, "The electric meter reading is not working properly." The user expresses anxiety.

[0807] 2. The device sends the audio data to the server.

[0808] 3. The server converts the voice data into text, typically using the Google Cloud Speech-to-Text API.

[0809] 4. Use AI models to analyze intent and emotion engines to analyze anxiety.

[0810] 5. The server retrieves guidance information on how to read the electricity meter from its internal database.

[0811] 6. The server generates a response that says "Take a photo of your electric meter and tap the camera icon in the chat window," with an additional message that says "If you have any concerns, please contact support."

[0812] 7. The response is sent to the terminal, which displays it to the user.

[0813] This allows the user to feel at ease and solve the problem.

[0814] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0815] Step 1:

[0816] The user inputs a request into the device, which can be in the form of voice, text, image, or video.

[0817] Example: A user texts a device asking, "What's my electricity bill this month?"

[0818] Input: User request (e.g., text format)

[0819] Output: Request data in the terminal

[0820] Step 2:

[0821] The terminal transmits the received request data to the server.

[0822] Specific operation: The terminal sends the request data to the server via the network.

[0823] Input: Request data in the terminal

[0824] Output: Request data to the server

[0825] Step 3:

[0826] The server preprocesses the request data received: if it is voice data, it converts it into text using a voice recognition system, and if it is image data, it uses an image analysis system to extract the necessary information.

[0827] What happens: The server uses a speech recognition API (e.g., Google Cloud Speech-to-Text) to convert the speech to text.

[0828] Input: Request data to the server (e.g., audio data)

[0829] Output: Preprocessed text data

[0830] Step 4:

[0831] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions, and also analyzes the user's emotions using an emotion engine.

[0832] How it works: The server uses a generative AI model (e.g., OpenAI GPT) to analyze the intent and evaluates the user's emotions with an emotion engine.

[0833] Input: Preprocessed text data

[0834] Output: Parsed intent and emotion data

[0835] Step 5:

[0836] The server retrieves the required information by referencing an internal database or an external API.

[0837] Specific operation: The server retrieves the user's account information from a database (e.g., AWS DynamoDB) and collects related information.

[0838] Input: Parsed intent and emotion data

[0839] Output: Required information (e.g. electricity bill information)

[0840] Step 6:

[0841] Based on the information obtained by the server, the user's emotions are taken into consideration and an appropriate response is generated.

[0842] What it does: The server uses a generative AI model to generate a response like "Your electricity bill this month is 5000 yen" and includes an additional message based on sentiment data: "Do you need help with payment methods?"

[0843] Input: Required information (e.g., electricity bill information) and emotion data

[0844] Output: The generated response

[0845] Step 7:

[0846] The server generates a response and sends it to the terminal.

[0847] Specific operation: The server sends response data to the terminal using the network.

[0848] Input: The generated response

[0849] Output: Response data to the terminal

[0850] Step 8:

[0851] The terminal displays the response received from the server to the user.

[0852] Specific operation: The terminal displays a response message on the display.

[0853] Input: Response data to the terminal

[0854] Output: The response message that is displayed to the user

[0855] (Application example 2)

[0856] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0857] Current customer service systems can only provide standard responses to user requests, making it difficult to provide a personalized user experience. Furthermore, they lack the ability to recognize user emotions and respond appropriately, resulting in insufficient responses when users feel dissatisfied or anxious. Therefore, improving the quality of the user experience is a challenge in areas such as food delivery, where fast and personalized responses are required.

[0858] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0859] In this invention, the server includes a means for analyzing user emotions, a means for analyzing user intentions using preprocessed data, and a means for obtaining necessary information by referencing an internal database or an external API, thereby enabling the server to generate an appropriate response according to the user's emotions.

[0860] "User request data" refers to information provided by a user to the system in the form of voice, text, images, or video.

[0861] A "server" is a computing system that receives, pre-processes, and generates a response from a user's request data.

[0862] A "voice recognition system" is a technology that converts voice data into text.

[0863] An "image analysis system" is a technology that extracts specific information from image data.

[0864] "Means for analyzing user intent" refers to a process for understanding the user's purpose and requirements from the user's request data.

[0865] "Means for analyzing user emotions" refers to technology that recognizes and classifies the user's emotional state based on request data.

[0866] An "internal database" is a database that stores internal information managed by the system.

[0867] An "External API" is an application programming interface for accessing services or data outside the system.

[0868] The "means for generating an appropriate response" is a process for generating an optimal response to the user based on the acquired information and the analyzed user's emotions.

[0869] A "terminal" is a device through which a user inputs requests and receives responses.

[0870] The present invention provides a customer service system capable of processing requests from multiple modalities, and in particular adds a function for recognizing a user's emotions and adjusting the content of responses, thereby enabling a more sophisticated and personalized user experience. An embodiment of this system will be described below.

[0871] System Configuration

[0872] The system consists of a server, a terminal, a user, and an emotion engine as its main components. The server is responsible for processing the AI ​​model and managing data, while the terminal provides the user interface, i.e., inputting requests and displaying responses.

[0873] Hardware and Software Specifications

[0874] Server: A high-performance computer system is required.

[0875] Device: Using a user interface device such as a smartphone, smart glasses, a head-mounted display, or a robot.

[0876] Software used:

[0877] AI models: Use Natural Language Processing (NLP) libraries and models for sentiment analysis.

[0878] Speech recognition system: Used to convert voice data into text.

[0879] Image analysis system: Used to analyze image data and extract necessary information.

[0880] Emotion Engine: Used to analyze emotions from user request data.

[0881] Program processing flow

[0882] Receiving a user request

[0883] The user makes a request through the device. The request can be input in the form of voice, text, image, or video. For example, the user can type "My order hasn't arrived" into their smartphone.

[0884] Submitting a Request

[0885] The terminal transmits the received request data to the server.

[0886] Data Preprocessing

[0887] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[0888] Intention analysis and emotion recognition

[0889] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions and also analyzes the user's emotions using an emotion engine.

[0890] Database Reference

[0891] The server references an internal database or external API to obtain the necessary information.

[0892] Response generation and emotional reflection

[0893] Based on the information acquired by the server, the server generates an appropriate response taking into account the user's emotions. For example, it generates a response that takes into account emotions, such as "We apologize for the delay in your order. We are currently checking the current status. Would you mind waiting?"

[0894] Sending a Response

[0895] The server generates a response and sends it to the terminal.

[0896] Viewing the response

[0897] The device displays the response it receives from the server to the user. For example, a message like "We apologize for the delay in your order. We are currently checking the current status. Would you mind waiting?" appears on the smartphone screen.

[0898] Specific examples

[0899] Let's take a concrete example: when a user types "My order hasn't arrived" into their smartphone, the request is sent to the server, which preprocesses the data using a speech recognition system and image analysis system. The emotion engine then analyzes the user's frustration and generates an appropriate response using a generative AI model.

[0900] Prompt Sentence Examples

[0901] Analyze the sentiment when a user inquires about the status of an order in a food delivery app. Analyze the following user input text and generate a response based on the sentiment:

[0902] User Input: "Order not received"

[0903] Expected output:

[0904] User Sentiment: Annoyance

[0905] Response: "Sorry for the delay in your order. We're currently checking the status. Would you mind waiting?"

[0906] By using prompt sentences in this way, the generative AI model can recognize emotions and generate responses appropriately.

[0907] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0908] Step 1:

[0909] The user enters a request

[0910] Input: Text or speak "My order hasn't arrived" into your smartphone or other device.

[0911] Action: The user interacts with the device and types a request, which may include expressions of frustration or annoyance.

[0912] Output: The device receives the user's request data.

[0913] Step 2:

[0914] Sends request data to the server

[0915] Input: User request data obtained by the device.

[0916] What happens: Your device sends data to a server over Wi-Fi or your mobile network.

[0917] Output: The server receives the request data.

[0918] Step 3:

[0919] Data Preprocessing

[0920] Input: The request data received by the server.

[0921] How it works: The server converts voice data into text using a voice recognition system, and analyzes image data using an image analysis system.

[0922] Output: Text data and analysis results are obtained.

[0923] Step 4:

[0924] Intention analysis and emotion recognition

[0925] Input: Text data preprocessed by the server and analysis results.

[0926] How it works: The server inputs text data into the AI ​​model, analyzes the user's intention, and recognizes the user's emotions using the emotion engine.

[0927] Output: Data is obtained that analyzes the user's intent and emotions.

[0928] Step 5:

[0929] Referencing a database or external API

[0930] Input: User intent and sentiment parsed by the server.

[0931] What happens: The server looks up an internal database or an external API to get the information it needs. For example, it looks up a database to check the status of an order.

[0932] Output: The required information (e.g., current order status) is obtained.

[0933] Step 6:

[0934] Response generation and emotional reflection

[0935] Input: Captured information and analyzed user sentiment.

[0936] How it works: Based on the information the server obtains, it generates a response that takes into account the user's feelings. For example, it generates a message like, "We're sorry your order is delayed. We're currently checking the current status. Would you mind waiting?"

[0937] Output: A response message is generated that takes the user's feelings into consideration.

[0938] Step 7:

[0939] Sending a Response

[0940] Input: The server-generated response message.

[0941] Action: The server sends a response message to the terminal.

[0942] Output: The terminal receives the response message.

[0943] Step 8:

[0944] Viewing the response

[0945] Input: Response message received by the terminal.

[0946] What it does: The device displays a response message to the user. For example, a message on a smartphone screen might say, "We apologize for the delay in your order. We're currently checking the current status. Would you mind waiting?"

[0947] Output: The user confirms the response message and takes further action if necessary.

[0948] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0949] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0950] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0951] [Third embodiment]

[0952] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0953] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0954] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0955] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0956] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0957] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0958] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0959] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0960] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[0961] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0962] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0963] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. 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."

[0964] The present invention relates to a customer service system that can process requests in various modalities. It supports each modality (text, voice, image, video) and provides appropriate responses to user requests in real time. Specific embodiments of this system are described below.

[0965] Program Overview

[0966] This system consists of three main components: a server, a terminal, and a user. The server is responsible for processing the AI ​​model and managing data, while the terminal provides an interface for users to send requests and receive responses, allowing users to ask questions or make requests. The cooperation of these components creates a mechanism that can appropriately respond to a variety of user requests.

[0967] Program Details

[0968] Receiving a user request

[0969] The user makes a request through the device, which can be in the form of text, voice, image, or video.

[0970] Example: A user types the text "I want to see my electricity bill details" into the terminal.

[0971] Submitting a Request

[0972] The terminal transmits the received request data to the server.

[0973] Example: The terminal sends text data to the server saying, "I want to see my electricity bill details."

[0974] Data Preprocessing

[0975] The server preprocesses the received data: voice data is converted into text using a voice recognition system, and image data is extracted using an image analysis system.

[0976] Example: When the server converts voice data to text, it uses speech recognition to generate the text "I would like to see my electricity bill details."

[0977] Intent Analysis

[0978] The server inputs the preprocessed data into the AI ​​model to understand the user's intent, and then proceeds with the process based on the user's intent.

[0979] Example: From the text "I want to see my electricity bill details," the user's intent is interpreted as "I want to check the details of my current electricity bill."

[0980] Database Reference

[0981] The server references an internal database or external API to obtain the necessary information.

[0982] Example: The server looks up the user's account information in a database and retrieves the latest electricity bill data.

[0983] Generating a response

[0984] Based on the information obtained, the server generates an appropriate response to the user.

[0985] Example: Generate a response in text format saying "Your latest electricity bill is 5000 yen."

[0986] Sending a Response

[0987] The server generates a response and sends it to the terminal.

[0988] Example: The server sends text data to the terminal saying, "The latest electricity bill is 5,000 yen."

[0989] Viewing the response

[0990] The terminal displays the response received from the server to the user.

[0991] Example: The device displays a text message to the user saying, "Your latest electricity bill is 5000 yen."

[0992] Specific examples

[0993] The following is an example of the process when a user verbally asks the terminal, "I'm having trouble reading the electricity meter."

[0994] 1. Receiving a user request

[0995] The user verbally asks the terminal, "I'm having trouble reading the electric meter."

[0996] The terminal receives the audio data.

[0997] 2. Submitting a Request

[0998] The terminal transmits the voice data to the server.

[0999] 3. Data Preprocessing

[1000] The server converts the audio data into text.

[1001] The speech recognition system generates the text "I'm having trouble reading my electric meter."

[1002] 4. Intent Analysis

[1003] The server inputs the text data into an AI model and analyzes the user's intent.

[1004] The user's intent is interpreted as "I want to know how to read my electric meter."

[1005] 5. Referencing the database

[1006] The server retrieves guide information on how to read the electricity meter from an internal database.

[1007] 6. Generating the Response

[1008] The server generates a response saying, "Take a photo of your electric meter and tap the camera icon in the chat window."

[1009] 7. Sending the Response

[1010] The server generates a response and sends it to the terminal.

[1011] 8. Displaying the Response

[1012] The device displays a text message to the user saying, "Take a photo of your electricity meter and tap the camera icon on the chat screen."

[1013] This system can handle a variety of data formats and provide users with prompt and appropriate responses in real time.

[1014] The processing flow will be explained below.

[1015] Step 1:

[1016] The user makes a request through the device, which can be input in the form of text, voice, image, or video.

[1017] Example: A user texts a device asking, "What's my electricity bill this month?"

[1018] Step 2:

[1019] The terminal transmits the received request data to the server.

[1020] Example: The device sends text data such as "What is the electricity bill this month?" to the server.

[1021] Step 3:

[1022] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[1023] Example: The server receives voice data and converts it into text using a speech recognition system: "What is the electricity bill this month?"

[1024] Step 4:

[1025] The server inputs the preprocessed data into an AI model to analyze the user's intent.

[1026] Example: An AI model interprets the text "What is the electricity bill this month?" as the user's intent to "I want to know the current bill amount."

[1027] Step 5:

[1028] The server references an internal database or external API to obtain the necessary information.

[1029] Example: The server retrieves a user's account information from a database and retrieves the latest electricity bill information.

[1030] Step 6:

[1031] The server generates an appropriate response based on the information it has obtained.

[1032] Example: The server generates a text response saying "This month's electricity bill is 5000 yen."

[1033] Step 7:

[1034] The server generates a response and sends it to the terminal.

[1035] Example: The server sends the text "This month's electricity bill is 5,000 yen" to the terminal.

[1036] Step 8:

[1037] The terminal displays the response received from the server to the user.

[1038] Example: The device displays a text message to the user saying, "This month's electricity bill is 5000 yen."

[1039] The above is the flow of processing for a user request.

[1040] Example 1

[1041] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1042] Conventional customer service systems have had difficulty providing appropriate responses to user requests in real time. Furthermore, few systems support multiple modalities (voice, text, images, and video), creating a need for improved service quality. Furthermore, accurately analyzing user intent and quickly providing appropriate information is a challenge.

[1043] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1044] In this invention, the server includes means for receiving request data from a user in at least one form of voice, text, image, or video, means for preprocessing the received request data, converting voice data into text, and analyzing image data, means for using a generative model to analyze the user's intention using the preprocessed data, means for obtaining necessary information by referring to an internal database or an external interface, means for generating an appropriate response based on the obtained information, means for transmitting the generated response to a terminal, and means for the terminal to display the response to the user. This makes it possible to support various modalities, accurately analyze the user's intention, and provide appropriate information in real time.

[1045] A "user" is a person or entity that submits requests to a system to obtain information.

[1046] A "terminal" is a device through which a user inputs requests and receives responses.

[1047] A "server" is a computing system that processes requests from users and generates and sends appropriate responses to terminals.

[1048] "Request data" refers to voice, text, image, or video data that a user sends to the system through a terminal.

[1049] "Preprocessing" is the process of converting received request data into a format that is easy to analyze.

[1050] A "voice recognition system" is a technology that converts voice data into text data.

[1051] An "image analysis system" is a technology that analyzes image data and extracts necessary information.

[1052] A "generative model" is an AI technology that analyzes user intent based on preprocessed data.

[1053] An "internal database" is a system that manages data within a server.

[1054] An "external interface" is a means by which a server interacts with external systems and services to obtain data.

[1055] An "appropriate response" is data that contains information or instructions that are most appropriate to the user's request.

[1056] The present invention relates to a multi-modal customer service system that receives voice, text, image, and video request data from users and generates appropriate responses in real time.

[1057] System configuration

[1058] The system consists of three main components: a server, a terminal, and a user. The server is responsible for processing the AI ​​model and managing data, while the terminal provides an interface where users can send requests and receive responses. Users can ask questions or make requests and use the system's services.

[1059] Hardware and Software

[1060] Server: A high-performance computer system that preprocesses data, runs generative models, manages databases, and generates responses. Specific software includes speech recognition engines (e.g., Google Cloud Speech-to-Text) and generative AI models (e.g., OpenAI's GPT-3).

[1061] Terminal: A device that provides a user interface, such as a smartphone, tablet, or PC. The terminal uses a communication protocol (e.g., HTTP / HTTPS) to enable sending and receiving requests.

[1062] Initial Setup and Operation

[1063] 1. User request input: The user uses the device to send a request via voice, text, image, or video.

[1064] Example: A user makes a voice request to their smartphone saying, "I want to see my electricity bill details."

[1065] 2. Sending request data: The terminal sends the request data to the server via a communication protocol.

[1066] Example: A smartphone sends voice data saying "I want to see my electricity bill details" to a server via an HTTP request.

[1067] 3. Data preprocessing: The server preprocesses the received data. Voice data is converted into text using a voice recognition system, and image data is analyzed using an image analysis system.

[1068] Example: The server uses a speech recognition engine to convert the voice data into text, generating the text "I would like to see my electricity bill details."

[1069] 4. Intent analysis: The server inputs the preprocessed data into the generative AI model to analyze the user's intent.

[1070] Example: The server inputs the text "I want to see my electricity bill details" into the generated AI model, and analyzes the user's intent as "I want to know the details of my current electricity bill."

[1071] 5. Data retrieval: The server retrieves the required information using an internal database or an external interface.

[1072] Example: A server queries an internal database to get the latest electricity bill (e.g., 5000 yen).

[1073] 6. Generate a response: The server uses the information it has obtained to generate an appropriate response for the user.

[1074] Example: The server generates a text response saying "The latest electricity bill is 5000 yen."

[1075] 7. Send and display the response: Send the generated response to the terminal, which displays it to the user.

[1076] Example: The server sends generated text data to the terminal, and the terminal displays the message "The latest electricity bill is 5,000 yen."

[1077] Example prompt

[1078] An example prompt for using a generative AI model is:

[1079] "Please tell me the details of your latest electricity bill."

[1080] How do I read an electric meter?

[1081] As a result, the system can handle requests in a variety of data formats and provide quick and appropriate responses.

[1082] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1083] Step 1:

[1084] A user uses a terminal to input a request in the form of voice, text, image, or video. For example, a user makes a voice request to a smartphone saying, "I want to see my electricity bill details." In this case, the input is voice data, and the voice data is saved in the terminal as output.

[1085] Step 2:

[1086] The device sends the request data received from the user to the server. At this time, the device sends the voice data to the server as an HTTP request. The input is the voice data stored on the device, and the voice data is transferred to the server as the output.

[1087] Step 3:

[1088] The server preprocesses the received voice data. The server uses a speech recognition engine to convert the voice data into text data. Specifically, it uses the Google Cloud Speech-to-Text API to convert the voice data, such as "I want to see my electricity bill details," into text. The input is voice data, and text data is generated as the output.

[1089] Step 4:

[1090] The server inputs the preprocessed text data into a generative AI model to analyze the user's intent. Using OpenAI's GPT-3, the server inputs the text "I want to see my electricity bill details" and analyzes the user's intent as "I want to know the details of my current electricity bill." The input is text data, and the analyzed user's intent is output.

[1091] Step 5:

[1092] Based on the user's intent, the server retrieves the necessary information by referencing an internal database or an external interface. The server queries the internal database and retrieves the latest electricity bill data (e.g., 5,000 yen). The input is the parsed user's intent, and the electricity bill data is obtained as the output.

[1093] Step 6:

[1094] Based on the information acquired, the server generates an appropriate response for the user. The server uses a generative AI model to generate a text response such as "The latest electricity bill is 5,000 yen." The input is the electricity bill data, and the generated text response is obtained as the output.

[1095] Step 7:

[1096] The server sends the generated response to the terminal. The server sends the generated text data to the terminal as an HTTP response. The input is the generated text response, and the text data is sent to the terminal as output.

[1097] Step 8:

[1098] The terminal displays the response received from the server to the user. The terminal displays a text message on the screen saying "The latest electricity bill is 5000 yen." The input is the text data sent from the server, and the output is the text message displayed on the user's screen.

[1099] The above steps show in detail the process from when a user sends a request to the system to when an appropriate response is received. This system realizes quick and appropriate responses to a variety of data formats.

[1100] (Application example 1)

[1101] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1102] Modern content delivery services require systems that allow users to make requests using a variety of modalities (voice, text, images, and videos) and provide prompt and appropriate responses. However, existing systems often only support specific modalities, and there are few general-purpose systems that support all modalities. Furthermore, there are also limited systems that utilize generative AI models to provide optimal recommendations for user requests and improve the quality of information provided. Therefore, a system that can respond to requests in different modalities and generate optimal responses using generative AI models is needed.

[1103] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1104] In this invention, the server includes: means for receiving request data from a user in at least one of the following formats: voice, text, image, and video; means for transmitting the received request data to the server; means for preprocessing the request data, converting voice data to text, and analyzing image data; means for analyzing the user's intent using the preprocessed data; means for obtaining necessary information by referencing an internal database or an external API; means for generating an appropriate response based on the obtained information; means for generating a prompt sentence in response to the user's request using a generative AI model; means for generating an optimal response using the generated prompt sentence; means for transmitting the generated response to a terminal; and means for the terminal to display the response to the user. This enables prompt and appropriate responses to be provided even when the user makes requests in various formats. Furthermore, the use of a generative AI model enables higher quality recommendations and information provision.

[1105] A "user" is a person using a terminal that makes requests and receives responses in a variety of forms.

[1106] "Request data" is data sent by a user in one or more formats: voice, text, image, video.

[1107] "Preprocessing" refers to the process in which the server analyzes the request data received, converts voice data into text, and extracts necessary information from image data.

[1108] A "voice recognition system" is a system that converts voice data into text data.

[1109] An "image analysis system" is a system that analyzes image data and extracts necessary information.

[1110] "Intent analysis" is the process of using preprocessed data to understand the purpose and intent of a user request.

[1111] An "internal database" is a database stored within the system that is referenced to obtain required information.

[1112] An "external API" is an interface for communicating with services or databases outside the system.

[1113] A "generative AI model" is an artificial intelligence model that generates appropriate responses to user requests.

[1114] A "prompt sentence" is an input sentence to a generative AI model, which is generated based on a user request.

[1115] A "terminal" is a device through which a user sends requests and receives responses from a server.

[1116] A "server" is a computer system that processes request data, analyzes intent, retrieves information, and generates a response.

[1117] The present invention relates to a customer service system capable of processing requests of various modalities, and a specific embodiment thereof will be described below. The system is composed of three main components: a server, a terminal, and a user.

[1118] Receiving a user request

[1119] 1. The user sends a request to the device. The request can be in the form of voice, text, image, video, etc. For example, a user can make a voice request such as, "Tell me the next action movie I should watch."

[1120] 2. The terminal sends the received request data to the server.

[1121] Data Preprocessing

[1122] 1. The server preprocesses the data it receives: if it is voice data, it converts it into text using a voice recognition system, and if it is image data, it uses an image analysis system to extract the necessary information.

[1123] 2. As a speech recognition system, for example, use the speech_recognition library.

[1124] 3. PIL (Python Imaging Library) and pytesseract are used as image analysis systems.

[1125] Intent Analysis

[1126] 1. The server analyzes the user's intent based on the preprocessed data. For example, it understands the intent of "what action movie should I watch next" from the speech data.

[1127] Obtaining information

[1128] 1. The server retrieves the required information from an internal database or an external API. For example, it retrieves a list of recommended action movies from an internal database.

[1129] Generating a response

[1130] 1. Based on the information obtained by the server, an appropriate prompt sentence is generated using a generative AI model.

[1131] 2. The generated prompt sentence will be something like, "What's the next action movie I should watch?"

[1132] 3. Using this prompt, the generative AI model generates an appropriate response, such as "The recommended action movie is 'Recommended Movie Title'."

[1133] Sending and Displaying Responses

[1134] 1. The server generates a response and sends it to the device.

[1135] 2. The device displays a response to the user, for example, "The recommended action movie is 'Recommended Movie Title'."

[1136] Specific examples

[1137] For example, if a user speaks to their device, "What's the next action movie I should watch?", the process goes like this:

[1138] 1. The user's voice data is sent by the terminal to the server.

[1139] 2. The server uses a voice recognition system to convert the voice data into text.

[1140] 3. The server analyzes the user's intent based on the converted text data and generates an appropriate prompt.

[1141] 4. The generative AI model generates a response based on the prompt, "What's the next action movie I should watch?"

[1142] 5. The generated response will be "The recommended action movie is 'Recommended Movie Title'."

[1143] 6. The terminal displays this response to the user.

[1144] This embodiment allows users to easily make different types of requests and receive a quick and appropriate response, an example of a prompt sentence could be "What's the next action movie I should watch?"

[1145] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1146] Step 1:

[1147] The user sends request data. The request can be in the form of voice, text, image, or video. For example, the user might request by voice, "Tell me the next action movie I should watch." This voice data is input into the terminal.

[1148] Step 2:

[1149] The terminal transmits the received voice data to the server. The terminal uses the network to transmit the voice data and sends this data to the server. This inputs the voice data to the server.

[1150] Step 3:

[1151] The server preprocesses the received voice data. Specifically, it converts the voice data into text using the speech_recognition library. This process converts the voice data into text data.

[1152] Step 4:

[1153] The server analyzes the user's intent based on the converted text data. For example, from the text "Tell me the next action movie I should watch," it extracts the intent that the user is looking for movie recommendations.

[1154] Step 5:

[1155] The server retrieves the required information from an internal database or an external API. For example, it retrieves a list of action movies from an internal database. This process retrieves the data corresponding to the user's request.

[1156] Step 6:

[1157] Based on the information acquired by the server, a prompt sentence is generated using a generative AI model. For example, the prompt sentence might be, "Tell me the next action movie I should watch." This generated prompt sentence is input into the generative AI model.

[1158] Step 7:

[1159] The server uses the generative AI model to generate the optimal response based on the prompt. For example, the response generated is "The recommended action movie is 'Recommended Movie Title'." This response becomes the output of the generative AI model.

[1160] Step 8:

[1161] The server sends the generated response to the terminal, and this response message is sent to the terminal via the network and is input to the terminal.

[1162] Step 9:

[1163] The device displays the response received from the server to the user. For example, a message such as "The recommended action movie is 'Recommended Movie Title'" is displayed on the device screen.

[1164] This series of processing steps allows users to make requests in a variety of formats and receive appropriate responses in real time.

[1165] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1166] The present invention provides a customer service system capable of processing requests from multiple modalities, and in particular adds a function for recognizing a user's emotions and adjusting the content of responses accordingly. This makes it possible to provide a more sophisticated and personalized user experience. Specific embodiments of this system are described below.

[1167] Program Overview

[1168] The system's main components are a server, a terminal, and a user, combined with an emotion engine. The emotion engine analyzes emotions from user request data and reflects them in the response. The server is responsible for processing the AI ​​model and managing data, while the terminal provides the user interface, i.e., inputting requests and displaying responses.

[1169] Program Details

[1170] Receiving a user request

[1171] The user makes a request through the device, which can be input in the form of text, voice, image, or video.

[1172] Example: A user texts a device asking, "What's my electricity bill this month?"

[1173] Submitting a Request

[1174] The terminal transmits the received request data to the server.

[1175] Example: The device sends text data such as "What is the electricity bill this month?" to the server.

[1176] Data Preprocessing

[1177] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[1178] Example: The server receives voice data and converts it into text using a speech recognition system: "What is the electricity bill this month?"

[1179] Intention analysis and emotion recognition

[1180] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions and also analyzes the user's emotions using an emotion engine.

[1181] Example: From the text "What's the electricity bill this month?", an AI model interprets the user's intent as "I want to know the current bill amount," and an emotion engine recognizes the user's tension or irritation.

[1182] Database Reference

[1183] The server references an internal database or external API to obtain the necessary information.

[1184] Example: The server retrieves a user's account information from a database and retrieves the latest electricity bill information.

[1185] Response generation and emotional reflection

[1186] Based on the information acquired by the server, an appropriate response is generated taking into account the user's emotions.

[1187] Example: The server generates "This month's electricity bill is 5000 yen," but recognizing the user's frustration, adds an additional "Do you need help with payment methods?"

[1188] Sending a Response

[1189] The server generates a response and sends it to the terminal.

[1190] Example: The server sends a text to the device saying, "This month's electricity bill is 5000 yen. Do you need help with payment?"

[1191] Viewing the response

[1192] The terminal displays the response received from the server to the user.

[1193] Example: The device displays a text message to the user saying, "Your electricity bill this month is 5000 yen. Do you need help paying?"

[1194] Specific examples

[1195] As an example, the process will be described when a user asks a question by voice, "The electric meter reading is not working properly," and feels uneasy.

[1196] 1. Receiving a user request

[1197] The user asks the device a question by voice, "The electric meter reading is not working properly." The voice contains a sense of anxiety.

[1198] The terminal receives the audio data.

[1199] 2. Submitting a Request

[1200] The terminal transmits the voice data to the server.

[1201] 3. Data Preprocessing

[1202] The server converts the audio data into text.

[1203] The speech recognition system generates the text "I'm having trouble reading my electric meter."

[1204] 4. Intention Analysis and Emotion Recognition

[1205] The server inputs the text data into an AI model and analyzes the user's intent.

[1206] The emotion engine analyzes the user's anxiety.

[1207] 5. Referencing the database

[1208] The server retrieves guide information on how to read the electricity meter from an internal database.

[1209] 6. Response generation and emotional reflection

[1210] The server generates a response saying, "Take a photo of your electric meter and tap the camera icon in the chat window," and includes an additional message to address the user's concerns: "If you have any concerns, please contact support."

[1211] 7. Sending the Response

[1212] The server generates a response and sends it to the terminal.

[1213] 8. Displaying the Response

[1214] The device will display a text message to the user saying, "Take a photo of your electric meter and tap the camera icon in the chat screen. If you have any concerns, please contact support."

[1215] This system enables us to provide detailed responses that take into account the user's emotions, improving the quality of the user experience.

[1216] The processing flow will be explained below.

[1217] Step 1:

[1218] The user makes a request through the device, which can be input in the form of text, voice, image, or video.

[1219] Example: A user texts a device asking, "What's my electricity bill this month?"

[1220] Step 2:

[1221] The terminal transmits the received request data to the server.

[1222] Example: The device sends text data such as "What is the electricity bill this month?" to the server.

[1223] Step 3:

[1224] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[1225] Example: The server receives voice data and converts it into text using a speech recognition system: "What is the electricity bill this month?"

[1226] Step 4:

[1227] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions, and then analyzes the user's emotions using an emotion engine.

[1228] Example: From the text "What's the electricity bill this month?", an AI model interprets the user's intent as "I want to know the current bill amount," and the emotion engine recognizes the user's tension or irritation.

[1229] Step 5:

[1230] The server references an internal database or external API to obtain the necessary information.

[1231] Example: The server retrieves a user's account information from a database and retrieves the latest electricity bill information.

[1232] Step 6:

[1233] The server generates an appropriate response based on the acquired information, taking into account the user's emotions based on the emotion engine.

[1234] Example: The server generates the text "This month's electricity bill is 5000 yen", but recognizes the user's frustration and adds "This month's electricity bill is 5000 yen. Do you need help paying?"

[1235] Step 7:

[1236] The server generates a response and sends it to the terminal.

[1237] Example: The server sends a text to the device saying, "This month's electricity bill is 5000 yen. Do you need help with payment?"

[1238] Step 8:

[1239] The terminal displays the response received from the server to the user.

[1240] Example: The device displays a text message to the user saying, "Your electricity bill this month is 5000 yen. Do you need help paying?"

[1241] The above is the specific flow of processing for a user request.

[1242] Example 2

[1243] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1244] Conventional customer service systems simply respond to user requests and do not take the user's emotions into consideration when responding. This does not alleviate user dissatisfaction or frustration, resulting in a poor user experience. The present invention aims to provide a more sophisticated and personalized user experience by analyzing the user's emotions and generating responses accordingly.

[1245] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1246] In this invention, the server includes means for analyzing the user's intention, means for analyzing the user's intention using preprocessed data, means for obtaining necessary information by referencing an internal database or an external API, means for analyzing the user's emotions and adjusting the response content to reflect the emotions, and means for transmitting the generated response to the terminal, thereby enabling the server to generate an appropriate response that takes the user's emotions into consideration and improve the quality of the user experience.

[1247] A "user" is a user who inputs a request to the system.

[1248] "Audio, text, images, video" are different forms of data that users use to make requests to the system.

[1249] "Request data" is data containing information sent by a user, and represents a question or request.

[1250] A "terminal" is a device used by a user to enter requests and receive responses.

[1251] A "server" is a central computer system that processes request data sent by users and generates and sends responses.

[1252] "Preprocessing" refers to initial processing for converting request data into a format that is easy to analyze, and includes, for example, converting voice data into text and analyzing image data.

[1253] A "voice recognition system" is a technology for converting voice data into text data.

[1254] An "image analysis system" is a technology for extracting necessary information from image data.

[1255] "Intent analysis" is the process of understanding the intent of a user request.

[1256] "Sentiment analysis" is the process of identifying a user's emotional state from their request data.

[1257] An "internal database" is an information storage location installed within the system, which contains information about users and data necessary for providing services.

[1258] An "external API" is an interface for accessing external services and databases and obtaining information.

[1259] "Response generation" is the process of creating an appropriate reply to the user based on the analyzed intent and sentiment.

[1260] "Adjusting the content of a response" refers to changing or complementing the content of a generated response to match the emotional state of the user.

[1261] The present invention is a customer service system that can process requests from multiple modalities, and in particular has the function of recognizing the user's emotions and adjusting the response content. This system is composed of major components: a server, a terminal, a user, and an emotion engine, and operates as follows.

[1262] The user makes a request through the device. The request can be input in the form of voice, text, image, or video. For example, the user might ask the device in text, "What is my electricity bill this month?" The device then sends the received request data to the server. If the data is voice, the server uses a voice recognition system to convert it into text, and if it is image data, it uses an image analysis system to extract the required information. Specifically, this could involve using a voice recognition service such as Google Cloud Speech-to-Text API or Amazon Transcribe.

[1263] The preprocessed data is input into an AI model on the server, where the user's intent is analyzed. The emotion engine also analyzes the user's emotions. For example, the server uses OpenAI's GPT model to analyze the intent from the text "What is my electricity bill this month?", and the emotion engine recognizes the user's tension or irritation. This allows the system to interpret the user's intent as "I want to know the current bill amount" and also take the user's emotions into account.

[1264] The server then references internal databases and external APIs to retrieve the necessary information. Specifically, it retrieves the user's account information and latest electricity bill information from databases such as AWS DynamoDB and Google Firestore. Based on this information, the server generates an appropriate response that takes the user's emotions into account. For example, the server uses a generative AI model to generate a response such as "This month's electricity bill is 5000 yen," but recognizes the user's frustration and includes an additional message such as "Do you need help with your payment method?"

[1265] The generated response is sent from the server to the terminal, which then displays this response to the user. For example, the terminal might display the following text message to the user: "This month's electricity bill is 5000 yen. Do you need help with payment?"

[1266] In this way, the present invention enables a finely tuned response that takes into account the user's emotions, improving the quality of the user experience.

[1267] Specific examples

[1268] Consider a case where a user voices a question saying, "I'm having trouble reading my electric meter," and feels anxious.

[1269] 1. The user asks the device a question by voice, "The electric meter reading is not working properly." The user expresses anxiety.

[1270] 2. The device sends the audio data to the server.

[1271] 3. The server converts the voice data into text, typically using the Google Cloud Speech-to-Text API.

[1272] 4. Use AI models to analyze intent and emotion engines to analyze anxiety.

[1273] 5. The server retrieves guidance information on how to read the electricity meter from its internal database.

[1274] 6. The server generates a response that says "Take a photo of your electric meter and tap the camera icon in the chat window," with an additional message that says "If you have any concerns, please contact support."

[1275] 7. The response is sent to the terminal, which displays it to the user.

[1276] This allows the user to feel at ease and solve the problem.

[1277] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1278] Step 1:

[1279] The user inputs a request into the device, which can be in the form of voice, text, image, or video.

[1280] Example: A user texts a device asking, "What's my electricity bill this month?"

[1281] Input: User request (e.g., text format)

[1282] Output: Request data in the terminal

[1283] Step 2:

[1284] The terminal transmits the received request data to the server.

[1285] Specific operation: The terminal sends the request data to the server via the network.

[1286] Input: Request data in the terminal

[1287] Output: Request data to the server

[1288] Step 3:

[1289] The server preprocesses the request data received: if it is voice data, it converts it into text using a voice recognition system, and if it is image data, it uses an image analysis system to extract the necessary information.

[1290] What happens: The server uses a speech recognition API (e.g., Google Cloud Speech-to-Text) to convert the speech to text.

[1291] Input: Request data to the server (e.g., audio data)

[1292] Output: Preprocessed text data

[1293] Step 4:

[1294] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions, and also analyzes the user's emotions using an emotion engine.

[1295] How it works: The server uses a generative AI model (e.g., OpenAI GPT) to analyze the intent and evaluates the user's emotions with an emotion engine.

[1296] Input: Preprocessed text data

[1297] Output: Parsed intent and emotion data

[1298] Step 5:

[1299] The server retrieves the required information by referencing an internal database or an external API.

[1300] Specific operation: The server retrieves the user's account information from a database (e.g., AWS DynamoDB) and collects related information.

[1301] Input: Parsed intent and emotion data

[1302] Output: Required information (e.g. electricity bill information)

[1303] Step 6:

[1304] Based on the information obtained by the server, the user's emotions are taken into consideration and an appropriate response is generated.

[1305] What it does: The server uses a generative AI model to generate a response like "Your electricity bill this month is 5000 yen" and includes an additional message based on sentiment data: "Do you need help with payment methods?"

[1306] Input: Required information (e.g., electricity bill information) and emotion data

[1307] Output: The generated response

[1308] Step 7:

[1309] The server generates a response and sends it to the terminal.

[1310] Specific operation: The server sends response data to the terminal using the network.

[1311] Input: The generated response

[1312] Output: Response data to the terminal

[1313] Step 8:

[1314] The terminal displays the response received from the server to the user.

[1315] Specific operation: The terminal displays a response message on the display.

[1316] Input: Response data to the terminal

[1317] Output: The response message that is displayed to the user

[1318] (Application example 2)

[1319] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1320] Current customer service systems can only provide standard responses to user requests, making it difficult to provide a personalized user experience. Furthermore, they lack the ability to recognize user emotions and respond appropriately, resulting in insufficient responses when users feel dissatisfied or anxious. Therefore, improving the quality of the user experience is a challenge in areas such as food delivery, where fast and personalized responses are required.

[1321] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1322] In this invention, the server includes a means for analyzing user emotions, a means for analyzing user intentions using preprocessed data, and a means for obtaining necessary information by referencing an internal database or an external API, thereby enabling the server to generate an appropriate response according to the user's emotions.

[1323] "User request data" refers to information provided by a user to the system in the form of voice, text, images, or video.

[1324] A "server" is a computing system that receives, pre-processes, and generates a response from a user's request data.

[1325] A "voice recognition system" is a technology that converts voice data into text.

[1326] An "image analysis system" is a technology that extracts specific information from image data.

[1327] "Means for analyzing user intent" refers to a process for understanding the user's purpose and requirements from the user's request data.

[1328] "Means for analyzing user emotions" refers to technology that recognizes and classifies the user's emotional state based on request data.

[1329] An "internal database" is a database that stores internal information managed by the system.

[1330] An "External API" is an application programming interface for accessing services or data outside the system.

[1331] The "means for generating an appropriate response" is a process for generating an optimal response to the user based on the acquired information and the analyzed user's emotions.

[1332] A "terminal" is a device through which a user inputs requests and receives responses.

[1333] The present invention provides a customer service system capable of processing requests from multiple modalities, and in particular adds a function for recognizing a user's emotions and adjusting the content of responses, thereby enabling a more sophisticated and personalized user experience. An embodiment of this system will be described below.

[1334] System Configuration

[1335] The system consists of a server, a terminal, a user, and an emotion engine as its main components. The server is responsible for processing the AI ​​model and managing data, while the terminal provides the user interface, i.e., inputting requests and displaying responses.

[1336] Hardware and Software Specifications

[1337] Server: A high-performance computer system is required.

[1338] Device: Using a user interface device such as a smartphone, smart glasses, a head-mounted display, or a robot.

[1339] Software used:

[1340] AI models: Use Natural Language Processing (NLP) libraries and models for sentiment analysis.

[1341] Speech recognition system: Used to convert voice data into text.

[1342] Image analysis system: Used to analyze image data and extract necessary information.

[1343] Emotion Engine: Used to analyze emotions from user request data.

[1344] Program processing flow

[1345] Receiving a user request

[1346] The user makes a request through the device. The request can be input in the form of voice, text, image, or video. For example, the user can type "My order hasn't arrived" into their smartphone.

[1347] Submitting a Request

[1348] The terminal transmits the received request data to the server.

[1349] Data Preprocessing

[1350] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[1351] Intention analysis and emotion recognition

[1352] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions and also analyzes the user's emotions using an emotion engine.

[1353] Database Reference

[1354] The server references an internal database or external API to obtain the necessary information.

[1355] Response generation and emotional reflection

[1356] Based on the information acquired by the server, the server generates an appropriate response taking into account the user's emotions. For example, it generates a response that takes into account emotions, such as "We apologize for the delay in your order. We are currently checking the current status. Would you mind waiting?"

[1357] Sending a Response

[1358] The server generates a response and sends it to the terminal.

[1359] Viewing the response

[1360] The device displays the response it receives from the server to the user. For example, a message like "We apologize for the delay in your order. We are currently checking the current status. Would you mind waiting?" appears on the smartphone screen.

[1361] Specific examples

[1362] Let's take a concrete example: when a user types "My order hasn't arrived" into their smartphone, the request is sent to the server, which preprocesses the data using a speech recognition system and image analysis system. The emotion engine then analyzes the user's frustration and generates an appropriate response using a generative AI model.

[1363] Prompt Sentence Examples

[1364] Analyze the sentiment when a user inquires about the status of an order in a food delivery app. Analyze the following user input text and generate a response based on the sentiment:

[1365] User Input: "Order not received"

[1366] Expected output:

[1367] User Sentiment: Annoyance

[1368] Response: "Sorry for the delay in your order. We're currently checking the status. Would you mind waiting?"

[1369] By using prompt sentences in this way, the generative AI model can recognize emotions and generate responses appropriately.

[1370] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1371] Step 1:

[1372] The user enters a request

[1373] Input: Text or speak "My order hasn't arrived" into your smartphone or other device.

[1374] Action: The user interacts with the device and types a request, which may include expressions of frustration or annoyance.

[1375] Output: The device receives the user's request data.

[1376] Step 2:

[1377] Sends request data to the server

[1378] Input: User request data obtained by the device.

[1379] What happens: Your device sends data to a server over Wi-Fi or your mobile network.

[1380] Output: The server receives the request data.

[1381] Step 3:

[1382] Data Preprocessing

[1383] Input: The request data received by the server.

[1384] How it works: The server converts voice data into text using a voice recognition system, and analyzes image data using an image analysis system.

[1385] Output: Text data and analysis results are obtained.

[1386] Step 4:

[1387] Intention analysis and emotion recognition

[1388] Input: Text data preprocessed by the server and analysis results.

[1389] How it works: The server inputs text data into the AI ​​model, analyzes the user's intention, and recognizes the user's emotions using the emotion engine.

[1390] Output: Data is obtained that analyzes the user's intent and emotions.

[1391] Step 5:

[1392] Referencing a database or external API

[1393] Input: User intent and sentiment parsed by the server.

[1394] What happens: The server looks up an internal database or an external API to get the information it needs. For example, it looks up a database to check the status of an order.

[1395] Output: The required information (e.g., current order status) is obtained.

[1396] Step 6:

[1397] Response generation and emotional reflection

[1398] Input: Captured information and analyzed user sentiment.

[1399] How it works: Based on the information the server obtains, it generates a response that takes into account the user's feelings. For example, it generates a message like, "We're sorry your order is delayed. We're currently checking the current status. Would you mind waiting?"

[1400] Output: A response message is generated that takes the user's feelings into consideration.

[1401] Step 7:

[1402] Sending a Response

[1403] Input: The server-generated response message.

[1404] Action: The server sends a response message to the terminal.

[1405] Output: The terminal receives the response message.

[1406] Step 8:

[1407] Viewing the response

[1408] Input: Response message received by the terminal.

[1409] What it does: The device displays a response message to the user. For example, a message on a smartphone screen might say, "We apologize for the delay in your order. We're currently checking the current status. Would you mind waiting?"

[1410] Output: The user confirms the response message and takes further action if necessary.

[1411] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1412] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1413] 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 the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1414] [Fourth embodiment]

[1415] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1416] 7, a 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.

[1417] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1418] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1419] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1420] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1421] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1422] The control object 443 includes a display device, LEDs in the eyes, and motors for driving 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 emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1423] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1424] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[1425] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1426] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1427] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1428] The present invention relates to a customer service system that can process requests in various modalities. It supports each modality (text, voice, image, video) and provides appropriate responses to user requests in real time. Specific embodiments of this system are described below.

[1429] Program Overview

[1430] This system consists of three main components: a server, a terminal, and a user. The server is responsible for processing the AI ​​model and managing data, while the terminal provides an interface for users to send requests and receive responses, allowing users to ask questions or make requests. The cooperation of these components creates a mechanism that can appropriately respond to a variety of user requests.

[1431] Program Details

[1432] Receiving a user request

[1433] The user makes a request through the device, which can be in the form of text, voice, image, or video.

[1434] Example: A user types the text "I want to see my electricity bill details" into the terminal.

[1435] Submitting a Request

[1436] The terminal transmits the received request data to the server.

[1437] Example: The terminal sends text data to the server saying, "I want to see my electricity bill details."

[1438] Data Preprocessing

[1439] The server preprocesses the received data: voice data is converted into text using a voice recognition system, and image data is extracted using an image analysis system.

[1440] Example: When the server converts voice data to text, it uses speech recognition to generate the text "I would like to see my electricity bill details."

[1441] Intent Analysis

[1442] The server inputs the preprocessed data into the AI ​​model to understand the user's intent, and then proceeds with the process based on the user's intent.

[1443] Example: From the text "I want to see my electricity bill details," the user's intent is interpreted as "I want to check the details of my current electricity bill."

[1444] Database Reference

[1445] The server references an internal database or external API to obtain the necessary information.

[1446] Example: The server looks up the user's account information in a database and retrieves the latest electricity bill data.

[1447] Generating a response

[1448] Based on the information obtained, the server generates an appropriate response to the user.

[1449] Example: Generate a response in text format saying "Your latest electricity bill is 5000 yen."

[1450] Sending a Response

[1451] The server generates a response and sends it to the terminal.

[1452] Example: The server sends text data to the terminal saying, "The latest electricity bill is 5,000 yen."

[1453] Viewing the response

[1454] The terminal displays the response received from the server to the user.

[1455] Example: The device displays a text message to the user saying, "Your latest electricity bill is 5000 yen."

[1456] Specific examples

[1457] The following is an example of the process when a user verbally asks the terminal, "I'm having trouble reading the electricity meter."

[1458] 1. Receiving a user request

[1459] The user verbally asks the terminal, "I'm having trouble reading the electric meter."

[1460] The terminal receives the audio data.

[1461] 2. Submitting a Request

[1462] The terminal transmits the voice data to the server.

[1463] 3. Data Preprocessing

[1464] The server converts the audio data into text.

[1465] The speech recognition system generates the text "I'm having trouble reading my electric meter."

[1466] 4. Intent Analysis

[1467] The server inputs the text data into an AI model and analyzes the user's intent.

[1468] The user's intent is interpreted as "I want to know how to read my electric meter."

[1469] 5. Referencing the database

[1470] The server retrieves guide information on how to read the electricity meter from an internal database.

[1471] 6. Generating the Response

[1472] The server generates a response saying, "Take a photo of your electric meter and tap the camera icon in the chat window."

[1473] 7. Sending the Response

[1474] The server generates a response and sends it to the terminal.

[1475] 8. Displaying the Response

[1476] The device displays a text message to the user saying, "Take a photo of your electricity meter and tap the camera icon on the chat screen."

[1477] This system can handle a variety of data formats and provide users with prompt and appropriate responses in real time.

[1478] The processing flow will be explained below.

[1479] Step 1:

[1480] The user makes a request through the device, which can be input in the form of text, voice, image, or video.

[1481] Example: A user texts a device asking, "What's my electricity bill this month?"

[1482] Step 2:

[1483] The terminal transmits the received request data to the server.

[1484] Example: The device sends text data such as "What is the electricity bill this month?" to the server.

[1485] Step 3:

[1486] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[1487] Example: The server receives voice data and converts it into text using a speech recognition system: "What is the electricity bill this month?"

[1488] Step 4:

[1489] The server inputs the preprocessed data into an AI model to analyze the user's intent.

[1490] Example: An AI model interprets the text "What is the electricity bill this month?" as the user's intent to "I want to know the current bill amount."

[1491] Step 5:

[1492] The server references an internal database or external API to obtain the necessary information.

[1493] Example: The server retrieves a user's account information from a database and retrieves the latest electricity bill information.

[1494] Step 6:

[1495] The server generates an appropriate response based on the information it has obtained.

[1496] Example: The server generates a text response saying "This month's electricity bill is 5000 yen."

[1497] Step 7:

[1498] The server generates a response and sends it to the terminal.

[1499] Example: The server sends the text "This month's electricity bill is 5,000 yen" to the terminal.

[1500] Step 8:

[1501] The terminal displays the response received from the server to the user.

[1502] Example: The device displays a text message to the user saying, "This month's electricity bill is 5000 yen."

[1503] The above is the flow of processing for a user request.

[1504] Example 1

[1505] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1506] Conventional customer service systems have had difficulty providing appropriate responses to user requests in real time. Furthermore, few systems support multiple modalities (voice, text, images, and video), creating a need for improved service quality. Furthermore, accurately analyzing user intent and quickly providing appropriate information is a challenge.

[1507] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1508] In this invention, the server includes means for receiving request data from a user in at least one form of voice, text, image, or video, means for preprocessing the received request data, converting voice data into text, and analyzing image data, means for using a generative model to analyze the user's intention using the preprocessed data, means for obtaining necessary information by referring to an internal database or an external interface, means for generating an appropriate response based on the obtained information, means for transmitting the generated response to a terminal, and means for the terminal to display the response to the user. This makes it possible to support various modalities, accurately analyze the user's intention, and provide appropriate information in real time.

[1509] A "user" is a person or entity that submits requests to a system to obtain information.

[1510] A "terminal" is a device through which a user inputs requests and receives responses.

[1511] A "server" is a computing system that processes requests from users and generates and sends appropriate responses to terminals.

[1512] "Request data" refers to voice, text, image, or video data that a user sends to the system through a terminal.

[1513] "Preprocessing" is the process of converting received request data into a format that is easy to analyze.

[1514] A "voice recognition system" is a technology that converts voice data into text data.

[1515] An "image analysis system" is a technology that analyzes image data and extracts necessary information.

[1516] A "generative model" is an AI technology that analyzes user intent based on preprocessed data.

[1517] An "internal database" is a system that manages data within a server.

[1518] An "external interface" is a means by which a server interacts with external systems and services to obtain data.

[1519] An "appropriate response" is data that contains information or instructions that are most appropriate to the user's request.

[1520] The present invention relates to a multi-modal customer service system that receives voice, text, image, and video request data from users and generates appropriate responses in real time.

[1521] System configuration

[1522] The system consists of three main components: a server, a terminal, and a user. The server is responsible for processing the AI ​​model and managing data, while the terminal provides an interface where users can send requests and receive responses. Users can ask questions or make requests and use the system's services.

[1523] Hardware and Software

[1524] Server: A high-performance computer system that preprocesses data, runs generative models, manages databases, and generates responses. Specific software includes speech recognition engines (e.g., Google Cloud Speech-to-Text) and generative AI models (e.g., OpenAI's GPT-3).

[1525] Terminal: A device that provides a user interface, such as a smartphone, tablet, or PC. The terminal uses a communication protocol (e.g., HTTP / HTTPS) to enable sending and receiving requests.

[1526] Initial Setup and Operation

[1527] 1. User request input: The user uses the device to send a request via voice, text, image, or video.

[1528] Example: A user makes a voice request to their smartphone saying, "I want to see my electricity bill details."

[1529] 2. Sending request data: The terminal sends the request data to the server via a communication protocol.

[1530] Example: A smartphone sends voice data saying "I want to see my electricity bill details" to a server via an HTTP request.

[1531] 3. Data preprocessing: The server preprocesses the received data. Voice data is converted into text using a voice recognition system, and image data is analyzed using an image analysis system.

[1532] Example: The server uses a speech recognition engine to convert the voice data into text, generating the text "I would like to see my electricity bill details."

[1533] 4. Intent analysis: The server inputs the preprocessed data into the generative AI model to analyze the user's intent.

[1534] Example: The server inputs the text "I want to see my electricity bill details" into the generated AI model, and analyzes the user's intent as "I want to know the details of my current electricity bill."

[1535] 5. Data retrieval: The server retrieves the required information using an internal database or an external interface.

[1536] Example: A server queries an internal database to get the latest electricity bill (e.g., 5000 yen).

[1537] 6. Generate a response: The server uses the information it has obtained to generate an appropriate response for the user.

[1538] Example: The server generates a text response saying "The latest electricity bill is 5000 yen."

[1539] 7. Send and display the response: Send the generated response to the terminal, which displays it to the user.

[1540] Example: The server sends generated text data to the terminal, and the terminal displays the message "The latest electricity bill is 5,000 yen."

[1541] Example prompt

[1542] An example prompt for using a generative AI model is:

[1543] "Please tell me the details of your latest electricity bill."

[1544] How do I read an electric meter?

[1545] As a result, the system can handle requests in a variety of data formats and provide quick and appropriate responses.

[1546] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1547] Step 1:

[1548] A user uses a terminal to input a request in the form of voice, text, image, or video. For example, a user makes a voice request to a smartphone saying, "I want to see my electricity bill details." In this case, the input is voice data, and the voice data is saved in the terminal as output.

[1549] Step 2:

[1550] The device sends the request data received from the user to the server. At this time, the device sends the voice data to the server as an HTTP request. The input is the voice data stored on the device, and the voice data is transferred to the server as the output.

[1551] Step 3:

[1552] The server preprocesses the received voice data. The server uses a speech recognition engine to convert the voice data into text data. Specifically, it uses the Google Cloud Speech-to-Text API to convert the voice data, such as "I want to see my electricity bill details," into text. The input is voice data, and text data is generated as the output.

[1553] Step 4:

[1554] The server inputs the preprocessed text data into a generative AI model to analyze the user's intent. Using OpenAI's GPT-3, the server inputs the text "I want to see my electricity bill details" and analyzes the user's intent as "I want to know the details of my current electricity bill." The input is text data, and the analyzed user's intent is output.

[1555] Step 5:

[1556] Based on the user's intent, the server retrieves the necessary information by referencing an internal database or an external interface. The server queries the internal database and retrieves the latest electricity bill data (e.g., 5,000 yen). The input is the parsed user's intent, and the electricity bill data is obtained as the output.

[1557] Step 6:

[1558] Based on the information acquired, the server generates an appropriate response for the user. The server uses a generative AI model to generate a text response such as "The latest electricity bill is 5,000 yen." The input is the electricity bill data, and the generated text response is obtained as the output.

[1559] Step 7:

[1560] The server sends the generated response to the terminal. The server sends the generated text data to the terminal as an HTTP response. The input is the generated text response, and the text data is sent to the terminal as output.

[1561] Step 8:

[1562] The terminal displays the response received from the server to the user. The terminal displays a text message on the screen saying "The latest electricity bill is 5000 yen." The input is the text data sent from the server, and the output is the text message displayed on the user's screen.

[1563] The above steps show in detail the process from when a user sends a request to the system to when an appropriate response is received. This system realizes quick and appropriate responses to a variety of data formats.

[1564] (Application example 1)

[1565] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1566] Modern content delivery services require systems that allow users to make requests using a variety of modalities (voice, text, images, and videos) and provide prompt and appropriate responses. However, existing systems often only support specific modalities, and there are few general-purpose systems that support all modalities. Furthermore, there are also limited systems that utilize generative AI models to provide optimal recommendations for user requests and improve the quality of information provided. Therefore, a system that can respond to requests in different modalities and generate optimal responses using generative AI models is needed.

[1567] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1568] In this invention, the server includes: means for receiving request data from a user in at least one of the following formats: voice, text, image, and video; means for transmitting the received request data to the server; means for preprocessing the request data, converting voice data to text, and analyzing image data; means for analyzing the user's intent using the preprocessed data; means for obtaining necessary information by referencing an internal database or an external API; means for generating an appropriate response based on the obtained information; means for generating a prompt sentence in response to the user's request using a generative AI model; means for generating an optimal response using the generated prompt sentence; means for transmitting the generated response to a terminal; and means for the terminal to display the response to the user. This enables prompt and appropriate responses to be provided even when the user makes requests in various formats. Furthermore, the use of a generative AI model enables higher quality recommendations and information provision.

[1569] A "user" is a person using a terminal that makes requests and receives responses in a variety of forms.

[1570] "Request data" is data sent by a user in one or more formats: voice, text, image, video.

[1571] "Preprocessing" refers to the process in which the server analyzes the request data received, converts voice data into text, and extracts necessary information from image data.

[1572] A "voice recognition system" is a system that converts voice data into text data.

[1573] An "image analysis system" is a system that analyzes image data and extracts necessary information.

[1574] "Intent analysis" is the process of using preprocessed data to understand the purpose and intent of a user request.

[1575] An "internal database" is a database stored within the system that is referenced to obtain required information.

[1576] An "external API" is an interface for communicating with services or databases outside the system.

[1577] A "generative AI model" is an artificial intelligence model that generates appropriate responses to user requests.

[1578] A "prompt sentence" is an input sentence to a generative AI model, which is generated based on a user request.

[1579] A "terminal" is a device through which a user sends requests and receives responses from a server.

[1580] A "server" is a computer system that processes request data, analyzes intent, retrieves information, and generates a response.

[1581] The present invention relates to a customer service system capable of processing requests of various modalities, and a specific embodiment thereof will be described below. The system is composed of three main components: a server, a terminal, and a user.

[1582] Receiving a user request

[1583] 1. The user sends a request to the device. The request can be in the form of voice, text, image, video, etc. For example, a user can make a voice request such as, "Tell me the next action movie I should watch."

[1584] 2. The terminal sends the received request data to the server.

[1585] Data Preprocessing

[1586] 1. The server preprocesses the data it receives: if it is voice data, it converts it into text using a voice recognition system, and if it is image data, it uses an image analysis system to extract the necessary information.

[1587] 2. As a speech recognition system, for example, use the speech_recognition library.

[1588] 3. PIL (Python Imaging Library) and pytesseract are used as image analysis systems.

[1589] Intent Analysis

[1590] 1. The server analyzes the user's intent based on the preprocessed data. For example, it understands the intent of "what action movie should I watch next" from the speech data.

[1591] Obtaining information

[1592] 1. The server retrieves the required information from an internal database or an external API. For example, it retrieves a list of recommended action movies from an internal database.

[1593] Generating a response

[1594] 1. Based on the information obtained by the server, an appropriate prompt sentence is generated using a generative AI model.

[1595] 2. The generated prompt sentence will be something like, "What's the next action movie I should watch?"

[1596] 3. Using this prompt, the generative AI model generates an appropriate response, such as "The recommended action movie is 'Recommended Movie Title'."

[1597] Sending and Displaying Responses

[1598] 1. The server generates a response and sends it to the device.

[1599] 2. The device displays a response to the user, for example, "The recommended action movie is 'Recommended Movie Title'."

[1600] Specific examples

[1601] For example, if a user speaks to their device, "What's the next action movie I should watch?", the process goes like this:

[1602] 1. The user's voice data is sent by the terminal to the server.

[1603] 2. The server uses a voice recognition system to convert the voice data into text.

[1604] 3. The server analyzes the user's intent based on the converted text data and generates an appropriate prompt.

[1605] 4. The generative AI model generates a response based on the prompt, "What's the next action movie I should watch?"

[1606] 5. The generated response will be "The recommended action movie is 'Recommended Movie Title'."

[1607] 6. The terminal displays this response to the user.

[1608] This embodiment allows users to easily make different types of requests and receive a quick and appropriate response, an example of a prompt sentence could be "What's the next action movie I should watch?"

[1609] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1610] Step 1:

[1611] The user sends request data. The request can be in the form of voice, text, image, or video. For example, the user might request by voice, "Tell me the next action movie I should watch." This voice data is input into the terminal.

[1612] Step 2:

[1613] The terminal transmits the received voice data to the server. The terminal uses the network to transmit the voice data and sends this data to the server. This inputs the voice data to the server.

[1614] Step 3:

[1615] The server preprocesses the received voice data. Specifically, it converts the voice data into text using the speech_recognition library. This process converts the voice data into text data.

[1616] Step 4:

[1617] The server analyzes the user's intent based on the converted text data. For example, from the text "Tell me the next action movie I should watch," it extracts the intent that the user is looking for movie recommendations.

[1618] Step 5:

[1619] The server retrieves the required information from an internal database or an external API. For example, it retrieves a list of action movies from an internal database. This process retrieves the data corresponding to the user's request.

[1620] Step 6:

[1621] Based on the information acquired by the server, a prompt sentence is generated using a generative AI model. For example, the prompt sentence might be, "Tell me the next action movie I should watch." This generated prompt sentence is input into the generative AI model.

[1622] Step 7:

[1623] The server uses the generative AI model to generate the optimal response based on the prompt. For example, the response generated is "The recommended action movie is 'Recommended Movie Title'." This response becomes the output of the generative AI model.

[1624] Step 8:

[1625] The server sends the generated response to the terminal, and this response message is sent to the terminal via the network and is input to the terminal.

[1626] Step 9:

[1627] The device displays the response received from the server to the user. For example, a message such as "The recommended action movie is 'Recommended Movie Title'" is displayed on the device screen.

[1628] This series of processing steps allows users to make requests in a variety of formats and receive appropriate responses in real time.

[1629] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1630] The present invention provides a customer service system capable of processing requests from multiple modalities, and in particular adds a function for recognizing a user's emotions and adjusting the content of responses accordingly. This makes it possible to provide a more sophisticated and personalized user experience. Specific embodiments of this system are described below.

[1631] Program Overview

[1632] The system's main components are a server, a terminal, and a user, combined with an emotion engine. The emotion engine analyzes emotions from user request data and reflects them in the response. The server is responsible for processing the AI ​​model and managing data, while the terminal provides the user interface, i.e., inputting requests and displaying responses.

[1633] Program Details

[1634] Receiving a user request

[1635] The user makes a request through the device, which can be input in the form of text, voice, image, or video.

[1636] Example: A user texts a device asking, "What's my electricity bill this month?"

[1637] Submitting a Request

[1638] The terminal transmits the received request data to the server.

[1639] Example: The device sends text data such as "What is the electricity bill this month?" to the server.

[1640] Data Preprocessing

[1641] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[1642] Example: The server receives voice data and converts it into text using a speech recognition system: "What is the electricity bill this month?"

[1643] Intention analysis and emotion recognition

[1644] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions and also analyzes the user's emotions using an emotion engine.

[1645] Example: From the text "What's the electricity bill this month?", an AI model interprets the user's intent as "I want to know the current bill amount," and an emotion engine recognizes the user's tension or irritation.

[1646] Database Reference

[1647] The server references an internal database or external API to obtain the necessary information.

[1648] Example: The server retrieves a user's account information from a database and retrieves the latest electricity bill information.

[1649] Response generation and emotional reflection

[1650] Based on the information acquired by the server, an appropriate response is generated taking into account the user's emotions.

[1651] Example: The server generates "This month's electricity bill is 5000 yen," but recognizing the user's frustration, adds an additional "Do you need help with payment methods?"

[1652] Sending a Response

[1653] The server generates a response and sends it to the terminal.

[1654] Example: The server sends a text to the device saying, "This month's electricity bill is 5000 yen. Do you need help with payment?"

[1655] Viewing the response

[1656] The terminal displays the response received from the server to the user.

[1657] Example: The device displays a text message to the user saying, "Your electricity bill this month is 5000 yen. Do you need help paying?"

[1658] Specific examples

[1659] As an example, the process will be described when a user asks a question by voice, "The electric meter reading is not working properly," and feels uneasy.

[1660] 1. Receiving a user request

[1661] The user asks the device a question by voice, "The electric meter reading is not working properly." The voice contains a sense of anxiety.

[1662] The terminal receives the audio data.

[1663] 2. Submitting a Request

[1664] The terminal transmits the voice data to the server.

[1665] 3. Data Preprocessing

[1666] The server converts the audio data into text.

[1667] The speech recognition system generates the text "I'm having trouble reading my electric meter."

[1668] 4. Intention Analysis and Emotion Recognition

[1669] The server inputs the text data into an AI model and analyzes the user's intent.

[1670] The emotion engine analyzes the user's anxiety.

[1671] 5. Referencing the database

[1672] The server retrieves guide information on how to read the electricity meter from an internal database.

[1673] 6. Response generation and emotional reflection

[1674] The server generates a response saying, "Take a photo of your electric meter and tap the camera icon in the chat window," and includes an additional message to address the user's concerns: "If you have any concerns, please contact support."

[1675] 7. Sending the Response

[1676] The server generates a response and sends it to the terminal.

[1677] 8. Displaying the Response

[1678] The device will display a text message to the user saying, "Take a photo of your electric meter and tap the camera icon in the chat screen. If you have any concerns, please contact support."

[1679] This system enables us to provide detailed responses that take into account the user's emotions, improving the quality of the user experience.

[1680] The processing flow will be explained below.

[1681] Step 1:

[1682] The user makes a request through the device, which can be input in the form of text, voice, image, or video.

[1683] Example: A user texts a device asking, "What's my electricity bill this month?"

[1684] Step 2:

[1685] The terminal transmits the received request data to the server.

[1686] Example: The device sends text data such as "What is the electricity bill this month?" to the server.

[1687] Step 3:

[1688] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[1689] Example: The server receives voice data and converts it into text using a speech recognition system: "What is the electricity bill this month?"

[1690] Step 4:

[1691] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions, and then analyzes the user's emotions using an emotion engine.

[1692] Example: From the text "What's the electricity bill this month?", an AI model interprets the user's intent as "I want to know the current bill amount," and the emotion engine recognizes the user's tension or irritation.

[1693] Step 5:

[1694] The server references an internal database or external API to obtain the necessary information.

[1695] Example: The server retrieves a user's account information from a database and retrieves the latest electricity bill information.

[1696] Step 6:

[1697] The server generates an appropriate response based on the acquired information, taking into account the user's emotions based on the emotion engine.

[1698] Example: The server generates the text "This month's electricity bill is 5000 yen", but recognizes the user's frustration and adds "This month's electricity bill is 5000 yen. Do you need help paying?"

[1699] Step 7:

[1700] The server generates a response and sends it to the terminal.

[1701] Example: The server sends a text to the device saying, "This month's electricity bill is 5000 yen. Do you need help with payment?"

[1702] Step 8:

[1703] The terminal displays the response received from the server to the user.

[1704] Example: The device displays a text message to the user saying, "Your electricity bill this month is 5000 yen. Do you need help paying?"

[1705] The above is the specific flow of processing for a user request.

[1706] Example 2

[1707] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1708] Conventional customer service systems simply respond to user requests and do not take the user's emotions into consideration when responding. This does not alleviate user dissatisfaction or frustration, resulting in a poor user experience. The present invention aims to provide a more sophisticated and personalized user experience by analyzing the user's emotions and generating responses accordingly.

[1709] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1710] In this invention, the server includes means for analyzing the user's intention, means for analyzing the user's intention using preprocessed data, means for obtaining necessary information by referencing an internal database or an external API, means for analyzing the user's emotions and adjusting the response content to reflect the emotions, and means for transmitting the generated response to the terminal, thereby enabling the server to generate an appropriate response that takes the user's emotions into consideration and improve the quality of the user experience.

[1711] A "user" is a user who inputs a request to the system.

[1712] "Audio, text, images, video" are different forms of data that users use to make requests to the system.

[1713] "Request data" is data containing information sent by a user, and represents a question or request.

[1714] A "terminal" is a device used by a user to enter requests and receive responses.

[1715] A "server" is a central computer system that processes request data sent by users and generates and sends responses.

[1716] "Preprocessing" refers to initial processing for converting request data into a format that is easy to analyze, and includes, for example, converting voice data into text and analyzing image data.

[1717] A "voice recognition system" is a technology for converting voice data into text data.

[1718] An "image analysis system" is a technology for extracting necessary information from image data.

[1719] "Intent analysis" is the process of understanding the intent of a user request.

[1720] "Sentiment analysis" is the process of identifying a user's emotional state from their request data.

[1721] An "internal database" is an information storage location installed within the system, which contains information about users and data necessary for providing services.

[1722] An "external API" is an interface for accessing external services and databases and obtaining information.

[1723] "Response generation" is the process of creating an appropriate reply to the user based on the analyzed intent and sentiment.

[1724] "Adjusting the content of a response" refers to changing or complementing the content of a generated response to match the emotional state of the user.

[1725] The present invention is a customer service system that can process requests from multiple modalities, and in particular has the function of recognizing the user's emotions and adjusting the response content. This system is composed of major components: a server, a terminal, a user, and an emotion engine, and operates as follows.

[1726] The user makes a request through the device. The request can be input in the form of voice, text, image, or video. For example, the user might ask the device in text, "What is my electricity bill this month?" The device then sends the received request data to the server. If the data is voice, the server uses a voice recognition system to convert it into text, and if it is image data, it uses an image analysis system to extract the required information. Specifically, this could involve using a voice recognition service such as Google Cloud Speech-to-Text API or Amazon Transcribe.

[1727] The preprocessed data is input into an AI model on the server, where the user's intent is analyzed. The emotion engine also analyzes the user's emotions. For example, the server uses OpenAI's GPT model to analyze the intent from the text "What is my electricity bill this month?", and the emotion engine recognizes the user's tension or irritation. This allows the system to interpret the user's intent as "I want to know the current bill amount" and also take the user's emotions into account.

[1728] The server then references internal databases and external APIs to retrieve the necessary information. Specifically, it retrieves the user's account information and latest electricity bill information from databases such as AWS DynamoDB and Google Firestore. Based on this information, the server generates an appropriate response that takes the user's emotions into account. For example, the server uses a generative AI model to generate a response such as "This month's electricity bill is 5000 yen," but recognizes the user's frustration and includes an additional message such as "Do you need help with your payment method?"

[1729] The generated response is sent from the server to the terminal, which then displays this response to the user. For example, the terminal might display the following text message to the user: "This month's electricity bill is 5000 yen. Do you need help with payment?"

[1730] In this way, the present invention enables a finely tuned response that takes into account the user's emotions, improving the quality of the user experience.

[1731] Specific examples

[1732] Consider a case where a user voices a question saying, "I'm having trouble reading my electric meter," and feels anxious.

[1733] 1. The user asks the device a question by voice, "The electric meter reading is not working properly." The user expresses anxiety.

[1734] 2. The device sends the audio data to the server.

[1735] 3. The server converts the voice data into text, typically using the Google Cloud Speech-to-Text API.

[1736] 4. Use AI models to analyze intent and emotion engines to analyze anxiety.

[1737] 5. The server retrieves guidance information on how to read the electricity meter from its internal database.

[1738] 6. The server generates a response that says "Take a photo of your electric meter and tap the camera icon in the chat window," with an additional message that says "If you have any concerns, please contact support."

[1739] 7. The response is sent to the terminal, which displays it to the user.

[1740] This allows the user to feel at ease and solve the problem.

[1741] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1742] Step 1:

[1743] The user inputs a request into the device, which can be in the form of voice, text, image, or video.

[1744] Example: A user texts a device asking, "What's my electricity bill this month?"

[1745] Input: User request (e.g., text format)

[1746] Output: Request data in the terminal

[1747] Step 2:

[1748] The terminal transmits the received request data to the server.

[1749] Specific operation: The terminal sends the request data to the server via the network.

[1750] Input: Request data in the terminal

[1751] Output: Request data to the server

[1752] Step 3:

[1753] The server preprocesses the request data received: if it is voice data, it converts it into text using a voice recognition system, and if it is image data, it uses an image analysis system to extract the necessary information.

[1754] What happens: The server uses a speech recognition API (e.g., Google Cloud Speech-to-Text) to convert the speech to text.

[1755] Input: Request data to the server (e.g., audio data)

[1756] Output: Preprocessed text data

[1757] Step 4:

[1758] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions, and also analyzes the user's emotions using an emotion engine.

[1759] How it works: The server uses a generative AI model (e.g., OpenAI GPT) to analyze the intent and evaluates the user's emotions with an emotion engine.

[1760] Input: Preprocessed text data

[1761] Output: Parsed intent and emotion data

[1762] Step 5:

[1763] The server retrieves the required information by referencing an internal database or an external API.

[1764] Specific operation: The server retrieves the user's account information from a database (e.g., AWS DynamoDB) and collects related information.

[1765] Input: Parsed intent and emotion data

[1766] Output: Required information (e.g. electricity bill information)

[1767] Step 6:

[1768] Based on the information obtained by the server, the user's emotions are taken into consideration and an appropriate response is generated.

[1769] What it does: The server uses a generative AI model to generate a response like "Your electricity bill this month is 5000 yen" and includes an additional message based on sentiment data: "Do you need help with payment methods?"

[1770] Input: Required information (e.g., electricity bill information) and emotion data

[1771] Output: The generated response

[1772] Step 7:

[1773] The server generates a response and sends it to the terminal.

[1774] Specific operation: The server sends response data to the terminal using the network.

[1775] Input: The generated response

[1776] Output: Response data to the terminal

[1777] Step 8:

[1778] The terminal displays the response received from the server to the user.

[1779] Specific operation: The terminal displays a response message on the display.

[1780] Input: Response data to the terminal

[1781] Output: The response message that is displayed to the user

[1782] (Application example 2)

[1783] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1784] Current customer service systems can only provide standard responses to user requests, making it difficult to provide a personalized user experience. Furthermore, they lack the ability to recognize user emotions and respond appropriately, resulting in insufficient responses when users feel dissatisfied or anxious. Therefore, improving the quality of the user experience is a challenge in areas such as food delivery, where fast and personalized responses are required.

[1785] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1786] In this invention, the server includes a means for analyzing user emotions, a means for analyzing user intentions using preprocessed data, and a means for obtaining necessary information by referencing an internal database or an external API, thereby enabling the server to generate an appropriate response according to the user's emotions.

[1787] "User request data" refers to information provided by a user to the system in the form of voice, text, images, or video.

[1788] A "server" is a computing system that receives, pre-processes, and generates a response from a user's request data.

[1789] A "voice recognition system" is a technology that converts voice data into text.

[1790] An "image analysis system" is a technology that extracts specific information from image data.

[1791] "Means for analyzing user intent" refers to a process for understanding the user's purpose and requirements from the user's request data.

[1792] "Means for analyzing user emotions" refers to technology that recognizes and classifies the user's emotional state based on request data.

[1793] An "internal database" is a database that stores internal information managed by the system.

[1794] An "External API" is an application programming interface for accessing services or data outside the system.

[1795] The "means for generating an appropriate response" is a process for generating an optimal response to the user based on the acquired information and the analyzed user's emotions.

[1796] A "terminal" is a device through which a user inputs requests and receives responses.

[1797] The present invention provides a customer service system capable of processing requests from multiple modalities, and in particular adds a function for recognizing a user's emotions and adjusting the content of responses, thereby enabling a more sophisticated and personalized user experience. An embodiment of this system will be described below.

[1798] System Configuration

[1799] The system consists of a server, a terminal, a user, and an emotion engine as its main components. The server is responsible for processing the AI ​​model and managing data, while the terminal provides the user interface, i.e., inputting requests and displaying responses.

[1800] Hardware and Software Specifications

[1801] Server: A high-performance computer system is required.

[1802] Device: Using a user interface device such as a smartphone, smart glasses, a head-mounted display, or a robot.

[1803] Software used:

[1804] AI models: Use Natural Language Processing (NLP) libraries and models for sentiment analysis.

[1805] Speech recognition system: Used to convert voice data into text.

[1806] Image analysis system: Used to analyze image data and extract necessary information.

[1807] Emotion Engine: Used to analyze emotions from user request data.

[1808] Program processing flow

[1809] Receiving a user request

[1810] The user makes a request through the device. The request can be input in the form of voice, text, image, or video. For example, the user can type "My order hasn't arrived" into their smartphone.

[1811] Submitting a Request

[1812] The terminal transmits the received request data to the server.

[1813] Data Preprocessing

[1814] The server preprocesses the request data received. For voice data, it converts it into text using a voice recognition system. For image data, it extracts the necessary information using an image analysis system.

[1815] Intention analysis and emotion recognition

[1816] The server inputs the preprocessed data into the AI ​​model to analyze the user's intentions and also analyzes the user's emotions using an emotion engine.

[1817] Database Reference

[1818] The server references an internal database or external API to obtain the necessary information.

[1819] Response generation and emotional reflection

[1820] Based on the information acquired by the server, the server generates an appropriate response taking into account the user's emotions. For example, it generates a response that takes into account emotions, such as "We apologize for the delay in your order. We are currently checking the current status. Would you mind waiting?"

[1821] Sending a Response

[1822] The server generates a response and sends it to the terminal.

[1823] Viewing the response

[1824] The device displays the response it receives from the server to the user. For example, a message like "We apologize for the delay in your order. We are currently checking the current status. Would you mind waiting?" appears on the smartphone screen.

[1825] Specific examples

[1826] Let's take a concrete example: when a user types "My order hasn't arrived" into their smartphone, the request is sent to the server, which preprocesses the data using a speech recognition system and image analysis system. The emotion engine then analyzes the user's frustration and generates an appropriate response using a generative AI model.

[1827] Prompt Sentence Examples

[1828] Analyze the sentiment when a user inquires about the status of an order in a food delivery app. Analyze the following user input text and generate a response based on the sentiment:

[1829] User Input: "Order not received"

[1830] Expected output:

[1831] User Sentiment: Annoyance

[1832] Response: "Sorry for the delay in your order. We're currently checking the status. Would you mind waiting?"

[1833] By using prompt sentences in this way, the generative AI model can recognize emotions and generate responses appropriately.

[1834] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1835] Step 1:

[1836] The user enters a request

[1837] Input: Text or speak "My order hasn't arrived" into your smartphone or other device.

[1838] Action: The user interacts with the device and types a request, which may include expressions of frustration or annoyance.

[1839] Output: The device receives the user's request data.

[1840] Step 2:

[1841] Sends request data to the server

[1842] Input: User request data obtained by the device.

[1843] What happens: Your device sends data to a server over Wi-Fi or your mobile network.

[1844] Output: The server receives the request data.

[1845] Step 3:

[1846] Data Preprocessing

[1847] Input: The request data received by the server.

[1848] How it works: The server converts voice data into text using a voice recognition system, and analyzes image data using an image analysis system.

[1849] Output: Text data and analysis results are obtained.

[1850] Step 4:

[1851] Intention analysis and emotion recognition

[1852] Input: Text data preprocessed by the server and analysis results.

[1853] How it works: The server inputs text data into the AI ​​model, analyzes the user's intention, and recognizes the user's emotions using the emotion engine.

[1854] Output: Data is obtained that analyzes the user's intent and emotions.

[1855] Step 5:

[1856] Referencing a database or external API

[1857] Input: User intent and sentiment parsed by the server.

[1858] What happens: The server looks up an internal database or an external API to get the information it needs. For example, it looks up a database to check the status of an order.

[1859] Output: The required information (e.g., current order status) is obtained.

[1860] Step 6:

[1861] Response generation and emotional reflection

[1862] Input: Captured information and analyzed user sentiment.

[1863] How it works: Based on the information the server obtains, it generates a response that takes into account the user's feelings. For example, it generates a message like, "We're sorry your order is delayed. We're currently checking the current status. Would you mind waiting?"

[1864] Output: A response message is generated that takes the user's feelings into consideration.

[1865] Step 7:

[1866] Sending a Response

[1867] Input: The server-generated response message.

[1868] Action: The server sends a response message to the terminal.

[1869] Output: The terminal receives the response message.

[1870] Step 8:

[1871] Viewing the response

[1872] Input: Response message received by the terminal.

[1873] What it does: The device displays a response message to the user. For example, a message on a smartphone screen might say, "We apologize for the delay in your order. We're currently checking the current status. Would you mind waiting?"

[1874] Output: The user confirms the response message and takes further action if necessary.

[1875] 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 control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1876] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1877] 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 the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1878] The emotion identification model 59 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 an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1879] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1880] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1881] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1882] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1883] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs 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 a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1884] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1885] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1886] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1887] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1888] 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.

[1889] It is not necessary to store all 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 all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1890] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1891] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with 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). Also, the hardware resource that executes the specific processing may be a single processor.

[1892] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1893] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1894] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1895] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1896] The following is further disclosed regarding the above embodiment.

[1897] (Claim 1)

[1898] means for receiving request data from a user in the form of at least one of voice, text, image, and video;

[1899] means for transmitting the received request data to a server;

[1900] means for preprocessing the request data, converting voice data to text, and analyzing image data;

[1901] means for analyzing user intent using the preprocessed data;

[1902] A means to retrieve the required information by referencing an internal database or external API;

[1903] means for generating an appropriate response based on the obtained information;

[1904] means for transmitting the generated response to the terminal;

[1905] means for the terminal to display the response to the user;

[1906] A system including:

[1907] (Claim 2)

[1908] 2. The system of claim 1, wherein the means for converting voice data to text is a voice recognition system.

[1909] (Claim 3)

[1910] 2. The system of claim 1, wherein the means for analyzing the image data is an image analysis system.

[1911] "Example 1"

[1912] (Claim 1)

[1913] means for receiving request data from a user in the form of at least one of voice, text, image, and video;

[1914] means for transmitting the received request data to a server;

[1915] means for preprocessing the request data, converting voice data to text, and analyzing image data;

[1916] means for using a generative model to analyze user intent using the preprocessed data;

[1917] A means of obtaining the required information by referencing an internal database or an external interface;

[1918] means for generating an appropriate response based on the obtained information;

[1919] means for transmitting the generated response to the terminal;

[1920] means for the terminal to display the response to the user;

[1921] A system including:

[1922] (Claim 2)

[1923] 2. The system of claim 1, wherein the means for converting voice data to text is a voice recognition system.

[1924] (Claim 3)

[1925] 2. The system of claim 1, wherein the means for analyzing the image data is an image analysis system.

[1926] "Application Example 1"

[1927] (Claim 1)

[1928] means for receiving request data from a user in the form of at least one of voice, text, image, and video;

[1929] means for transmitting the received request data to a server;

[1930] means for preprocessing the request data, converting voice data to text, and analyzing image data;

[1931] means for analyzing user intent using the preprocessed data;

[1932] A means to retrieve the required information by referencing an internal database or external API;

[1933] means for generating an appropriate response based on the obtained information;

[1934] a means for generating a prompt sentence in response to a user request using a generative AI model;

[1935] a means for generating an optimal response using the generated prompt sentence;

[1936] means for transmitting the generated response to the terminal;

[1937] means for the terminal to display the response to the user;

[1938] A system including:

[1939] (Claim 2)

[1940] 2. The system of claim 1, wherein the means for converting voice data to text is a voice recognition system.

[1941] (Claim 3)

[1942] 2. The system of claim 1, wherein the means for analyzing the image data is an image analysis system.

[1943] "Example 2: Combining Emotion Engines"

[1944] (Claim 1)

[1945] means for receiving request data from a user in the form of at least one of voice, text, image, and video;

[1946] means for transmitting the received request data to a server;

[1947] means for preprocessing the request data, converting voice data to text, and analyzing image data;

[1948] means for analyzing user intent using the preprocessed data;

[1949] A means to retrieve the required information by referencing an internal database or external API;

[1950] means for generating an appropriate response based on the obtained information;

[1951] A means for analyzing the user's emotions and adjusting the response content to reflect the emotions;

[1952] means for transmitting the generated response to the terminal;

[1953] means for the terminal to display the response to the user;

[1954] A system including:

[1955] (Claim 2)

[1956] 2. The system of claim 1, wherein the means for converting voice data to text is a voice recognition system.

[1957] (Claim 3)

[1958] 2. The system of claim 1, wherein the means for analyzing the image data is an image analysis system.

[1959] "Application example 2 when combining emotion engines"

[1960] (Claim 1)

[1961] means for receiving request data from a user in the form of at least one of voice, text, image, and video;

[1962] means for transmitting the received request data to a server;

[1963] means for preprocessing the request data, converting voice data to text, and analyzing image data;

[1964] means for analyzing user intent using the preprocessed data;

[1965] means for analyzing user emotions;

[1966] A means to retrieve the required information by referencing an internal database or external API;

[1967] means for generating an appropriate response based on the acquired information and the analyzed emotions;

[1968] means for transmitting the generated response to the terminal;

[1969] means for the terminal to display the response to the user;

[1970] A system including:

[1971] (Claim 2)

[1972] 2. The system of claim 1, wherein the means for converting voice data to text is a voice recognition system.

[1973] (Claim 3)

[1974] 2. The system of claim 1, wherein the means for analyzing the image data is an image analysis system. [Explanation of symbols]

[1975] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for receiving request data from a user in the form of at least one of voice, text, image, and video; means for transmitting the received request data to a server; means for preprocessing the request data, converting voice data to text, and analyzing image data; means for analyzing user intent using the preprocessed data; A means to retrieve the required information by referencing an internal database or external API; means for generating an appropriate response based on the obtained information; means for transmitting the generated response to the terminal; means for the terminal to display the response to the user; A system including:

2. 2. The system of claim 1, wherein said means for converting voice data to text is a voice recognition system.

3. 2. The system of claim 1, wherein the means for analyzing the image data is an image analysis system.

Images