System

Abstract

A system is provided.SOLUTION: A system includes a means for receiving an inquiry from a customer, a means for transmitting the received inquiry content to a generative AI model to generate an answer, a means for providing the generated answer to the customer, a means for analyzing an emotion of the inquiry content of the customer, and a means for optimizing a response content based on a result of the emotion analysis.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] As the number of customer inquiries increases, companies are demanding faster and more accurate responses, which requires more time and manpower. However, there are limits to what can be done manually, resulting in problems such as a decline in customer satisfaction and increased operational costs. It is also difficult to properly assess the emotions expressed in customer inquiries and provide the optimal response. To solve these issues, it is necessary to introduce an automated response system that uses advanced natural language processing technology. [Means for solving the problem]

[0005] The present invention provides a system including a means for receiving customer inquiries, a means for generating answers by passing the received inquiry content to a generative AI model, a means for providing the generated answers to customers, a means for analyzing the sentiment of the customer's inquiry content, and a means for optimizing the response content based on the results of the sentiment analysis. This enables quick and accurate responses to customer inquiries and improves customer satisfaction. Furthermore, optimizing the response method based on the results of the sentiment analysis enables more appropriate responses and reduces operational costs.

[0006] "Means for receiving inquiries from customers" refers to an interface or function that receives the inquiry content entered by the customer and transmits that information to the processing system.

[0007] A "generative AI model" refers to an algorithm or software that uses natural language processing technology to generate answers in natural language in response to input queries.

[0008] "Means for generating an answer" refers to the process or mechanism that passes the received query to a generative AI model and generates the resulting answer.

[0009] "Means for providing answers to customers" refers to an interface or system that displays or provides the generated answers to customers by voice.

[0010] "Means for analyzing emotions" refers to technologies and algorithms that analyze customer inquiries and identify and evaluate the emotions contained therein.

[0011] "Means for optimizing response content" refers to the process or function of determining the most appropriate response based on the results of sentiment analysis and adjusting the response method.

[0012] "Natural language processing technology" refers to technologies and methods that enable computers to understand, generate, and analyze natural human language.

[0013] A "sentiment analysis pipeline" refers to a series of analysis processes and algorithms used to identify the emotional state (e.g., joy, sadness, anger, etc.) of natural language data such as sentences or utterances. [Brief explanation of the drawings]

[0014] [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

[0015] 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.

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

[0017] 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).

[0018] 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.

[0019] 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.

[0020] 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.

[0021] 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."

[0022] [First embodiment]

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

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form 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.

[0029] 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.

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

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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."

[0035] The present invention is a system for responding to customer inquiries quickly and accurately, and in particular utilizes generative AI models and sentiment analysis techniques. Specific embodiments of this system are described below.

[0036] First, the terminal provides an interface for receiving queries entered by the user. For example, suppose a user enters a query such as "My Internet connection is slow. What should I do?" This query is sent to the system through the input field of the terminal.

[0037] The server then processes the received query. The server analyzes the query and passes it to a generative AI model (e.g., a model using natural language processing technology). This generative AI model generates an appropriate answer to the query. For example, it might say, "If your Internet connection is slow, try restarting your router."

[0038] The generated answer is returned to the terminal and displayed to the user, who can then take action based on the answer.

[0039] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it uses a sentiment analysis pipeline (e.g., an algorithm that identifies the sentiment in a sentence). If the user's query contains negative sentiment, such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server assigns a "NEGATIVE" label to the query.

[0040] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, it can present additional support options. The message displayed to the user might say, "It seems you're having trouble. Would you like us to connect you to our support team?"

[0041] As can be seen from this example, this system not only responds to customer inquiries quickly and accurately, but also aims to improve customer satisfaction by providing optimal responses through sentiment analysis. This system also reduces operational costs.

[0042] The processing flow will be explained below.

[0043] Step 1:

[0044] The user inputs a query. For example, the user might input "My Internet connection is slow. What should I do?" The user's input is entered into an input field on the device.

[0045] Step 2:

[0046] The terminal receives the query entered by the user, which is then temporarily stored and prepared for transmission to the server.

[0047] Step 3:

[0048] The terminal sends the received query content to the server. Specifically, the terminal creates an HTTP POST request, includes the query content in JSON format in the request body, and sends it to the specified endpoint of the server.

[0049] Step 4:

[0050] The server analyzes the received query, extracts the query content from the received request, and passes it to a generative AI model to generate an appropriate answer.

[0051] Step 5:

[0052] The server generates an answer using a generative AI model. The generative AI model generates the best answer based on the query and returns that answer to the server. For example, if the query is "My Internet connection is slow. What should I do?", the generated answer might be "If your Internet connection is slow, try restarting your router."

[0053] Step 6:

[0054] The server returns the generated answer to the terminal. The server creates an HTTP response that includes the generated answer and sends the response to the terminal.

[0055] Step 7:

[0056] The device displays the answer received from the server to the user. The device analyzes the response from the server and displays the generated answer on the screen. For example, the answer "If your Internet connection is slow, try restarting your router" may be presented to the user.

[0057] Step 8:

[0058] The server analyzes the sentiment of the inquiry. Using a sentiment analysis pipeline, the server analyzes the received inquiry and identifies the sentiment. For example, if the inquiry is "I can't connect to the Internet at all and I'm really in trouble. Please help me," the result of sentiment analysis will be assigned a "NEGATIVE" label.

[0059] Step 9:

[0060] The server optimizes the response based on the results of sentiment analysis. If the result of sentiment analysis is negative, it generates a message suggesting additional support. For example, it generates a message saying, "It seems you're having trouble. Would you like me to connect you to our support desk?"

[0061] Step 10:

[0062] The device displays the emotion analysis results and additional support suggestions sent from the server to the user. The user can see a message on the screen saying, "It seems you're having trouble. Would you like us to direct you to our support desk?"

[0063] Step 11:

[0064] The user can select additional support options or follow the generated answer to resolve the issue, thus providing the user with a timely and appropriate response.

[0065] Example 1

[0066] 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."

[0067] Conventional customer support systems often lacked timely and accurate responses to customer inquiries, leading to poor customer satisfaction. Furthermore, they lacked the ability to consider customer emotions, making it difficult to provide appropriate support, especially to customers with negative feelings. Additionally, these systems also had the problem of high operating costs.

[0068] 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.

[0069] In this invention, the server includes a means for passing the received inquiry content to a generative AI model to generate an answer, a means for providing the generated answer to the customer via a terminal, a means for analyzing the sentiment of the customer's inquiry content, and a means for optimizing the response content based on the results of the sentiment analysis. This enables quick and accurate responses to customer inquiries, and by realizing optimal responses through sentiment analysis, it is possible to improve customer satisfaction. It also realizes efficient operations and enables cost reduction.

[0070] A "terminal" is an input device that allows a customer to input an inquiry and transmit it to the system.

[0071] The "server" is a central processing unit that analyzes the content of the received inquiry, generates a response using a generative AI model, and then performs sentiment analysis to optimize the response content.

[0072] An "inquiry" is a question or inquiry that a customer inputs into the system via a terminal.

[0073] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to generate appropriate answers to customer inquiries.

[0074] An "answer" is a response generated by a generative AI model and provided to a customer.

[0075] "Sentiment analysis" is a technology that analyzes the content of customer inquiries and identifies their emotional state.

[0076] A "sentiment analysis pipeline" is a series of processing steps for performing sentiment analysis, including algorithms for extracting sentiment from text data.

[0077] "Response optimization" is a technique that provides optimal answers and additional support options to customers based on the results of sentiment analysis.

[0078] The present invention is a system for responding to customer inquiries quickly and accurately, and in particular utilizes generative AI models and sentiment analysis techniques. Specific embodiments of this system are described below.

[0079] First, the terminal provides an interface for receiving queries entered by the user. For example, suppose a user enters a query such as "My Internet connection is slow. What should I do?" This query is sent to the system through the input field of the terminal.

[0080] The server then processes the received query. The server analyzes the query and passes it to a generative AI model (e.g., a model using natural language processing technology). This generative AI model generates an appropriate answer to the query. For example, it might say, "If your Internet connection is slow, try restarting your router."

[0081] The generated answer is returned to the terminal and displayed to the user, who can then take action based on the answer.

[0082] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it uses a sentiment analysis pipeline (e.g., an algorithm that identifies the sentiment in a sentence). If the user's query contains negative sentiment, such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server assigns a "NEGATIVE" label to the query.

[0083] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, additional support options can be presented. The message displayed to the user is, "It seems you're having trouble. Would you like us to connect you to our support team?"

[0084] As can be seen from this example, this system not only responds to customer inquiries quickly and accurately, but also aims to improve customer satisfaction by providing optimal responses through sentiment analysis. This system also reduces operational costs.

[0085] Below are some examples of prompt sentences.

[0086] Example prompt: "My internet connection is slow. What should I do?"

[0087] The system aims to achieve a high level of customer satisfaction by quickly and accurately processing customer inquiries and optimizing responses through sentiment analysis, so it is important to effectively utilize generative AI models and sentiment analysis pipelines.

[0088] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0089] Step 1:

[0090] A user inputs a query through the terminal, for example, "My Internet connection is slow. What should I do?" The terminal receives this query and sends it to the server.

[0091] Input: User's inquiry

[0092] Output: Query sent to server

[0093] What happens: The user uses the device interface to enter a question into the text box and clicks "Submit."

[0094] Step 2:

[0095] The server receives the query, normalizes and parses it, which includes removing unnecessary whitespace and special characters from the text, and then passes the query to a generative AI model.

[0096] Input: Query sent from the terminal

[0097] Output: Normalized and parsed text

[0098] Specific operation: After receiving the query, the server cleans and preprocesses the text, preparing it for input into the generative AI model.

[0099] Step 3:

[0100] The server uses a generative AI model to generate an appropriate answer to the query. The query is passed to the generative AI model (for example, a model using natural language processing technology) to generate an answer. For example, the server may receive a response such as, "If your Internet connection is slow, try restarting your router."

[0101] Input: Normalized text query

[0102] Output: The answer generated by the generative AI model

[0103] How it works: The server inputs a query into a natural language processing model and retrieves the best answer provided by the model.

[0104] Step 4:

[0105] The server sends the generated answer to the terminal, which displays the answer to the user.

[0106] Input: The answer generated by the generative AI model

[0107] Output: The answer that is displayed to the user

[0108] Specific operation: The device displays the response received from the server on the screen for the user to see. For example, it may display "If your Internet connection is slow, try restarting your router."

[0109] Step 5:

[0110] The server passes the query content to a sentiment analysis pipeline for sentiment analysis, which detects negative sentiment such as "troubled."

[0111] Input: Normalized text query

[0112] Output: Sentiment analysis result (e.g., NEGATIVE)

[0113] What it does: The server uses a sentiment analysis algorithm to identify positive, negative, neutral, or other sentiment from the input text.

[0114] Step 6:

[0115] The server optimizes the response based on the results of sentiment analysis. If negative sentiment is detected, it generates additional support options, such as "It seems like you're having trouble. Would you like me to connect you to our support team?"

[0116] Input: Sentiment analysis results

[0117] Output: Optimized response (e.g., "Sounds like you're having trouble. Would you like me to connect you to support?")

[0118] What happens: The server runs logic to generate an appropriate follow-up message for the user based on the results of the sentiment analysis.

[0119] Step 7:

[0120] The server transmits the optimized response content to the terminal, which then displays the content to the user.

[0121] Input: Optimized response

[0122] Output: The optimized greeting that is displayed to the user.

[0123] Specific operation: The device displays the optimized message received from the server on the screen and guides the user, for example, "It seems you are having trouble. Would you like us to direct you to our support center?"

[0124] Through these steps, the system can respond to customer inquiries quickly and accurately, and even provide optimal responses through sentiment analysis.

[0125] (Application example 1)

[0126] 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."

[0127] Modern online shopping sites are required to respond quickly and accurately to inquiries from a large number of customers. However, conventional systems make it difficult to respond while taking into account customer emotions, making it difficult to improve customer satisfaction and streamline customer support. Furthermore, there is a lack of support tools that enable customer support staff to respond optimally in real time, making it difficult to maintain high response quality.

[0128] 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.

[0129] In this invention, the server includes means for receiving inquiries from customers, means for passing the received inquiry content to a generative AI model to generate a response, means for providing the generated response to the customer, means for analyzing the sentiment of the customer's inquiry content, means for optimizing the response content based on the result of the sentiment analysis, and means for displaying the inquiry content, the response of the generative AI model, and the result of the sentiment analysis in real time using smart glasses. This enables quick and accurate responses that take customer sentiment into consideration, thereby improving the efficiency of customer support and customer satisfaction.

[0130] "Customer" refers to a consumer or entity that uses a product or service.

[0131] "Inquiry" refers to the act of a customer communicating a question, complaint, or request regarding a product or service to customer support.

[0132] "Means of receiving" refers to the interface or mechanism for receiving inquiries from customers.

[0133] "Generative AI model" refers to an artificial intelligence model that uses natural language processing technology to generate appropriate answers to input inquiries.

[0134] "Answer" refers to the answer or advice that a generative AI model generates based on a query.

[0135] "Means for analyzing emotions" refers to technology that analyzes emotions from customer inquiries and classifies them into emotional labels such as positive or negative.

[0136] "Means for optimizing response content" refers to technology that determines and provides the optimal response method for customers based on the results of sentiment analysis.

[0137] "Smart glasses" are a wearable device in the form of glasses equipped with a display device, allowing customer support representatives to view information in real time.

[0138] "Real-time" refers to information processing occurring simultaneously with real time, with almost no delay.

[0139] The present invention is a system for responding to customer inquiries quickly and accurately, and in particular utilizes generative AI models and sentiment analysis techniques. A specific embodiment of this system is described below.

[0140] First, the server and terminal are the main hardware components that realize this system. The terminal is implemented as smart glasses and is worn by a customer support representative. The terminal provides an interface for receiving customer inquiries using input or voice recognition technology. For example, a user can make an inquiry such as, "My product delivery is delayed. What should I do?" This inquiry is sent to the server through the input field on the terminal.

[0141] The server then processes the received inquiry. The server analyzes the inquiry and passes it to a generative AI model (for example, a model using natural language processing technology). This generative AI model generates an appropriate answer to the inquiry. For example, it might respond, "Please wait a moment while we check the delivery status." The generated answer is then sent back to the device and displayed on the customer support representative's smart glasses. The representative can then explain the situation to the customer while looking at the answer.

[0142] The server then analyzes the sentiment of the inquiry. Sentiment analysis uses a sentiment analysis pipeline (e.g., an algorithm that identifies emotions in sentences). If the user's inquiry contains a negative sentiment, such as "I'm very frustrated because my item hasn't arrived yet," the sentiment analysis will assign a "NEGATIVE" label. Based on the results of this sentiment analysis, the server can optimize the response. For example, if a negative sentiment is detected, additional support options can be presented. A message such as "It seems you're having trouble. Would you like me to direct you to our support desk for further assistance?" will be displayed on the customer support representative's smart glasses.

[0143] Hardware and software used

[0144] Hardware: Smart glasses (built-in microphone, display)

[0145] software:

[0146] Speech Recognition Library:SpeechRecognition Library

[0147] Generative AI model: a natural language processing model (e.g., OpenAI's GPT)

[0148] Sentiment Analysis Pipeline: Transformer Library and Japanese Model

[0149] For example, the following prompt sentences may be used when a customer asks, "My product delivery is delayed. What should I do?"

[0150] Prompt Sentence Examples

[0151] Customer enquiry: "My item delivery is delayed. What should I do?"

[0152] Please provide an appropriate response.

[0153] With the above configuration, the present invention enables quick and accurate responses that take into consideration the customer's feelings, thereby realizing more efficient customer support and improved customer satisfaction.

[0154] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0155] Step 1:

[0156] The user makes a query.

[0157] For example, a user may make a voice inquiry through the built-in microphone of the smart glasses, saying, "My product delivery is delayed. What should I do?" The voice data is input into the terminal.

[0158] Step 2:

[0159] The device converts the voice data into text.

[0160] The device's speech recognition library (for example, the SpeechRecognition library) is used to convert the voice data into text data. The converted text becomes the string "My product delivery is delayed. What should I do?"

[0161] Step 3:

[0162] The text data is sent to the server.

[0163] The terminal transmits the converted character string data to the server, which receives the text data.

[0164] Step 4:

[0165] The server generates an answer using a generative AI model.

[0166] The server passes the received text data to a generative AI model (for example, a model using natural language processing) to generate an appropriate answer. Using the prompt sentence, the answer to "My product delivery is delayed. What should I do?" is output as "We will check the delivery status, so please wait a moment."

[0167] Step 5:

[0168] The generated answer is sent to the terminal.

[0169] The server sends the generated response "We will check the delivery status, so please wait a moment" to the terminal.

[0170] Step 6:

[0171] The device will display the answer.

[0172] The terminal displays the received answer on the display of the smart glasses, where a customer support representative can view the answer and communicate it to the user verbally or via text.

[0173] Step 7:

[0174] The server performs sentiment analysis.

[0175] The server passes the received text data to a sentiment analysis pipeline (e.g., the Transformer library and a Japanese model) to analyze the sentiment. If the input is "My product delivery is delayed. What should I do?", a negative sentiment is detected. The analysis results in a label of "NEGATIVE."

[0176] Step 8:

[0177] Optimize responses based on the results of sentiment analysis.

[0178] The server generates additional response options based on the results of sentiment analysis, such as a suggestion like, "It sounds like you're having trouble. Would you like me to direct you to our support team if you need further assistance?"

[0179] Step 9:

[0180] Suggestions for what to do are displayed on the device.

[0181] The server sends the generated suggestions to the device, which displays them on the smart glasses display. The customer support representative then uses the suggestions to provide the optimal response.

[0182] 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.

[0183] The present invention provides a system for quickly and accurately responding to customer inquiries, and in particular utilizes generative AI models, sentiment analysis techniques, and an emotion engine. Specific embodiments of this system are described below.

[0184] First, the terminal provides an interface to receive queries entered by the user. For example, a user may enter a query such as "My Internet connection is slow. What should I do?" The user's input is sent to the system through the input field on the terminal.

[0185] The server then processes the received query. The server analyzes the query and passes it to a generative AI model (for example, a model using natural language processing technology) to generate an appropriate answer. The generated answer is then returned to the server and sent to the device. For example, the answer generated might be, "If your Internet connection is slow, try restarting your router."

[0186] The generated answer is returned to the terminal and displayed to the user, who can then take action based on the answer.

[0187] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it uses a sentiment analysis pipeline (e.g., an algorithm that identifies the sentiment in a sentence). If the user's query contains negative sentiment, such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server assigns a "NEGATIVE" label to the query.

[0188] The server also uses an emotion engine to recognize the user's emotions. The emotion engine analyzes voice data and text data to identify the user's emotions. For example, when a user makes a voice inquiry, the emotion engine analyzes the voice data to determine whether it contains negative emotions. Similarly, the emotion engine also analyzes text data to identify emotions.

[0189] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, it can present additional support options. The message displayed to the user might say, "It seems you're having trouble. Would you like us to connect you to our support team?"

[0190] As can be seen from this example, this system not only responds to customer inquiries quickly and accurately, but also aims to improve customer satisfaction by providing optimal responses through emotion analysis using an emotion engine. This system also reduces operational costs.

[0191] The processing flow will be explained below.

[0192] Step 1:

[0193] The user inputs the inquiry, for example, "My Internet connection is slow. What should I do?", and clicks the send button.

[0194] Step 2:

[0195] The terminal receives the inquiry entered by the user, temporarily stores the inquiry internally, and sends it to the server when it is ready.

[0196] Step 3:

[0197] The device sends the received query content to the server. Specifically, it creates an HTTP POST request, includes the query content in JSON format, and sends it to the server's API endpoint.

[0198] Step 4:

[0199] The server analyzes the received query, extracts the query content from the HTTP request, and passes it to a generative AI model to generate an appropriate answer.

[0200] Step 5:

[0201] The server generates an answer using a generative AI model, which generates an answer based on the query, such as "If your internet connection is slow, try restarting your router," and sends the answer back to the server.

[0202] Step 6:

[0203] The server returns the generated answer to the terminal, creates an HTTP response including the generated answer, and sends it to the terminal.

[0204] Step 7:

[0205] The device displays the answer received from the server to the user. It analyzes the response from the server and displays the generated answer on the screen. For example, it may say, "If your Internet connection is slow, try restarting your router."

[0206] Step 8:

[0207] The server analyzes the sentiment of the inquiry. Using a sentiment analysis pipeline, it analyzes the inquiry and identifies the sentiment. For example, if the inquiry says, "I can't connect to the Internet at all and I'm really in trouble. Please help me," the sentiment analysis will assign a "NEGATIVE" label.

[0208] Step 9:

[0209] The server uses an emotion engine to recognize the user's emotions. The emotion engine analyzes voice and text data to identify the user's emotions. For example, when a user makes a voice inquiry, the server analyzes the voice data to determine whether the user is expressing negative emotions.

[0210] Step 10:

[0211] The server optimizes the response based on the results of the sentiment analysis. If negative sentiment is detected, it will offer additional support options, such as generating an additional message like, "It seems like you're having trouble. Would you like us to connect you to our support team?"

[0212] Step 11:

[0213] The device displays the emotion analysis results and additional support suggestions sent from the server to the user. The message "It seems you're having trouble. Would you like us to direct you to our support desk?" is displayed on the screen.

[0214] Step 12:

[0215] The user can select additional support options or follow the initial response to resolve the issue, thus providing the user with a proper and prompt response.

[0216] Example 2

[0217] 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."

[0218] While it is necessary to respond to customer inquiries quickly and accurately, it is also important to appropriately understand and respond to customer emotions. However, conventional systems have had the problem of difficulty not only responding to the content of the inquiry but also identifying the customer's emotions and providing an optimal response. The present invention aims to provide a system that responds to customer inquiries quickly and accurately, performs emotion analysis, and realizes an optimal response based on the customer's emotions.

[0219] 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.

[0220] In this invention, the server includes a means for a terminal to receive an inquiry from a customer, a means for analyzing the received inquiry, a means for passing the analysis result to a generative AI model to generate an answer, a means for returning the generated answer to the terminal and providing it to the customer, a means for analyzing the sentiment of the customer's inquiry and the generated answer, a means for identifying emotions from the customer's voice data and text data using an emotion engine, and a means for optimizing the response content based on the results of the emotion analysis and presenting additional support options as necessary. This enables a fast and accurate response to an inquiry as well as an optimal response that takes into account the customer's emotions.

[0221] A "terminal" is an interface device through which a user inputs inquiry details and transmits them to a server.

[0222] A "server" is an information processing device that analyzes the content of the received inquiry and generates a response using a generative AI model.

[0223] An "inquiry" is information about a question or problem that a user sends via a terminal.

[0224] "Analysis" is the process of breaking down the content of a received inquiry into information and meaning and understanding it.

[0225] A "generative AI model" is an algorithm that uses natural language processing technology to generate appropriate answers to input prompts.

[0226] An "answer" is a response to a query generated by a generative AI model.

[0227] "Sentiment analysis" is the process of identifying and classifying the sentiments contained in inquiries and responses.

[0228] An "emotion engine" is an algorithm for identifying a user's emotions from voice data and text data.

[0229] "Response details" refers to the specific details of the response to the user's inquiry.

[0230] "Optimization" is the process of adjusting the response content based on the results of sentiment analysis to create the best possible version.

[0231] "Support Options" means additional methods or means of assistance provided to meet Customer needs.

[0232] The present invention is a system that utilizes generative AI models, sentiment analysis technology, and an emotion engine to respond quickly and accurately to customer inquiries. Specific embodiments of this system are described below.

[0233] First, the terminal provides an interface for receiving queries entered by the user. For example, if a user enters a query such as "My Internet connection is slow. What should I do?", the user's input is sent to the system through the terminal's input field. The terminal then sends the received query to the server using the HTTPS protocol.

[0234] The server analyzes the received inquiry. The analysis can use Google Cloud's natural language processing API. The analysis results include the subject of the inquiry and important keywords. For example, in the case of an inquiry such as "My internet connection is slow. What should I do?", keywords such as "internet connection" and "slow" are extracted.

[0235] The server then passes the analysis results to a generative AI model (e.g., OpenAI's GPT-4 (registered trademark)) to generate an appropriate answer. The generative AI model generates an appropriate answer based on the prompt text. This prompt text includes the user's inquiry and the analysis results. The generated answer is then returned to the server and sent to the device. For example, the answer generated might be, "If your Internet connection is slow, try restarting your router."

[0236] The generated answer is then returned to the device and displayed to the user, allowing the user to take action based on the information provided to resolve the problem.

[0237] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, the IBM Watson (registered trademark) sentiment analysis API may be used. For example, if a user sends a query containing negative sentiment such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server analyzes this sentence and assigns a "NEGATIVE" label.

[0238] The server also uses Microsoft® Azure® voice sentiment analysis API to enable the emotion engine to analyze the customer's voice and text data and identify emotions. In the case of voice data, the emotion engine detects negative emotions based on the analysis results and considers appropriate action. Similarly, it also identifies emotions in text data.

[0239] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, the server can offer additional support options. Specifically, the message displayed to the user is, "It seems you're having trouble. Would you like us to connect you to our support desk?"

[0240] This system's series of processes not only enables quick and accurate responses to customer inquiries, but also enables optimal responses that take into account the customer's emotions, improving customer satisfaction. This is also expected to reduce operational costs. The system aims to automate customer service while responding to customer needs in a detailed manner.

[0241] Example prompt sentence:

[0242] "User Asks: 'My Internet connection is slow. What should I do?'

[0243] Prompt for the generative AI model:

[0244] Please provide appropriate responses to the following inquiries, including specific steps to take, and, if the user has negative feelings, suggest additional support options that take those feelings into consideration.

[0245] Question: "My internet connection is slow. What should I do?"

[0246] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0247] Step 1:

[0248] The user inputs a query into the terminal interface, such as "My Internet connection is slow. What should I do?" The input is sent to the system through the terminal's input field.

[0249] Step 2:

[0250] The terminal sends the query content to the server. The terminal then securely sends the received query content to the server using the HTTPS protocol. The input is in text format, and the output is the data sent to the server.

[0251] Step 3:

[0252] The server analyzes the query content. The server uses Google Cloud's natural language processing API to analyze the received query content. The input is the received text data, and the output is the analysis results of important keywords and content. For example, from a query such as "My internet connection is slow," keywords such as "internet connection" and "slow" are extracted.

[0253] Step 4:

[0254] The server inputs the analysis results into the generative AI model. The server passes the analysis results to the generative AI model (e.g., generative AI model) as a prompt. The input is the analyzed query content and the prompt, and the output is the input prompt for the generative AI model.

[0255] Step 5:

[0256] The generative AI model generates an appropriate answer. The generative AI model generates an appropriate answer to a query based on the prompt text. The input is the prompt text, and the output is the generated answer. For example, the generated answer might be, "If your Internet connection is slow, try restarting your router."

[0257] Step 6:

[0258] The server receives the generated answer. The server receives the answer returned from the generative AI model and then prepares to send it to the device. The input is the answer from the generative AI model, and the output is the data to be sent to the device.

[0259] Step 7:

[0260] The server performs sentiment analysis. Using IBM Watson's sentiment analysis API, the server analyzes both the received inquiry and the generated answer. The input is the inquiry and the generated answer, and the output is the result of the sentiment analysis. For example, an inquiry such as "I can't connect to the Internet at all and I'm really having trouble. Please help me" is classified as "NEGATIVE."

[0261] Step 8:

[0262] The server uses an emotion engine to identify the customer's emotions. The server uses Microsoft Azure's voice emotion analysis API to analyze the customer's voice data and text data to identify the customer's emotions. The input is voice data or text data, and the output is the emotion identification result.

[0263] Step 9:

[0264] The server optimizes the response content. The server optimizes the response content based on the results of sentiment analysis. In particular, if negative sentiment is detected, additional support options are presented. The input is the result of sentiment analysis, and the output is the optimized response content. For example, a message such as "It seems you're having trouble. Would you like me to direct you to our support desk?" is added.

[0265] Step 10:

[0266] The terminal displays the optimal response content to the user. The terminal displays the optimized response content received from the server to the user. The input is the response data from the server, and the output is the display to the user. As a specific example, the user terminal might display the message, "If your Internet connection is slow, try restarting your router. It seems you're having trouble. Would you like me to direct you to support?"

[0267] (Application example 2)

[0268] 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."

[0269] In today's world, where fast and accurate responses to customer inquiries are required, improving customer satisfaction and reducing operational costs are key challenges, especially in the online shopping field. Conventional systems often lack the ability to adequately analyze emotions, resulting in inappropriate responses. The present invention aims to solve these challenges by utilizing generative AI models, emotion analysis technology, and smartphones.

[0270] 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.

[0271] In this invention, the server includes means for receiving inquiries from customers, means for passing the received inquiry content to a generative AI model to generate a response, means for providing the generated response to the customer, means for analyzing the sentiment of the customer's inquiry content, means for optimizing the response content based on the results of the sentiment analysis, and means for receiving inquiries using a smartphone and displaying the response, thereby enabling improved customer satisfaction and reduced operation costs.

[0272] The "means for receiving inquiries from customers" refers to the means for delivering the contents of inquiries entered by customers to the server via a terminal or system.

[0273] "A means of generating a response by passing the received inquiry content to a generative AI model" refers to a means of analyzing the content of a customer inquiry and using a generative AI model to create a response appropriate to that content.

[0274] "Means for providing the generated answer to the customer" refers to means for providing the answer generated by the generative AI model to the customer in an easy-to-understand manner.

[0275] The "means for analyzing the emotions of the customer's inquiry content" is a means for extracting and analyzing the customer's emotions from the inquiry content entered by the customer.

[0276] "Means for optimizing response content based on the results of sentiment analysis" refers to means for optimizing response content based on the results of sentiment analysis, enabling appropriate responses in a variety of situations.

[0277] The "means for receiving an inquiry using a smartphone and displaying a response" refers to a means for a customer to input an inquiry using a smartphone and display the result.

[0278] This invention is a system for responding quickly and accurately to customer inquiries on an online shopping site. The system includes a generative AI model, sentiment analysis technology, and a smartphone-based response display.

[0279] Hardware and software configuration used

[0280] Hardware:

[0281] Smartphone: A device for receiving customer inquiries and displaying responses.

[0282] Server: Analyzes the inquiry content, generates answers using a generation AI model, and performs sentiment analysis.

[0283] software:

[0284] openai library: An API for using generative AI models to generate appropriate answers to inquiries.

[0285] textblob library: Natural language processing technology for performing sentiment analysis of customer inquiries.

[0286] Sentiment analysis pipeline: Algorithms to dig deep into customer sentiment and help optimize responses.

[0287] Processing Description

[0288] 1. Receiving Inquiries

[0289] A user uses a chat interface on their smartphone to input an inquiry. For example, they might type, "I'm having trouble because my product hasn't arrived yet. When will it arrive?" This input is then sent to the server via the device.

[0290] 2. Analysis of inquiry content

[0291] The server passes the received query to the generative AI model, which generates an appropriate answer to the query.

[0292] 3. Sentiment analysis

[0293] At the same time, the server uses the textblob library to analyze the sentiment of the query, and based on the results, identifies whether the user's sentiment is positive or negative.

[0294] 4. Response optimization

[0295] Optimize the generated answer based on the results of sentiment analysis. For example, suggest additional support options if negative sentiment is detected. Add a message to the generated answer such as "I see you're having trouble. Would you like me to connect you to support?"

[0296] 5. Providing answers

[0297] The optimized response is then sent back to the smartphone and displayed to the user, for example, "We usually ship within 3-5 business days of your order. If you need further assistance, please contact customer service."

[0298] Examples and prompts

[0299] Examples:

[0300] If a user types an inquiry such as, "I'm having trouble receiving my item. When will it arrive?", the AI ​​will respond with, "Items are usually shipped within 3 to 5 business days of your order. If you need further support, please contact customer service." If sentiment analysis determines that the user is expressing negative emotions, it will additionally display the message, "It seems you're having trouble. Would you like me to direct you to support?"

[0301] Example prompt for a generative AI model:

[0302] "A customer asks: 'I'm having trouble receiving my item. When will it arrive?' Please provide a suitable response to this question."

[0303] The present invention can improve the efficiency and quality of customer support at online shopping sites, thereby improving customer satisfaction.

[0304] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0305] Step 1:

[0306] User enters a query

[0307] The user uses the chat interface on their smartphone to input an inquiry. This input is used as the initial data for the system. For example, the user might input, "I'm having trouble because my product hasn't arrived yet. When will it arrive?"

[0308] Step 2:

[0309] The device sends a query to the server

[0310] The device sends the query entered by the user to the server, which then sends the text data to the server and prepares it for analysis.

[0311] Step 3:

[0312] The server analyzes the query

[0313] The server passes the received query content to the generative AI model. Specifically, it uses the openai library to provide the query content to the AI ​​model as a prompt.

[0314] Step 4:

[0315] Generative AI models generate answers

[0316] The generative AI model generates an appropriate answer based on the prompt text, and this data is returned to the server as generated text. For example, the answer generated might be, "We usually ship within 3-5 business days of your order."

[0317] Step 5:

[0318] The server performs sentiment analysis

[0319] At the same time, the server analyzes the sentiment of the received query using the textblob library, extracting positive or negative sentiment from the input text and providing the result (sentiment score).

[0320] Step 6:

[0321] The server optimizes the response

[0322] Based on the results of the sentiment analysis, the server optimizes the response. For example, if a negative sentiment is detected, the generated response will include a message such as, "It sounds like you're having trouble. Would you like me to connect you to support?"

[0323] Step 7:

[0324] Optimized answers sent to your device

[0325] The optimized answer is sent from the server to the device, where it is ready to be displayed to the user as the final answer.

[0326] Step 8:

[0327] The device displays the answer to the user

[0328] The device will display the optimized answer received from the server to the user, who can then refer to it to resolve the issue. For example, the answer might read, "We usually ship within 3-5 business days of your order. If you need further assistance, please contact customer service."

[0329] 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.

[0330] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (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.

[0331] 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.

[0332] [Second embodiment]

[0333] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0334] 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.

[0335] 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).

[0336] 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.

[0337] 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.

[0338] 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).

[0339] 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. 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.

[0340] 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.

[0341] 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.

[0342] 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.

[0343] 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.

[0344] 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."

[0345] The present invention is a system for responding to customer inquiries quickly and accurately, and in particular utilizes generative AI models and sentiment analysis techniques. Specific embodiments of this system are described below.

[0346] First, the terminal provides an interface for receiving queries entered by the user. For example, suppose a user enters a query such as "My Internet connection is slow. What should I do?" This query is sent to the system through the input field of the terminal.

[0347] The server then processes the received query. The server analyzes the query and passes it to a generative AI model (e.g., a model using natural language processing technology). This generative AI model generates an appropriate answer to the query. For example, it might say, "If your Internet connection is slow, try restarting your router."

[0348] The generated answer is returned to the terminal and displayed to the user, who can then take action based on the answer.

[0349] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it uses a sentiment analysis pipeline (e.g., an algorithm that identifies the sentiment in a sentence). If the user's query contains negative sentiment, such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server assigns a "NEGATIVE" label to the query.

[0350] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, it can present additional support options. The message displayed to the user might say, "It seems you're having trouble. Would you like us to connect you to our support team?"

[0351] As can be seen from this example, this system not only responds to customer inquiries quickly and accurately, but also aims to improve customer satisfaction by providing optimal responses through sentiment analysis. This system also reduces operational costs.

[0352] The processing flow will be explained below.

[0353] Step 1:

[0354] The user inputs a query. For example, the user might input "My Internet connection is slow. What should I do?" The user's input is entered into an input field on the device.

[0355] Step 2:

[0356] The terminal receives the query entered by the user, which is then temporarily stored and prepared for transmission to the server.

[0357] Step 3:

[0358] The terminal sends the received query content to the server. Specifically, the terminal creates an HTTP POST request, includes the query content in JSON format in the request body, and sends it to the specified endpoint of the server.

[0359] Step 4:

[0360] The server analyzes the received query, extracts the query content from the received request, and passes it to a generative AI model to generate an appropriate answer.

[0361] Step 5:

[0362] The server generates an answer using a generative AI model. The generative AI model generates the best answer based on the query and returns that answer to the server. For example, if the query is "My Internet connection is slow. What should I do?", the generated answer might be "If your Internet connection is slow, try restarting your router."

[0363] Step 6:

[0364] The server returns the generated answer to the terminal. The server creates an HTTP response that includes the generated answer and sends the response to the terminal.

[0365] Step 7:

[0366] The device displays the answer received from the server to the user. The device analyzes the response from the server and displays the generated answer on the screen. For example, the answer "If your Internet connection is slow, try restarting your router" may be presented to the user.

[0367] Step 8:

[0368] The server analyzes the sentiment of the inquiry. Using a sentiment analysis pipeline, the server analyzes the received inquiry and identifies the sentiment. For example, if the inquiry is "I can't connect to the Internet at all and I'm really in trouble. Please help me," the result of sentiment analysis will be assigned a "NEGATIVE" label.

[0369] Step 9:

[0370] The server optimizes the response based on the results of sentiment analysis. If the result of sentiment analysis is negative, it generates a message suggesting additional support. For example, it generates a message saying, "It seems you're having trouble. Would you like me to connect you to our support desk?"

[0371] Step 10:

[0372] The device displays the emotion analysis results and additional support suggestions sent from the server to the user. The user can see a message on the screen saying, "It seems you're having trouble. Would you like us to direct you to our support desk?"

[0373] Step 11:

[0374] The user can select additional support options or follow the generated answer to resolve the issue, thus providing the user with a timely and appropriate response.

[0375] Example 1

[0376] 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."

[0377] Conventional customer support systems often lacked timely and accurate responses to customer inquiries, leading to poor customer satisfaction. Furthermore, they lacked the ability to consider customer emotions, making it difficult to provide appropriate support, especially to customers with negative feelings. Additionally, these systems also had the problem of high operating costs.

[0378] 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.

[0379] In this invention, the server includes a means for passing the received inquiry content to a generative AI model to generate an answer, a means for providing the generated answer to the customer via a terminal, a means for analyzing the sentiment of the customer's inquiry content, and a means for optimizing the response content based on the results of the sentiment analysis. This enables quick and accurate responses to customer inquiries, and by realizing optimal responses through sentiment analysis, it is possible to improve customer satisfaction. It also realizes efficient operations and enables cost reduction.

[0380] A "terminal" is an input device that allows a customer to input an inquiry and transmit it to the system.

[0381] The "server" is a central processing unit that analyzes the content of the received inquiry, generates a response using a generative AI model, and then performs sentiment analysis to optimize the response content.

[0382] An "inquiry" is a question or inquiry that a customer inputs into the system via a terminal.

[0383] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to generate appropriate answers to customer inquiries.

[0384] An "answer" is a response generated by a generative AI model and provided to a customer.

[0385] "Sentiment analysis" is a technology that analyzes the content of customer inquiries and identifies their emotional state.

[0386] A "sentiment analysis pipeline" is a series of processing steps for performing sentiment analysis, including algorithms for extracting sentiment from text data.

[0387] "Response optimization" is a technique that provides optimal answers and additional support options to customers based on the results of sentiment analysis.

[0388] The present invention is a system for responding to customer inquiries quickly and accurately, and in particular utilizes generative AI models and sentiment analysis techniques. Specific embodiments of this system are described below.

[0389] First, the terminal provides an interface for receiving queries entered by the user. For example, suppose a user enters a query such as "My Internet connection is slow. What should I do?" This query is sent to the system through the input field of the terminal.

[0390] The server then processes the received query. The server analyzes the query and passes it to a generative AI model (e.g., a model using natural language processing technology). This generative AI model generates an appropriate answer to the query. For example, it might say, "If your Internet connection is slow, try restarting your router."

[0391] The generated answer is returned to the terminal and displayed to the user, who can then take action based on the answer.

[0392] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it uses a sentiment analysis pipeline (e.g., an algorithm that identifies the sentiment in a sentence). If the user's query contains negative sentiment, such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server assigns a "NEGATIVE" label to the query.

[0393] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, additional support options can be presented. The message displayed to the user is, "It seems you're having trouble. Would you like us to connect you to our support team?"

[0394] As can be seen from this example, this system not only responds to customer inquiries quickly and accurately, but also aims to improve customer satisfaction by providing optimal responses through sentiment analysis. This system also reduces operational costs.

[0395] Below are some examples of prompt sentences.

[0396] Example prompt: "My internet connection is slow. What should I do?"

[0397] The system aims to achieve a high level of customer satisfaction by quickly and accurately processing customer inquiries and optimizing responses through sentiment analysis, so it is important to effectively utilize generative AI models and sentiment analysis pipelines.

[0398] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0399] Step 1:

[0400] A user inputs a query through the terminal, for example, "My Internet connection is slow. What should I do?" The terminal receives this query and sends it to the server.

[0401] Input: User's inquiry

[0402] Output: Query sent to server

[0403] What happens: The user uses the device interface to enter a question into the text box and clicks "Submit."

[0404] Step 2:

[0405] The server receives the query, normalizes and parses it, which includes removing unnecessary whitespace and special characters from the text, and then passes the query to a generative AI model.

[0406] Input: Query sent from the terminal

[0407] Output: Normalized and parsed text

[0408] Specific operation: After receiving the query, the server cleans and preprocesses the text, preparing it for input into the generative AI model.

[0409] Step 3:

[0410] The server uses a generative AI model to generate an appropriate answer to the query. The query is passed to the generative AI model (for example, a model using natural language processing technology) to generate an answer. For example, the server may receive a response such as, "If your Internet connection is slow, try restarting your router."

[0411] Input: Normalized text query

[0412] Output: The answer generated by the generative AI model

[0413] How it works: The server inputs a query into a natural language processing model and retrieves the best answer provided by the model.

[0414] Step 4:

[0415] The server sends the generated answer to the terminal, which displays the answer to the user.

[0416] Input: The answer generated by the generative AI model

[0417] Output: The answer that is displayed to the user

[0418] Specific operation: The device displays the response received from the server on the screen for the user to see. For example, it may display "If your Internet connection is slow, try restarting your router."

[0419] Step 5:

[0420] The server passes the query content to a sentiment analysis pipeline for sentiment analysis, which detects negative sentiment such as "troubled."

[0421] Input: Normalized text query

[0422] Output: Sentiment analysis result (e.g., NEGATIVE)

[0423] What it does: The server uses a sentiment analysis algorithm to identify positive, negative, neutral, or other sentiment from the input text.

[0424] Step 6:

[0425] The server optimizes the response based on the results of sentiment analysis. If negative sentiment is detected, it generates additional support options, such as "It seems like you're having trouble. Would you like me to connect you to our support team?"

[0426] Input: Sentiment analysis results

[0427] Output: Optimized response (e.g., "Sounds like you're having trouble. Would you like me to connect you to support?")

[0428] What happens: The server runs logic to generate an appropriate follow-up message for the user based on the results of the sentiment analysis.

[0429] Step 7:

[0430] The server transmits the optimized response content to the terminal, which then displays the content to the user.

[0431] Input: Optimized response

[0432] Output: The optimized greeting that is displayed to the user.

[0433] Specific operation: The device displays the optimized message received from the server on the screen and guides the user, for example, "It seems you are having trouble. Would you like us to direct you to our support center?"

[0434] Through these steps, the system can respond to customer inquiries quickly and accurately, and even provide optimal responses through sentiment analysis.

[0435] (Application example 1)

[0436] 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."

[0437] Modern online shopping sites are required to respond quickly and accurately to inquiries from a large number of customers. However, conventional systems make it difficult to respond while taking into account customer emotions, making it difficult to improve customer satisfaction and streamline customer support. Furthermore, there is a lack of support tools that enable customer support staff to respond optimally in real time, making it difficult to maintain high response quality.

[0438] 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.

[0439] In this invention, the server includes means for receiving inquiries from customers, means for passing the received inquiry content to a generative AI model to generate a response, means for providing the generated response to the customer, means for analyzing the sentiment of the customer's inquiry content, means for optimizing the response content based on the result of the sentiment analysis, and means for displaying the inquiry content, the response of the generative AI model, and the result of the sentiment analysis in real time using smart glasses. This enables quick and accurate responses that take customer sentiment into consideration, thereby improving the efficiency of customer support and customer satisfaction.

[0440] "Customer" refers to a consumer or entity that uses a product or service.

[0441] "Inquiry" refers to the act of a customer communicating a question, complaint, or request regarding a product or service to customer support.

[0442] "Means of receiving" refers to the interface or mechanism for receiving inquiries from customers.

[0443] "Generative AI model" refers to an artificial intelligence model that uses natural language processing technology to generate appropriate answers to input inquiries.

[0444] "Answer" refers to the answer or advice that a generative AI model generates based on a query.

[0445] "Means for analyzing emotions" refers to technology that analyzes emotions from customer inquiries and classifies them into emotional labels such as positive or negative.

[0446] "Means for optimizing response content" refers to technology that determines and provides the optimal response method for customers based on the results of sentiment analysis.

[0447] "Smart glasses" are a wearable device in the form of glasses equipped with a display device, allowing customer support representatives to view information in real time.

[0448] "Real-time" refers to information processing occurring simultaneously with real time, with almost no delay.

[0449] The present invention is a system for responding to customer inquiries quickly and accurately, and in particular utilizes generative AI models and sentiment analysis techniques. A specific embodiment of this system is described below.

[0450] First, the server and terminal are the main hardware components that realize this system. The terminal is implemented as smart glasses and is worn by a customer support representative. The terminal provides an interface for receiving customer inquiries using input or voice recognition technology. For example, a user can make an inquiry such as, "My product delivery is delayed. What should I do?" This inquiry is sent to the server through the input field on the terminal.

[0451] The server then processes the received inquiry. The server analyzes the inquiry and passes it to a generative AI model (for example, a model using natural language processing technology). This generative AI model generates an appropriate answer to the inquiry. For example, it might respond, "Please wait a moment while we check the delivery status." The generated answer is then sent back to the device and displayed on the customer support representative's smart glasses. The representative can then explain the situation to the customer while looking at the answer.

[0452] The server then analyzes the sentiment of the inquiry. Sentiment analysis uses a sentiment analysis pipeline (e.g., an algorithm that identifies emotions in sentences). If the user's inquiry contains a negative sentiment, such as "I'm very frustrated because my item hasn't arrived yet," the sentiment analysis will assign a "NEGATIVE" label. Based on the results of this sentiment analysis, the server can optimize the response. For example, if a negative sentiment is detected, additional support options can be presented. A message such as "It seems you're having trouble. Would you like me to direct you to our support desk for further assistance?" will be displayed on the customer support representative's smart glasses.

[0453] Hardware and software used

[0454] Hardware: Smart glasses (built-in microphone, display)

[0455] software:

[0456] Speech Recognition Library:SpeechRecognition Library

[0457] Generative AI models: Natural language processing models (e.g., OpenAI's GPT)

[0458] Sentiment Analysis Pipeline: Transformer Library and Japanese Model

[0459] For example, the following prompt sentences may be used when a customer asks, "My product delivery is delayed. What should I do?"

[0460] Prompt Sentence Examples

[0461] Customer enquiry: "My item delivery is delayed. What should I do?"

[0462] Please provide an appropriate response.

[0463] With the above configuration, the present invention enables quick and accurate responses that take into consideration the customer's feelings, thereby realizing more efficient customer support and improved customer satisfaction.

[0464] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0465] Step 1:

[0466] The user makes a query.

[0467] For example, a user may make a voice inquiry through the built-in microphone of the smart glasses, saying, "My product delivery is delayed. What should I do?" The voice data is input into the terminal.

[0468] Step 2:

[0469] The device converts the voice data into text.

[0470] The device's speech recognition library (for example, the SpeechRecognition library) is used to convert the voice data into text data. The converted text becomes the string "My product delivery is delayed. What should I do?"

[0471] Step 3:

[0472] The text data is sent to the server.

[0473] The terminal transmits the converted character string data to the server, which receives the text data.

[0474] Step 4:

[0475] The server generates an answer using a generative AI model.

[0476] The server passes the received text data to a generative AI model (for example, a model using natural language processing) to generate an appropriate answer. Using the prompt sentence, the answer to "My product delivery is delayed. What should I do?" is output as "We will check the delivery status, so please wait a moment."

[0477] Step 5:

[0478] The generated answer is sent to the terminal.

[0479] The server sends the generated response "We will check the delivery status, so please wait a moment" to the terminal.

[0480] Step 6:

[0481] The device will display the answer.

[0482] The terminal displays the received answer on the display of the smart glasses, where a customer support representative can view the answer and communicate it to the user verbally or via text.

[0483] Step 7:

[0484] The server performs sentiment analysis.

[0485] The server passes the received text data to a sentiment analysis pipeline (e.g., the Transformer library and a Japanese model) to analyze the sentiment. If the input is "My product delivery is delayed. What should I do?", a negative sentiment is detected. The analysis results in a label of "NEGATIVE."

[0486] Step 8:

[0487] Optimize responses based on the results of sentiment analysis.

[0488] The server generates additional response options based on the results of sentiment analysis, such as a suggestion like, "It sounds like you're having trouble. Would you like me to direct you to our support team if you need further assistance?"

[0489] Step 9:

[0490] Suggestions for what to do are displayed on the device.

[0491] The server sends the generated suggestions to the device, which displays them on the smart glasses display. The customer support representative then uses the suggestions to provide the optimal response.

[0492] 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.

[0493] The present invention provides a system for quickly and accurately responding to customer inquiries, and in particular utilizes generative AI models, sentiment analysis techniques, and an emotion engine. Specific embodiments of this system are described below.

[0494] First, the terminal provides an interface to receive queries entered by the user. For example, a user may enter a query such as "My Internet connection is slow. What should I do?" The user's input is sent to the system through the input field on the terminal.

[0495] The server then processes the received query. The server analyzes the query and passes it to a generative AI model (for example, a model using natural language processing technology) to generate an appropriate answer. The generated answer is then returned to the server and sent to the device. For example, the answer generated might be, "If your Internet connection is slow, try restarting your router."

[0496] The generated answer is returned to the terminal and displayed to the user, who can then take action based on the answer.

[0497] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it uses a sentiment analysis pipeline (e.g., an algorithm that identifies the sentiment in a sentence). If the user's query contains negative sentiment, such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server assigns a "NEGATIVE" label to the query.

[0498] The server also uses an emotion engine to recognize the user's emotions. The emotion engine analyzes voice data and text data to identify the user's emotions. For example, when a user makes a voice inquiry, the emotion engine analyzes the voice data to determine whether it contains negative emotions. Similarly, the emotion engine also analyzes text data to identify emotions.

[0499] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, it can present additional support options. The message displayed to the user might say, "It seems you're having trouble. Would you like us to connect you to our support team?"

[0500] As can be seen from this example, this system not only responds to customer inquiries quickly and accurately, but also aims to improve customer satisfaction by providing optimal responses through emotion analysis using an emotion engine. This system also reduces operational costs.

[0501] The processing flow will be explained below.

[0502] Step 1:

[0503] The user inputs the inquiry, for example, "My Internet connection is slow. What should I do?", and clicks the send button.

[0504] Step 2:

[0505] The terminal receives the inquiry entered by the user, temporarily stores the inquiry internally, and sends it to the server when it is ready.

[0506] Step 3:

[0507] The device sends the received query content to the server. Specifically, it creates an HTTP POST request, includes the query content in JSON format, and sends it to the server's API endpoint.

[0508] Step 4:

[0509] The server analyzes the received query, extracts the query content from the HTTP request, and passes it to a generative AI model to generate an appropriate answer.

[0510] Step 5:

[0511] The server generates an answer using a generative AI model, which generates an answer based on the query, such as "If your internet connection is slow, try restarting your router," and sends the answer back to the server.

[0512] Step 6:

[0513] The server returns the generated answer to the terminal, creates an HTTP response including the generated answer, and sends it to the terminal.

[0514] Step 7:

[0515] The device displays the answer received from the server to the user. It analyzes the response from the server and displays the generated answer on the screen. For example, it may say, "If your Internet connection is slow, try restarting your router."

[0516] Step 8:

[0517] The server analyzes the sentiment of the inquiry. Using a sentiment analysis pipeline, it analyzes the inquiry and identifies the sentiment. For example, if the inquiry says, "I can't connect to the Internet at all and I'm really in trouble. Please help me," the sentiment analysis will assign a "NEGATIVE" label.

[0518] Step 9:

[0519] The server uses an emotion engine to recognize the user's emotions. The emotion engine analyzes voice and text data to identify the user's emotions. For example, when a user makes a voice inquiry, the server analyzes the voice data to determine whether the user is expressing negative emotions.

[0520] Step 10:

[0521] The server optimizes the response based on the results of the sentiment analysis. If negative sentiment is detected, it will offer additional support options, such as generating an additional message like, "It seems like you're having trouble. Would you like us to connect you to our support team?"

[0522] Step 11:

[0523] The device displays the emotion analysis results and additional support suggestions sent from the server to the user. The message "It seems you're having trouble. Would you like us to direct you to our support desk?" is displayed on the screen.

[0524] Step 12:

[0525] The user can select additional support options or follow the initial response to resolve the issue, thus providing the user with a proper and prompt response.

[0526] Example 2

[0527] 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."

[0528] While it is necessary to respond to customer inquiries quickly and accurately, it is also important to appropriately understand and respond to customer emotions. However, conventional systems have had the problem of difficulty not only responding to the content of the inquiry but also identifying the customer's emotions and providing an optimal response. The present invention aims to provide a system that responds to customer inquiries quickly and accurately, performs emotion analysis, and realizes an optimal response based on the customer's emotions.

[0529] 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.

[0530] In this invention, the server includes a means for a terminal to receive an inquiry from a customer, a means for analyzing the received inquiry, a means for passing the analysis result to a generative AI model to generate an answer, a means for returning the generated answer to the terminal and providing it to the customer, a means for analyzing the sentiment of the customer's inquiry and the generated answer, a means for identifying emotions from the customer's voice data and text data using an emotion engine, and a means for optimizing the response content based on the results of the emotion analysis and presenting additional support options as necessary. This enables a fast and accurate response to an inquiry as well as an optimal response that takes into account the customer's emotions.

[0531] A "terminal" is an interface device through which a user inputs inquiry details and transmits them to a server.

[0532] A "server" is an information processing device that analyzes the content of the received inquiry and generates a response using a generative AI model.

[0533] An "inquiry" is information about a question or problem that a user sends via a terminal.

[0534] "Analysis" is the process of breaking down the content of a received inquiry into information and meaning and understanding it.

[0535] A "generative AI model" is an algorithm that uses natural language processing technology to generate appropriate answers to input prompts.

[0536] An "answer" is a response to a query generated by a generative AI model.

[0537] "Sentiment analysis" is the process of identifying and classifying the sentiments contained in inquiries and responses.

[0538] An "emotion engine" is an algorithm for identifying a user's emotions from voice data and text data.

[0539] "Response details" refers to the specific details of the response to the user's inquiry.

[0540] "Optimization" is the process of adjusting the response content based on the results of sentiment analysis to create the best possible version.

[0541] "Support Options" means additional methods or means of assistance provided to meet Customer needs.

[0542] The present invention is a system that utilizes generative AI models, sentiment analysis technology, and an emotion engine to respond quickly and accurately to customer inquiries. Specific embodiments of this system are described below.

[0543] First, the terminal provides an interface for receiving queries entered by the user. For example, if a user enters a query such as "My Internet connection is slow. What should I do?", the user's input is sent to the system through the terminal's input field. The terminal then sends the received query to the server using the HTTPS protocol.

[0544] The server analyzes the received inquiry. This analysis can be performed using Google Cloud's natural language processing API. The analysis results include the subject of the inquiry and important keywords. For example, in the case of an inquiry such as "My internet connection is slow. What should I do?", keywords such as "internet connection" and "slow" are extracted.

[0545] The server then passes the analysis results to a generative AI model (e.g., OpenAI's GPT-4) to generate an appropriate answer. The generative AI model generates an appropriate answer based on the prompt text. This prompt text includes the user's query and the analysis results. The generated answer is then returned to the server and sent to the device. For example, the answer generated might be, "If your Internet connection is slow, try restarting your router."

[0546] The generated answer is then returned to the device and displayed to the user, allowing the user to take action based on the information provided to resolve the problem.

[0547] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it may use IBM Watson's sentiment analysis API. For example, if a user sends a query containing negative sentiment such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server analyzes this sentence and assigns a "NEGATIVE" label.

[0548] The server also uses Microsoft Azure's voice sentiment analysis API to enable the emotion engine to analyze customer voice and text data and identify emotions. For voice data, the emotion engine uses the analysis results to detect negative emotions and consider appropriate responses. Similarly, it also identifies emotions for text data.

[0549] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, the server can offer additional support options. Specifically, the message displayed to the user is, "It seems you're having trouble. Would you like us to connect you to our support desk?"

[0550] This system's series of processes not only enables quick and accurate responses to customer inquiries, but also enables optimal responses that take into account the customer's emotions, improving customer satisfaction. This is also expected to reduce operational costs. The system aims to automate customer service while responding to customer needs in a detailed manner.

[0551] Example prompt sentence:

[0552] "User Asks: 'My Internet connection is slow. What should I do?'

[0553] Prompt for the generative AI model:

[0554] Please provide appropriate responses to the following inquiries, including specific steps to take, and, if the user has negative feelings, suggest additional support options that take those feelings into consideration.

[0555] Question: "My internet connection is slow. What should I do?"

[0556] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0557] Step 1:

[0558] The user inputs a query into the terminal interface, such as "My Internet connection is slow. What should I do?" The input is sent to the system through the terminal's input field.

[0559] Step 2:

[0560] The terminal sends the query content to the server. The terminal then securely sends the received query content to the server using the HTTPS protocol. The input is in text format, and the output is the data sent to the server.

[0561] Step 3:

[0562] The server analyzes the query content. The server uses Google Cloud's natural language processing API to analyze the received query content. The input is the received text data, and the output is the analysis results of important keywords and content. For example, from a query such as "My internet connection is slow," keywords such as "internet connection" and "slow" are extracted.

[0563] Step 4:

[0564] The server inputs the analysis results into the generative AI model. The server passes the analysis results to the generative AI model (e.g., generative AI model) as a prompt. The input is the analyzed query content and the prompt, and the output is the input prompt for the generative AI model.

[0565] Step 5:

[0566] The generative AI model generates an appropriate answer. The generative AI model generates an appropriate answer to a query based on the prompt text. The input is the prompt text, and the output is the generated answer. For example, the generated answer might be, "If your Internet connection is slow, try restarting your router."

[0567] Step 6:

[0568] The server receives the generated answer. The server receives the answer returned from the generative AI model and then prepares to send it to the device. The input is the answer from the generative AI model, and the output is the data to be sent to the device.

[0569] Step 7:

[0570] The server performs sentiment analysis. Using IBM Watson's sentiment analysis API, the server analyzes both the received inquiry and the generated answer. The input is the inquiry and the generated answer, and the output is the result of the sentiment analysis. For example, an inquiry such as "I can't connect to the Internet at all and I'm really having trouble. Please help me" is classified as "NEGATIVE."

[0571] Step 8:

[0572] The server uses an emotion engine to identify the customer's emotions. The server uses Microsoft Azure's voice emotion analysis API to analyze the customer's voice data and text data to identify the customer's emotions. The input is voice data or text data, and the output is the emotion identification result.

[0573] Step 9:

[0574] The server optimizes the response content. The server optimizes the response content based on the results of sentiment analysis. In particular, if negative sentiment is detected, additional support options are presented. The input is the result of sentiment analysis, and the output is the optimized response content. For example, a message such as "It seems you're having trouble. Would you like me to direct you to our support desk?" is added.

[0575] Step 10:

[0576] The terminal displays the optimal response content to the user. The terminal displays the optimized response content received from the server to the user. The input is the response data from the server, and the output is the display to the user. As a specific example, the user terminal might display the message, "If your Internet connection is slow, try restarting your router. It seems you're having trouble. Would you like me to direct you to support?"

[0577] (Application example 2)

[0578] 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."

[0579] In today's world, where fast and accurate responses to customer inquiries are required, improving customer satisfaction and reducing operational costs are key challenges, especially in the online shopping field. Conventional systems often lack the ability to adequately analyze emotions, resulting in inappropriate responses. The present invention aims to solve these challenges by utilizing generative AI models, emotion analysis technology, and smartphones.

[0580] 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.

[0581] In this invention, the server includes means for receiving inquiries from customers, means for passing the received inquiry content to a generative AI model to generate a response, means for providing the generated response to the customer, means for analyzing the sentiment of the customer's inquiry content, means for optimizing the response content based on the results of the sentiment analysis, and means for receiving inquiries using a smartphone and displaying the response, thereby enabling improved customer satisfaction and reduced operation costs.

[0582] The "means for receiving inquiries from customers" refers to the means for delivering the contents of inquiries entered by customers to the server via a terminal or system.

[0583] "A means of generating a response by passing the received inquiry content to a generative AI model" refers to a means of analyzing the content of a customer inquiry and using a generative AI model to create a response appropriate to that content.

[0584] "Means for providing the generated answer to the customer" refers to means for providing the answer generated by the generative AI model to the customer in an easy-to-understand manner.

[0585] The "means for analyzing the emotions of the customer's inquiry content" is a means for extracting and analyzing the customer's emotions from the inquiry content entered by the customer.

[0586] "Means for optimizing response content based on the results of sentiment analysis" refers to means for optimizing response content based on the results of sentiment analysis, enabling appropriate responses in a variety of situations.

[0587] The "means for receiving an inquiry using a smartphone and displaying a response" refers to a means for a customer to input an inquiry using a smartphone and display the result.

[0588] This invention is a system for responding quickly and accurately to customer inquiries on an online shopping site. The system includes a generative AI model, sentiment analysis technology, and a smartphone-based response display.

[0589] Hardware and software configuration used

[0590] Hardware:

[0591] Smartphone: A device for receiving customer inquiries and displaying responses.

[0592] Server: Analyzes the inquiry content, generates answers using a generation AI model, and performs sentiment analysis.

[0593] software:

[0594] openai library: An API for using generative AI models to generate appropriate answers to inquiries.

[0595] textblob library: Natural language processing technology for performing sentiment analysis of customer inquiries.

[0596] Sentiment analysis pipeline: Algorithms to dig deep into customer sentiment and help optimize responses.

[0597] Processing Description

[0598] 1. Receiving Inquiries

[0599] A user uses a chat interface on their smartphone to input an inquiry. For example, they might type, "I'm having trouble because my product hasn't arrived yet. When will it arrive?" This input is then sent to the server via the device.

[0600] 2. Analysis of inquiry content

[0601] The server passes the received query to the generative AI model, which generates an appropriate answer to the query.

[0602] 3. Sentiment analysis

[0603] At the same time, the server uses the textblob library to analyze the sentiment of the query, and based on the results, identifies whether the user's sentiment is positive or negative.

[0604] 4. Response optimization

[0605] Optimize the generated answer based on the results of sentiment analysis. For example, suggest additional support options if negative sentiment is detected. Add a message to the generated answer such as "I see you're having trouble. Would you like me to connect you to support?"

[0606] 5. Providing answers

[0607] The optimized response is then sent back to the smartphone and displayed to the user, for example, "We usually ship within 3-5 business days of your order. If you need further assistance, please contact customer service."

[0608] Examples and prompts

[0609] Examples:

[0610] If a user types an inquiry such as, "I'm having trouble receiving my item. When will it arrive?", the AI ​​will respond with, "Items are usually shipped within 3 to 5 business days of your order. If you need further support, please contact customer service." If sentiment analysis determines that the user is expressing negative emotions, it will additionally display the message, "It seems you're having trouble. Would you like me to direct you to support?"

[0611] Example prompt for a generative AI model:

[0612] "A customer asks: 'I'm having trouble receiving my item. When will it arrive?' Please provide a suitable response to this question."

[0613] The present invention can improve the efficiency and quality of customer support at online shopping sites, thereby improving customer satisfaction.

[0614] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0615] Step 1:

[0616] User enters a query

[0617] The user uses the chat interface on their smartphone to input an inquiry. This input is used as the initial data for the system. For example, the user might input, "I'm having trouble because my product hasn't arrived yet. When will it arrive?"

[0618] Step 2:

[0619] The device sends a query to the server

[0620] The device sends the query entered by the user to the server, which then sends the text data to the server and prepares it for analysis.

[0621] Step 3:

[0622] The server analyzes the query

[0623] The server passes the received query content to the generative AI model. Specifically, it uses the openai library to provide the query content to the AI ​​model as a prompt.

[0624] Step 4:

[0625] Generative AI models generate answers

[0626] The generative AI model generates an appropriate answer based on the prompt text, and this data is returned to the server as generated text. For example, the answer generated might be, "We usually ship within 3-5 business days of your order."

[0627] Step 5:

[0628] The server performs sentiment analysis

[0629] At the same time, the server analyzes the sentiment of the received query using the textblob library, extracting positive or negative sentiment from the input text and providing the result (sentiment score).

[0630] Step 6:

[0631] The server optimizes the response

[0632] Based on the results of the sentiment analysis, the server optimizes the response. For example, if a negative sentiment is detected, the generated response will include a message such as, "It sounds like you're having trouble. Would you like me to connect you to support?"

[0633] Step 7:

[0634] Optimized answers sent to your device

[0635] The optimized answer is sent from the server to the device, where it is ready to be displayed to the user as the final answer.

[0636] Step 8:

[0637] The device displays the answer to the user

[0638] The device will display the optimized answer received from the server to the user, who can then refer to it to resolve the issue. For example, the answer might read, "We usually ship within 3-5 business days of your order. If you need further assistance, please contact customer service."

[0639] 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.

[0640] 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.

[0641] 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.

[0642] [Third embodiment]

[0643] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0644] 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.

[0645] 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).

[0646] 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.

[0647] 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.

[0648] 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).

[0649] 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. 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.

[0650] 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.

[0651] 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.

[0652] 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.

[0653] 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.

[0654] 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."

[0655] The present invention is a system for responding to customer inquiries quickly and accurately, and in particular utilizes generative AI models and sentiment analysis techniques. Specific embodiments of this system are described below.

[0656] First, the terminal provides an interface for receiving queries entered by the user. For example, suppose a user enters a query such as "My Internet connection is slow. What should I do?" This query is sent to the system through the input field of the terminal.

[0657] The server then processes the received query. The server analyzes the query and passes it to a generative AI model (e.g., a model using natural language processing technology). This generative AI model generates an appropriate answer to the query. For example, it might say, "If your Internet connection is slow, try restarting your router."

[0658] The generated answer is returned to the terminal and displayed to the user, who can then take action based on the answer.

[0659] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it uses a sentiment analysis pipeline (e.g., an algorithm that identifies the sentiment in a sentence). If the user's query contains negative sentiment, such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server assigns a "NEGATIVE" label to the query.

[0660] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, it can present additional support options. The message displayed to the user might say, "It seems you're having trouble. Would you like us to connect you to our support team?"

[0661] As can be seen from this example, this system not only responds to customer inquiries quickly and accurately, but also aims to improve customer satisfaction by providing optimal responses through sentiment analysis. This system also reduces operational costs.

[0662] The processing flow will be explained below.

[0663] Step 1:

[0664] The user inputs a query. For example, the user might input "My Internet connection is slow. What should I do?" The user's input is entered into an input field on the device.

[0665] Step 2:

[0666] The terminal receives the query entered by the user, which is then temporarily stored and prepared for transmission to the server.

[0667] Step 3:

[0668] The terminal sends the received query content to the server. Specifically, the terminal creates an HTTP POST request, includes the query content in JSON format in the request body, and sends it to the specified endpoint of the server.

[0669] Step 4:

[0670] The server analyzes the received query, extracts the query content from the received request, and passes it to a generative AI model to generate an appropriate answer.

[0671] Step 5:

[0672] The server generates an answer using a generative AI model. The generative AI model generates the best answer based on the query and returns that answer to the server. For example, if the query is "My Internet connection is slow. What should I do?", the generated answer might be "If your Internet connection is slow, try restarting your router."

[0673] Step 6:

[0674] The server returns the generated answer to the terminal. The server creates an HTTP response that includes the generated answer and sends the response to the terminal.

[0675] Step 7:

[0676] The device displays the answer received from the server to the user. The device analyzes the response from the server and displays the generated answer on the screen. For example, the answer "If your Internet connection is slow, try restarting your router" may be presented to the user.

[0677] Step 8:

[0678] The server analyzes the sentiment of the inquiry. Using a sentiment analysis pipeline, the server analyzes the received inquiry and identifies the sentiment. For example, if the inquiry is "I can't connect to the Internet at all and I'm really in trouble. Please help me," the result of sentiment analysis will be assigned a "NEGATIVE" label.

[0679] Step 9:

[0680] The server optimizes the response based on the results of sentiment analysis. If the result of sentiment analysis is negative, it generates a message suggesting additional support. For example, it generates a message saying, "It seems you're having trouble. Would you like me to connect you to our support desk?"

[0681] Step 10:

[0682] The device displays the emotion analysis results and additional support suggestions sent from the server to the user. The user can see a message on the screen saying, "It seems you're having trouble. Would you like us to direct you to our support desk?"

[0683] Step 11:

[0684] The user can select additional support options or follow the generated answer to resolve the issue, thus providing the user with a timely and appropriate response.

[0685] Example 1

[0686] 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."

[0687] Conventional customer support systems often lacked timely and accurate responses to customer inquiries, leading to poor customer satisfaction. Furthermore, they lacked the ability to consider customer emotions, making it difficult to provide appropriate support, especially to customers with negative feelings. Additionally, these systems also had the problem of high operating costs.

[0688] 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.

[0689] In this invention, the server includes a means for passing the received inquiry content to a generative AI model to generate an answer, a means for providing the generated answer to the customer via a terminal, a means for analyzing the sentiment of the customer's inquiry content, and a means for optimizing the response content based on the results of the sentiment analysis. This enables quick and accurate responses to customer inquiries, and by realizing optimal responses through sentiment analysis, it is possible to improve customer satisfaction. It also realizes efficient operations and enables cost reduction.

[0690] A "terminal" is an input device that allows a customer to input an inquiry and transmit it to the system.

[0691] The "server" is a central processing unit that analyzes the content of the received inquiry, generates a response using a generative AI model, and then performs sentiment analysis to optimize the response content.

[0692] An "inquiry" is a question or inquiry that a customer inputs into the system via a terminal.

[0693] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to generate appropriate answers to customer inquiries.

[0694] An "answer" is a response generated by a generative AI model and provided to a customer.

[0695] "Sentiment analysis" is a technology that analyzes the content of customer inquiries and identifies their emotional state.

[0696] A "sentiment analysis pipeline" is a series of processing steps for performing sentiment analysis, including algorithms for extracting sentiment from text data.

[0697] "Response optimization" is a technique that provides optimal answers and additional support options to customers based on the results of sentiment analysis.

[0698] The present invention is a system for responding to customer inquiries quickly and accurately, and in particular utilizes generative AI models and sentiment analysis techniques. Specific embodiments of this system are described below.

[0699] First, the terminal provides an interface for receiving queries entered by the user. For example, suppose a user enters a query such as "My Internet connection is slow. What should I do?" This query is sent to the system through the input field of the terminal.

[0700] The server then processes the received query. The server analyzes the query and passes it to a generative AI model (e.g., a model using natural language processing technology). This generative AI model generates an appropriate answer to the query. For example, it might say, "If your Internet connection is slow, try restarting your router."

[0701] The generated answer is returned to the terminal and displayed to the user, who can then take action based on the answer.

[0702] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it uses a sentiment analysis pipeline (e.g., an algorithm that identifies the sentiment in a sentence). If the user's query contains negative sentiment, such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server assigns a "NEGATIVE" label to the query.

[0703] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, additional support options can be presented. The message displayed to the user is, "It seems you're having trouble. Would you like us to connect you to our support team?"

[0704] As can be seen from this example, this system not only responds to customer inquiries quickly and accurately, but also aims to improve customer satisfaction by providing optimal responses through sentiment analysis. This system also reduces operational costs.

[0705] Below are some examples of prompt sentences.

[0706] Example prompt: "My internet connection is slow. What should I do?"

[0707] The system aims to achieve a high level of customer satisfaction by quickly and accurately processing customer inquiries and optimizing responses through sentiment analysis, so it is important to effectively utilize generative AI models and sentiment analysis pipelines.

[0708] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0709] Step 1:

[0710] A user inputs a query through the terminal, for example, "My Internet connection is slow. What should I do?" The terminal receives this query and sends it to the server.

[0711] Input: User's inquiry

[0712] Output: Query sent to server

[0713] What happens: The user uses the device interface to enter a question into the text box and clicks "Submit."

[0714] Step 2:

[0715] The server receives the query, normalizes and parses it, which includes removing unnecessary whitespace and special characters from the text, and then passes the query to a generative AI model.

[0716] Input: Query sent from the terminal

[0717] Output: Normalized and parsed text

[0718] Specific operation: After receiving the query, the server cleans and preprocesses the text, preparing it for input into the generative AI model.

[0719] Step 3:

[0720] The server uses a generative AI model to generate an appropriate answer to the query. The query is passed to the generative AI model (for example, a model using natural language processing technology) to generate an answer. For example, the server may receive a response such as, "If your Internet connection is slow, try restarting your router."

[0721] Input: Normalized text query

[0722] Output: The answer generated by the generative AI model

[0723] How it works: The server inputs a query into a natural language processing model and retrieves the best answer provided by the model.

[0724] Step 4:

[0725] The server sends the generated answer to the terminal, which displays the answer to the user.

[0726] Input: The answer generated by the generative AI model

[0727] Output: The answer that is displayed to the user

[0728] Specific operation: The device displays the response received from the server on the screen for the user to see. For example, it may display "If your Internet connection is slow, try restarting your router."

[0729] Step 5:

[0730] The server passes the query content to a sentiment analysis pipeline for sentiment analysis, which detects negative sentiment such as "troubled."

[0731] Input: Normalized text query

[0732] Output: Sentiment analysis result (e.g., NEGATIVE)

[0733] What it does: The server uses a sentiment analysis algorithm to identify positive, negative, neutral, or other sentiment from the input text.

[0734] Step 6:

[0735] The server optimizes the response based on the results of sentiment analysis. If negative sentiment is detected, it generates additional support options, such as "It seems like you're having trouble. Would you like me to connect you to our support team?"

[0736] Input: Sentiment analysis results

[0737] Output: Optimized response (e.g., "Sounds like you're having trouble. Would you like me to connect you to support?")

[0738] What happens: The server runs logic to generate an appropriate follow-up message for the user based on the results of the sentiment analysis.

[0739] Step 7:

[0740] The server transmits the optimized response content to the terminal, which then displays the content to the user.

[0741] Input: Optimized response

[0742] Output: The optimized greeting that is displayed to the user.

[0743] Specific operation: The device displays the optimized message received from the server on the screen and guides the user, for example, "It seems you are having trouble. Would you like us to direct you to our support center?"

[0744] Through these steps, the system can respond to customer inquiries quickly and accurately, and even provide optimal responses through sentiment analysis.

[0745] (Application example 1)

[0746] 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."

[0747] Modern online shopping sites are required to respond quickly and accurately to inquiries from a large number of customers. However, conventional systems make it difficult to respond while taking into account customer emotions, making it difficult to improve customer satisfaction and streamline customer support. Furthermore, there is a lack of support tools that enable customer support staff to respond optimally in real time, making it difficult to maintain high response quality.

[0748] 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.

[0749] In this invention, the server includes means for receiving inquiries from customers, means for passing the received inquiry content to a generative AI model to generate a response, means for providing the generated response to the customer, means for analyzing the sentiment of the customer's inquiry content, means for optimizing the response content based on the result of the sentiment analysis, and means for displaying the inquiry content, the response of the generative AI model, and the result of the sentiment analysis in real time using smart glasses. This enables quick and accurate responses that take customer sentiment into consideration, thereby improving the efficiency of customer support and customer satisfaction.

[0750] "Customer" refers to a consumer or entity that uses a product or service.

[0751] "Inquiry" refers to the act of a customer communicating a question, complaint, or request regarding a product or service to customer support.

[0752] "Means of receiving" refers to the interface or mechanism for receiving inquiries from customers.

[0753] "Generative AI model" refers to an artificial intelligence model that uses natural language processing technology to generate appropriate answers to input inquiries.

[0754] "Answer" refers to the answer or advice that a generative AI model generates based on a query.

[0755] "Means for analyzing emotions" refers to technology that analyzes emotions from customer inquiries and classifies them into emotional labels such as positive or negative.

[0756] "Means for optimizing response content" refers to technology that determines and provides the optimal response method for customers based on the results of sentiment analysis.

[0757] "Smart glasses" are a wearable device in the form of glasses equipped with a display device, allowing customer support representatives to view information in real time.

[0758] "Real-time" refers to information processing occurring simultaneously with real time, with almost no delay.

[0759] The present invention is a system for responding to customer inquiries quickly and accurately, and in particular utilizes generative AI models and sentiment analysis techniques. A specific embodiment of this system is described below.

[0760] First, the server and terminal are the main hardware components that realize this system. The terminal is implemented as smart glasses and is worn by a customer support representative. The terminal provides an interface for receiving customer inquiries using input or voice recognition technology. For example, a user can make an inquiry such as, "My product delivery is delayed. What should I do?" This inquiry is sent to the server through the input field on the terminal.

[0761] The server then processes the received inquiry. The server analyzes the inquiry and passes it to a generative AI model (for example, a model using natural language processing technology). This generative AI model generates an appropriate answer to the inquiry. For example, it might respond, "Please wait a moment while we check the delivery status." The generated answer is then sent back to the device and displayed on the customer support representative's smart glasses. The representative can then explain the situation to the customer while looking at the answer.

[0762] The server then analyzes the sentiment of the inquiry. Sentiment analysis uses a sentiment analysis pipeline (e.g., an algorithm that identifies emotions in sentences). If the user's inquiry contains a negative sentiment, such as "I'm very frustrated because my item hasn't arrived yet," the sentiment analysis will assign a "NEGATIVE" label. Based on the results of this sentiment analysis, the server can optimize the response. For example, if a negative sentiment is detected, additional support options can be presented. A message such as "It seems you're having trouble. Would you like me to direct you to our support desk for further assistance?" will be displayed on the customer support representative's smart glasses.

[0763] Hardware and software used

[0764] Hardware: Smart glasses (built-in microphone, display)

[0765] software:

[0766] Speech Recognition Library:SpeechRecognition Library

[0767] Generative AI models: Natural language processing models (e.g., OpenAI's GPT)

[0768] Sentiment Analysis Pipeline: Transformer Library and Japanese Model

[0769] For example, the following prompt sentences may be used when a customer asks, "My product delivery is delayed. What should I do?"

[0770] Prompt Sentence Examples

[0771] Customer enquiry: "My item delivery is delayed. What should I do?"

[0772] Please provide an appropriate response.

[0773] With the above configuration, the present invention enables quick and accurate responses that take into consideration the customer's feelings, thereby realizing more efficient customer support and improved customer satisfaction.

[0774] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0775] Step 1:

[0776] The user makes a query.

[0777] For example, a user may make a voice inquiry through the built-in microphone of the smart glasses, saying, "My product delivery is delayed. What should I do?" The voice data is input into the terminal.

[0778] Step 2:

[0779] The device converts the voice data into text.

[0780] The device's speech recognition library (for example, the SpeechRecognition library) is used to convert the voice data into text data. The converted text becomes the string "My product delivery is delayed. What should I do?"

[0781] Step 3:

[0782] The text data is sent to the server.

[0783] The terminal transmits the converted character string data to the server, which receives the text data.

[0784] Step 4:

[0785] The server generates an answer using a generative AI model.

[0786] The server passes the received text data to a generative AI model (for example, a model using natural language processing) to generate an appropriate answer. Using the prompt sentence, the answer to "My product delivery is delayed. What should I do?" is output as "We will check the delivery status, so please wait a moment."

[0787] Step 5:

[0788] The generated answer is sent to the terminal.

[0789] The server sends the generated response "We will check the delivery status, so please wait a moment" to the terminal.

[0790] Step 6:

[0791] The device will display the answer.

[0792] The terminal displays the received answer on the display of the smart glasses, where a customer support representative can view the answer and communicate it to the user verbally or via text.

[0793] Step 7:

[0794] The server performs sentiment analysis.

[0795] The server passes the received text data to a sentiment analysis pipeline (e.g., the Transformer library and a Japanese model) to analyze the sentiment. If the input is "My product delivery is delayed. What should I do?", a negative sentiment is detected. The analysis results in a label of "NEGATIVE."

[0796] Step 8:

[0797] Optimize responses based on the results of sentiment analysis.

[0798] The server generates additional response options based on the results of sentiment analysis, such as a suggestion like, "It sounds like you're having trouble. Would you like me to direct you to our support team if you need further assistance?"

[0799] Step 9:

[0800] Suggestions for what to do are displayed on the device.

[0801] The server sends the generated suggestions to the device, which displays them on the smart glasses display. The customer support representative then uses the suggestions to provide the optimal response.

[0802] 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.

[0803] The present invention provides a system for quickly and accurately responding to customer inquiries, and in particular utilizes generative AI models, sentiment analysis techniques, and an emotion engine. Specific embodiments of this system are described below.

[0804] First, the terminal provides an interface to receive queries entered by the user. For example, a user may enter a query such as "My Internet connection is slow. What should I do?" The user's input is sent to the system through the input field on the terminal.

[0805] The server then processes the received query. The server analyzes the query and passes it to a generative AI model (for example, a model using natural language processing technology) to generate an appropriate answer. The generated answer is then returned to the server and sent to the device. For example, the answer generated might be, "If your Internet connection is slow, try restarting your router."

[0806] The generated answer is returned to the terminal and displayed to the user, who can then take action based on the answer.

[0807] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it uses a sentiment analysis pipeline (e.g., an algorithm that identifies the sentiment in a sentence). If the user's query contains negative sentiment, such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server assigns a "NEGATIVE" label to the query.

[0808] The server also uses an emotion engine to recognize the user's emotions. The emotion engine analyzes voice data and text data to identify the user's emotions. For example, when a user makes a voice inquiry, the emotion engine analyzes the voice data to determine whether it contains negative emotions. Similarly, the emotion engine also analyzes text data to identify emotions.

[0809] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, it can present additional support options. The message displayed to the user might say, "It seems you're having trouble. Would you like us to connect you to our support team?"

[0810] As can be seen from this example, this system not only responds to customer inquiries quickly and accurately, but also aims to improve customer satisfaction by providing optimal responses through emotion analysis using an emotion engine. This system also reduces operational costs.

[0811] The processing flow will be explained below.

[0812] Step 1:

[0813] The user inputs the inquiry, for example, "My Internet connection is slow. What should I do?", and clicks the send button.

[0814] Step 2:

[0815] The terminal receives the inquiry entered by the user, temporarily stores the inquiry internally, and sends it to the server when it is ready.

[0816] Step 3:

[0817] The device sends the received query content to the server. Specifically, it creates an HTTP POST request, includes the query content in JSON format, and sends it to the server's API endpoint.

[0818] Step 4:

[0819] The server analyzes the received query, extracts the query content from the HTTP request, and passes it to a generative AI model to generate an appropriate answer.

[0820] Step 5:

[0821] The server generates an answer using a generative AI model, which generates an answer based on the query, such as "If your internet connection is slow, try restarting your router," and sends the answer back to the server.

[0822] Step 6:

[0823] The server returns the generated answer to the terminal, creates an HTTP response including the generated answer, and sends it to the terminal.

[0824] Step 7:

[0825] The device displays the answer received from the server to the user. It analyzes the response from the server and displays the generated answer on the screen. For example, it may say, "If your Internet connection is slow, try restarting your router."

[0826] Step 8:

[0827] The server analyzes the sentiment of the inquiry. Using a sentiment analysis pipeline, it analyzes the inquiry and identifies the sentiment. For example, if the inquiry says, "I can't connect to the Internet at all and I'm really in trouble. Please help me," the sentiment analysis will assign a "NEGATIVE" label.

[0828] Step 9:

[0829] The server uses an emotion engine to recognize the user's emotions. The emotion engine analyzes voice and text data to identify the user's emotions. For example, when a user makes a voice inquiry, the server analyzes the voice data to determine whether the user is expressing negative emotions.

[0830] Step 10:

[0831] The server optimizes the response based on the results of the sentiment analysis. If negative sentiment is detected, it will offer additional support options, such as generating an additional message like, "It seems like you're having trouble. Would you like us to connect you to our support team?"

[0832] Step 11:

[0833] The device displays the emotion analysis results and additional support suggestions sent from the server to the user. The message "It seems you're having trouble. Would you like us to direct you to our support desk?" is displayed on the screen.

[0834] Step 12:

[0835] The user can select additional support options or follow the initial response to resolve the issue, thus providing the user with a proper and prompt response.

[0836] Example 2

[0837] 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."

[0838] While it is necessary to respond to customer inquiries quickly and accurately, it is also important to appropriately understand and respond to customer emotions. However, conventional systems have had the problem of difficulty not only responding to the content of the inquiry but also identifying the customer's emotions and providing an optimal response. The present invention aims to provide a system that responds to customer inquiries quickly and accurately, performs emotion analysis, and realizes an optimal response based on the customer's emotions.

[0839] 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.

[0840] In this invention, the server includes a means for a terminal to receive an inquiry from a customer, a means for analyzing the received inquiry, a means for passing the analysis result to a generative AI model to generate an answer, a means for returning the generated answer to the terminal and providing it to the customer, a means for analyzing the sentiment of the customer's inquiry and the generated answer, a means for identifying emotions from the customer's voice data and text data using an emotion engine, and a means for optimizing the response content based on the results of the emotion analysis and presenting additional support options as necessary. This enables a fast and accurate response to an inquiry as well as an optimal response that takes into account the customer's emotions.

[0841] A "terminal" is an interface device through which a user inputs inquiry details and transmits them to a server.

[0842] A "server" is an information processing device that analyzes the content of the received inquiry and generates a response using a generative AI model.

[0843] An "inquiry" is information about a question or problem that a user sends via a terminal.

[0844] "Analysis" is the process of breaking down the content of a received inquiry into information and meaning and understanding it.

[0845] A "generative AI model" is an algorithm that uses natural language processing technology to generate appropriate answers to input prompts.

[0846] An "answer" is a response to a query generated by a generative AI model.

[0847] "Sentiment analysis" is the process of identifying and classifying the sentiments contained in inquiries and responses.

[0848] An "emotion engine" is an algorithm for identifying a user's emotions from voice data and text data.

[0849] "Response details" refers to the specific details of the response to the user's inquiry.

[0850] "Optimization" is the process of adjusting the response content based on the results of sentiment analysis to create the best possible version.

[0851] "Support Options" means additional methods or means of assistance provided to meet Customer needs.

[0852] The present invention is a system that utilizes generative AI models, sentiment analysis technology, and an emotion engine to respond quickly and accurately to customer inquiries. Specific embodiments of this system are described below.

[0853] First, the terminal provides an interface for receiving queries entered by the user. For example, if a user enters a query such as "My Internet connection is slow. What should I do?", the user's input is sent to the system through the terminal's input field. The terminal then sends the received query to the server using the HTTPS protocol.

[0854] The server analyzes the received inquiry. This analysis can be performed using Google Cloud's natural language processing API. The analysis results include the subject of the inquiry and important keywords. For example, in the case of an inquiry such as "My internet connection is slow. What should I do?", keywords such as "internet connection" and "slow" are extracted.

[0855] The server then passes the analysis results to a generative AI model (e.g., OpenAI's GPT-4) to generate an appropriate answer. The generative AI model generates an appropriate answer based on the prompt text. This prompt text includes the user's query and the analysis results. The generated answer is then returned to the server and sent to the device. For example, the answer generated might be, "If your Internet connection is slow, try restarting your router."

[0856] The generated answer is then returned to the device and displayed to the user, allowing the user to take action based on the information provided to resolve the problem.

[0857] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it may use IBM Watson's sentiment analysis API. For example, if a user sends a query containing negative sentiment such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server analyzes this sentence and assigns a "NEGATIVE" label.

[0858] The server also uses Microsoft Azure's voice sentiment analysis API to enable the emotion engine to analyze customer voice and text data and identify emotions. For voice data, the emotion engine uses the analysis results to detect negative emotions and consider appropriate responses. Similarly, it also identifies emotions for text data.

[0859] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, the server can offer additional support options. Specifically, the message displayed to the user is, "It seems you're having trouble. Would you like us to connect you to our support desk?"

[0860] This system's series of processes not only enables quick and accurate responses to customer inquiries, but also enables optimal responses that take into account the customer's emotions, improving customer satisfaction. This is also expected to reduce operational costs. The system aims to automate customer service while responding to customer needs in a detailed manner.

[0861] Example prompt sentence:

[0862] "User Asks: 'My Internet connection is slow. What should I do?'

[0863] Prompt for the generative AI model:

[0864] Please provide appropriate responses to the following inquiries, including specific steps to take, and, if the user has negative feelings, suggest additional support options that take those feelings into consideration.

[0865] Question: "My internet connection is slow. What should I do?"

[0866] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0867] Step 1:

[0868] The user inputs a query into the terminal interface, such as "My Internet connection is slow. What should I do?" The input is sent to the system through the terminal's input field.

[0869] Step 2:

[0870] The terminal sends the query content to the server. The terminal then securely sends the received query content to the server using the HTTPS protocol. The input is in text format, and the output is the data sent to the server.

[0871] Step 3:

[0872] The server analyzes the query content. The server uses Google Cloud's natural language processing API to analyze the received query content. The input is the received text data, and the output is the analysis results of important keywords and content. For example, from a query such as "My internet connection is slow," keywords such as "internet connection" and "slow" are extracted.

[0873] Step 4:

[0874] The server inputs the analysis results into the generative AI model. The server passes the analysis results to the generative AI model (e.g., generative AI model) as a prompt. The input is the analyzed query content and the prompt, and the output is the input prompt for the generative AI model.

[0875] Step 5:

[0876] The generative AI model generates an appropriate answer. The generative AI model generates an appropriate answer to a query based on the prompt text. The input is the prompt text, and the output is the generated answer. For example, the generated answer might be, "If your Internet connection is slow, try restarting your router."

[0877] Step 6:

[0878] The server receives the generated answer. The server receives the answer returned from the generative AI model and then prepares to send it to the device. The input is the answer from the generative AI model, and the output is the data to be sent to the device.

[0879] Step 7:

[0880] The server performs sentiment analysis. Using IBM Watson's sentiment analysis API, the server analyzes both the received inquiry and the generated answer. The input is the inquiry and the generated answer, and the output is the result of the sentiment analysis. For example, an inquiry such as "I can't connect to the Internet at all and I'm really having trouble. Please help me" is classified as "NEGATIVE."

[0881] Step 8:

[0882] The server uses an emotion engine to identify the customer's emotions. The server uses Microsoft Azure's voice emotion analysis API to analyze the customer's voice data and text data to identify the customer's emotions. The input is voice data or text data, and the output is the emotion identification result.

[0883] Step 9:

[0884] The server optimizes the response content. The server optimizes the response content based on the results of sentiment analysis. In particular, if negative sentiment is detected, additional support options are presented. The input is the result of sentiment analysis, and the output is the optimized response content. For example, a message such as "It seems you're having trouble. Would you like me to direct you to our support desk?" is added.

[0885] Step 10:

[0886] The terminal displays the optimal response content to the user. The terminal displays the optimized response content received from the server to the user. The input is the response data from the server, and the output is the display to the user. As a specific example, the user terminal might display the message, "If your Internet connection is slow, try restarting your router. It seems you're having trouble. Would you like me to direct you to support?"

[0887] (Application example 2)

[0888] 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."

[0889] In today's world, where fast and accurate responses to customer inquiries are required, improving customer satisfaction and reducing operational costs are key challenges, especially in the online shopping field. Conventional systems often lack the ability to adequately analyze emotions, resulting in inappropriate responses. The present invention aims to solve these challenges by utilizing generative AI models, emotion analysis technology, and smartphones.

[0890] 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.

[0891] In this invention, the server includes means for receiving inquiries from customers, means for passing the received inquiry content to a generative AI model to generate a response, means for providing the generated response to the customer, means for analyzing the sentiment of the customer's inquiry content, means for optimizing the response content based on the results of the sentiment analysis, and means for receiving inquiries using a smartphone and displaying the response, thereby enabling improved customer satisfaction and reduced operation costs.

[0892] The "means for receiving inquiries from customers" refers to the means for delivering the contents of inquiries entered by customers to the server via a terminal or system.

[0893] "A means of generating a response by passing the received inquiry content to a generative AI model" refers to a means of analyzing the content of a customer inquiry and using a generative AI model to create a response appropriate to that content.

[0894] "Means for providing the generated answer to the customer" refers to means for providing the answer generated by the generative AI model to the customer in an easy-to-understand manner.

[0895] The "means for analyzing the emotions of the customer's inquiry content" is a means for extracting and analyzing the customer's emotions from the inquiry content entered by the customer.

[0896] "Means for optimizing response content based on the results of sentiment analysis" refers to means for optimizing response content based on the results of sentiment analysis, enabling appropriate responses in a variety of situations.

[0897] The "means for receiving an inquiry using a smartphone and displaying a response" refers to a means for a customer to input an inquiry using a smartphone and display the result.

[0898] This invention is a system for responding quickly and accurately to customer inquiries on an online shopping site. The system includes a generative AI model, sentiment analysis technology, and a smartphone-based response display.

[0899] Hardware and software configuration used

[0900] Hardware:

[0901] Smartphone: A device for receiving customer inquiries and displaying responses.

[0902] Server: Analyzes the inquiry content, generates answers using a generation AI model, and performs sentiment analysis.

[0903] software:

[0904] openai library: An API for using generative AI models to generate appropriate answers to inquiries.

[0905] textblob library: Natural language processing technology for performing sentiment analysis of customer inquiries.

[0906] Sentiment analysis pipeline: Algorithms to dig deep into customer sentiment and help optimize responses.

[0907] Processing Description

[0908] 1. Receiving Inquiries

[0909] A user uses a chat interface on their smartphone to input an inquiry. For example, they might type, "I'm having trouble because my product hasn't arrived yet. When will it arrive?" This input is then sent to the server via the device.

[0910] 2. Analysis of inquiry content

[0911] The server passes the received query to the generative AI model, which generates an appropriate answer to the query.

[0912] 3. Sentiment analysis

[0913] At the same time, the server uses the textblob library to analyze the sentiment of the query, and based on the results, identifies whether the user's sentiment is positive or negative.

[0914] 4. Response optimization

[0915] Optimize the generated answer based on the results of sentiment analysis. For example, suggest additional support options if negative sentiment is detected. Add a message to the generated answer such as "I see you're having trouble. Would you like me to connect you to support?"

[0916] 5. Providing answers

[0917] The optimized response is then sent back to the smartphone and displayed to the user, for example, "We usually ship within 3-5 business days of your order. If you need further assistance, please contact customer service."

[0918] Examples and prompts

[0919] Examples:

[0920] If a user types an inquiry such as, "I'm having trouble receiving my item. When will it arrive?", the AI ​​will respond with, "Items are usually shipped within 3 to 5 business days of your order. If you need further support, please contact customer service." If sentiment analysis determines that the user is expressing negative emotions, it will additionally display the message, "It seems you're having trouble. Would you like me to direct you to support?"

[0921] Example prompt for a generative AI model:

[0922] "A customer asks: 'I'm having trouble receiving my item. When will it arrive?' Please provide a suitable response to this question."

[0923] The present invention can improve the efficiency and quality of customer support at online shopping sites, thereby improving customer satisfaction.

[0924] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0925] Step 1:

[0926] User enters a query

[0927] The user uses the chat interface on their smartphone to input an inquiry. This input is used as the initial data for the system. For example, the user might input, "I'm having trouble because my product hasn't arrived yet. When will it arrive?"

[0928] Step 2:

[0929] The device sends a query to the server

[0930] The device sends the query entered by the user to the server, which then sends the text data to the server and prepares it for analysis.

[0931] Step 3:

[0932] The server analyzes the query

[0933] The server passes the received query content to the generative AI model. Specifically, it uses the openai library to provide the query content to the AI ​​model as a prompt.

[0934] Step 4:

[0935] Generative AI models generate answers

[0936] The generative AI model generates an appropriate answer based on the prompt text, and this data is returned to the server as generated text. For example, the answer generated might be, "We usually ship within 3-5 business days of your order."

[0937] Step 5:

[0938] The server performs sentiment analysis

[0939] At the same time, the server analyzes the sentiment of the received query using the textblob library, extracting positive or negative sentiment from the input text and providing the result (sentiment score).

[0940] Step 6:

[0941] The server optimizes the response

[0942] Based on the results of the sentiment analysis, the server optimizes the response. For example, if a negative sentiment is detected, the generated response will include a message such as, "It sounds like you're having trouble. Would you like me to connect you to support?"

[0943] Step 7:

[0944] Optimized answers sent to your device

[0945] The optimized answer is sent from the server to the device, where it is ready to be displayed to the user as the final answer.

[0946] Step 8:

[0947] The device displays the answer to the user

[0948] The device will display the optimized answer received from the server to the user, who can then refer to it to resolve the issue. For example, the answer might read, "We usually ship within 3-5 business days of your order. If you need further assistance, please contact customer service."

[0949] 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.

[0950] 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.

[0951] 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.

[0952] [Fourth embodiment]

[0953] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0954] 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.

[0955] 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).

[0956] 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.

[0957] 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.

[0958] 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).

[0959] 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. 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.

[0960] 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.

[0961] 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.

[0962] 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.

[0963] 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.

[0964] 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.

[0965] 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."

[0966] The present invention is a system for responding to customer inquiries quickly and accurately, and in particular utilizes generative AI models and sentiment analysis techniques. Specific embodiments of this system are described below.

[0967] First, the terminal provides an interface for receiving queries entered by the user. For example, suppose a user enters a query such as "My Internet connection is slow. What should I do?" This query is sent to the system through the input field of the terminal.

[0968] The server then processes the received query. The server analyzes the query and passes it to a generative AI model (e.g., a model using natural language processing technology). This generative AI model generates an appropriate answer to the query. For example, it might say, "If your Internet connection is slow, try restarting your router."

[0969] The generated answer is returned to the terminal and displayed to the user, who can then take action based on the answer.

[0970] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it uses a sentiment analysis pipeline (e.g., an algorithm that identifies the sentiment in a sentence). If the user's query contains negative sentiment, such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server assigns a "NEGATIVE" label to the query.

[0971] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, it can present additional support options. The message displayed to the user might say, "It seems you're having trouble. Would you like us to connect you to our support team?"

[0972] As can be seen from this example, this system not only responds to customer inquiries quickly and accurately, but also aims to improve customer satisfaction by providing optimal responses through sentiment analysis. This system also reduces operational costs.

[0973] The processing flow will be explained below.

[0974] Step 1:

[0975] The user inputs a query. For example, the user might input "My Internet connection is slow. What should I do?" The user's input is entered into an input field on the device.

[0976] Step 2:

[0977] The terminal receives the query entered by the user, which is then temporarily stored and prepared for transmission to the server.

[0978] Step 3:

[0979] The terminal sends the received query content to the server. Specifically, the terminal creates an HTTP POST request, includes the query content in JSON format in the request body, and sends it to the specified endpoint of the server.

[0980] Step 4:

[0981] The server analyzes the received query, extracts the query content from the received request, and passes it to a generative AI model to generate an appropriate answer.

[0982] Step 5:

[0983] The server generates an answer using a generative AI model. The generative AI model generates the best answer based on the query and returns that answer to the server. For example, if the query is "My Internet connection is slow. What should I do?", the generated answer might be "If your Internet connection is slow, try restarting your router."

[0984] Step 6:

[0985] The server returns the generated answer to the terminal. The server creates an HTTP response that includes the generated answer and sends the response to the terminal.

[0986] Step 7:

[0987] The device displays the answer received from the server to the user. The device analyzes the response from the server and displays the generated answer on the screen. For example, the answer "If your Internet connection is slow, try restarting your router" may be presented to the user.

[0988] Step 8:

[0989] The server analyzes the sentiment of the inquiry. Using a sentiment analysis pipeline, the server analyzes the received inquiry and identifies the sentiment. For example, if the inquiry is "I can't connect to the Internet at all and I'm really in trouble. Please help me," the result of sentiment analysis will be assigned a "NEGATIVE" label.

[0990] Step 9:

[0991] The server optimizes the response based on the results of sentiment analysis. If the result of sentiment analysis is negative, it generates a message suggesting additional support. For example, it generates a message saying, "It seems you're having trouble. Would you like me to connect you to our support desk?"

[0992] Step 10:

[0993] The device displays the emotion analysis results and additional support suggestions sent from the server to the user. The user can see a message on the screen saying, "It seems you're having trouble. Would you like us to direct you to our support desk?"

[0994] Step 11:

[0995] The user can select additional support options or follow the generated answer to resolve the issue, thus providing the user with a timely and appropriate response.

[0996] Example 1

[0997] 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."

[0998] Conventional customer support systems often lacked timely and accurate responses to customer inquiries, leading to poor customer satisfaction. Furthermore, they lacked the ability to consider customer emotions, making it difficult to provide appropriate support, especially to customers with negative feelings. Additionally, these systems also had the problem of high operating costs.

[0999] 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.

[1000] In this invention, the server includes a means for passing the received inquiry content to a generative AI model to generate an answer, a means for providing the generated answer to the customer via a terminal, a means for analyzing the sentiment of the customer's inquiry content, and a means for optimizing the response content based on the results of the sentiment analysis. This enables quick and accurate responses to customer inquiries, and by realizing optimal responses through sentiment analysis, it is possible to improve customer satisfaction. It also realizes efficient operations and enables cost reduction.

[1001] A "terminal" is an input device that allows a customer to input an inquiry and transmit it to the system.

[1002] The "server" is a central processing unit that analyzes the content of the received inquiry, generates a response using a generative AI model, and then performs sentiment analysis to optimize the response content.

[1003] An "inquiry" is a question or inquiry that a customer inputs into the system via a terminal.

[1004] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to generate appropriate answers to customer inquiries.

[1005] An "answer" is a response generated by a generative AI model and provided to a customer.

[1006] "Sentiment analysis" is a technology that analyzes the content of customer inquiries and identifies their emotional state.

[1007] A "sentiment analysis pipeline" is a series of processing steps for performing sentiment analysis, including algorithms for extracting sentiment from text data.

[1008] "Response optimization" is a technique that provides optimal answers and additional support options to customers based on the results of sentiment analysis.

[1009] The present invention is a system for responding to customer inquiries quickly and accurately, and in particular utilizes generative AI models and sentiment analysis techniques. Specific embodiments of this system are described below.

[1010] First, the terminal provides an interface for receiving queries entered by the user. For example, suppose a user enters a query such as "My Internet connection is slow. What should I do?" This query is sent to the system through the input field of the terminal.

[1011] The server then processes the received query. The server analyzes the query and passes it to a generative AI model (e.g., a model using natural language processing technology). This generative AI model generates an appropriate answer to the query. For example, it might say, "If your Internet connection is slow, try restarting your router."

[1012] The generated answer is returned to the terminal and displayed to the user, who can then take action based on the answer.

[1013] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it uses a sentiment analysis pipeline (e.g., an algorithm that identifies the sentiment in a sentence). If the user's query contains negative sentiment, such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server assigns a "NEGATIVE" label to the query.

[1014] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, additional support options can be presented. The message displayed to the user is, "It seems you're having trouble. Would you like us to connect you to our support team?"

[1015] As can be seen from this example, this system not only responds to customer inquiries quickly and accurately, but also aims to improve customer satisfaction by providing optimal responses through sentiment analysis. This system also reduces operational costs.

[1016] Below are some examples of prompt sentences.

[1017] Example prompt: "My internet connection is slow. What should I do?"

[1018] The system aims to achieve a high level of customer satisfaction by quickly and accurately processing customer inquiries and optimizing responses through sentiment analysis, so it is important to effectively utilize generative AI models and sentiment analysis pipelines.

[1019] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1020] Step 1:

[1021] A user inputs a query through the terminal, for example, "My Internet connection is slow. What should I do?" The terminal receives this query and sends it to the server.

[1022] Input: User's inquiry

[1023] Output: Query sent to server

[1024] What happens: The user uses the device interface to enter a question into the text box and clicks "Submit."

[1025] Step 2:

[1026] The server receives the query, normalizes and parses it, which includes removing unnecessary whitespace and special characters from the text, and then passes the query to a generative AI model.

[1027] Input: Query sent from the terminal

[1028] Output: Normalized and parsed text

[1029] Specific operation: After receiving the query, the server cleans and preprocesses the text, preparing it for input into the generative AI model.

[1030] Step 3:

[1031] The server uses a generative AI model to generate an appropriate answer to the query. The query is passed to the generative AI model (for example, a model using natural language processing technology) to generate an answer. For example, the server may receive a response such as, "If your Internet connection is slow, try restarting your router."

[1032] Input: Normalized text query

[1033] Output: The answer generated by the generative AI model

[1034] How it works: The server inputs a query into a natural language processing model and retrieves the best answer provided by the model.

[1035] Step 4:

[1036] The server sends the generated answer to the terminal, which displays the answer to the user.

[1037] Input: The answer generated by the generative AI model

[1038] Output: The answer that is displayed to the user

[1039] Specific operation: The device displays the response received from the server on the screen for the user to see. For example, it may display "If your Internet connection is slow, try restarting your router."

[1040] Step 5:

[1041] The server passes the query content to a sentiment analysis pipeline for sentiment analysis, which detects negative sentiment such as "troubled."

[1042] Input: Normalized text query

[1043] Output: Sentiment analysis result (e.g., NEGATIVE)

[1044] What it does: The server uses a sentiment analysis algorithm to identify positive, negative, neutral, or other sentiment from the input text.

[1045] Step 6:

[1046] The server optimizes the response based on the results of sentiment analysis. If negative sentiment is detected, it generates additional support options, such as "It seems like you're having trouble. Would you like me to connect you to our support team?"

[1047] Input: Sentiment analysis results

[1048] Output: Optimized response (e.g., "Sounds like you're having trouble. Would you like me to connect you to support?")

[1049] What happens: The server runs logic to generate an appropriate follow-up message for the user based on the results of the sentiment analysis.

[1050] Step 7:

[1051] The server transmits the optimized response content to the terminal, which then displays the content to the user.

[1052] Input: Optimized response

[1053] Output: The optimized greeting that is displayed to the user.

[1054] Specific operation: The device displays the optimized message received from the server on the screen and guides the user, for example, "It seems you are having trouble. Would you like us to direct you to our support center?"

[1055] Through these steps, the system can respond to customer inquiries quickly and accurately, and even provide optimal responses through sentiment analysis.

[1056] (Application example 1)

[1057] 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."

[1058] Modern online shopping sites are required to respond quickly and accurately to inquiries from a large number of customers. However, conventional systems make it difficult to respond while taking into account customer emotions, making it difficult to improve customer satisfaction and streamline customer support. Furthermore, there is a lack of support tools that enable customer support staff to respond optimally in real time, making it difficult to maintain high response quality.

[1059] 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.

[1060] In this invention, the server includes means for receiving inquiries from customers, means for passing the received inquiry content to a generative AI model to generate a response, means for providing the generated response to the customer, means for analyzing the sentiment of the customer's inquiry content, means for optimizing the response content based on the result of the sentiment analysis, and means for displaying the inquiry content, the response of the generative AI model, and the result of the sentiment analysis in real time using smart glasses. This enables quick and accurate responses that take customer sentiment into consideration, thereby improving the efficiency of customer support and customer satisfaction.

[1061] "Customer" refers to a consumer or entity that uses a product or service.

[1062] "Inquiry" refers to the act of a customer communicating a question, complaint, or request regarding a product or service to customer support.

[1063] "Means of receiving" refers to the interface or mechanism for receiving inquiries from customers.

[1064] "Generative AI model" refers to an artificial intelligence model that uses natural language processing technology to generate appropriate answers to input inquiries.

[1065] "Answer" refers to the answer or advice that a generative AI model generates based on a query.

[1066] "Means for analyzing emotions" refers to technology that analyzes emotions from customer inquiries and classifies them into emotional labels such as positive or negative.

[1067] "Means for optimizing response content" refers to technology that determines and provides the optimal response method for customers based on the results of sentiment analysis.

[1068] "Smart glasses" are a wearable device in the form of glasses equipped with a display device, allowing customer support representatives to view information in real time.

[1069] "Real-time" refers to information processing occurring simultaneously with real time, with almost no delay.

[1070] The present invention is a system for responding to customer inquiries quickly and accurately, and in particular utilizes generative AI models and sentiment analysis techniques. A specific embodiment of this system is described below.

[1071] First, the server and terminal are the main hardware components that realize this system. The terminal is implemented as smart glasses and is worn by a customer support representative. The terminal provides an interface for receiving customer inquiries using input or voice recognition technology. For example, a user can make an inquiry such as, "My product delivery is delayed. What should I do?" This inquiry is sent to the server through the input field on the terminal.

[1072] The server then processes the received inquiry. The server analyzes the inquiry and passes it to a generative AI model (for example, a model using natural language processing technology). This generative AI model generates an appropriate answer to the inquiry. For example, it might respond, "Please wait a moment while we check the delivery status." The generated answer is then sent back to the device and displayed on the customer support representative's smart glasses. The representative can then explain the situation to the customer while looking at the answer.

[1073] The server then analyzes the sentiment of the inquiry. Sentiment analysis uses a sentiment analysis pipeline (e.g., an algorithm that identifies emotions in sentences). If the user's inquiry contains a negative sentiment, such as "I'm very frustrated because my item hasn't arrived yet," the sentiment analysis will assign a "NEGATIVE" label. Based on the results of this sentiment analysis, the server can optimize the response. For example, if a negative sentiment is detected, additional support options can be presented. A message such as "It seems you're having trouble. Would you like me to direct you to our support desk for further assistance?" will be displayed on the customer support representative's smart glasses.

[1074] Hardware and software used

[1075] Hardware: Smart glasses (built-in microphone, display)

[1076] software:

[1077] Speech Recognition Library:SpeechRecognition Library

[1078] Generative AI models: Natural language processing models (e.g., OpenAI's GPT)

[1079] Sentiment Analysis Pipeline: Transformer Library and Japanese Model

[1080] For example, the following prompt sentences may be used when a customer asks, "My product delivery is delayed. What should I do?"

[1081] Prompt Sentence Examples

[1082] Customer enquiry: "My item delivery is delayed. What should I do?"

[1083] Please provide an appropriate response.

[1084] With the above configuration, the present invention enables quick and accurate responses that take into consideration the customer's feelings, thereby realizing more efficient customer support and improved customer satisfaction.

[1085] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1086] Step 1:

[1087] The user makes a query.

[1088] For example, a user may make a voice inquiry through the built-in microphone of the smart glasses, saying, "My product delivery is delayed. What should I do?" The voice data is input into the terminal.

[1089] Step 2:

[1090] The device converts the voice data into text.

[1091] The device's speech recognition library (for example, the SpeechRecognition library) is used to convert the voice data into text data. The converted text becomes the string "My product delivery is delayed. What should I do?"

[1092] Step 3:

[1093] The text data is sent to the server.

[1094] The terminal transmits the converted character string data to the server, which receives the text data.

[1095] Step 4:

[1096] The server generates an answer using a generative AI model.

[1097] The server passes the received text data to a generative AI model (for example, a model using natural language processing) to generate an appropriate answer. Using the prompt sentence, the answer to "My product delivery is delayed. What should I do?" is output as "We will check the delivery status, so please wait a moment."

[1098] Step 5:

[1099] The generated answer is sent to the terminal.

[1100] The server sends the generated response "We will check the delivery status, so please wait a moment" to the terminal.

[1101] Step 6:

[1102] The device will display the answer.

[1103] The terminal displays the received answer on the display of the smart glasses, where a customer support representative can view the answer and communicate it to the user verbally or via text.

[1104] Step 7:

[1105] The server performs sentiment analysis.

[1106] The server passes the received text data to a sentiment analysis pipeline (e.g., the Transformer library and a Japanese model) to analyze the sentiment. If the input is "My product delivery is delayed. What should I do?", a negative sentiment is detected. The analysis results in a label of "NEGATIVE."

[1107] Step 8:

[1108] Optimize responses based on the results of sentiment analysis.

[1109] The server generates additional response options based on the results of sentiment analysis, such as a suggestion like, "It sounds like you're having trouble. Would you like me to direct you to our support team if you need further assistance?"

[1110] Step 9:

[1111] Suggestions for what to do are displayed on the device.

[1112] The server sends the generated suggestions to the device, which displays them on the smart glasses display. The customer support representative then uses the suggestions to provide the optimal response.

[1113] 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.

[1114] The present invention provides a system for quickly and accurately responding to customer inquiries, and in particular utilizes generative AI models, sentiment analysis techniques, and an emotion engine. Specific embodiments of this system are described below.

[1115] First, the terminal provides an interface to receive queries entered by the user. For example, a user may enter a query such as "My Internet connection is slow. What should I do?" The user's input is sent to the system through the input field on the terminal.

[1116] The server then processes the received query. The server analyzes the query and passes it to a generative AI model (for example, a model using natural language processing technology) to generate an appropriate answer. The generated answer is then returned to the server and sent to the device. For example, the answer generated might be, "If your Internet connection is slow, try restarting your router."

[1117] The generated answer is returned to the terminal and displayed to the user, who can then take action based on the answer.

[1118] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it uses a sentiment analysis pipeline (e.g., an algorithm that identifies the sentiment in a sentence). If the user's query contains negative sentiment, such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server assigns a "NEGATIVE" label to the query.

[1119] The server also uses an emotion engine to recognize the user's emotions. The emotion engine analyzes voice data and text data to identify the user's emotions. For example, when a user makes a voice inquiry, the emotion engine analyzes the voice data to determine whether it contains negative emotions. Similarly, the emotion engine also analyzes text data to identify emotions.

[1120] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, it can present additional support options. The message displayed to the user might say, "It seems you're having trouble. Would you like us to connect you to our support team?"

[1121] As can be seen from this example, this system not only responds to customer inquiries quickly and accurately, but also aims to improve customer satisfaction by providing optimal responses through emotion analysis using an emotion engine. This system also reduces operational costs.

[1122] The processing flow will be explained below.

[1123] Step 1:

[1124] The user inputs the inquiry, for example, "My Internet connection is slow. What should I do?", and clicks the send button.

[1125] Step 2:

[1126] The terminal receives the inquiry entered by the user, temporarily stores the inquiry internally, and sends it to the server when it is ready.

[1127] Step 3:

[1128] The device sends the received query content to the server. Specifically, it creates an HTTP POST request, includes the query content in JSON format, and sends it to the server's API endpoint.

[1129] Step 4:

[1130] The server analyzes the received query, extracts the query content from the HTTP request, and passes it to a generative AI model to generate an appropriate answer.

[1131] Step 5:

[1132] The server generates an answer using a generative AI model, which generates an answer based on the query, such as "If your internet connection is slow, try restarting your router," and sends the answer back to the server.

[1133] Step 6:

[1134] The server returns the generated answer to the terminal, creates an HTTP response including the generated answer, and sends it to the terminal.

[1135] Step 7:

[1136] The device displays the answer received from the server to the user. It analyzes the response from the server and displays the generated answer on the screen. For example, it may say, "If your Internet connection is slow, try restarting your router."

[1137] Step 8:

[1138] The server analyzes the sentiment of the inquiry. Using a sentiment analysis pipeline, it analyzes the inquiry and identifies the sentiment. For example, if the inquiry says, "I can't connect to the Internet at all and I'm really in trouble. Please help me," the sentiment analysis will assign a "NEGATIVE" label.

[1139] Step 9:

[1140] The server uses an emotion engine to recognize the user's emotions. The emotion engine analyzes voice and text data to identify the user's emotions. For example, when a user makes a voice inquiry, the server analyzes the voice data to determine whether the user is expressing negative emotions.

[1141] Step 10:

[1142] The server optimizes the response based on the results of the sentiment analysis. If negative sentiment is detected, it will offer additional support options, such as generating an additional message like, "It seems like you're having trouble. Would you like us to connect you to our support team?"

[1143] Step 11:

[1144] The device displays the emotion analysis results and additional support suggestions sent from the server to the user. The message "It seems you're having trouble. Would you like us to direct you to our support desk?" is displayed on the screen.

[1145] Step 12:

[1146] The user can select additional support options or follow the initial response to resolve the issue, thus providing the user with a proper and prompt response.

[1147] Example 2

[1148] 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."

[1149] While it is necessary to respond to customer inquiries quickly and accurately, it is also important to appropriately understand and respond to customer emotions. However, conventional systems have had the problem of difficulty not only responding to the content of the inquiry but also identifying the customer's emotions and providing an optimal response. The present invention aims to provide a system that responds to customer inquiries quickly and accurately, performs emotion analysis, and realizes an optimal response based on the customer's emotions.

[1150] 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.

[1151] In this invention, the server includes a means for a terminal to receive an inquiry from a customer, a means for analyzing the received inquiry, a means for passing the analysis result to a generative AI model to generate an answer, a means for returning the generated answer to the terminal and providing it to the customer, a means for analyzing the sentiment of the customer's inquiry and the generated answer, a means for identifying emotions from the customer's voice data and text data using an emotion engine, and a means for optimizing the response content based on the results of the emotion analysis and presenting additional support options as necessary. This enables a fast and accurate response to an inquiry as well as an optimal response that takes into account the customer's emotions.

[1152] A "terminal" is an interface device through which a user inputs inquiry details and transmits them to a server.

[1153] A "server" is an information processing device that analyzes the content of the received inquiry and generates a response using a generative AI model.

[1154] An "inquiry" is information about a question or problem that a user sends via a terminal.

[1155] "Analysis" is the process of breaking down the content of a received inquiry into information and meaning and understanding it.

[1156] A "generative AI model" is an algorithm that uses natural language processing technology to generate appropriate answers to input prompts.

[1157] An "answer" is a response to a query generated by a generative AI model.

[1158] "Sentiment analysis" is the process of identifying and classifying the sentiments contained in inquiries and responses.

[1159] An "emotion engine" is an algorithm for identifying a user's emotions from voice data and text data.

[1160] "Response details" refers to the specific details of the response to the user's inquiry.

[1161] "Optimization" is the process of adjusting the response content based on the results of sentiment analysis to create the best possible version.

[1162] "Support Options" means additional methods or means of assistance provided to meet Customer needs.

[1163] The present invention is a system that utilizes generative AI models, sentiment analysis technology, and an emotion engine to respond quickly and accurately to customer inquiries. Specific embodiments of this system are described below.

[1164] First, the terminal provides an interface for receiving queries entered by the user. For example, if a user enters a query such as "My Internet connection is slow. What should I do?", the user's input is sent to the system through the terminal's input field. The terminal then sends the received query to the server using the HTTPS protocol.

[1165] The server analyzes the received inquiry. This analysis can be performed using Google Cloud's natural language processing API. The analysis results include the subject of the inquiry and important keywords. For example, in the case of an inquiry such as "My internet connection is slow. What should I do?", keywords such as "internet connection" and "slow" are extracted.

[1166] The server then passes the analysis results to a generative AI model (e.g., OpenAI's GPT-4) to generate an appropriate answer. The generative AI model generates an appropriate answer based on the prompt text. This prompt text includes the user's query and the analysis results. The generated answer is then returned to the server and sent to the device. For example, the answer generated might be, "If your Internet connection is slow, try restarting your router."

[1167] The generated answer is then returned to the device and displayed to the user, allowing the user to take action based on the information provided to resolve the problem.

[1168] Furthermore, the server analyzes the sentiment of the query. For sentiment analysis, it may use IBM Watson's sentiment analysis API. For example, if a user sends a query containing negative sentiment such as "I can't connect to the Internet at all and I'm really in trouble. Please help me," the server analyzes this sentence and assigns a "NEGATIVE" label.

[1169] The server also uses Microsoft Azure's voice sentiment analysis API to enable the emotion engine to analyze customer voice and text data and identify emotions. For voice data, the emotion engine uses the analysis results to detect negative emotions and consider appropriate responses. Similarly, it also identifies emotions for text data.

[1170] Based on the results of this sentiment analysis, the server can optimize the response. For example, if negative sentiment is detected, the server can offer additional support options. Specifically, the message displayed to the user is, "It seems you're having trouble. Would you like us to connect you to our support desk?"

[1171] This system's series of processes not only enables quick and accurate responses to customer inquiries, but also enables optimal responses that take into account the customer's emotions, improving customer satisfaction. This is also expected to reduce operational costs. The system aims to automate customer service while responding to customer needs in a detailed manner.

[1172] Example prompt sentence:

[1173] "User Asks: 'My Internet connection is slow. What should I do?'

[1174] Prompt for the generative AI model:

[1175] Please provide appropriate responses to the following inquiries, including specific steps to take, and, if the user has negative feelings, suggest additional support options that take those feelings into consideration.

[1176] Question: "My internet connection is slow. What should I do?"

[1177] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1178] Step 1:

[1179] The user inputs a query into the terminal interface, such as "My Internet connection is slow. What should I do?" The input is sent to the system through the terminal's input field.

[1180] Step 2:

[1181] The terminal sends the query content to the server. The terminal then securely sends the received query content to the server using the HTTPS protocol. The input is in text format, and the output is the data sent to the server.

[1182] Step 3:

[1183] The server analyzes the query content. The server uses Google Cloud's natural language processing API to analyze the received query content. The input is the received text data, and the output is the analysis results of important keywords and content. For example, from a query such as "My internet connection is slow," keywords such as "internet connection" and "slow" are extracted.

[1184] Step 4:

[1185] The server inputs the analysis results into the generative AI model. The server passes the analysis results to the generative AI model (e.g., generative AI model) as a prompt. The input is the analyzed query content and the prompt, and the output is the input prompt for the generative AI model.

[1186] Step 5:

[1187] The generative AI model generates an appropriate answer. The generative AI model generates an appropriate answer to a query based on the prompt text. The input is the prompt text, and the output is the generated answer. For example, the generated answer might be, "If your Internet connection is slow, try restarting your router."

[1188] Step 6:

[1189] The server receives the generated answer. The server receives the answer returned from the generative AI model and then prepares to send it to the device. The input is the answer from the generative AI model, and the output is the data to be sent to the device.

[1190] Step 7:

[1191] The server performs sentiment analysis. Using IBM Watson's sentiment analysis API, the server analyzes both the received inquiry and the generated answer. The input is the inquiry and the generated answer, and the output is the result of the sentiment analysis. For example, an inquiry such as "I can't connect to the Internet at all and I'm really having trouble. Please help me" is classified as "NEGATIVE."

[1192] Step 8:

[1193] The server uses an emotion engine to identify the customer's emotions. The server uses Microsoft Azure's voice emotion analysis API to analyze the customer's voice data and text data to identify the customer's emotions. The input is voice data or text data, and the output is the emotion identification result.

[1194] Step 9:

[1195] The server optimizes the response content. The server optimizes the response content based on the results of sentiment analysis. In particular, if negative sentiment is detected, additional support options are presented. The input is the result of sentiment analysis, and the output is the optimized response content. For example, a message such as "It seems you're having trouble. Would you like me to direct you to our support desk?" is added.

[1196] Step 10:

[1197] The terminal displays the optimal response content to the user. The terminal displays the optimized response content received from the server to the user. The input is the response data from the server, and the output is the display to the user. As a specific example, the user terminal might display the message, "If your Internet connection is slow, try restarting your router. It seems you're having trouble. Would you like me to direct you to support?"

[1198] (Application example 2)

[1199] 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."

[1200] In today's world, where fast and accurate responses to customer inquiries are required, improving customer satisfaction and reducing operational costs are key challenges, especially in the online shopping field. Conventional systems often lack the ability to adequately analyze emotions, resulting in inappropriate responses. The present invention aims to solve these challenges by utilizing generative AI models, emotion analysis technology, and smartphones.

[1201] 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.

[1202] In this invention, the server includes means for receiving inquiries from customers, means for passing the received inquiry content to a generative AI model to generate a response, means for providing the generated response to the customer, means for analyzing the sentiment of the customer's inquiry content, means for optimizing the response content based on the results of the sentiment analysis, and means for receiving inquiries using a smartphone and displaying the response, thereby enabling improved customer satisfaction and reduced operation costs.

[1203] The "means for receiving inquiries from customers" refers to the means for delivering the contents of inquiries entered by customers to the server via a terminal or system.

[1204] "A means of generating a response by passing the received inquiry content to a generative AI model" refers to a means of analyzing the content of a customer inquiry and using a generative AI model to create a response appropriate to that content.

[1205] "Means for providing the generated answer to the customer" refers to means for providing the answer generated by the generative AI model to the customer in an easy-to-understand manner.

[1206] The "means for analyzing the emotions of the customer's inquiry content" is a means for extracting and analyzing the customer's emotions from the inquiry content entered by the customer.

[1207] "Means for optimizing response content based on the results of sentiment analysis" refers to means for optimizing response content based on the results of sentiment analysis, enabling appropriate responses in a variety of situations.

[1208] The "means for receiving an inquiry using a smartphone and displaying a response" refers to a means for a customer to input an inquiry using a smartphone and display the result.

[1209] This invention is a system for responding quickly and accurately to customer inquiries on an online shopping site. The system includes a generative AI model, sentiment analysis technology, and a smartphone-based response display.

[1210] Hardware and software configuration used

[1211] Hardware:

[1212] Smartphone: A device for receiving customer inquiries and displaying responses.

[1213] Server: Analyzes the inquiry content, generates answers using a generation AI model, and performs sentiment analysis.

[1214] software:

[1215] openai library: An API for using generative AI models to generate appropriate answers to inquiries.

[1216] textblob library: Natural language processing technology for performing sentiment analysis of customer inquiries.

[1217] Sentiment analysis pipeline: Algorithms to dig deep into customer sentiment and help optimize responses.

[1218] Processing Description

[1219] 1. Receiving Inquiries

[1220] A user uses a chat interface on their smartphone to input an inquiry. For example, they might type, "I'm having trouble because my product hasn't arrived yet. When will it arrive?" This input is then sent to the server via the device.

[1221] 2. Analysis of inquiry content

[1222] The server passes the received query to the generative AI model, which generates an appropriate answer to the query.

[1223] 3. Sentiment analysis

[1224] At the same time, the server uses the textblob library to analyze the sentiment of the query, and based on the results, identifies whether the user's sentiment is positive or negative.

[1225] 4. Response optimization

[1226] Optimize the generated answer based on the results of sentiment analysis. For example, suggest additional support options if negative sentiment is detected. Add a message to the generated answer such as "I see you're having trouble. Would you like me to connect you to support?"

[1227] 5. Providing answers

[1228] The optimized response is then sent back to the smartphone and displayed to the user, for example, "We usually ship within 3-5 business days of your order. If you need further assistance, please contact customer service."

[1229] Examples and prompts

[1230] Examples:

[1231] If a user types an inquiry such as, "I'm having trouble receiving my item. When will it arrive?", the AI ​​will respond with, "Items are usually shipped within 3 to 5 business days of your order. If you need further support, please contact customer service." If sentiment analysis determines that the user is expressing negative emotions, it will additionally display the message, "It seems you're having trouble. Would you like me to direct you to support?"

[1232] Example prompt for a generative AI model:

[1233] "A customer asks: 'I'm having trouble receiving my item. When will it arrive?' Please provide a suitable response to this question."

[1234] The present invention can improve the efficiency and quality of customer support at online shopping sites, thereby improving customer satisfaction.

[1235] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1236] Step 1:

[1237] User enters a query

[1238] The user uses the chat interface on their smartphone to input an inquiry. This input is used as the initial data for the system. For example, the user might input, "I'm having trouble because my product hasn't arrived yet. When will it arrive?"

[1239] Step 2:

[1240] The device sends a query to the server

[1241] The device sends the query entered by the user to the server, which then sends the text data to the server and prepares it for analysis.

[1242] Step 3:

[1243] The server analyzes the query

[1244] The server passes the received query content to the generative AI model. Specifically, it uses the openai library to provide the query content to the AI ​​model as a prompt.

[1245] Step 4:

[1246] Generative AI models generate answers

[1247] The generative AI model generates an appropriate answer based on the prompt text, and this data is returned to the server as generated text. For example, the answer generated might be, "We usually ship within 3-5 business days of your order."

[1248] Step 5:

[1249] The server performs sentiment analysis

[1250] At the same time, the server analyzes the sentiment of the received query using the textblob library, extracting positive or negative sentiment from the input text and providing the result (sentiment score).

[1251] Step 6:

[1252] The server optimizes the response

[1253] Based on the results of the sentiment analysis, the server optimizes the response. For example, if a negative sentiment is detected, the generated response will include a message such as, "It sounds like you're having trouble. Would you like me to connect you to support?"

[1254] Step 7:

[1255] Optimized answers sent to your device

[1256] The optimized answer is sent from the server to the device, where it is ready to be displayed to the user as the final answer.

[1257] Step 8:

[1258] The device displays the answer to the user

[1259] The device will display the optimized answer received from the server to the user, who can then refer to it to resolve the issue. For example, the answer might read, "We usually ship within 3-5 business days of your order. If you need further assistance, please contact customer service."

[1260] 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.

[1261] 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.

[1262] 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.

[1263] 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.

[1264] FIG. 9 illustrates 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 behaviors 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.

[1265] 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.

[1266] 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).

[1267] 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.

[1268] 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."

[1269] 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.

[1270] 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).

[1271] 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.

[1272] 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.

[1273] 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.

[1274] 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.

[1275] 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.

[1276] 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.

[1277] 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.

[1278] 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.

[1279] 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.

[1280] 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.

[1281] The following is further disclosed regarding the above embodiment.

[1282] (Claim 1)

[1283] a means for receiving customer inquiries;

[1284] A means for passing the received inquiry content to a generative AI model to generate an answer;

[1285] a means for providing the generated answer to the customer;

[1286] A means of analyzing the sentiment of customer inquiries,

[1287] A means for optimizing response content based on the results of sentiment analysis;

[1288] A system including:

[1289] (Claim 2)

[1290] 10. The system of claim 1,

[1291] A system in which the generative AI model uses natural language processing technology.

[1292] (Claim 3)

[1293] 10. The system of claim 1,

[1294] The system wherein the sentiment analysis means uses a sentiment analysis pipeline.

[1295] "Example 1"

[1296] (Claim 1)

[1297] a means for the terminal to receive inquiries from customers;

[1298] A means for passing the query received by the server to a generative AI model to generate an answer;

[1299] means for the server to provide the generated answer to the customer through the terminal;

[1300] A means for the server to analyze the sentiment of the customer's inquiry content;

[1301] A means for the server to optimize the response content based on the result of the sentiment analysis;

[1302] A system including:

[1303] (Claim 2)

[1304] 2. The system of claim 1, wherein the generative AI model uses natural language processing technology.

[1305] (Claim 3)

[1306] 10. The system of claim 1, wherein the means for sentiment analysis uses a sentiment analysis pipeline.

[1307] "Application Example 1"

[1308] (Claim 1)

[1309] a means for receiving customer inquiries;

[1310] A means for passing the received inquiry content to a generative AI model to generate an answer;

[1311] a means for providing the generated answer to the customer;

[1312] A means of analyzing the sentiment of customer inquiries,

[1313] A means for optimizing response content based on the results of sentiment analysis;

[1314] A means for displaying the inquiry content, the generated AI model's response, and the sentiment analysis results in real time using smart glasses;

[1315] A system including:

[1316] (Claim 2)

[1317] 2. The system of claim 1, wherein the generative AI model uses natural language processing technology.

[1318] (Claim 3)

[1319] 10. The system of claim 1, wherein the means for sentiment analysis uses a sentiment analysis pipeline.

[1320] "Example 2: Combining Emotion Engines"

[1321] (Claim 1)

[1322] a means for the terminal to receive inquiries from customers;

[1323] means for analyzing the received inquiry;

[1324] A means of passing the analysis results to a generative AI model to generate an answer;

[1325] means for transmitting the generated answer back to the terminal for presentation to the customer;

[1326] a means for analyzing the sentiment of customer inquiries and generated responses;

[1327] means for identifying emotions from the customer's voice data and text data using an emotion engine;

[1328] A means to optimize interactions based on sentiment analysis and offer additional support options when needed; and

[1329] A system including:

[1330] (Claim 2)

[1331] The system of claim 1, wherein the generative AI model uses natural language processing techniques.

[1332] (Claim 3)

[1333] 10. The system of claim 1, wherein the means for sentiment analysis uses a sentiment analysis algorithm.

[1334] "Application example 2 when combining emotion engines"

[1335] (Claim 1)

[1336] a means for receiving customer inquiries;

[1337] A means for passing the received inquiry content to a generative AI model to generate an answer;

[1338] a means for providing the generated answer to the customer;

[1339] A means of analyzing the sentiment of customer inquiries,

[1340] A means for optimizing response content based on the results of sentiment analysis;

[1341] means for receiving queries and displaying responses using a smartphone;

[1342] A system including:

[1343] (Claim 2)

[1344] 2. The system of claim 1, wherein the generative AI model uses natural language processing technology.

[1345] (Claim 3)

[1346] 10. The system of claim 1, wherein the means for sentiment analysis uses a sentiment analysis pipeline. [Explanation of symbols]

[1347] 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. a means for receiving customer inquiries; A means for passing the received inquiry content to a generative AI model to generate an answer; a means for providing the generated answer to the customer; A means of analyzing the sentiment of customer inquiries, A means for optimizing response content based on the results of sentiment analysis; A system including:

2. 10. The system of claim 1, A system in which the generative AI model uses natural language processing technology.

3. 10. The system of claim 1, The system wherein the sentiment analysis means uses a sentiment analysis pipeline.

Images