system

Abstract

We provide the system. [Solution] A means for receiving customer inquiries, A means for analyzing the aforementioned query content using natural language processing, A means for searching a database based on the analyzed query content, A means for generating an appropriate answer from the aforementioned search results, A means for sending the generated response to the customer, A means of escalating a customer inquiry to the support team if a suitable answer cannot be found, A system that includes this.

Description

Technical Field

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

Background Art

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

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In conventional customer support systems, it has been difficult to respond quickly and accurately to customer inquiries. In particular, when the analysis of the inquiry content and the generation of appropriate answers are performed manually, a lot of time and labor are required, imposing a heavy burden on the support team. Also, when appropriate information cannot be found, the escalation of the inquiry may be delayed, leading to a decrease in customer satisfaction. The present invention aims to provide a system that automatically analyzes the inquiry content, quickly generates appropriate answers, and further efficiently performs escalation as needed to solve these problems.

Means for Solving the Problems

[0005] The present invention provides a system comprising: means for receiving customer inquiries; means for analyzing the inquiry content using natural language processing; means for searching a database based on the analyzed inquiry content; means for generating an appropriate answer from the search results; means for sending the generated answer to the customer; and means for escalating the customer inquiry to a support team if no suitable answer is found. The natural language processing is characterized by extracting keywords from the inquiry content and determining the inquiry category based on these keywords. Furthermore, the escalation means includes notification means for providing detailed information about the inquiry to the support team. This enables prompt and accurate inquiry response, reduces the burden on the support team, and improves customer satisfaction.

[0006] "Customer inquiries" refer to questions and problems that customers submit through the system.

[0007] "Means of receiving" refers to the functions or devices that allow the system to receive inquiries sent by customers.

[0008] "Natural language processing" refers to the technologies and algorithms that enable computers to understand and analyze human language.

[0009] "Means of analysis" refers to the processes and functions used to understand and interpret the meaning of received inquiries.

[0010] A "database" refers to a collection of stored information that a system uses to search for answers to queries.

[0011] "Searching methods" refer to the functions and processes used to find appropriate information within a database based on the analyzed query content.

[0012] "Means of generating responses" refers to the functions and processes used to create response texts to provide to customers based on information obtained from a database.

[0013] "Means of transmission" refers to the function or device that sends data from the system to the customer in order to provide the customer with the generated response.

[0014] "Means of escalation" refers to functions or processes for notifying the support team of an issue when the inquiry cannot be resolved within the system.

[0015] "Keyword extraction" refers to the process of identifying and extracting important words and phrases from the content of an inquiry.

[0016] "Determining a category" refers to the process of assigning the inquiry content to the appropriate classification based on the extracted keywords.

[0017] "Notification means" refers to functions or devices that provide the support team with information about inquiries escalated from the system. [Brief explanation of the drawing]

[0018] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7]It is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] It is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] It shows an emotion map to which a plurality of emotions are mapped. [Figure 10] It shows an emotion map to which a plurality of emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

Mode for Carrying Out the Invention

[0019] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0020] First, the language used in the following description will be explained.

[0021] In the following embodiments, the signed processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Furthermore, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include CPU (Central Processing Unit), GPU (Graphics Processing Unit), GPGPU (General-Purpose computing on Graphics Processing Units), and APU (Accelerated Processing Unit).

[0022] In the following embodiments, signed RAM (Random Access Memory) is a memory that temporarily stores information and is used as work memory by the processor.

[0023] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0024] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0025] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0026] [First Embodiment]

[0027] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0028] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0029] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0030] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0031] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0032] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0033] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0034] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0035] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0036] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0037] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0038] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0039] As an example of a system according to the present invention, a system for automatically processing customer inquiries and generating appropriate responses will be described.

[0040] System Overview

[0041] This system includes the following main features:

[0042] 1. Receipt of inquiry

[0043] 2. Analysis of the inquiry content using natural language processing (NLP)

[0044] 3. Database Search

[0045] 4. Generating the answer

[0046] 5. Submit your response

[0047] 6. Escalation as needed

[0048] Processing flow

[0049] First, the user submits an inquiry using the support chat. For example, let's say the inquiry is, "How do I return a product?"

[0050] The server receives the inquiry sent by the user. It then analyzes the received string using a natural language processing (NLP) library. In this case, keywords such as "product," "return," and "method" are extracted from the text.

[0051] Next, the server uses the extracted keywords to search its internal database. This search utilizes an indexed database for faster information retrieval.

[0052] Once search results are obtained, the server generates a response to provide to the user based on them. For example, it might retrieve information from the database such as, "To return a product, please return it in its original packaging within 30 days of purchase," and then format this into a response.

[0053] The generated response is sent from the server to the user. The user can then view the response displayed on the support chat screen.

[0054] If the server cannot find a suitable answer, or if the inquiry is too complex, the system will use an escalation mechanism. In this case, the server will notify the support team of the details of the inquiry.

[0055] Specific example

[0056] 1. User inquiry: "How do I return an item?"

[0057] 2. Server processing:

[0058] Inquiry received

[0059] Analysis using NLP

[0060] Extract the keywords "product," "return," and "method."

[0061] Search database

[0062] The response generated reads: "To return an item, please return it in its original packaging within 30 days of purchase."

[0063] 3. Server submission: Send the generated response to the user.

[0064] 4. Escalation (additional example):

[0065] User inquiry: "How do I return an item from overseas?"

[0066] Server processing:

[0067] Inquiry received

[0068] Analysis using NLP

[0069] I searched the database but couldn't find a suitable answer.

[0070] Server escalation: Notify the support team of the inquiry details.

[0071] Support team: We will review your inquiry and contact you directly to provide further details.

[0072] This embodiment allows the system to efficiently respond to customer inquiries. Furthermore, it can improve customer satisfaction by promptly escalating cases when a suitable answer cannot be found.

[0073] The following describes the processing flow.

[0074] Step 1:

[0075] User: Open the support chat screen and enter your inquiry into the text box. Type "How do I return a product?" and click the send button.

[0076] Step 2:

[0077] Server: Receives queries sent by users. The server stores the received query content in an appropriate format.

[0078] Step 3:

[0079] Server: Passes the received query content to a natural language processing (NLP) library. The NLP library performs grammatical analysis and word segmentation of the text and extracts the main keywords (in this example, "product," "return," and "method").

[0080] Step 4:

[0081] Server: Searches the internal database based on the extracted keywords. Generates search queries and quickly accesses the database index to retrieve the appropriate information.

[0082] Step 5:

[0083] Server: Based on information retrieved from the database, it generates a response to provide to the user. This response is formatted as a natural-sounding sentence. For example, it might generate a response such as, "To return an item, please return it in its original packaging within 30 days of purchase."

[0084] Step 6:

[0085] Server: Sends the generated response to the user's support chat screen. Displays the response appropriately so that the user can review it.

[0086] Step 7:

[0087] User: Review the response displayed on the support chat screen and make additional inquiries if necessary.

[0088] Step 8:

[0089] Server: If a suitable answer cannot be found, or if the inquiry is complex and difficult to handle automatically, the server initiates escalation. It sets an escalation flag and logs the inquiry details and analysis results.

[0090] Step 9:

[0091] Server: Generates an escalation request and sends a notification to the support team. The notification includes detailed information about the inquiry.

[0092] Step 10:

[0093] Terminal (e.g., support team's PC): An escalation notification will appear on the terminal of a support team member. The support team will review the notification and begin taking detailed action.

[0094] Step 11:

[0095] Support Team: Review the details of the escalated inquiry and contact the user directly to resolve the issue.

[0096] (Example 1)

[0097] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0098] Traditional customer support systems consume significant time and resources by manually analyzing inquiries and generating appropriate responses. This can lead to delays in customer service and decreased customer satisfaction. Furthermore, escalation processes are often manual when appropriate answers cannot be found, adding to the overall time burden. A system is needed to address these challenges and provide rapid and accurate responses to customer inquiries.

[0099] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0100] In this invention, the server includes means for receiving user inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the user, and means for escalating the user's inquiry to a support team if no suitable answer is found. This makes it possible to respond to user inquiries quickly and automatically.

[0101] "User" refers to a person who makes inquiries to the system or an end-user.

[0102] "Inquiry content" refers to messages containing questions and requests that users send to the system.

[0103] A "server" refers to a hardware or software system that receives inquiries from users and performs processes such as analysis, searching, response generation, transmission, and escalation.

[0104] Natural Language Processing (NLP) refers to the techniques and algorithms used to analyze human language and understand its meaning.

[0105] A "database" refers to a collection of structured data designed for efficient searching and management of information.

[0106] "Means of searching" refers to a mechanism or process for retrieving information from a database based on a specific query.

[0107] "Means of generating answers" refers to the process of creating appropriate answer text to provide to the user based on the searched information.

[0108] "Means of transmission" refers to the communication technologies and protocols used to send the generated response to the user's device.

[0109] "Escalation means" refers to the process of forwarding an inquiry to other resources, such as a support team, when the system cannot find a suitable answer.

[0110] A "support team" refers to a group of individuals or individuals who provide additional human support in response to user inquiries.

[0111] As an example of a system according to the present invention, a system for automatically processing customer inquiries and generating appropriate responses will be described.

[0112] System Overview

[0113] This system includes the following main components:

[0114] 1. Means for receiving inquiries from users

[0115] 2. Means for analyzing received inquiry content using natural language processing (NLP)

[0116] 3. Means for searching the database based on the analyzed query content

[0117] 4. Means for generating appropriate answers from search results

[0118] 5. Means for sending the generated response to the user

[0119] 6. A means of escalating a user's inquiry to the support team if a suitable answer cannot be found.

[0120] Hardware and software usage examples

[0121] 1. Receiving means

[0122] Users submit inquiries via support chat using devices such as smartphones or computers. The server receives these inquiries over the internet.

[0123] 2. Natural Language Processing (NLP)

[0124] The server analyzes the received query using a natural language processing (NLP) library (e.g., Python's NLTK or SpaCy). Specifically, it divides the text into tokens and extracts keywords.

[0125] 3. Database Search

[0126] The server searches indexed databases (e.g., ElasticSearch® or MySQL®) using the extracted keywords. This search allows for the rapid retrieval of appropriate information in response to user inquiries.

[0127] 4. Answer generation means

[0128] The server generates answers to provide to the user based on the search results. For example, it organizes information obtained from the database into an easily understandable format.

[0129] 5. Transmission method

[0130] The server sends the generated response to the user's device. The user can then view the response on the support chat screen.

[0131] 6. Escalation measures

[0132] If the server cannot find a suitable answer, it will notify the support team of the inquiry details. This notification allows the support team to handle the inquiry manually.

[0133] Specific example

[0134] The following shows a specific example of operation.

[0135] 1. User inquiry: "How do I return an item?"

[0136] 2. Server processing:

[0137] Inquiry received

[0138] Analysis using NLP

[0139] Extract the keywords "product," "return," and "method."

[0140] Search database

[0141] The response generated reads: "To return an item, please return it in its original packaging within 30 days of purchase."

[0142] 3. Server submission: Send the generated response to the user.

[0143] If it is incorrect, we will escalate it as follows:

[0144] 1. User inquiry: "How do I return an item from overseas?"

[0145] 2. Server processing:

[0146] Inquiry received

[0147] Analysis using NLP

[0148] I searched the database but couldn't find a suitable answer.

[0149] 3. Server escalation: Notify the support team of the inquiry details.

[0150] 4. Support Team: They will review the inquiry and contact the user directly to resolve the issue.

[0151] Example of a prompt

[0152] By inputting prompt messages like the following into the AI ​​model, it generates appropriate answers to inquiries.

[0153] "Please tell me how to return an item."

[0154] "Please tell me how to return items from overseas."

[0155] This allows the system to respond to customer inquiries efficiently and automatically. Furthermore, if a suitable answer cannot be found, it can quickly escalate the issue, improving customer satisfaction.

[0156] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0157] Step 1:

[0158] A user submits an inquiry using the support chat. An example of the inquiry is, "How do I return a product?"

[0159] Input: User inquiry message (text format)

[0160] Output: JSON data containing the query details

[0161] Specific action: The user enters their inquiry into the chat window and presses the send button.

[0162] Step 2:

[0163] The server receives the user's inquiry.

[0164] Input: JSON data containing the user's inquiry message

[0165] Output: Received inquiry content (string format)

[0166] Specific operation: The server receives the query in real time and extracts the query string for analysis.

[0167] Step 3:

[0168] The server analyzes the received query content using a natural language processing (NLP) library.

[0169] Input: Received inquiry content (string format)

[0170] Output: Extracted keywords (list format)

[0171] Specific operation: The server uses a Python NLP library (e.g., NLTK, SpaCy) to tokenize the text and extract important keywords (e.g., "product", "return", "method").

[0172] Step 4:

[0173] The server searches the database using the extracted keywords.

[0174] Input: Extracted keywords (list format)

[0175] Output: Search results (information from the database)

[0176] Specific operation: The server uses an indexed database (e.g., Elasticsearch, MySQL) to execute queries to retrieve information that matches the keyword.

[0177] Step 5:

[0178] The server generates answers to provide to the user based on the search results.

[0179] Input: Search results (information from the database)

[0180] Output: Generated response (text format)

[0181] Specific operation: The server generates an appropriate response based on information retrieved from the database (e.g., "To return the product, please return it in its original packaging within 30 days of purchase").

[0182] Step 6:

[0183] The server sends the generated response to the user.

[0184] Input: Generated response (text format)

[0185] Output: Display of response to user terminal

[0186] Specific operation: The server sends the response to the user's chat window, where it is displayed on the user's device.

[0187] Step 7:

[0188] If the server cannot find a suitable answer, the user's inquiry will be escalated.

[0189] Input: If the search result is empty

[0190] Output: Notification to the support team

[0191] Specific actions: The server detects that there is no suitable answer and sends a notification to the support team containing details of the inquiry. It also sends a message to the user such as, "We're sorry, but we were unable to find a suitable answer at this time."

[0192] In this way, the system can respond to user inquiries efficiently and automatically, and can quickly escalate cases if a suitable answer cannot be found.

[0193] (Application Example 1)

[0194] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0195] Traditional customer inquiry systems often rely on manual processes for analyzing inquiries and generating responses, resulting in low efficiency. Furthermore, the difficulty of easily submitting inquiries via smartphones hinders the provision of adequate support in today's fast-paced customer service environment. In addition, escalation options for complex customer inquiries are limited, highlighting the need for further operational efficiency and improved customer satisfaction.

[0196] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0197] In this invention, the server includes means for receiving customer inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the customer, means for escalating the customer's inquiry to a support team if no suitable answer is found, means for notifying the customer of detailed information about the inquiry during the escalation, means for easily inputting customer inquiries using a smartphone, means for using a generative AI model to automatically extract keywords from the inquiry and generate an answer based on them, and means for displaying the generated answer as a prompt. This enables the automation and efficiency of inquiry processing, allows customers to make inquiries intuitively using their smartphones, and enhances the escalation means, thereby realizing prompt and accurate customer support.

[0198] A "customer" refers to someone who purchases or uses a product or service.

[0199] "Inquiry content" refers to questions and requests provided by customers.

[0200] "Means of receiving information" refers to the methods and mechanisms by which the system receives customer inquiries.

[0201] "Natural language processing" refers to the technology that allows computers to analyze and understand natural language data.

[0202] A "database" refers to a system in which information is systematically stored and can be searched and managed.

[0203] "Means of searching" refers to the methods and mechanisms for obtaining necessary information from a database.

[0204] "Means of generating answers" refers to methods and mechanisms for creating appropriate answers to provide to customers based on search results.

[0205] "Means of transmission" refers to the methods and mechanisms used to deliver the generated responses to customers.

[0206] "Means of escalation" refers to methods and mechanisms for transferring inquiries that the system cannot handle to the support team.

[0207] "Means of notification" refers to the methods and mechanisms for informing the support team of the details of an inquiry during an escalation.

[0208] A "smartphone" refers to a portable information device that has telephone functionality while also being able to run applications and connect to the internet.

[0209] "Means for easily submitting inquiries" refers to methods and systems that allow customers to easily submit inquiries using devices such as smartphones.

[0210] "Methods for automatically extracting keywords" refers to methods and mechanisms that use natural language processing to extract important terms from query content.

[0211] A "generative AI model" refers to a model that uses artificial intelligence technology to automatically generate answers based on the content of an inquiry.

[0212] A "prompt message" refers to the guidance or response message provided to the customer based on the generated answer.

[0213] System Overview

[0214] The system for implementing this invention automates a series of processes and can respond quickly and accurately to customer inquiries. It mainly uses a server, a smartphone terminal, a natural language processing (NLP) library, a generative AI model, and a database.

[0215] Hardware and software

[0216] Server: The central hardware that receives, processes, parses, generates answers for, and escalates queries.

[0217] Smartphone: A device used by customers to enter inquiries.

[0218] NLP libraries: Software used to analyze natural language. A specific example is spaCy ("ja_core_news_sm" model).

[0219] Generative AI Model: An artificial intelligence model that generates appropriate answers from the content of an inquiry.

[0220] Database: Stores frequently occurring queries and their answers, enabling rapid data retrieval.

[0221] System Implementation Method

[0222] First, the user enters their inquiry using their smartphone. For example, the user might ask, "When will the product I ordered arrive?"

[0223] The server receives this query and analyzes it using an NLP library. Specifically, it extracts keywords such as "order," "product," and "deliver" from the query text. This clarifies the category and intent of the query.

[0224] Next, the database is searched based on the extracted keywords. The database contains pre-indexed response data, which can be searched efficiently. For example, suppose the response "Ordered items are usually shipped within 3-5 business days" is found.

[0225] Based on the responses found, a generative AI model generates answers. This process uses AI to generate more specific and appropriate responses to the original inquiry.

[0226] The generated response is sent from the server to the smartphone device, where the user confirms it. For example, the response might say, "Your order will usually be shipped within 3-5 business days."

[0227] If the server cannot find a suitable answer, the system escalates the inquiry details to the support team. The support team is also notified of the detailed inquiry, allowing them to respond quickly.

[0228] Specific example

[0229] 1. User inquiry: "How do I return an item?"

[0230] 2. Server processing:

[0231] Inquiry received

[0232] Analysis using NLP

[0233] Extract the keywords "product," "return," and "method."

[0234] Search database

[0235] I received the response, "To return the product, please return it in its original packaging within 30 days of purchase."

[0236] 3. Submit the generated response:

[0237] The response sent to the smartphone was: "To return the product, please return it in its original packaging within 30 days of purchase."

[0238] 4. Examples of escalation:

[0239] User inquiry: "How do I return an item from overseas?"

[0240] Server processing:

[0241] Inquiry received

[0242] Analysis using NLP

[0243] I searched the database but couldn't find a suitable answer.

[0244] Server escalation:

[0245] Please notify the support team of the details of your inquiry.

[0246] Support team:

[0247] We will review the inquiry and contact the user directly to provide further details.

[0248] Example of a prompt

[0249] "Question to the Generative AI Model: A user has submitted the following inquiry: 'When will my ordered item arrive?' The keywords analyzed using NLP are 'order,' 'item,' and 'arrive.' Please provide as much specific information as possible regarding the delivery timing."

[0250] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0251] Step 1:

[0252] The user enters their inquiry using their smartphone and presses the send button. Specifically, the user might enter "When will my ordered items arrive?". This input is the user's inquiry.

[0253] Step 2:

[0254] The smartphone device sends the entered inquiry content to the server. The input is the user's inquiry, and the output is the data sent to the server.

[0255] Step 3:

[0256] The server uses an NLP library to analyze the received query content. Specifically, it uses the spaCy "ja_core_news_sm" model to extract keywords from the query content. In this case, the input is the query content, and the output is the extracted keywords (such as "order," "product," and "deliver").

[0257] Step 4:

[0258] The server searches the database based on the extracted keywords. The database stores pre-indexed response data. The input is the extracted keywords, and the output is the corresponding response data (e.g., "Your order will be shipped within 3-5 business days").

[0259] Step 5:

[0260] The server generates an answer using a generative AI model based on the search results. In this process, the AI ​​model generates the optimal answer based on the query and search results. The input is the search results and the query, and the output is the generated answer.

[0261] Step 6:

[0262] The server sends the generated response text to the smartphone device. The input is the generated response text, and the output is the data sent to the smartphone device.

[0263] Step 7:

[0264] The smartphone displays the response text received from the server to the user. The input is the response data from the server, and the output is the response text displayed on the smartphone screen.

[0265] Step 8:

[0266] If the server cannot find a suitable answer, the issue will be escalated to the support team. Specifically, a notification containing detailed information about the inquiry will be sent to the support team. The input is the inquiry requiring support, and the output is the notification data sent to the support team.

[0267] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0268] The system according to the present invention is a system for automatically processing user inquiries and generating appropriate responses. Furthermore, by combining it with an emotion engine that recognizes user emotions and adjusts responses based on those emotions, it achieves more accurate customer service.

[0269] System Overview

[0270] This system includes the following main features:

[0271] 1. Receipt of inquiry

[0272] 2. Analysis of the inquiry content using natural language processing (NLP)

[0273] 3. User emotion recognition by an emotion engine

[0274] 4. Database Search

[0275] 5. Generating and adjusting responses

[0276] 6. Submit your response

[0277] 7. Escalation as needed

[0278] Processing flow

[0279] First, the user submits an inquiry using the support chat. For example, let's say the inquiry is, "How do I return a product?"

[0280] The server receives inquiries sent by users. It then passes the received inquiry content to a natural language processing (NLP) library for analysis. Here, keywords such as "product," "return," and "method" are extracted from the text.

[0281] Next, the server uses its built-in emotion engine to analyze the user's emotions in the inquiry. This analysis determines whether the user is feeling, for example, dissatisfied ("negative") or grateful ("positive").

[0282] Based on the parsed keywords and sentiment data, the server searches the internal database to obtain appropriate information. Based on the retrieved information, an answer text to be provided to the user is generated. At this time, the sentiment engine adjusts the answer based on the user's sentiment. For example, when the user is angry, a more polite and intimate style is used.

[0283] The generated answer is sent from the server to the user and displayed on the support chat screen. The user checks the presented answer and makes additional inquiries if necessary.

[0284] If the server cannot find an appropriate answer or if the user's sentiment is very negative, the system uses escalation means. In this case, the server notifies the support team of the details of the inquiry and the sentiment analysis results.

[0285] Specific Example

[0286] 1. User's inquiry: "Please tell me how to return the product."

[0287] 2. Server's processing:

[0288] Receive the inquiry

[0289] Analyze using NLP

[0290] Extract keywords "product", "return", "method"

[0291] Recognize that the user is "confused" by the sentiment engine

[0292] Search the database

[0293] Generate an answer text "The method to return the product is to send it back in the original package within 30 days of purchase."

[0294] Add a supplement such as "We apologize for the inconvenience" to the answer text

[0295] 3. Server Transmission: Send the adjusted response to the user

[0296] 4. Escalation (Additional Example):

[0297] User Inquiry: "The quality of this product is too bad. I want to return it."

[0298] Server Processing:

[0299] Receive the inquiry

[0300] Analyze with NLP

[0301] Extract keywords "product", "quality", "return"

[0302] Recognize that the user is feeling "angry" with the emotion engine

[0303] Search the database but no appropriate answer is found

[0304] Server Escalation: Notify the support team of the inquiry details and the emotion recognition result of "anger"

[0305] Support Team: Check the inquiry and contact the user directly to resolve the problem

[0306] With this embodiment, the system can handle inquiries efficiently and consider emotions. Also, by quickly escalating when no appropriate answer is found or when the user's emotion is negative, customer satisfaction can be improved.

[0307] The following describes the processing flow.

[0308] Step 1:

[0309] User: Open the support chat screen and enter your inquiry into the text box. Type "How do I return a product?" and click the send button.

[0310] Step 2:

[0311] Server: Receives queries sent by users. The server stores the received query content in an appropriate format.

[0312] Step 3:

[0313] Server: Passes the received query content to a natural language processing (NLP) library. The NLP library performs grammatical analysis and word segmentation of the text and extracts the main keywords (in this example, "product," "return," and "method").

[0314] Step 4:

[0315] Server: Based on the extracted keywords, the server passes the query to the emotion engine. The emotion engine analyzes the user's emotions from the text and identifies emotions such as confusion, anger, and joy. In this example, the user's emotion is recognized as "confused".

[0316] Step 5:

[0317] Server: Searches the internal database based on extracted keywords and sentiment recognition results. Generates search queries and accesses the database index to retrieve relevant information.

[0318] Step 6:

[0319] Server: Generates a response to provide to the user based on information retrieved from the database. This response is formatted in a natural style and adjusted based on the results of the sentiment engine. For example, polite expressions may be added, such as, "To return the product, please return it in its original packaging within 30 days of purchase. We apologize for any inconvenience this may cause."

[0320] Step 7:

[0321] Server: Sends the generated response to the user's support chat screen. Displays the response appropriately so that the user can review it.

[0322] Step 8:

[0323] User: Review the response displayed on the support chat screen and make additional inquiries if necessary.

[0324] Step 9:

[0325] Server: If a suitable answer cannot be found, or if the inquiry is complex and difficult to handle automatically, the server will initiate escalation. Escalation will also occur if the user's emotions are very negative.

[0326] Step 10:

[0327] Server: Set the escalation flag and log the query content and sentiment analysis results.

[0328] Step 11:

[0329] Server: Generates an escalation request and sends a notification to the support team. The notification includes the inquiry details and sentiment analysis results.

[0330] Step 12:

[0331] Terminal (e.g., support team's PC): An escalation notification will appear on the terminal of a support team member. The support team will review the notification and begin taking detailed action.

[0332] Step 13:

[0333] Support Team: Review the details of the escalated inquiry and contact the user directly to resolve the issue.

[0334] (Example 2)

[0335] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0336] Traditional customer support systems analyze inquiries and generate appropriate responses, but they fail to consider customer emotions, resulting in insufficient improvement in customer satisfaction. Furthermore, delays in escalation when an appropriate answer cannot be found can lead to service delays.

[0337] The identification processing performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for receiving customer inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the customer, means for escalating the customer's inquiry to the support team if no suitable answer is found, means for analyzing the customer's emotions included in the inquiry, and means for adjusting the answer based on the analyzed customer emotions. This makes it possible to respond in a way that takes customer emotions into consideration, thereby improving customer satisfaction. In addition, if an appropriate answer cannot be found, it is quickly escalated to the support team, preventing delays in response.

[0338] "Customer" refers to anyone who uses the system to receive services or support.

[0339] "Inquiry details" refers to the text information of questions or requests that customers enter or submit when they require support.

[0340] "Means of receiving information" refers to the function for importing customer inquiries into the system.

[0341] "Natural language processing" refers to the technology that enables computers to understand and analyze text written in human language.

[0342] "Means of analysis" refers to the function that uses natural language processing to analyze the meaning and intent of received inquiries.

[0343] "Means of searching the database" refers to the function of searching for relevant information from the internal database based on the analyzed query content.

[0344] "Means of generating responses" refers to a function that creates responses tailored to the customer based on information obtained from a database.

[0345] "Means of transmission" refers to the function used to communicate the generated response to the customer.

[0346] "Escalation options" refer to the function that allows you to transfer your inquiry to the support team if you cannot find a suitable answer or under certain conditions.

[0347] "Methods for analyzing emotions" refers to technologies used to identify customer emotions from the content of inquiries.

[0348] "Means of adjusting responses" refers to a function that optimizes the content and style of responses based on analyzed customer sentiment.

[0349] The system according to the present invention automatically processes customer inquiries and generates appropriate responses. Furthermore, by having a function to recognize customer emotions and adjust responses accordingly, it enables more accurate customer service.

[0350] System Configuration

[0351] 1. Receiving customer inquiries

[0352] Users access the support chat using their device and enter their questions or requests. At this time, the customer's inquiry is sent to the system server.

[0353] 2. Natural Language Processing (NLP) of the inquiry content

[0354] The server uses an NLP library (e.g., SpaCy or NLTK) to analyze the received query. Specifically, it performs tokenization of the text (splitting it into individual words), part-of-speech tagging, and keyword extraction. This analysis extracts the important keywords from the query.

[0355] 3. Recognizing customer emotions

[0356] The server uses an emotion engine (such as Google® Cloud Natural Language API or IBM Watson® Tone Analyzer) to analyze the customer's emotions from the inquiry. For example, emotions such as "confused," "angry," and "grateful" may be recognized.

[0357] 4. Searching the internal database

[0358] The server searches its internal database for relevant information based on the analyzed keywords and sentiment data. This process efficiently retrieves information using SQL queries.

[0359] 5. Generating and adjusting responses

[0360] The server generates appropriate responses based on database search results using a generative AI model (e.g., OpenAI® GPT-3®). It also adjusts the style and content of the responses based on the results of the sentiment engine. For example, if a customer is confused, a polite clarification such as "We apologize for the inconvenience" may be added.

[0361] 6. Submit your response

[0362] The server sends the generated response to the user and displays it on the support chat screen. This allows the user to see the answer immediately.

[0363] 7. Escalation as needed

[0364] If the server cannot find a suitable answer or if the customer's sentiment is very negative, it will notify the support team of the inquiry details and sentiment analysis results. This allows the support team to respond quickly.

[0365] Specific example

[0366] Specific Example 1

[0367] User inquiry: "How do I return an item?"

[0368] Processing flow:

[0369] The user enters and submits their inquiry.

[0370] The server receives the query.

[0371] The server uses NLP to extract keywords such as "product," "return," and "method."

[0372] The server uses an emotion engine to recognize that the user is "confused."

[0373] The server searches the database and generates a response message stating, "To return an item, please return it in its original packaging within 30 days of purchase."

[0374] The server adds a note to the response such as, "We apologize for any inconvenience this may cause."

[0375] The server sends the adjusted response to the user.

[0376] Specific example 2 (escalation)

[0377] User inquiry: "The quality of this product is terrible. I want to return it."

[0378] Processing flow:

[0379] The user enters and submits their inquiry.

[0380] The server receives the query.

[0381] The server uses NLP to analyze the data and extract information on "products," "quality," and "returns."

[0382] The server uses an emotion engine to recognize that the user is feeling "anger."

[0383] The server searches the database but cannot find a suitable answer.

[0384] The server notifies the support team of the inquiry details and the result of the "anger" emotion recognition.

[0385] The support team will review the inquiry and contact the user directly to resolve the issue.

[0386] This invention enables the system to handle inquiries efficiently and with consideration for the user's emotions. By promptly escalating cases when an appropriate answer cannot be found or when the user's emotions are negative, customer satisfaction can be improved.

[0387] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0388] Step 1:

[0389] The user accesses the support chat using their device, enters their inquiry, and submits it. For example, the text entered in the inquiry might be, "How do I return a product?" The entered data is then sent to the server.

[0390] Step 2:

[0391] The server receives the query. The received data is logged as is and passed on to the next analysis process. The input is text data, and the output is similarly text data for analysis.

[0392] Step 3:

[0393] The server uses an NLP library (e.g., SpaCy or NLTK) to analyze the received text data. Specifically, it performs text tokenization, part-of-speech tagging, and keyword extraction. The input is the received text data, and the output is a list of keywords (e.g., "product," "return," "method").

[0394] Step 4:

[0395] The server uses an internal sentiment engine (e.g., Google Cloud Natural Language API or IBM Watson Tone Analyzer) to analyze the user's sentiment in the query. The input is received text data, and the output is sentiment tags (e.g., "confused").

[0396] Step 5:

[0397] The server searches its internal database based on the analyzed keywords and sentiment data. Specifically, it uses SQL queries to retrieve relevant information. The input is a list of keywords and sentiment tags, and the output is the appropriate response data.

[0398] Step 6:

[0399] The server uses a generative AI model (e.g., OpenAI GPT-3) to generate appropriate responses based on the search results. Here, the responses are refined based on the results of the sentiment engine. The input is the response data and sentiment tags, and the output is the refined response text (e.g., "To return the product, please return it in its original packaging within 30 days of purchase. We apologize for any inconvenience.").

[0400] Step 7:

[0401] The server sends the generated response to the user. The response is displayed on the support chat screen, where the user can review it. The input is the edited response, and the output is what is displayed on the user's screen.

[0402] Step 8:

[0403] If a suitable answer cannot be found, or if the user's sentiment is very negative, the server escalates the inquiry details and sentiment analysis results to the support team. The input is the inquiry details and sentiment tags obtained from the previous processing, and the output is a notification to the support team.

[0404] (Application Example 2)

[0405] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0406] While conventional customer service systems can automatically generate appropriate responses to inquiries, they struggle to improve customer satisfaction because they cannot consider customer emotions. Furthermore, the system lacks the flexibility to adapt to situations requiring changes in writing style based on emotions or, in some cases, rapid escalation.

[0407] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for receiving customer inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the customer, means for escalating the customer's inquiry to a support team if no suitable answer is found, means for recognizing the customer's emotions regarding the inquiry, and means for adjusting the generated answer based on the recognized emotions. This enables automated responses that take emotions into consideration and rapid escalation.

[0408] A "customer" is a user who makes an inquiry using this system.

[0409] "Inquiry content" refers to questions and requests that customers submit through this system.

[0410] "Natural language processing" is a technology that enables computers to understand and analyze human language.

[0411] "Analyzing" means converting input data into a format that can be understood, either partially or entirely.

[0412] A "database" is an electronic information storage system that systematically accumulates information.

[0413] "Generating" means creating new data or information according to a certain algorithm or rule.

[0414] "To send" means to transfer specific data from one location to another.

[0415] "Escalation" is the act of passing on a difficult problem to a higher-ranking or specialized person.

[0416] A "support team" is a group of specialists responsible for handling customer inquiries and resolving problems.

[0417] "Recognizing emotions" means analyzing customer inquiries and identifying the types of emotions they contain.

[0418] "Adjusting" means appropriately changing the information and responses provided to match the perceived emotions.

[0419] A "system" is a mechanism that provides a certain function through the coordinated action of multiple components or means.

[0420] The system for implementing this invention analyzes customer inquiries, recognizes their emotions, and generates and adjusts appropriate responses based on those emotions in a content distribution service using smartphones. The specific operation of the system will be described below.

[0421] First, the user submits an inquiry using a smartphone app. The user's inquiry is in the format of, for example, "Please tell me about the latest TV dramas." The server receives this inquiry and analyzes it using a natural language processing (NLP) library. During this process, SpaCy is used to extract keywords from the inquiry and identify the subject of the inquiry.

[0422] Next, the server uses VADER Sentiment to perform sentiment analysis on the inquiry and recognizes the user's sentiment as either "positive," "negative," or "neutral." Based on this sentiment data and extracted keywords, the server searches the SQLite database and retrieves the relevant response. The retrieved response is then adjusted by the sentiment engine based on the user's sentiment. For example, if the user is "positive," the response will be written in a friendly style.

[0423] The adjusted response is sent from the server to the user's smartphone and displayed on the app screen. If a suitable response does not exist in the database, or if the user's sentiment is very strong and negative, the server escalates the inquiry to the support team. During this escalation process, the support team is notified of the detailed inquiry and the results of the sentiment analysis.

[0424] Hardware and software to be used

[0425] Hardware: Smartphone (iOS or Android®)

[0426] Software: Flask (Python framework), SpaCy (natural language processing library), VADER Sentiment (sentiment analysis library), SQLite (database)

[0427] Specific example

[0428] For example, if a user sends an inquiry asking "Please tell me the latest TV dramas," the server will process it as follows:

[0429] 1. Receive user inquiries.

[0430] 2. Analyze the inquiries and extract the keywords "latest" and "drama".

[0431] 3. Conduct sentiment analysis and recognize the user's emotions as "neutral."

[0432] 4. Search the database and obtain the answer, "The latest drama is XX."

[0433] 5. Adjust the response based on the user's emotions and generate an answer in the format "The latest drama is XX. (User's emotion: Neutral)".

[0434] 6. Send the generated response to the user and display it on the smartphone app screen.

[0435] Example of a prompt

[0436] 1. Inquiry: "Please tell me about the latest TV dramas."

[0437] 2. Answer: "The latest drama is XX. (User sentiment: Neutral)"

[0438] This makes it possible to provide automated responses to user inquiries that take emotions into consideration.

[0439] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0440] Step 1:

[0441] Users submit their inquiries through a smartphone app. The inquiry content is provided in text format. At this point, users ask specific questions such as, "Please tell me about the latest TV dramas." The output is the text data of the inquiry.

[0442] Step 2:

[0443] The server processes the received query. First, it passes the query to a natural language processing (NLP) library. The SpaCy library is used for this purpose. The input is the query text received in step 1, and the output is a list of keywords. Specifically, keywords such as "latest" and "drama" are extracted from the query.

[0444] Step 3:

[0445] The server performs sentiment analysis on the extracted keywords. The VADER Sentiment library is used for this purpose. The input consists of the keywords extracted in step 2 and the original query. The output is a sentiment classification (positive, negative, or neutral). For example, the query "Please tell me the latest TV dramas" would be classified as "neutral."

[0446] Step 4:

[0447] The server searches the database based on the extracted keywords and sentiment analysis results. It executes queries against the SQLite database to retrieve the relevant answers. The inputs are the keywords from step 2 and the sentiment classification results from step 3. The output is the corresponding answer text. Specific answers such as "The latest drama is XX" can be obtained.

[0448] Step 5:

[0449] The server adjusts the acquired responses based on the sentiment analysis results. This process involves changing the writing style and adding supplementary information according to the sentiment. The input is the response text acquired in step 4 and the sentiment classification result obtained in step 3. The output is the adjusted response text. For example, the response might reflect a "neutral" sentiment recognition, resulting in "The latest drama is XX. (User's sentiment: Neutral)."

[0450] Step 6:

[0451] The server sends the adjusted response to the user. To do this, it uses the Flask framework to return the response as an HTTP response. The input is the response text adjusted in step 5, and the output is the response displayed on the user's smartphone.

[0452] Step 7:

[0453] If a suitable answer cannot be retrieved from the database, or if the sentiment analysis indicates very strong negative emotions, the server escalates the inquiry to the support team. The input consists of the sentiment analysis results obtained in step 3 and the database search results that were deemed irrelevant in step 4. The output is a notification to the support team, which includes the inquiry details and the sentiment analysis results.

[0454] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0455] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0456] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0457] [Second Embodiment]

[0458] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0459] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0460] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0461] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0462] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0463] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0464] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0465] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0466] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0467] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0468] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0469] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0470] As an example of a system according to the present invention, a system for automatically processing customer inquiries and generating appropriate responses will be described.

[0471] System Overview

[0472] This system includes the following main features:

[0473] 1. Receipt of inquiry

[0474] 2. Analysis of the inquiry content using natural language processing (NLP)

[0475] 3. Database Search

[0476] 4. Generating the answer

[0477] 5. Submit your response

[0478] 6. Escalation as needed

[0479] Processing flow

[0480] First, the user submits an inquiry using the support chat. For example, let's say the inquiry is, "How do I return a product?"

[0481] The server receives the inquiry sent by the user. It then analyzes the received string using a natural language processing (NLP) library. In this case, keywords such as "product," "return," and "method" are extracted from the text.

[0482] Next, the server uses the extracted keywords to search its internal database. This search utilizes an indexed database for faster information retrieval.

[0483] Once search results are obtained, the server generates a response to provide to the user based on them. For example, it might retrieve information from the database such as, "To return a product, please return it in its original packaging within 30 days of purchase," and then format this into a response.

[0484] The generated response is sent from the server to the user. The user can then view the response displayed on the support chat screen.

[0485] If the server cannot find a suitable answer, or if the inquiry is too complex, the system will use an escalation mechanism. In this case, the server will notify the support team of the details of the inquiry.

[0486] Specific example

[0487] 1. User inquiry: "How do I return an item?"

[0488] 2. Server processing:

[0489] Inquiry received

[0490] Analysis using NLP

[0491] Extract the keywords "product," "return," and "method."

[0492] Search database

[0493] The response generated reads: "To return an item, please return it in its original packaging within 30 days of purchase."

[0494] 3. Server submission: Send the generated response to the user.

[0495] 4. Escalation (additional example):

[0496] User inquiry: "How do I return an item from overseas?"

[0497] Server processing:

[0498] Inquiry received

[0499] Analysis using NLP

[0500] I searched the database but couldn't find a suitable answer.

[0501] Server escalation: Notify the support team of the inquiry details.

[0502] Support team: We will review your inquiry and contact you directly to provide further details.

[0503] This embodiment allows the system to efficiently respond to customer inquiries. Furthermore, it can improve customer satisfaction by promptly escalating cases when a suitable answer cannot be found.

[0504] The following describes the processing flow.

[0505] Step 1:

[0506] User: Open the support chat screen and enter your inquiry into the text box. Type "How do I return a product?" and click the send button.

[0507] Step 2:

[0508] Server: Receives queries sent by users. The server stores the received query content in an appropriate format.

[0509] Step 3:

[0510] Server: Passes the received query content to a natural language processing (NLP) library. The NLP library performs grammatical analysis and word segmentation of the text and extracts the main keywords (in this example, "product," "return," and "method").

[0511] Step 4:

[0512] Server: Searches the internal database based on the extracted keywords. Generates search queries and quickly accesses the database index to retrieve the appropriate information.

[0513] Step 5:

[0514] Server: Based on information retrieved from the database, it generates a response to provide to the user. This response is formatted as a natural-sounding sentence. For example, it might generate a response such as, "To return an item, please return it in its original packaging within 30 days of purchase."

[0515] Step 6:

[0516] Server: Sends the generated response to the user's support chat screen. Displays the response appropriately so that the user can review it.

[0517] Step 7:

[0518] User: Review the response displayed on the support chat screen and make additional inquiries if necessary.

[0519] Step 8:

[0520] Server: If a suitable answer cannot be found, or if the inquiry is complex and difficult to handle automatically, the server initiates escalation. It sets an escalation flag and logs the inquiry details and analysis results.

[0521] Step 9:

[0522] Server: Generates an escalation request and sends a notification to the support team. The notification includes detailed information about the inquiry.

[0523] Step 10:

[0524] Terminal (e.g., support team's PC): An escalation notification will appear on the terminal of a support team member. The support team will review the notification and begin taking detailed action.

[0525] Step 11:

[0526] Support Team: Review the details of the escalated inquiry and contact the user directly to resolve the issue.

[0527] (Example 1)

[0528] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0529] Traditional customer support systems consume significant time and resources by manually analyzing inquiries and generating appropriate responses. This can lead to delays in customer service and decreased customer satisfaction. Furthermore, escalation processes are often manual when appropriate answers cannot be found, adding to the overall time burden. A system is needed to address these challenges and provide rapid and accurate responses to customer inquiries.

[0530] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0531] In this invention, the server includes means for receiving user inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the user, and means for escalating the user's inquiry to a support team if no suitable answer is found. This makes it possible to respond to user inquiries quickly and automatically.

[0532] "User" refers to a person who makes inquiries to the system or an end-user.

[0533] "Inquiry content" refers to messages containing questions and requests that users send to the system.

[0534] A "server" refers to a hardware or software system that receives inquiries from users and performs processes such as analysis, searching, response generation, transmission, and escalation.

[0535] Natural Language Processing (NLP) refers to the techniques and algorithms used to analyze human language and understand its meaning.

[0536] A "database" refers to a collection of structured data designed for efficient searching and management of information.

[0537] "Means of searching" refers to a mechanism or process for retrieving information from a database based on a specific query.

[0538] "Means of generating answers" refers to the process of creating appropriate answer text to provide to the user based on the searched information.

[0539] "Means of transmission" refers to the communication technologies and protocols used to send the generated response to the user's device.

[0540] "Escalation means" refers to the process of forwarding an inquiry to other resources, such as a support team, when the system cannot find a suitable answer.

[0541] A "support team" refers to a group of individuals or individuals who provide additional human support in response to user inquiries.

[0542] As an example of a system according to the present invention, a system for automatically processing customer inquiries and generating appropriate responses will be described.

[0543] System Overview

[0544] This system includes the following main components:

[0545] 1. Means for receiving inquiries from users

[0546] 2. Means for analyzing received inquiry content using natural language processing (NLP)

[0547] 3. Means for searching the database based on the analyzed query content

[0548] 4. Means for generating appropriate answers from search results

[0549] 5. Means for sending the generated response to the user

[0550] 6. A means of escalating a user's inquiry to the support team if a suitable answer cannot be found.

[0551] Hardware and software usage examples

[0552] 1. Receiving means

[0553] Users submit inquiries via support chat using devices such as smartphones or computers. The server receives these inquiries over the internet.

[0554] 2. Natural Language Processing (NLP)

[0555] The server analyzes the received query using a natural language processing (NLP) library (e.g., Python's NLTK or SpaCy). Specifically, it divides the text into tokens and extracts keywords.

[0556] 3. Database Search

[0557] The server searches indexed databases (e.g., Elasticsearch or MySQL) using the extracted keywords. This search allows for the rapid retrieval of appropriate information in response to user queries.

[0558] 4. Answer generation means

[0559] The server generates answers to provide to the user based on the search results. For example, it organizes information obtained from the database into an easily understandable format.

[0560] 5. Transmission method

[0561] The server sends the generated response to the user's device. The user can then view the response on the support chat screen.

[0562] 6. Escalation measures

[0563] If the server cannot find a suitable answer, it will notify the support team of the inquiry details. This notification allows the support team to handle the inquiry manually.

[0564] Specific example

[0565] The following shows a specific example of operation.

[0566] 1. User inquiry: "How do I return an item?"

[0567] 2. Server processing:

[0568] Inquiry received

[0569] Analysis using NLP

[0570] Extract the keywords "product," "return," and "method."

[0571] Search database

[0572] The response generated reads: "To return an item, please return it in its original packaging within 30 days of purchase."

[0573] 3. Server submission: Send the generated response to the user.

[0574] If it is incorrect, we will escalate it as follows:

[0575] 1. User inquiry: "How do I return an item from overseas?"

[0576] 2. Server processing:

[0577] Inquiry received

[0578] Analysis using NLP

[0579] I searched the database but couldn't find a suitable answer.

[0580] 3. Server escalation: Notify the support team of the inquiry details.

[0581] 4. Support Team: They will review the inquiry and contact the user directly to resolve the issue.

[0582] Example of a prompt

[0583] By inputting prompt messages like the following into the AI ​​model, it generates appropriate answers to inquiries.

[0584] "Please tell me how to return an item."

[0585] "Please tell me how to return items from overseas."

[0586] This allows the system to respond to customer inquiries efficiently and automatically. Furthermore, if a suitable answer cannot be found, it can quickly escalate the issue, improving customer satisfaction.

[0587] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0588] Step 1:

[0589] A user submits an inquiry using the support chat. An example of the inquiry is, "How do I return a product?"

[0590] Input: User inquiry message (text format)

[0591] Output: JSON data containing the query details

[0592] Specific action: The user enters their inquiry into the chat window and presses the send button.

[0593] Step 2:

[0594] The server receives the user's inquiry.

[0595] Input: JSON data containing the user's inquiry message

[0596] Output: Received inquiry content (string format)

[0597] Specific operation: The server receives the query in real time and extracts the query string for analysis.

[0598] Step 3:

[0599] The server analyzes the received query content using a natural language processing (NLP) library.

[0600] Input: Received inquiry content (string format)

[0601] Output: Extracted keywords (list format)

[0602] Specific operation: The server uses a Python NLP library (e.g., NLTK, SpaCy) to tokenize the text and extract important keywords (e.g., "product", "return", "method").

[0603] Step 4:

[0604] The server searches the database using the extracted keywords.

[0605] Input: Extracted keywords (list format)

[0606] Output: Search results (information from the database)

[0607] Specific operation: The server uses an indexed database (e.g., Elasticsearch, MySQL) to execute queries to retrieve information that matches the keyword.

[0608] Step 5:

[0609] The server generates answers to provide to the user based on the search results.

[0610] Input: Search results (information from the database)

[0611] Output: Generated response (text format)

[0612] Specific operation: The server generates an appropriate response based on information retrieved from the database (e.g., "To return the product, please return it in its original packaging within 30 days of purchase").

[0613] Step 6:

[0614] The server sends the generated response to the user.

[0615] Input: Generated response (text format)

[0616] Output: Display of response to user terminal

[0617] Specific operation: The server sends the response to the user's chat window, where it is displayed on the user's device.

[0618] Step 7:

[0619] If the server cannot find a suitable answer, the user's inquiry will be escalated.

[0620] Input: If the search result is empty

[0621] Output: Notification to the support team

[0622] Specific actions: The server detects that there is no suitable answer and sends a notification to the support team containing details of the inquiry. It also sends a message to the user such as, "We're sorry, but we were unable to find a suitable answer at this time."

[0623] In this way, the system can respond to user inquiries efficiently and automatically, and can quickly escalate cases if a suitable answer cannot be found.

[0624] (Application Example 1)

[0625] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0626] Traditional customer inquiry systems often rely on manual processes for analyzing inquiries and generating responses, resulting in low efficiency. Furthermore, the difficulty of easily submitting inquiries via smartphones hinders the provision of adequate support in today's fast-paced customer service environment. In addition, escalation options for complex customer inquiries are limited, highlighting the need for further operational efficiency and improved customer satisfaction.

[0627] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0628] In this invention, the server includes means for receiving customer inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the customer, means for escalating the customer's inquiry to a support team if no suitable answer is found, means for notifying the customer of detailed information about the inquiry during the escalation, means for easily inputting customer inquiries using a smartphone, means for using a generative AI model to automatically extract keywords from the inquiry and generate an answer based on them, and means for displaying the generated answer as a prompt. This enables the automation and efficiency of inquiry processing, allows customers to make inquiries intuitively using their smartphones, and enhances the escalation means, thereby realizing prompt and accurate customer support.

[0629] A "customer" refers to someone who purchases or uses a product or service.

[0630] "Inquiry content" refers to questions and requests provided by customers.

[0631] "Means of receiving information" refers to the methods and mechanisms by which the system receives customer inquiries.

[0632] "Natural language processing" refers to the technology that allows computers to analyze and understand natural language data.

[0633] A "database" refers to a system in which information is systematically stored and can be searched and managed.

[0634] "Means of searching" refers to the methods and mechanisms for obtaining necessary information from a database.

[0635] "Means of generating answers" refers to methods and mechanisms for creating appropriate answers to provide to customers based on search results.

[0636] "Means of transmission" refers to the methods and mechanisms used to deliver the generated responses to customers.

[0637] "Means of escalation" refers to methods and mechanisms for transferring inquiries that the system cannot handle to the support team.

[0638] "Means of notification" refers to the methods and mechanisms for informing the support team of the details of an inquiry during an escalation.

[0639] A "smartphone" refers to a portable information device that has telephone functionality while also being able to run applications and connect to the internet.

[0640] "Means for easily submitting inquiries" refers to methods and systems that allow customers to easily submit inquiries using devices such as smartphones.

[0641] "Methods for automatically extracting keywords" refers to methods and mechanisms that use natural language processing to extract important terms from query content.

[0642] A "generative AI model" refers to a model that uses artificial intelligence technology to automatically generate answers based on the content of an inquiry.

[0643] A "prompt message" refers to the guidance or response message provided to the customer based on the generated answer.

[0644] System Overview

[0645] The system for implementing this invention automates a series of processes and can respond quickly and accurately to customer inquiries. It mainly uses a server, a smartphone terminal, a natural language processing (NLP) library, a generative AI model, and a database.

[0646] Hardware and software

[0647] Server: The central hardware that receives, processes, parses, generates answers for, and escalates queries.

[0648] Smartphone: A device used by customers to enter inquiries.

[0649] NLP libraries: Software used to analyze natural language. A specific example is spaCy ("ja_core_news_sm" model).

[0650] Generative AI Model: An artificial intelligence model that generates appropriate answers from the content of an inquiry.

[0651] Database: Stores frequently occurring queries and their answers, enabling rapid data retrieval.

[0652] System Implementation Method

[0653] First, the user enters their inquiry using their smartphone. For example, the user might ask, "When will the product I ordered arrive?"

[0654] The server receives this query and analyzes it using an NLP library. Specifically, it extracts keywords such as "order," "product," and "deliver" from the query text. This clarifies the category and intent of the query.

[0655] Next, the database is searched based on the extracted keywords. The database contains pre-indexed response data, which can be searched efficiently. For example, suppose the response "Ordered items are usually shipped within 3-5 business days" is found.

[0656] Based on the responses found, a generative AI model generates answers. This process uses AI to generate more specific and appropriate responses to the original inquiry.

[0657] The generated response is sent from the server to the smartphone device, where the user confirms it. For example, the response might say, "Your order will usually be shipped within 3-5 business days."

[0658] If the server cannot find a suitable answer, the system escalates the inquiry details to the support team. The support team is also notified of the detailed inquiry, allowing them to respond quickly.

[0659] Specific example

[0660] 1. User inquiry: "How do I return an item?"

[0661] 2. Server processing:

[0662] Inquiry received

[0663] Analysis using NLP

[0664] Extract the keywords "product," "return," and "method."

[0665] Search database

[0666] I received the response, "To return the product, please return it in its original packaging within 30 days of purchase."

[0667] 3. Submit the generated response:

[0668] The response sent to the smartphone was: "To return the product, please return it in its original packaging within 30 days of purchase."

[0669] 4. Examples of escalation:

[0670] User inquiry: "How do I return an item from overseas?"

[0671] Server processing:

[0672] Inquiry received

[0673] Analysis using NLP

[0674] I searched the database but couldn't find a suitable answer.

[0675] Server escalation:

[0676] Please notify the support team of the details of your inquiry.

[0677] Support team:

[0678] We will review the inquiry and contact the user directly to provide further details.

[0679] Example of a prompt

[0680] "Question to the Generative AI Model: A user has submitted the following inquiry: 'When will my ordered item arrive?' The keywords analyzed using NLP are 'order,' 'item,' and 'arrive.' Please provide as much specific information as possible regarding the delivery timing."

[0681] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0682] Step 1:

[0683] The user enters their inquiry using their smartphone and presses the send button. Specifically, the user enters "When will my ordered items arrive?". This input is the user's inquiry.

[0684] Step 2:

[0685] The smartphone device sends the entered inquiry content to the server. The input is the user's inquiry, and the output is the data sent to the server.

[0686] Step 3:

[0687] The server uses an NLP library to analyze the received query content. Specifically, it uses the spaCy "ja_core_news_sm" model to extract keywords from the query content. In this case, the input is the query content, and the output is the extracted keywords (such as "order," "product," and "deliver").

[0688] Step 4:

[0689] The server searches the database based on the extracted keywords. The database stores pre-indexed response data. The input is the extracted keywords, and the output is the corresponding response data (e.g., "Your order will be shipped within 3-5 business days").

[0690] Step 5:

[0691] The server generates an answer using a generative AI model based on the search results. In this process, the AI ​​model generates the optimal answer based on the query and search results. The input is the search results and the query, and the output is the generated answer.

[0692] Step 6:

[0693] The server sends the generated response text to the smartphone device. The input is the generated response text, and the output is the data sent to the smartphone device.

[0694] Step 7:

[0695] The smartphone displays the response text received from the server to the user. The input is the response data from the server, and the output is the response text displayed on the smartphone screen.

[0696] Step 8:

[0697] If the server cannot find a suitable answer, the issue will be escalated to the support team. Specifically, a notification containing detailed information about the inquiry will be sent to the support team. The input is the inquiry requiring support, and the output is the notification data sent to the support team.

[0698] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0699] The system according to the present invention is a system for automatically processing user inquiries and generating appropriate responses. Furthermore, by combining it with an emotion engine that recognizes user emotions and adjusts responses based on those emotions, it achieves more accurate customer service.

[0700] System Overview

[0701] This system includes the following main features:

[0702] 1. Receipt of inquiry

[0703] 2. Analysis of the inquiry content using natural language processing (NLP)

[0704] 3. User emotion recognition by an emotion engine

[0705] 4. Database Search

[0706] 5. Generating and adjusting responses

[0707] 6. Submit your response

[0708] 7. Escalation as needed

[0709] Processing flow

[0710] First, the user submits an inquiry using the support chat. For example, let's say the inquiry is, "How do I return a product?"

[0711] The server receives inquiries sent by users. It then passes the received inquiry content to a natural language processing (NLP) library for analysis. Here, keywords such as "product," "return," and "method" are extracted from the text.

[0712] Next, the server uses its built-in emotion engine to analyze the user's emotions in the inquiry. This analysis determines whether the user is feeling, for example, "negative" if they are dissatisfied, or "positive" if they are grateful.

[0713] Based on the analyzed keywords and sentiment data, the server searches its internal database to retrieve appropriate information. Based on the retrieved information, it generates a response to provide to the user. During this process, the sentiment engine adjusts the response based on the user's emotions. For example, if the user is angry, a more polite and intimate tone will be used.

[0714] The generated response is sent from the server to the user and displayed on the support chat screen. The user reviews the provided response and asks further questions if necessary.

[0715] If the server cannot find a suitable answer, or if the user's sentiment is very negative, the system will use escalation mechanisms. In this case, the server will notify the support team of the details of the inquiry and the sentiment analysis results.

[0716] Specific example

[0717] 1. User inquiry: "How do I return an item?"

[0718] 2. Server processing:

[0719] Inquiry received

[0720] Analysis using NLP

[0721] Extract the keywords "product," "return," and "method."

[0722] The emotion engine recognizes that the user is "confused".

[0723] Search database

[0724] The response text generated reads: "To return an item, please return it in its original packaging within 30 days of purchase."

[0725] Add a supplement to your response such as, "We apologize for any inconvenience this may cause."

[0726] 3. Server submission: Send the adjusted response to the user.

[0727] 4. Escalation (additional example):

[0728] User inquiry: "The quality of this product is terrible. I want to return it."

[0729] Server processing:

[0730] Inquiry received

[0731] Analysis using NLP

[0732] Extract the keywords "product," "quality," and "returns."

[0733] The emotion engine recognizes that the user is feeling "anger".

[0734] I searched the database but couldn't find a suitable answer.

[0735] Server escalation: Notify the support team of the inquiry details and the result of the "anger" emotion recognition.

[0736] Support Team: We review inquiries and contact users directly to resolve issues.

[0737] This embodiment allows the system to handle inquiries efficiently and with consideration for the user's feelings. Furthermore, by promptly escalating cases when an appropriate answer cannot be found or when the user's feelings are negative, customer satisfaction can be improved.

[0738] The following describes the processing flow.

[0739] Step 1:

[0740] User: Open the support chat screen and enter your inquiry into the text box. Type "How do I return a product?" and click the send button.

[0741] Step 2:

[0742] Server: Receives queries sent by users. The server stores the received query content in an appropriate format.

[0743] Step 3:

[0744] Server: Passes the received query content to a natural language processing (NLP) library. The NLP library performs grammatical analysis and word segmentation of the text and extracts the main keywords (in this example, "product," "return," and "method").

[0745] Step 4:

[0746] Server: Based on the extracted keywords, the server passes the query to the emotion engine. The emotion engine analyzes the user's emotions from the text and identifies emotions such as confusion, anger, and joy. In this example, the user's emotion is recognized as "confused".

[0747] Step 5:

[0748] Server: Searches the internal database based on extracted keywords and sentiment recognition results. Generates search queries and accesses the database index to retrieve relevant information.

[0749] Step 6:

[0750] Server: Generates a response to provide to the user based on information retrieved from the database. This response is formatted in a natural style and adjusted based on the results of the sentiment engine. For example, polite expressions may be added, such as, "To return the product, please return it in its original packaging within 30 days of purchase. We apologize for any inconvenience this may cause."

[0751] Step 7:

[0752] Server: Sends the generated response to the user's support chat screen. Displays the response appropriately so that the user can review it.

[0753] Step 8:

[0754] User: Review the response displayed on the support chat screen and make additional inquiries if necessary.

[0755] Step 9:

[0756] Server: If a suitable answer cannot be found, or if the inquiry is complex and difficult to handle automatically, the server will initiate escalation. Escalation will also occur if the user's emotions are very negative.

[0757] Step 10:

[0758] Server: Set the escalation flag and log the query content and sentiment analysis results.

[0759] Step 11:

[0760] Server: Generates an escalation request and sends a notification to the support team. The notification includes the inquiry details and sentiment analysis results.

[0761] Step 12:

[0762] Terminal (e.g., support team's PC): An escalation notification will appear on the terminal of a support team member. The support team will review the notification and begin taking detailed action.

[0763] Step 13:

[0764] Support Team: Review the details of the escalated inquiry and contact the user directly to resolve the issue.

[0765] (Example 2)

[0766] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0767] Traditional customer support systems analyze inquiries and generate appropriate responses, but they fail to consider customer emotions, resulting in insufficient improvement in customer satisfaction. Furthermore, delays in escalation when an appropriate answer cannot be found can lead to service delays.

[0768] The identification processing performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for receiving customer inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the customer, means for escalating the customer's inquiry to the support team if no suitable answer is found, means for analyzing the customer's emotions included in the inquiry, and means for adjusting the answer based on the analyzed customer emotions. This makes it possible to respond in a way that takes customer emotions into consideration, thereby improving customer satisfaction. In addition, if an appropriate answer cannot be found, it is quickly escalated to the support team, preventing delays in response.

[0769] "Customer" refers to anyone who uses the system to receive services or support.

[0770] "Inquiry details" refers to the text information of questions or requests that customers enter or submit when they require support.

[0771] "Means of receiving information" refers to the function for importing customer inquiries into the system.

[0772] "Natural language processing" refers to the technology that enables computers to understand and analyze text written in human language.

[0773] "Means of analysis" refers to the function that uses natural language processing to analyze the meaning and intent of received inquiries.

[0774] "Means of searching the database" refers to the function of searching for relevant information from the internal database based on the analyzed query content.

[0775] "Means of generating responses" refers to a function that creates responses tailored to the customer based on information obtained from a database.

[0776] "Means of transmission" refers to the function used to communicate the generated response to the customer.

[0777] "Escalation options" refer to the function that allows you to transfer your inquiry to the support team if you cannot find a suitable answer or under certain conditions.

[0778] "Methods for analyzing emotions" refers to technologies used to identify customer emotions from the content of inquiries.

[0779] "Means of adjusting responses" refers to a function that optimizes the content and style of responses based on analyzed customer sentiment.

[0780] The system according to the present invention automatically processes customer inquiries and generates appropriate responses. Furthermore, by having a function to recognize customer emotions and adjust responses accordingly, it enables more accurate customer service.

[0781] System Configuration

[0782] 1. Receiving customer inquiries

[0783] Users access the support chat using their device and enter their questions or requests. At this time, the customer's inquiry is sent to the system server.

[0784] 2. Natural Language Processing (NLP) of the inquiry content

[0785] The server uses an NLP library (e.g., SpaCy or NLTK) to analyze the received query. Specifically, it performs tokenization of the text (splitting it into individual words), part-of-speech tagging, and keyword extraction. This analysis extracts the important keywords from the query.

[0786] 3. Recognizing customer emotions

[0787] The server uses an emotion engine (such as Google Cloud Natural Language API or IBM Watson Tone Analyzer) to analyze customer emotions from inquiries. For example, emotions such as "confused," "angry," and "grateful" may be recognized.

[0788] 4. Searching the internal database

[0789] The server searches its internal database for relevant information based on the analyzed keywords and sentiment data. This process efficiently retrieves information using SQL queries.

[0790] 5. Generating and adjusting responses

[0791] The server generates appropriate responses using a generative AI model (e.g., OpenAI GPT-3) based on database search results. It also adjusts the style and content of the responses based on the results of the sentiment engine. For example, if a customer is confused, a polite clarification such as "We apologize for the inconvenience" might be added.

[0792] 6. Submit your response

[0793] The server sends the generated response to the user and displays it on the support chat screen. This allows the user to see the answer immediately.

[0794] 7. Escalation as needed

[0795] If the server cannot find a suitable answer or if the customer's sentiment is very negative, it will notify the support team of the inquiry details and sentiment analysis results. This allows the support team to respond quickly.

[0796] Specific example

[0797] Specific Example 1

[0798] User inquiry: "How do I return an item?"

[0799] Processing flow:

[0800] The user enters and submits their inquiry.

[0801] The server receives the query.

[0802] The server uses NLP to extract keywords such as "product," "return," and "method."

[0803] The server uses an emotion engine to recognize that the user is "confused."

[0804] The server searches the database and generates a response message stating, "To return an item, please return it in its original packaging within 30 days of purchase."

[0805] The server adds a note to the response such as, "We apologize for any inconvenience this may cause."

[0806] The server sends the adjusted response to the user.

[0807] Specific example 2 (escalation)

[0808] User inquiry: "The quality of this product is terrible. I want to return it."

[0809] Processing flow:

[0810] The user enters and submits their inquiry.

[0811] The server receives the query.

[0812] The server uses NLP to analyze the data and extract information on "products," "quality," and "returns."

[0813] The server uses an emotion engine to recognize that the user is feeling "anger."

[0814] The server searches the database but cannot find a suitable answer.

[0815] The server notifies the support team of the inquiry details and the result of the "anger" emotion recognition.

[0816] The support team will review the inquiry and contact the user directly to resolve the issue.

[0817] This invention enables the system to handle inquiries efficiently and with consideration for the user's emotions. By promptly escalating cases when an appropriate answer cannot be found or when the user's emotions are negative, customer satisfaction can be improved.

[0818] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0819] Step 1:

[0820] The user accesses the support chat using their device, enters their inquiry, and submits it. For example, the text entered in the inquiry might be, "How do I return a product?" The entered data is then sent to the server.

[0821] Step 2:

[0822] The server receives the query. The received data is logged as is and passed on to the next analysis process. The input is text data, and the output is similarly text data for analysis.

[0823] Step 3:

[0824] The server uses an NLP library (e.g., SpaCy or NLTK) to analyze the received text data. Specifically, it performs text tokenization, part-of-speech tagging, and keyword extraction. The input is the received text data, and the output is a list of keywords (e.g., "product," "return," "method").

[0825] Step 4:

[0826] The server uses an internal sentiment engine (e.g., Google Cloud Natural Language API or IBM Watson Tone Analyzer) to analyze the user's sentiment in the query. The input is received text data, and the output is sentiment tags (e.g., "confused").

[0827] Step 5:

[0828] The server searches its internal database based on the analyzed keywords and sentiment data. Specifically, it uses SQL queries to retrieve relevant information. The input is a list of keywords and sentiment tags, and the output is the appropriate response data.

[0829] Step 6:

[0830] The server uses a generative AI model (e.g., OpenAI GPT-3) to generate appropriate responses based on the search results. Here, the responses are refined based on the results of the sentiment engine. The input is the response data and sentiment tags, and the output is the refined response text (e.g., "To return the product, please return it in its original packaging within 30 days of purchase. We apologize for any inconvenience.").

[0831] Step 7:

[0832] The server sends the generated response to the user. The response is displayed on the support chat screen, where the user can review it. The input is the edited response, and the output is what is displayed on the user's screen.

[0833] Step 8:

[0834] If a suitable answer cannot be found, or if the user's sentiment is very negative, the server escalates the inquiry details and sentiment analysis results to the support team. The input is the inquiry details and sentiment tags obtained from the previous processing, and the output is a notification to the support team.

[0835] (Application Example 2)

[0836] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0837] While conventional customer service systems can automatically generate appropriate responses to inquiries, they struggle to improve customer satisfaction because they cannot consider customer emotions. Furthermore, systems often lack the flexibility to adapt to situations requiring changes in writing style based on emotions or, in some cases, rapid escalation.

[0838] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for receiving customer inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the customer, means for escalating the customer's inquiry to a support team if no suitable answer is found, means for recognizing the customer's emotions regarding the inquiry, and means for adjusting the generated answer based on the recognized emotions. This enables automated responses that take emotions into consideration and rapid escalation.

[0839] A "customer" is a user who makes an inquiry using this system.

[0840] "Inquiry content" refers to questions and requests that customers submit through this system.

[0841] "Natural language processing" is a technology that enables computers to understand and analyze human language.

[0842] "Analyzing" means converting input data into a format that can be understood, either partially or entirely.

[0843] A "database" is an electronic information storage system that systematically accumulates information.

[0844] "Generating" means creating new data or information according to a certain algorithm or rule.

[0845] "To send" means to transfer specific data from one location to another.

[0846] "Escalation" is the act of passing on a difficult problem to a higher-ranking or specialized person.

[0847] A "support team" is a group of specialists responsible for handling customer inquiries and resolving problems.

[0848] "Recognizing emotions" means analyzing customer inquiries and identifying the types of emotions they contain.

[0849] "Adjusting" means appropriately changing the information and responses provided to match the perceived emotions.

[0850] A "system" is a mechanism that provides a certain function through the coordinated action of multiple components or means.

[0851] The system for implementing this invention analyzes customer inquiries, recognizes their emotions, and generates and adjusts appropriate responses based on those emotions in a content distribution service using smartphones. The specific operation of the system will be described below.

[0852] First, the user submits an inquiry using a smartphone app. The user's inquiry is in the format of, for example, "Please tell me about the latest TV dramas." The server receives this inquiry and analyzes it using a natural language processing (NLP) library. During this process, SpaCy is used to extract keywords from the inquiry and identify the subject of the inquiry.

[0853] Next, the server uses VADER Sentiment to perform sentiment analysis on the inquiry and recognizes the user's sentiment as either "positive," "negative," or "neutral." Based on this sentiment data and extracted keywords, the server searches the SQLite database and retrieves the relevant response. The retrieved response is then adjusted by the sentiment engine based on the user's sentiment. For example, if the user is "positive," the response will be written in a friendly style.

[0854] The adjusted response is sent from the server to the user's smartphone and displayed on the app screen. If a suitable response does not exist in the database, or if the user's sentiment is very strong and negative, the server escalates the inquiry to the support team. During this escalation process, the support team is notified of the detailed inquiry and the results of the sentiment analysis.

[0855] Hardware and software to be used

[0856] Hardware: Smartphone (iOS or Android)

[0857] Software: Flask (Python framework), SpaCy (natural language processing library), VADER Sentiment (sentiment analysis library), SQLite (database)

[0858] Specific example

[0859] For example, if a user sends an inquiry asking "Please tell me the latest TV dramas," the server will process it as follows:

[0860] 1. Receive user inquiries.

[0861] 2. Analyze the inquiries and extract the keywords "latest" and "drama".

[0862] 3. Conduct sentiment analysis and recognize the user's emotions as "neutral."

[0863] 4. Search the database and obtain the answer, "The latest drama is XX."

[0864] 5. Adjust the response based on the user's emotions and generate an answer in the format "The latest drama is XX. (User's emotion: Neutral)".

[0865] 6. Send the generated response to the user and display it on the smartphone app screen.

[0866] Example of a prompt

[0867] 1. Inquiry: "Please tell me about the latest TV dramas."

[0868] 2. Answer: "The latest drama is XX. (User sentiment: Neutral)"

[0869] This makes it possible to provide automated responses to user inquiries that take emotions into consideration.

[0870] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0871] Step 1:

[0872] Users submit their inquiries through a smartphone app. The inquiry content is provided in text format. At this point, users ask specific questions such as, "Please tell me about the latest TV dramas." The output is the text data of the inquiry.

[0873] Step 2:

[0874] The server processes the received query. First, it passes the query to a natural language processing (NLP) library. The SpaCy library is used for this purpose. The input is the query text received in step 1, and the output is a list of keywords. Specifically, keywords such as "latest" and "drama" are extracted from the query.

[0875] Step 3:

[0876] The server performs sentiment analysis on the extracted keywords. The VADER Sentiment library is used for this purpose. The input consists of the keywords extracted in step 2 and the original query. The output is a sentiment classification (positive, negative, or neutral). For example, the query "Please tell me the latest TV dramas" would be classified as "neutral."

[0877] Step 4:

[0878] The server searches the database based on the extracted keywords and sentiment analysis results. It executes queries against the SQLite database to retrieve the relevant answers. The inputs are the keywords from step 2 and the sentiment classification results from step 3. The output is the corresponding answer text. Specific answers such as "The latest drama is XX" can be obtained.

[0879] Step 5:

[0880] The server adjusts the acquired responses based on the sentiment analysis results. This process involves changing the writing style and adding supplementary information according to the sentiment. The input is the response text acquired in step 4 and the sentiment classification result obtained in step 3. The output is the adjusted response text. For example, the response might reflect a "neutral" sentiment recognition, resulting in "The latest drama is XX. (User's sentiment: Neutral)."

[0881] Step 6:

[0882] The server sends the adjusted response to the user. To do this, it uses the Flask framework to return the response as an HTTP response. The input is the response text adjusted in step 5, and the output is the response displayed on the user's smartphone.

[0883] Step 7:

[0884] If a suitable answer cannot be retrieved from the database, or if the sentiment analysis indicates very strong negative emotions, the server escalates the inquiry to the support team. The input consists of the sentiment analysis results obtained in step 3 and the database search results that were deemed irrelevant in step 4. The output is a notification to the support team, which includes the inquiry details and the sentiment analysis results.

[0885] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0886] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0887] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0888] [Third Embodiment]

[0889] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0890] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0891] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0892] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0893] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0894] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0895] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0896] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0897] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0898] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0899] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0900] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0901] As an example of a system according to the present invention, a system for automatically processing customer inquiries and generating appropriate responses will be described.

[0902] System Overview

[0903] This system includes the following main features:

[0904] 1. Receipt of inquiry

[0905] 2. Analysis of the inquiry content using natural language processing (NLP)

[0906] 3. Database Search

[0907] 4. Generating the answer

[0908] 5. Submit your response

[0909] 6. Escalation as needed

[0910] Processing flow

[0911] First, the user submits an inquiry using the support chat. For example, let's say the inquiry is, "How do I return a product?"

[0912] The server receives the inquiry sent by the user. It then analyzes the received string using a natural language processing (NLP) library. In this case, keywords such as "product," "return," and "method" are extracted from the text.

[0913] Next, the server uses the extracted keywords to search its internal database. This search utilizes an indexed database for faster information retrieval.

[0914] Once search results are obtained, the server generates a response to provide to the user based on them. For example, it might retrieve information from the database such as, "To return a product, please return it in its original packaging within 30 days of purchase," and then format this into a response.

[0915] The generated response is sent from the server to the user. The user can then view the response displayed on the support chat screen.

[0916] If the server cannot find a suitable answer, or if the inquiry is too complex, the system will use an escalation mechanism. In this case, the server will notify the support team of the details of the inquiry.

[0917] Specific example

[0918] 1. User inquiry: "How do I return an item?"

[0919] 2. Server processing:

[0920] Inquiry received

[0921] Analysis using NLP

[0922] Extract the keywords "product," "return," and "method."

[0923] Search database

[0924] The response generated reads: "To return an item, please return it in its original packaging within 30 days of purchase."

[0925] 3. Server submission: Send the generated response to the user.

[0926] 4. Escalation (additional example):

[0927] User inquiry: "How do I return an item from overseas?"

[0928] Server processing:

[0929] Inquiry received

[0930] Analysis using NLP

[0931] I searched the database but couldn't find a suitable answer.

[0932] Server escalation: Notify the support team of the inquiry details.

[0933] Support team: We will review your inquiry and contact you directly to provide further details.

[0934] This embodiment allows the system to efficiently respond to customer inquiries. Furthermore, it can improve customer satisfaction by promptly escalating cases when a suitable answer cannot be found.

[0935] The following describes the processing flow.

[0936] Step 1:

[0937] User: Open the support chat screen and enter your inquiry into the text box. Type "How do I return a product?" and click the send button.

[0938] Step 2:

[0939] Server: Receives queries sent by users. The server stores the received query content in an appropriate format.

[0940] Step 3:

[0941] Server: Passes the received query content to a natural language processing (NLP) library. The NLP library performs grammatical analysis and word segmentation of the text and extracts the main keywords (in this example, "product," "return," and "method").

[0942] Step 4:

[0943] Server: Searches the internal database based on the extracted keywords. Generates search queries and quickly accesses the database index to retrieve the appropriate information.

[0944] Step 5:

[0945] Server: Based on information retrieved from the database, it generates a response to provide to the user. This response is formatted as a natural-sounding sentence. For example, it might generate a response such as, "To return an item, please return it in its original packaging within 30 days of purchase."

[0946] Step 6:

[0947] Server: Sends the generated response to the user's support chat screen. Displays the response appropriately so that the user can review it.

[0948] Step 7:

[0949] User: Review the response displayed on the support chat screen and make additional inquiries if necessary.

[0950] Step 8:

[0951] Server: If a suitable answer cannot be found, or if the inquiry is complex and difficult to handle automatically, the server initiates escalation. It sets an escalation flag and logs the inquiry details and analysis results.

[0952] Step 9:

[0953] Server: Generates an escalation request and sends a notification to the support team. The notification includes detailed information about the inquiry.

[0954] Step 10:

[0955] Terminal (e.g., support team's PC): An escalation notification will appear on the terminal of a support team member. The support team will review the notification and begin taking detailed action.

[0956] Step 11:

[0957] Support Team: Review the details of the escalated inquiry and contact the user directly to resolve the issue.

[0958] (Example 1)

[0959] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0960] Traditional customer support systems consume significant time and resources by manually analyzing inquiries and generating appropriate responses. This can lead to delays in customer service and decreased customer satisfaction. Furthermore, escalation processes are often manual when appropriate answers cannot be found, adding to the overall time burden. A system is needed to address these challenges and provide rapid and accurate responses to customer inquiries.

[0961] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0962] In this invention, the server includes means for receiving user inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the user, and means for escalating the user's inquiry to a support team if no suitable answer is found. This makes it possible to respond to user inquiries quickly and automatically.

[0963] "User" refers to a person who makes inquiries to the system or an end-user.

[0964] "Inquiry content" refers to messages containing questions and requests that users send to the system.

[0965] A "server" refers to a hardware or software system that receives inquiries from users and performs processes such as analysis, searching, response generation, transmission, and escalation.

[0966] Natural Language Processing (NLP) refers to the techniques and algorithms used to analyze human language and understand its meaning.

[0967] A "database" refers to a collection of structured data designed for efficient searching and management of information.

[0968] "Means of searching" refers to a mechanism or process for retrieving information from a database based on a specific query.

[0969] "Means of generating answers" refers to the process of creating appropriate answer text to provide to the user based on the searched information.

[0970] "Means of transmission" refers to the communication technologies and protocols used to send the generated response to the user's device.

[0971] "Escalation means" refers to the process of forwarding an inquiry to other resources, such as a support team, when the system cannot find a suitable answer.

[0972] A "support team" refers to a group of individuals or individuals who provide additional human support in response to user inquiries.

[0973] As an example of a system according to the present invention, a system for automatically processing customer inquiries and generating appropriate responses will be described.

[0974] System Overview

[0975] This system includes the following main components:

[0976] 1. Means for receiving inquiries from users

[0977] 2. Means for analyzing received inquiry content using natural language processing (NLP)

[0978] 3. Means for searching the database based on the analyzed query content

[0979] 4. Means for generating appropriate answers from search results

[0980] 5. Means for sending the generated response to the user

[0981] 6. A means of escalating a user's inquiry to the support team if a suitable answer cannot be found.

[0982] Hardware and software usage examples

[0983] 1. Receiving means

[0984] Users submit inquiries via support chat using devices such as smartphones or computers. The server receives these inquiries over the internet.

[0985] 2. Natural Language Processing (NLP)

[0986] The server analyzes the received query using a natural language processing (NLP) library (e.g., Python's NLTK or SpaCy). Specifically, it divides the text into tokens and extracts keywords.

[0987] 3. Database Search

[0988] The server searches indexed databases (e.g., Elasticsearch or MySQL) using the extracted keywords. This search allows for the rapid retrieval of appropriate information in response to user queries.

[0989] 4. Answer generation means

[0990] The server generates answers to provide to the user based on the search results. For example, it organizes information obtained from the database into an easily understandable format.

[0991] 5. Transmission method

[0992] The server sends the generated response to the user's device. The user can then view the response on the support chat screen.

[0993] 6. Escalation measures

[0994] If the server cannot find a suitable answer, it will notify the support team of the inquiry details. This notification allows the support team to handle the inquiry manually.

[0995] Specific example

[0996] The following shows a specific example of operation.

[0997] 1. User inquiry: "How do I return an item?"

[0998] 2. Server processing:

[0999] Inquiry received

[1000] Analysis using NLP

[1001] Extract the keywords "product," "return," and "method."

[1002] Search database

[1003] The response generated reads: "To return an item, please return it in its original packaging within 30 days of purchase."

[1004] 3. Server submission: Send the generated response to the user.

[1005] If it is incorrect, we will escalate it as follows:

[1006] 1. User inquiry: "How do I return an item from overseas?"

[1007] 2. Server processing:

[1008] Inquiry received

[1009] Analysis using NLP

[1010] I searched the database but couldn't find a suitable answer.

[1011] 3. Server escalation: Notify the support team of the inquiry details.

[1012] 4. Support Team: They will review the inquiry and contact the user directly to resolve the issue.

[1013] Example of a prompt

[1014] By inputting prompt messages like the following into the AI ​​model, it generates appropriate answers to inquiries.

[1015] "Please tell me how to return an item."

[1016] "Please tell me how to return items from overseas."

[1017] This allows the system to respond to customer inquiries efficiently and automatically. Furthermore, if a suitable answer cannot be found, it can quickly escalate the issue, improving customer satisfaction.

[1018] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1019] Step 1:

[1020] A user submits an inquiry using the support chat. An example of the inquiry is, "How do I return a product?"

[1021] Input: User inquiry message (text format)

[1022] Output: JSON data containing the query details

[1023] Specific action: The user enters their inquiry into the chat window and presses the send button.

[1024] Step 2:

[1025] The server receives the user's inquiry.

[1026] Input: JSON data containing the user's inquiry message

[1027] Output: Received inquiry content (string format)

[1028] Specific operation: The server receives the query in real time and extracts the query string for analysis.

[1029] Step 3:

[1030] The server analyzes the received query content using a natural language processing (NLP) library.

[1031] Input: Received inquiry content (string format)

[1032] Output: Extracted keywords (list format)

[1033] Specific operation: The server uses a Python NLP library (e.g., NLTK, SpaCy) to tokenize the text and extract important keywords (e.g., "product", "return", "method").

[1034] Step 4:

[1035] The server searches the database using the extracted keywords.

[1036] Input: Extracted keywords (list format)

[1037] Output: Search results (information from the database)

[1038] Specific operation: The server uses an indexed database (e.g., Elasticsearch, MySQL) to execute queries to retrieve information that matches the keyword.

[1039] Step 5:

[1040] The server generates answers to provide to the user based on the search results.

[1041] Input: Search results (information from the database)

[1042] Output: Generated response (text format)

[1043] Specific operation: The server generates an appropriate response based on information retrieved from the database (e.g., "To return the product, please return it in its original packaging within 30 days of purchase").

[1044] Step 6:

[1045] The server sends the generated response to the user.

[1046] Input: Generated response (text format)

[1047] Output: Display of response to user terminal

[1048] Specific operation: The server sends the response to the user's chat window, where it is displayed on the user's device.

[1049] Step 7:

[1050] If the server cannot find a suitable answer, the user's inquiry will be escalated.

[1051] Input: If the search result is empty

[1052] Output: Notification to the support team

[1053] Specific actions: The server detects that there is no suitable answer and sends a notification to the support team containing details of the inquiry. It also sends a message to the user such as, "We're sorry, but we were unable to find a suitable answer at this time."

[1054] In this way, the system can respond to user inquiries efficiently and automatically, and can quickly escalate cases if a suitable answer cannot be found.

[1055] (Application Example 1)

[1056] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1057] Traditional customer inquiry systems often rely on manual processes for analyzing inquiries and generating responses, resulting in low efficiency. Furthermore, the difficulty of easily submitting inquiries via smartphones hinders the provision of adequate support in today's fast-paced customer service environment. In addition, escalation options for complex customer inquiries are limited, highlighting the need for further operational efficiency and improved customer satisfaction.

[1058] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1059] In this invention, the server includes means for receiving customer inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the customer, means for escalating the customer's inquiry to a support team if no suitable answer is found, means for notifying the customer of detailed information about the inquiry during the escalation, means for easily inputting customer inquiries using a smartphone, means for using a generative AI model to automatically extract keywords from the inquiry and generate an answer based on them, and means for displaying the generated answer as a prompt. This enables the automation and efficiency of inquiry processing, allows customers to make inquiries intuitively using their smartphones, and enhances the escalation means, thereby realizing prompt and accurate customer support.

[1060] A "customer" refers to someone who purchases or uses a product or service.

[1061] "Inquiry content" refers to questions and requests provided by customers.

[1062] "Means of receiving information" refers to the methods and mechanisms by which the system receives customer inquiries.

[1063] "Natural language processing" refers to the technology that allows computers to analyze and understand natural language data.

[1064] A "database" refers to a system in which information is systematically stored and can be searched and managed.

[1065] "Means of searching" refers to the methods and mechanisms for obtaining necessary information from a database.

[1066] "Means of generating answers" refers to methods and mechanisms for creating appropriate answers to provide to customers based on search results.

[1067] "Means of transmission" refers to the methods and mechanisms used to deliver the generated responses to customers.

[1068] "Means of escalation" refers to methods and mechanisms for transferring inquiries that the system cannot handle to the support team.

[1069] "Means of notification" refers to the methods and mechanisms for informing the support team of the details of an inquiry during an escalation.

[1070] A "smartphone" refers to a portable information device that has telephone functionality while also being able to run applications and connect to the internet.

[1071] "Means for easily submitting inquiries" refers to methods and systems that allow customers to easily submit inquiries using devices such as smartphones.

[1072] "Methods for automatically extracting keywords" refers to methods and mechanisms that use natural language processing to extract important terms from query content.

[1073] A "generative AI model" refers to a model that uses artificial intelligence technology to automatically generate answers based on the content of an inquiry.

[1074] A "prompt message" refers to the guidance or response message provided to the customer based on the generated answer.

[1075] System Overview

[1076] The system for implementing this invention automates a series of processes and can respond quickly and accurately to customer inquiries. It mainly uses a server, a smartphone terminal, a natural language processing (NLP) library, a generative AI model, and a database.

[1077] Hardware and software

[1078] Server: The central hardware that receives, processes, parses, generates answers for, and escalates queries.

[1079] Smartphone: A device used by customers to enter inquiries.

[1080] NLP libraries: Software used to analyze natural language. A specific example is spaCy ("ja_core_news_sm" model).

[1081] Generative AI Model: An artificial intelligence model that generates appropriate answers from the content of an inquiry.

[1082] Database: Stores frequently occurring queries and their answers, enabling rapid data retrieval.

[1083] System Implementation Method

[1084] First, the user enters their inquiry using their smartphone. For example, the user might ask, "When will the product I ordered arrive?"

[1085] The server receives this query and analyzes it using an NLP library. Specifically, it extracts keywords such as "order," "product," and "deliver" from the query text. This clarifies the category and intent of the query.

[1086] Next, the database is searched based on the extracted keywords. The database contains pre-indexed response data, which can be searched efficiently. For example, suppose the response "Ordered items are usually shipped within 3-5 business days" is found.

[1087] Based on the responses found, a generative AI model generates answers. This process uses AI to generate more specific and appropriate responses to the original inquiry.

[1088] The generated response is sent from the server to the smartphone device, where the user confirms it. For example, the response might say, "Your order will usually be shipped within 3-5 business days."

[1089] If the server cannot find a suitable answer, the system escalates the inquiry details to the support team. The support team is also notified of the detailed inquiry, allowing them to respond quickly.

[1090] Specific example

[1091] 1. User inquiry: "How do I return an item?"

[1092] 2. Server processing:

[1093] Inquiry received

[1094] Analysis using NLP

[1095] Extract the keywords "product," "return," and "method."

[1096] Search database

[1097] I received the response, "To return the product, please return it in its original packaging within 30 days of purchase."

[1098] 3. Submit the generated response:

[1099] The response sent to the smartphone was: "To return the product, please return it in its original packaging within 30 days of purchase."

[1100] 4. Examples of escalation:

[1101] User inquiry: "How do I return an item from overseas?"

[1102] Server processing:

[1103] Inquiry received

[1104] Analysis using NLP

[1105] I searched the database but couldn't find a suitable answer.

[1106] Server escalation:

[1107] Please notify the support team of the details of your inquiry.

[1108] Support team:

[1109] We will review the inquiry and contact the user directly to provide further details.

[1110] Example of a prompt

[1111] "Question to the Generative AI Model: A user has submitted the following inquiry: 'When will my ordered item arrive?' The keywords analyzed using NLP are 'order,' 'item,' and 'arrive.' Please provide as much specific information as possible regarding the delivery timing."

[1112] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1113] Step 1:

[1114] The user enters their inquiry using their smartphone and presses the send button. Specifically, the user enters "When will my ordered items arrive?". This input is the user's inquiry.

[1115] Step 2:

[1116] The smartphone device sends the entered inquiry content to the server. The input is the user's inquiry, and the output is the data sent to the server.

[1117] Step 3:

[1118] The server uses an NLP library to analyze the received query content. Specifically, it uses the spaCy "ja_core_news_sm" model to extract keywords from the query content. In this case, the input is the query content, and the output is the extracted keywords (such as "order," "product," and "deliver").

[1119] Step 4:

[1120] The server searches the database based on the extracted keywords. The database stores pre-indexed response data. The input is the extracted keywords, and the output is the corresponding response data (e.g., "Your order will be shipped within 3-5 business days").

[1121] Step 5:

[1122] The server generates an answer using a generative AI model based on the search results. In this process, the AI ​​model generates the optimal answer based on the query and search results. The input is the search results and the query, and the output is the generated answer.

[1123] Step 6:

[1124] The server sends the generated response text to the smartphone device. The input is the generated response text, and the output is the data sent to the smartphone device.

[1125] Step 7:

[1126] The smartphone displays the response text received from the server to the user. The input is the response data from the server, and the output is the response text displayed on the smartphone screen.

[1127] Step 8:

[1128] If the server cannot find a suitable answer, the issue will be escalated to the support team. Specifically, a notification containing detailed information about the inquiry will be sent to the support team. The input is the inquiry requiring support, and the output is the notification data sent to the support team.

[1129] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1130] The system according to the present invention is a system for automatically processing user inquiries and generating appropriate responses. Furthermore, by combining it with an emotion engine that recognizes user emotions and adjusts responses based on those emotions, it achieves more accurate customer service.

[1131] System Overview

[1132] This system includes the following main features:

[1133] 1. Receipt of inquiry

[1134] 2. Analysis of the inquiry content using natural language processing (NLP)

[1135] 3. User emotion recognition by an emotion engine

[1136] 4. Database Search

[1137] 5. Generating and adjusting responses

[1138] 6. Submit your response

[1139] 7. Escalation as needed

[1140] Processing flow

[1141] First, the user submits an inquiry using the support chat. For example, let's say the inquiry is, "How do I return a product?"

[1142] The server receives inquiries sent by users. It then passes the received inquiry content to a natural language processing (NLP) library for analysis. Here, keywords such as "product," "return," and "method" are extracted from the text.

[1143] Next, the server uses its built-in emotion engine to analyze the user's emotions in the inquiry. This analysis determines whether the user is feeling, for example, "negative" if they are dissatisfied, or "positive" if they are grateful.

[1144] Based on the analyzed keywords and sentiment data, the server searches its internal database to retrieve appropriate information. Based on the retrieved information, it generates a response to provide to the user. During this process, the sentiment engine adjusts the response based on the user's emotions. For example, if the user is angry, a more polite and intimate tone will be used.

[1145] The generated response is sent from the server to the user and displayed on the support chat screen. The user reviews the provided response and asks further questions if necessary.

[1146] If the server cannot find a suitable answer, or if the user's sentiment is very negative, the system will use escalation mechanisms. In this case, the server will notify the support team of the details of the inquiry and the sentiment analysis results.

[1147] Specific example

[1148] 1. User inquiry: "How do I return an item?"

[1149] 2. Server processing:

[1150] Inquiry received

[1151] Analysis using NLP

[1152] Extract the keywords "product," "return," and "method."

[1153] The emotion engine recognizes that the user is "confused".

[1154] Search database

[1155] The response text generated reads: "To return an item, please return it in its original packaging within 30 days of purchase."

[1156] Add a supplement to your response such as, "We apologize for any inconvenience this may cause."

[1157] 3. Server submission: Send the adjusted response to the user.

[1158] 4. Escalation (additional example):

[1159] User inquiry: "The quality of this product is terrible. I want to return it."

[1160] Server processing:

[1161] Inquiry received

[1162] Analysis using NLP

[1163] Extract the keywords "product," "quality," and "returns."

[1164] The emotion engine recognizes that the user is feeling "anger".

[1165] I searched the database but couldn't find a suitable answer.

[1166] Server escalation: Notify the support team of the inquiry details and the result of the "anger" emotion recognition.

[1167] Support Team: We review inquiries and contact users directly to resolve issues.

[1168] This embodiment allows the system to handle inquiries efficiently and with consideration for the user's feelings. Furthermore, by promptly escalating cases when an appropriate answer cannot be found or when the user's feelings are negative, customer satisfaction can be improved.

[1169] The following describes the processing flow.

[1170] Step 1:

[1171] User: Open the support chat screen and enter your inquiry into the text box. Type "How do I return a product?" and click the send button.

[1172] Step 2:

[1173] Server: Receives queries sent by users. The server stores the received query content in an appropriate format.

[1174] Step 3:

[1175] Server: Passes the received query content to a natural language processing (NLP) library. The NLP library performs grammatical analysis and word segmentation of the text and extracts the main keywords (in this example, "product," "return," and "method").

[1176] Step 4:

[1177] Server: Based on the extracted keywords, the server passes the query to the emotion engine. The emotion engine analyzes the user's emotions from the text and identifies emotions such as confusion, anger, and joy. In this example, the user's emotion is recognized as "confused".

[1178] Step 5:

[1179] Server: Searches the internal database based on extracted keywords and sentiment recognition results. Generates search queries and accesses the database index to retrieve relevant information.

[1180] Step 6:

[1181] Server: Generates a response to provide to the user based on information retrieved from the database. This response is formatted in a natural style and adjusted based on the results of the sentiment engine. For example, polite expressions may be added, such as, "To return the product, please return it in its original packaging within 30 days of purchase. We apologize for any inconvenience this may cause."

[1182] Step 7:

[1183] Server: Sends the generated response to the user's support chat screen. Displays the response appropriately so that the user can review it.

[1184] Step 8:

[1185] User: Review the response displayed on the support chat screen and make additional inquiries if necessary.

[1186] Step 9:

[1187] Server: If a suitable answer cannot be found, or if the inquiry is complex and difficult to handle automatically, the server will initiate escalation. Escalation will also occur if the user's emotions are very negative.

[1188] Step 10:

[1189] Server: Set the escalation flag and log the query content and sentiment analysis results.

[1190] Step 11:

[1191] Server: Generates an escalation request and sends a notification to the support team. The notification includes the inquiry details and sentiment analysis results.

[1192] Step 12:

[1193] Terminal (e.g., support team's PC): An escalation notification will appear on the terminal of a support team member. The support team will review the notification and begin taking detailed action.

[1194] Step 13:

[1195] Support Team: Review the details of the escalated inquiry and contact the user directly to resolve the issue.

[1196] (Example 2)

[1197] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1198] Traditional customer support systems analyze inquiries and generate appropriate responses, but they fail to consider customer emotions, resulting in insufficient improvement in customer satisfaction. Furthermore, delays in escalation when an appropriate answer cannot be found can lead to service delays.

[1199] The identification processing performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for receiving customer inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the customer, means for escalating the customer's inquiry to the support team if no suitable answer is found, means for analyzing the customer's emotions included in the inquiry, and means for adjusting the answer based on the analyzed customer emotions. This makes it possible to respond in a way that takes customer emotions into consideration, thereby improving customer satisfaction. In addition, if an appropriate answer cannot be found, it is quickly escalated to the support team, preventing delays in response.

[1200] "Customer" refers to anyone who uses the system to receive services or support.

[1201] "Inquiry details" refers to the text information of questions or requests that customers enter or submit when they require support.

[1202] "Means of receiving information" refers to the function for importing customer inquiries into the system.

[1203] "Natural language processing" refers to the technology that enables computers to understand and analyze text written in human language.

[1204] "Means of analysis" refers to the function that uses natural language processing to analyze the meaning and intent of received inquiries.

[1205] "Means of searching the database" refers to the function of searching for relevant information from the internal database based on the analyzed query content.

[1206] "Means of generating responses" refers to a function that creates responses tailored to the customer based on information obtained from a database.

[1207] "Means of transmission" refers to the function used to communicate the generated response to the customer.

[1208] "Escalation options" refer to the function that allows you to transfer your inquiry to the support team if you cannot find a suitable answer or under certain conditions.

[1209] "Methods for analyzing emotions" refers to technologies used to identify customer emotions from the content of inquiries.

[1210] "Means of adjusting responses" refers to a function that optimizes the content and style of responses based on analyzed customer sentiment.

[1211] The system according to the present invention automatically processes customer inquiries and generates appropriate responses. Furthermore, by having a function to recognize customer emotions and adjust responses accordingly, it enables more accurate customer service.

[1212] System Configuration

[1213] 1. Receiving customer inquiries

[1214] Users access the support chat using their device and enter their questions or requests. At this time, the customer's inquiry is sent to the system server.

[1215] 2. Natural Language Processing (NLP) of the inquiry content

[1216] The server uses an NLP library (e.g., SpaCy or NLTK) to analyze the received query. Specifically, it performs tokenization of the text (splitting it into individual words), part-of-speech tagging, and keyword extraction. This analysis extracts the important keywords from the query.

[1217] 3. Recognizing customer emotions

[1218] The server uses an emotion engine (such as Google Cloud Natural Language API or IBM Watson Tone Analyzer) to analyze customer emotions from inquiries. For example, emotions such as "confused," "angry," and "grateful" may be recognized.

[1219] 4. Searching the internal database

[1220] The server searches its internal database for relevant information based on the analyzed keywords and sentiment data. This process efficiently retrieves information using SQL queries.

[1221] 5. Generating and adjusting responses

[1222] The server generates appropriate responses using a generative AI model (e.g., OpenAI GPT-3) based on database search results. It also adjusts the style and content of the responses based on the results of the sentiment engine. For example, if a customer is confused, a polite clarification such as "We apologize for the inconvenience" might be added.

[1223] 6. Submit your response

[1224] The server sends the generated response to the user and displays it on the support chat screen. This allows the user to see the answer immediately.

[1225] 7. Escalation as needed

[1226] If the server cannot find a suitable answer or if the customer's sentiment is very negative, it will notify the support team of the inquiry details and sentiment analysis results. This allows the support team to respond quickly.

[1227] Specific example

[1228] Specific Example 1

[1229] User inquiry: "How do I return an item?"

[1230] Processing flow:

[1231] The user enters and submits their inquiry.

[1232] The server receives the query.

[1233] The server uses NLP to extract keywords such as "product," "return," and "method."

[1234] The server uses an emotion engine to recognize that the user is "confused."

[1235] The server searches the database and generates a response message stating, "To return an item, please return it in its original packaging within 30 days of purchase."

[1236] The server adds a note to the response such as, "We apologize for any inconvenience this may cause."

[1237] The server sends the adjusted response to the user.

[1238] Specific example 2 (escalation)

[1239] User inquiry: "The quality of this product is terrible. I want to return it."

[1240] Processing flow:

[1241] The user enters and submits their inquiry.

[1242] The server receives the query.

[1243] The server uses NLP to analyze the data and extract information on "products," "quality," and "returns."

[1244] The server uses an emotion engine to recognize that the user is feeling "anger."

[1245] The server searches the database but cannot find a suitable answer.

[1246] The server notifies the support team of the inquiry details and the result of the "anger" emotion recognition.

[1247] The support team will review the inquiry and contact the user directly to resolve the issue.

[1248] This invention enables the system to handle inquiries efficiently and with consideration for the user's emotions. By promptly escalating cases when an appropriate answer cannot be found or when the user's emotions are negative, customer satisfaction can be improved.

[1249] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1250] Step 1:

[1251] The user accesses the support chat using their device, enters their inquiry, and submits it. For example, the text entered in the inquiry might be, "How do I return a product?" The entered data is then sent to the server.

[1252] Step 2:

[1253] The server receives the query. The received data is logged as is and passed on to the next analysis process. The input is text data, and the output is similarly text data for analysis.

[1254] Step 3:

[1255] The server uses an NLP library (e.g., SpaCy or NLTK) to analyze the received text data. Specifically, it performs text tokenization, part-of-speech tagging, and keyword extraction. The input is the received text data, and the output is a list of keywords (e.g., "product," "return," "method").

[1256] Step 4:

[1257] The server uses an internal sentiment engine (e.g., Google Cloud Natural Language API or IBM Watson Tone Analyzer) to analyze the user's sentiment in the query. The input is received text data, and the output is sentiment tags (e.g., "confused").

[1258] Step 5:

[1259] The server searches its internal database based on the analyzed keywords and sentiment data. Specifically, it uses SQL queries to retrieve relevant information. The input is a list of keywords and sentiment tags, and the output is the appropriate response data.

[1260] Step 6:

[1261] The server uses a generative AI model (e.g., OpenAI GPT-3) to generate appropriate responses based on the search results. Here, the responses are refined based on the results of the sentiment engine. The input is the response data and sentiment tags, and the output is the refined response text (e.g., "To return the product, please return it in its original packaging within 30 days of purchase. We apologize for any inconvenience.").

[1262] Step 7:

[1263] The server sends the generated response to the user. The response is displayed on the support chat screen, where the user can review it. The input is the edited response, and the output is what is displayed on the user's screen.

[1264] Step 8:

[1265] If a suitable answer cannot be found, or if the user's sentiment is very negative, the server escalates the inquiry details and sentiment analysis results to the support team. The input is the inquiry details and sentiment tags obtained from the previous processing, and the output is a notification to the support team.

[1266] (Application Example 2)

[1267] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1268] While conventional customer service systems can automatically generate appropriate responses to inquiries, they struggle to improve customer satisfaction because they cannot consider customer emotions. Furthermore, systems often lack the flexibility to adapt to situations requiring changes in writing style based on emotions or, in some cases, rapid escalation.

[1269] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for receiving customer inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the customer, means for escalating the customer's inquiry to a support team if no suitable answer is found, means for recognizing the customer's emotions regarding the inquiry, and means for adjusting the generated answer based on the recognized emotions. This enables automated responses that take emotions into consideration and rapid escalation.

[1270] A "customer" is a user who makes an inquiry using this system.

[1271] "Inquiry content" refers to questions and requests that customers submit through this system.

[1272] "Natural language processing" is a technology that enables computers to understand and analyze human language.

[1273] "Analyzing" means converting input data into a format that can be understood, either partially or entirely.

[1274] A "database" is an electronic information storage system that systematically accumulates information.

[1275] "Generating" means creating new data or information according to a certain algorithm or rule.

[1276] "To send" means to transfer specific data from one location to another.

[1277] "Escalation" is the act of passing on a difficult problem to a higher-ranking or specialized person.

[1278] A "support team" is a group of specialists responsible for handling customer inquiries and resolving problems.

[1279] "Recognizing emotions" means analyzing customer inquiries and identifying the types of emotions they contain.

[1280] "Adjusting" means appropriately changing the information and responses provided to match the perceived emotions.

[1281] A "system" is a mechanism that provides a certain function through the coordinated action of multiple components or means.

[1282] The system for implementing this invention analyzes customer inquiries, recognizes their emotions, and generates and adjusts appropriate responses based on those emotions in a content distribution service using smartphones. The specific operation of the system will be described below.

[1283] First, the user submits an inquiry using a smartphone app. The user's inquiry is in the format of, for example, "Please tell me about the latest TV dramas." The server receives this inquiry and analyzes it using a natural language processing (NLP) library. During this process, SpaCy is used to extract keywords from the inquiry and identify the subject of the inquiry.

[1284] Next, the server uses VADER Sentiment to perform sentiment analysis on the inquiry and recognizes the user's sentiment as either "positive," "negative," or "neutral." Based on this sentiment data and extracted keywords, the server searches the SQLite database and retrieves the relevant response. The retrieved response is then adjusted by the sentiment engine based on the user's sentiment. For example, if the user is "positive," the response will be written in a friendly style.

[1285] The adjusted response is sent from the server to the user's smartphone and displayed on the app screen. If a suitable response does not exist in the database, or if the user's sentiment is very strong and negative, the server escalates the inquiry to the support team. During this escalation process, the support team is notified of the detailed inquiry and the results of the sentiment analysis.

[1286] Hardware and software to be used

[1287] Hardware: Smartphone (iOS or Android)

[1288] Software: Flask (Python framework), SpaCy (natural language processing library), VADER Sentiment (sentiment analysis library), SQLite (database)

[1289] Specific example

[1290] For example, if a user sends an inquiry asking "Please tell me the latest TV dramas," the server will process it as follows:

[1291] 1. Receive user inquiries.

[1292] 2. Analyze the inquiries and extract the keywords "latest" and "drama".

[1293] 3. Conduct sentiment analysis and recognize the user's emotions as "neutral."

[1294] 4. Search the database and obtain the answer, "The latest drama is XX."

[1295] 5. Adjust the response based on the user's emotions and generate an answer in the format "The latest drama is XX. (User's emotion: Neutral)".

[1296] 6. Send the generated response to the user and display it on the smartphone app screen.

[1297] Example of a prompt

[1298] 1. Inquiry: "Please tell me about the latest TV dramas."

[1299] 2. Answer: "The latest drama is XX. (User sentiment: Neutral)"

[1300] This makes it possible to provide automated responses to user inquiries that take emotions into consideration.

[1301] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1302] Step 1:

[1303] Users submit their inquiries through a smartphone app. The inquiry content is provided in text format. At this point, users ask specific questions such as, "Please tell me about the latest TV dramas." The output is the text data of the inquiry.

[1304] Step 2:

[1305] The server processes the received query. First, it passes the query to a natural language processing (NLP) library. The SpaCy library is used for this purpose. The input is the query text received in step 1, and the output is a list of keywords. Specifically, keywords such as "latest" and "drama" are extracted from the query.

[1306] Step 3:

[1307] The server performs sentiment analysis on the extracted keywords. The VADER Sentiment library is used for this purpose. The input consists of the keywords extracted in step 2 and the original query. The output is a sentiment classification (positive, negative, or neutral). For example, the query "Please tell me the latest TV dramas" would be classified as "neutral."

[1308] Step 4:

[1309] The server searches the database based on the extracted keywords and sentiment analysis results. It executes queries against the SQLite database to retrieve the relevant answers. The inputs are the keywords from step 2 and the sentiment classification results from step 3. The output is the corresponding answer text. Specific answers such as "The latest drama is XX" can be obtained.

[1310] Step 5:

[1311] The server adjusts the acquired responses based on the sentiment analysis results. This process involves changing the writing style and adding supplementary information according to the sentiment. The input is the response text acquired in step 4 and the sentiment classification result obtained in step 3. The output is the adjusted response text. For example, the response might reflect a "neutral" sentiment recognition, resulting in "The latest drama is XX. (User's sentiment: Neutral)."

[1312] Step 6:

[1313] The server sends the adjusted response to the user. To do this, it uses the Flask framework to return the response as an HTTP response. The input is the response text adjusted in step 5, and the output is the response displayed on the user's smartphone.

[1314] Step 7:

[1315] If a suitable answer cannot be retrieved from the database, or if the sentiment analysis indicates very strong negative emotions, the server escalates the inquiry to the support team. The input consists of the sentiment analysis results obtained in step 3 and the database search results that were deemed irrelevant in step 4. The output is a notification to the support team, which includes the inquiry details and the sentiment analysis results.

[1316] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1317] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1318] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[1319] [Fourth Embodiment]

[1320] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1321] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1322] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1323] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[1324] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1325] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1326] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1327] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[1328] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1329] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1330] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1331] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1332] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1333] As an example of a system according to the present invention, a system for automatically processing customer inquiries and generating appropriate responses will be described.

[1334] System Overview

[1335] This system includes the following main features:

[1336] 1. Receipt of inquiry

[1337] 2. Analysis of the inquiry content using natural language processing (NLP)

[1338] 3. Database Search

[1339] 4. Generating the answer

[1340] 5. Submit your response

[1341] 6. Escalation as needed

[1342] Processing flow

[1343] First, the user submits an inquiry using the support chat. For example, let's say the inquiry is, "How do I return a product?"

[1344] The server receives the inquiry sent by the user. It then analyzes the received string using a natural language processing (NLP) library. In this case, keywords such as "product," "return," and "method" are extracted from the text.

[1345] Next, the server uses the extracted keywords to search its internal database. This search utilizes an indexed database for faster information retrieval.

[1346] Once search results are obtained, the server generates a response to provide to the user based on them. For example, it might retrieve information from the database such as, "To return a product, please return it in its original packaging within 30 days of purchase," and then format this into a response.

[1347] The generated response is sent from the server to the user. The user can then view the response displayed on the support chat screen.

[1348] If the server cannot find a suitable answer, or if the inquiry is too complex, the system will use an escalation mechanism. In this case, the server will notify the support team of the details of the inquiry.

[1349] Specific example

[1350] 1. User inquiry: "How do I return an item?"

[1351] 2. Server processing:

[1352] Inquiry received

[1353] Analysis using NLP

[1354] Extract the keywords "product," "return," and "method."

[1355] Search database

[1356] The response generated reads: "To return an item, please return it in its original packaging within 30 days of purchase."

[1357] 3. Server submission: Send the generated response to the user.

[1358] 4. Escalation (additional example):

[1359] User inquiry: "How do I return an item from overseas?"

[1360] Server processing:

[1361] Inquiry received

[1362] Analysis using NLP

[1363] I searched the database but couldn't find a suitable answer.

[1364] Server escalation: Notify the support team of the inquiry details.

[1365] Support team: We will review your inquiry and contact you directly to provide further details.

[1366] This embodiment allows the system to efficiently respond to customer inquiries. Furthermore, it can improve customer satisfaction by promptly escalating cases when a suitable answer cannot be found.

[1367] The following describes the processing flow.

[1368] Step 1:

[1369] User: Open the support chat screen and enter your inquiry into the text box. Type "How do I return a product?" and click the send button.

[1370] Step 2:

[1371] Server: Receives queries sent by users. The server stores the received query content in an appropriate format.

[1372] Step 3:

[1373] Server: Passes the received query content to a natural language processing (NLP) library. The NLP library performs grammatical analysis and word segmentation of the text and extracts the main keywords (in this example, "product," "return," and "method").

[1374] Step 4:

[1375] Server: Searches the internal database based on the extracted keywords. Generates search queries and quickly accesses the database index to retrieve the appropriate information.

[1376] Step 5:

[1377] Server: Based on information retrieved from the database, it generates a response to provide to the user. This response is formatted as a natural-sounding sentence. For example, it might generate a response such as, "To return an item, please return it in its original packaging within 30 days of purchase."

[1378] Step 6:

[1379] Server: Sends the generated response to the user's support chat screen. Displays the response appropriately so that the user can review it.

[1380] Step 7:

[1381] User: Review the response displayed on the support chat screen and make additional inquiries if necessary.

[1382] Step 8:

[1383] Server: If a suitable answer cannot be found, or if the inquiry is complex and difficult to handle automatically, the server initiates escalation. It sets an escalation flag and logs the inquiry details and analysis results.

[1384] Step 9:

[1385] Server: Generates an escalation request and sends a notification to the support team. The notification includes detailed information about the inquiry.

[1386] Step 10:

[1387] Terminal (e.g., support team's PC): An escalation notification will appear on the terminal of a support team member. The support team will review the notification and begin taking detailed action.

[1388] Step 11:

[1389] Support Team: Review the details of the escalated inquiry and contact the user directly to resolve the issue.

[1390] (Example 1)

[1391] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1392] Traditional customer support systems consume significant time and resources by manually analyzing inquiries and generating appropriate responses. This can lead to delays in customer service and decreased customer satisfaction. Furthermore, escalation processes are often manual when appropriate answers cannot be found, adding to the overall time burden. A system is needed to address these challenges and provide rapid and accurate responses to customer inquiries.

[1393] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1394] In this invention, the server includes means for receiving user inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the user, and means for escalating the user's inquiry to a support team if no suitable answer is found. This makes it possible to respond to user inquiries quickly and automatically.

[1395] "User" refers to a person who makes inquiries to the system or an end-user.

[1396] "Inquiry content" refers to messages containing questions and requests that users send to the system.

[1397] A "server" refers to a hardware or software system that receives inquiries from users and performs processes such as analysis, searching, response generation, transmission, and escalation.

[1398] Natural Language Processing (NLP) refers to the techniques and algorithms used to analyze human language and understand its meaning.

[1399] A "database" refers to a collection of structured data designed for efficient searching and management of information.

[1400] "Means of searching" refers to a mechanism or process for retrieving information from a database based on a specific query.

[1401] "Means of generating answers" refers to the process of creating appropriate answer text to provide to the user based on the searched information.

[1402] "Means of transmission" refers to the communication technologies and protocols used to send the generated response to the user's device.

[1403] "Escalation means" refers to the process of forwarding an inquiry to other resources, such as a support team, when the system cannot find a suitable answer.

[1404] A "support team" refers to a group of individuals or individuals who provide additional human support in response to user inquiries.

[1405] As an example of a system according to the present invention, a system for automatically processing customer inquiries and generating appropriate responses will be described.

[1406] System Overview

[1407] This system includes the following main components:

[1408] 1. Means for receiving inquiries from users

[1409] 2. Means for analyzing received inquiry content using natural language processing (NLP)

[1410] 3. Means for searching the database based on the analyzed query content

[1411] 4. Means for generating appropriate answers from search results

[1412] 5. Means for sending the generated response to the user

[1413] 6. A means of escalating a user's inquiry to the support team if a suitable answer cannot be found.

[1414] Hardware and software usage examples

[1415] 1. Receiving means

[1416] Users submit inquiries via support chat using devices such as smartphones or computers. The server receives these inquiries over the internet.

[1417] 2. Natural Language Processing (NLP)

[1418] The server analyzes the received query using a natural language processing (NLP) library (e.g., Python's NLTK or SpaCy). Specifically, it divides the text into tokens and extracts keywords.

[1419] 3. Database Search

[1420] The server searches indexed databases (e.g., Elasticsearch or MySQL) using the extracted keywords. This search allows for the rapid retrieval of appropriate information in response to user queries.

[1421] 4. Answer generation means

[1422] The server generates answers to provide to the user based on the search results. For example, it organizes information obtained from the database into an easily understandable format.

[1423] 5. Transmission method

[1424] The server sends the generated response to the user's device. The user can then view the response on the support chat screen.

[1425] 6. Escalation measures

[1426] If the server cannot find a suitable answer, it will notify the support team of the inquiry details. This notification allows the support team to handle the inquiry manually.

[1427] Specific example

[1428] The following shows a specific example of operation.

[1429] 1. User inquiry: "How do I return an item?"

[1430] 2. Server processing:

[1431] Inquiry received

[1432] Analysis using NLP

[1433] Extract the keywords "product," "return," and "method."

[1434] Search database

[1435] The response generated reads: "To return an item, please return it in its original packaging within 30 days of purchase."

[1436] 3. Server submission: Send the generated response to the user.

[1437] If it is incorrect, we will escalate it as follows:

[1438] 1. User inquiry: "How do I return an item from overseas?"

[1439] 2. Server processing:

[1440] Inquiry received

[1441] Analysis using NLP

[1442] I searched the database but couldn't find a suitable answer.

[1443] 3. Server escalation: Notify the support team of the inquiry details.

[1444] 4. Support Team: They will review the inquiry and contact the user directly to resolve the issue.

[1445] Example of a prompt

[1446] By inputting prompt messages like the following into the AI ​​model, it generates appropriate answers to inquiries.

[1447] "Please tell me how to return an item."

[1448] "Please tell me how to return items from overseas."

[1449] This allows the system to respond to customer inquiries efficiently and automatically. Furthermore, if a suitable answer cannot be found, it can quickly escalate the issue, improving customer satisfaction.

[1450] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1451] Step 1:

[1452] A user submits an inquiry using the support chat. An example of the inquiry is, "How do I return a product?"

[1453] Input: User inquiry message (text format)

[1454] Output: JSON data containing the query details

[1455] Specific action: The user enters their inquiry into the chat window and presses the send button.

[1456] Step 2:

[1457] The server receives the user's inquiry.

[1458] Input: JSON data containing the user's inquiry message

[1459] Output: Received inquiry content (string format)

[1460] Specific operation: The server receives the query in real time and extracts the query string for analysis.

[1461] Step 3:

[1462] The server analyzes the received query content using a natural language processing (NLP) library.

[1463] Input: Received inquiry content (string format)

[1464] Output: Extracted keywords (list format)

[1465] Specific operation: The server uses a Python NLP library (e.g., NLTK, SpaCy) to tokenize the text and extract important keywords (e.g., "product", "return", "method").

[1466] Step 4:

[1467] The server searches the database using the extracted keywords.

[1468] Input: Extracted keywords (list format)

[1469] Output: Search results (information from the database)

[1470] Specific operation: The server uses an indexed database (e.g., Elasticsearch, MySQL) to execute queries to retrieve information that matches the keyword.

[1471] Step 5:

[1472] The server generates answers to provide to the user based on the search results.

[1473] Input: Search results (information from the database)

[1474] Output: Generated response (text format)

[1475] Specific operation: The server generates an appropriate response based on information retrieved from the database (e.g., "To return the product, please return it in its original packaging within 30 days of purchase").

[1476] Step 6:

[1477] The server sends the generated response to the user.

[1478] Input: Generated response (text format)

[1479] Output: Display of response to user terminal

[1480] Specific operation: The server sends the response to the user's chat window, where it is displayed on the user's device.

[1481] Step 7:

[1482] If the server cannot find a suitable answer, the user's inquiry will be escalated.

[1483] Input: If the search result is empty

[1484] Output: Notification to the support team

[1485] Specific actions: The server detects that there is no suitable answer and sends a notification to the support team containing details of the inquiry. It also sends a message to the user such as, "We're sorry, but we were unable to find a suitable answer at this time."

[1486] In this way, the system can respond to user inquiries efficiently and automatically, and can quickly escalate cases if a suitable answer cannot be found.

[1487] (Application Example 1)

[1488] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1489] Traditional customer inquiry systems often rely on manual processes for analyzing inquiries and generating responses, resulting in low efficiency. Furthermore, the difficulty of easily submitting inquiries via smartphones hinders the provision of adequate support in today's fast-paced customer service environment. In addition, escalation options for complex customer inquiries are limited, highlighting the need for further operational efficiency and improved customer satisfaction.

[1490] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1491] In this invention, the server includes means for receiving customer inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the customer, means for escalating the customer's inquiry to a support team if no suitable answer is found, means for notifying the customer of detailed information about the inquiry during the escalation, means for easily inputting customer inquiries using a smartphone, means for using a generative AI model to automatically extract keywords from the inquiry and generate an answer based on them, and means for displaying the generated answer as a prompt. This enables the automation and efficiency of inquiry processing, allows customers to make inquiries intuitively using their smartphones, and enhances the escalation means, thereby realizing prompt and accurate customer support.

[1492] A "customer" refers to someone who purchases or uses a product or service.

[1493] "Inquiry content" refers to questions and requests provided by customers.

[1494] "Means of receiving information" refers to the methods and mechanisms by which the system receives customer inquiries.

[1495] "Natural language processing" refers to the technology that allows computers to analyze and understand natural language data.

[1496] A "database" refers to a system in which information is systematically stored and can be searched and managed.

[1497] "Means of searching" refers to the methods and mechanisms for obtaining necessary information from a database.

[1498] "Means of generating answers" refers to methods and mechanisms for creating appropriate answers to provide to customers based on search results.

[1499] "Means of transmission" refers to the methods and mechanisms used to deliver the generated responses to customers.

[1500] "Means of escalation" refers to methods and mechanisms for transferring inquiries that the system cannot handle to the support team.

[1501] "Means of notification" refers to the methods and mechanisms for informing the support team of the details of an inquiry during an escalation.

[1502] A "smartphone" refers to a portable information device that has telephone functionality while also being able to run applications and connect to the internet.

[1503] "Means for easily submitting inquiries" refers to methods and systems that allow customers to easily submit inquiries using devices such as smartphones.

[1504] "Methods for automatically extracting keywords" refers to methods and mechanisms that use natural language processing to extract important terms from query content.

[1505] A "generative AI model" refers to a model that uses artificial intelligence technology to automatically generate answers based on the content of an inquiry.

[1506] A "prompt message" refers to the guidance or response message provided to the customer based on the generated answer.

[1507] System Overview

[1508] The system for implementing this invention automates a series of processes and can respond quickly and accurately to customer inquiries. It mainly uses a server, a smartphone terminal, a natural language processing (NLP) library, a generative AI model, and a database.

[1509] Hardware and software

[1510] Server: The central hardware that receives, processes, parses, generates answers for, and escalates queries.

[1511] Smartphone: A device used by customers to enter inquiries.

[1512] NLP libraries: Software used to analyze natural language. A specific example is spaCy ("ja_core_news_sm" model).

[1513] Generative AI Model: An artificial intelligence model that generates appropriate answers from the content of an inquiry.

[1514] Database: Stores frequently occurring queries and their answers, enabling rapid data retrieval.

[1515] System Implementation Method

[1516] First, the user enters their inquiry using their smartphone. For example, the user might ask, "When will the product I ordered arrive?"

[1517] The server receives this query and analyzes it using an NLP library. Specifically, it extracts keywords such as "order," "product," and "deliver" from the query text. This clarifies the category and intent of the query.

[1518] Next, the database is searched based on the extracted keywords. The database contains pre-indexed response data, which can be searched efficiently. For example, suppose the response "Ordered items are usually shipped within 3-5 business days" is found.

[1519] Based on the responses found, a generative AI model generates answers. This process uses AI to generate more specific and appropriate responses to the original inquiry.

[1520] The generated response is sent from the server to the smartphone device, where the user confirms it. For example, the response might say, "Your order will usually be shipped within 3-5 business days."

[1521] If the server cannot find a suitable answer, the system escalates the inquiry details to the support team. The support team is also notified of the detailed inquiry, allowing them to respond quickly.

[1522] Specific example

[1523] 1. User inquiry: "How do I return an item?"

[1524] 2. Server processing:

[1525] Inquiry received

[1526] Analysis using NLP

[1527] Extract the keywords "product," "return," and "method."

[1528] Search database

[1529] I received the response, "To return the product, please return it in its original packaging within 30 days of purchase."

[1530] 3. Submit the generated response:

[1531] The response sent to the smartphone was: "To return the product, please return it in its original packaging within 30 days of purchase."

[1532] 4. Examples of escalation:

[1533] User inquiry: "How do I return an item from overseas?"

[1534] Server processing:

[1535] Inquiry received

[1536] Analysis using NLP

[1537] I searched the database but couldn't find a suitable answer.

[1538] Server escalation:

[1539] Please notify the support team of the details of your inquiry.

[1540] Support team:

[1541] We will review the inquiry and contact the user directly to provide further details.

[1542] Example of a prompt

[1543] "Question to the Generative AI Model: A user has submitted the following inquiry: 'When will my ordered item arrive?' The keywords analyzed using NLP are 'order,' 'item,' and 'arrive.' Please provide as much specific information as possible regarding the delivery timing."

[1544] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1545] Step 1:

[1546] The user enters their inquiry using their smartphone and presses the send button. Specifically, the user enters "When will my ordered items arrive?". This input is the user's inquiry.

[1547] Step 2:

[1548] The smartphone device sends the entered inquiry content to the server. The input is the user's inquiry, and the output is the data sent to the server.

[1549] Step 3:

[1550] The server uses an NLP library to analyze the received query content. Specifically, it uses the spaCy "ja_core_news_sm" model to extract keywords from the query content. In this case, the input is the query content, and the output is the extracted keywords (such as "order," "product," and "deliver").

[1551] Step 4:

[1552] The server searches the database based on the extracted keywords. The database stores pre-indexed response data. The input is the extracted keywords, and the output is the corresponding response data (e.g., "Your order will be shipped within 3-5 business days").

[1553] Step 5:

[1554] The server generates an answer using a generative AI model based on the search results. In this process, the AI ​​model generates the optimal answer based on the query and search results. The input is the search results and the query, and the output is the generated answer.

[1555] Step 6:

[1556] The server sends the generated response text to the smartphone device. The input is the generated response text, and the output is the data sent to the smartphone device.

[1557] Step 7:

[1558] The smartphone displays the response text received from the server to the user. The input is the response data from the server, and the output is the response text displayed on the smartphone screen.

[1559] Step 8:

[1560] If the server cannot find a suitable answer, the issue will be escalated to the support team. Specifically, a notification containing detailed information about the inquiry will be sent to the support team. The input is the inquiry requiring support, and the output is the notification data sent to the support team.

[1561] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1562] The system according to the present invention is a system for automatically processing user inquiries and generating appropriate responses. Furthermore, by combining it with an emotion engine that recognizes user emotions and adjusts responses based on those emotions, it achieves more accurate customer service.

[1563] System Overview

[1564] This system includes the following main features:

[1565] 1. Receipt of inquiry

[1566] 2. Analysis of the inquiry content using natural language processing (NLP)

[1567] 3. User emotion recognition by an emotion engine

[1568] 4. Database Search

[1569] 5. Generating and adjusting responses

[1570] 6. Submit your response

[1571] 7. Escalation as needed

[1572] Processing flow

[1573] First, the user submits an inquiry using the support chat. For example, let's say the inquiry is, "How do I return a product?"

[1574] The server receives inquiries sent by users. It then passes the received inquiry content to a natural language processing (NLP) library for analysis. Here, keywords such as "product," "return," and "method" are extracted from the text.

[1575] Next, the server uses its built-in emotion engine to analyze the user's emotions in the inquiry. This analysis determines whether the user is feeling, for example, "negative" if they are dissatisfied, or "positive" if they are grateful.

[1576] Based on the analyzed keywords and sentiment data, the server searches its internal database to retrieve appropriate information. Based on the retrieved information, it generates a response to provide to the user. During this process, the sentiment engine adjusts the response based on the user's emotions. For example, if the user is angry, a more polite and intimate tone will be used.

[1577] The generated response is sent from the server to the user and displayed on the support chat screen. The user reviews the provided response and asks further questions if necessary.

[1578] If the server cannot find a suitable answer, or if the user's sentiment is very negative, the system will use escalation mechanisms. In this case, the server will notify the support team of the details of the inquiry and the sentiment analysis results.

[1579] Specific example

[1580] 1. User inquiry: "How do I return an item?"

[1581] 2. Server processing:

[1582] Inquiry received

[1583] Analysis using NLP

[1584] Extract the keywords "product," "return," and "method."

[1585] The emotion engine recognizes that the user is "confused".

[1586] Search database

[1587] The response text generated reads: "To return an item, please return it in its original packaging within 30 days of purchase."

[1588] Add a supplement to your response such as, "We apologize for any inconvenience this may cause."

[1589] 3. Server submission: Send the adjusted response to the user.

[1590] 4. Escalation (additional example):

[1591] User inquiry: "The quality of this product is terrible. I want to return it."

[1592] Server processing:

[1593] Inquiry received

[1594] Analysis using NLP

[1595] Extract the keywords "product," "quality," and "returns."

[1596] The emotion engine recognizes that the user is feeling "anger".

[1597] I searched the database but couldn't find a suitable answer.

[1598] Server escalation: Notify the support team of the inquiry details and the result of the "anger" emotion recognition.

[1599] Support Team: We review inquiries and contact users directly to resolve issues.

[1600] This embodiment allows the system to handle inquiries efficiently and with consideration for the user's feelings. Furthermore, by promptly escalating cases when an appropriate answer cannot be found or when the user's feelings are negative, customer satisfaction can be improved.

[1601] The following describes the processing flow.

[1602] Step 1:

[1603] User: Open the support chat screen and enter your inquiry into the text box. Type "How do I return a product?" and click the send button.

[1604] Step 2:

[1605] Server: Receives queries sent by users. The server stores the received query content in an appropriate format.

[1606] Step 3:

[1607] Server: Passes the received query content to a natural language processing (NLP) library. The NLP library performs grammatical analysis and word segmentation of the text and extracts the main keywords (in this example, "product," "return," and "method").

[1608] Step 4:

[1609] Server: Based on the extracted keywords, the server passes the query to the emotion engine. The emotion engine analyzes the user's emotions from the text and identifies emotions such as confusion, anger, and joy. In this example, the user's emotion is recognized as "confused".

[1610] Step 5:

[1611] Server: Searches the internal database based on extracted keywords and sentiment recognition results. Generates search queries and accesses the database index to retrieve relevant information.

[1612] Step 6:

[1613] Server: Generates a response to provide to the user based on information retrieved from the database. This response is formatted in a natural style and adjusted based on the results of the sentiment engine. For example, polite expressions may be added, such as, "To return the product, please return it in its original packaging within 30 days of purchase. We apologize for any inconvenience this may cause."

[1614] Step 7:

[1615] Server: Sends the generated response to the user's support chat screen. Displays the response appropriately so that the user can review it.

[1616] Step 8:

[1617] User: Review the response displayed on the support chat screen and make additional inquiries if necessary.

[1618] Step 9:

[1619] Server: If a suitable answer cannot be found, or if the inquiry is complex and difficult to handle automatically, the server will initiate escalation. Escalation will also occur if the user's emotions are very negative.

[1620] Step 10:

[1621] Server: Set the escalation flag and log the query content and sentiment analysis results.

[1622] Step 11:

[1623] Server: Generates an escalation request and sends a notification to the support team. The notification includes the inquiry details and sentiment analysis results.

[1624] Step 12:

[1625] Terminal (e.g., support team's PC): An escalation notification will appear on the terminal of a support team member. The support team will review the notification and begin taking detailed action.

[1626] Step 13:

[1627] Support Team: Review the details of the escalated inquiry and contact the user directly to resolve the issue.

[1628] (Example 2)

[1629] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1630] Traditional customer support systems analyze inquiries and generate appropriate responses, but they fail to consider customer emotions, resulting in insufficient improvement in customer satisfaction. Furthermore, delays in escalation when an appropriate answer cannot be found can lead to service delays.

[1631] The identification processing performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for receiving customer inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the customer, means for escalating the customer's inquiry to the support team if no suitable answer is found, means for analyzing the customer's emotions included in the inquiry, and means for adjusting the answer based on the analyzed customer emotions. This makes it possible to respond in a way that takes customer emotions into consideration, thereby improving customer satisfaction. In addition, if an appropriate answer cannot be found, it is quickly escalated to the support team, preventing delays in response.

[1632] "Customer" refers to anyone who uses the system to receive services or support.

[1633] "Inquiry details" refers to the text information of questions or requests that customers enter or submit when they require support.

[1634] "Means of receiving information" refers to the function for importing customer inquiries into the system.

[1635] "Natural language processing" refers to the technology that enables computers to understand and analyze text written in human language.

[1636] "Means of analysis" refers to the function that uses natural language processing to analyze the meaning and intent of received inquiries.

[1637] "Means of searching the database" refers to the function of searching for relevant information from the internal database based on the analyzed query content.

[1638] "Means of generating responses" refers to a function that creates responses tailored to the customer based on information obtained from a database.

[1639] "Means of transmission" refers to the function used to communicate the generated response to the customer.

[1640] "Escalation options" refer to the function that allows you to transfer your inquiry to the support team if you cannot find a suitable answer or under certain conditions.

[1641] "Methods for analyzing emotions" refers to technologies used to identify customer emotions from the content of inquiries.

[1642] "Means of adjusting responses" refers to a function that optimizes the content and style of responses based on analyzed customer sentiment.

[1643] The system according to the present invention automatically processes customer inquiries and generates appropriate responses. Furthermore, by having a function to recognize customer emotions and adjust responses accordingly, it enables more accurate customer service.

[1644] System Configuration

[1645] 1. Receiving customer inquiries

[1646] Users access the support chat using their device and enter their questions or requests. At this time, the customer's inquiry is sent to the system server.

[1647] 2. Natural Language Processing (NLP) of the inquiry content

[1648] The server uses an NLP library (e.g., SpaCy or NLTK) to analyze the received query. Specifically, it performs tokenization of the text (splitting it into individual words), part-of-speech tagging, and keyword extraction. This analysis extracts the important keywords from the query.

[1649] 3. Recognizing customer emotions

[1650] The server uses an emotion engine (such as Google Cloud Natural Language API or IBM Watson Tone Analyzer) to analyze customer emotions from inquiries. For example, emotions such as "confused," "angry," and "grateful" may be recognized.

[1651] 4. Searching the internal database

[1652] The server searches its internal database for relevant information based on the analyzed keywords and sentiment data. This process efficiently retrieves information using SQL queries.

[1653] 5. Generating and adjusting responses

[1654] The server generates appropriate responses using a generative AI model (e.g., OpenAI GPT-3) based on database search results. It also adjusts the style and content of the responses based on the results of the sentiment engine. For example, if a customer is confused, a polite clarification such as "We apologize for the inconvenience" might be added.

[1655] 6. Submit your response

[1656] The server sends the generated response to the user and displays it on the support chat screen. This allows the user to see the answer immediately.

[1657] 7. Escalation as needed

[1658] If the server cannot find a suitable answer or if the customer's sentiment is very negative, it will notify the support team of the inquiry details and sentiment analysis results. This allows the support team to respond quickly.

[1659] Specific example

[1660] Specific Example 1

[1661] User inquiry: "How do I return an item?"

[1662] Processing flow:

[1663] The user enters and submits their inquiry.

[1664] The server receives the query.

[1665] The server uses NLP to extract keywords such as "product," "return," and "method."

[1666] The server uses an emotion engine to recognize that the user is "confused."

[1667] The server searches the database and generates a response message stating, "To return an item, please return it in its original packaging within 30 days of purchase."

[1668] The server adds a note to the response such as, "We apologize for any inconvenience this may cause."

[1669] The server sends the adjusted response to the user.

[1670] Specific example 2 (escalation)

[1671] User inquiry: "The quality of this product is terrible. I want to return it."

[1672] Processing flow:

[1673] The user enters and submits their inquiry.

[1674] The server receives the query.

[1675] The server uses NLP to analyze the data and extract information on "products," "quality," and "returns."

[1676] The server uses an emotion engine to recognize that the user is feeling "anger."

[1677] The server searches the database but cannot find a suitable answer.

[1678] The server notifies the support team of the inquiry details and the result of the "anger" emotion recognition.

[1679] The support team will review the inquiry and contact the user directly to resolve the issue.

[1680] This invention enables the system to handle inquiries efficiently and with consideration for the user's emotions. By promptly escalating cases when an appropriate answer cannot be found or when the user's emotions are negative, customer satisfaction can be improved.

[1681] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1682] Step 1:

[1683] The user accesses the support chat using their device, enters their inquiry, and submits it. For example, the text entered in the inquiry might be, "How do I return a product?" The entered data is then sent to the server.

[1684] Step 2:

[1685] The server receives the query. The received data is logged as is and passed on to the next analysis process. The input is text data, and the output is similarly text data for analysis.

[1686] Step 3:

[1687] The server uses an NLP library (e.g., SpaCy or NLTK) to analyze the received text data. Specifically, it performs text tokenization, part-of-speech tagging, and keyword extraction. The input is the received text data, and the output is a list of keywords (e.g., "product," "return," "method").

[1688] Step 4:

[1689] The server uses an internal sentiment engine (e.g., Google Cloud Natural Language API or IBM Watson Tone Analyzer) to analyze the user's sentiment in the query. The input is received text data, and the output is sentiment tags (e.g., "confused").

[1690] Step 5:

[1691] The server searches its internal database based on the analyzed keywords and sentiment data. Specifically, it uses SQL queries to retrieve relevant information. The input is a list of keywords and sentiment tags, and the output is the appropriate response data.

[1692] Step 6:

[1693] The server uses a generative AI model (e.g., OpenAI GPT-3) to generate appropriate responses based on the search results. Here, the responses are refined based on the results of the sentiment engine. The input is the response data and sentiment tags, and the output is the refined response text (e.g., "To return the product, please return it in its original packaging within 30 days of purchase. We apologize for any inconvenience.").

[1694] Step 7:

[1695] The server sends the generated response to the user. The response is displayed on the support chat screen, where the user can review it. The input is the edited response, and the output is what is displayed on the user's screen.

[1696] Step 8:

[1697] If a suitable answer cannot be found, or if the user's sentiment is very negative, the server escalates the inquiry details and sentiment analysis results to the support team. The input is the inquiry details and sentiment tags obtained from the previous processing, and the output is a notification to the support team.

[1698] (Application Example 2)

[1699] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1700] While conventional customer service systems can automatically generate appropriate responses to inquiries, they struggle to improve customer satisfaction because they cannot consider customer emotions. Furthermore, systems often lack the flexibility to adapt to situations requiring changes in writing style based on emotions or, in some cases, rapid escalation.

[1701] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for receiving customer inquiries, means for analyzing the inquiry using natural language processing, means for searching a database based on the analyzed inquiry, means for generating an appropriate answer from the search results, means for sending the generated answer to the customer, means for escalating the customer's inquiry to a support team if no suitable answer is found, means for recognizing the customer's emotions regarding the inquiry, and means for adjusting the generated answer based on the recognized emotions. This enables automated responses that take emotions into consideration and rapid escalation.

[1702] A "customer" is a user who makes an inquiry using this system.

[1703] "Inquiry content" refers to questions and requests that customers submit through this system.

[1704] "Natural language processing" is a technology that enables computers to understand and analyze human language.

[1705] "Analyzing" means converting input data into a format that can be understood, either partially or entirely.

[1706] A "database" is an electronic information storage system that systematically accumulates information.

[1707] "Generating" means creating new data or information according to a certain algorithm or rule.

[1708] "To send" means to transfer specific data from one location to another.

[1709] "Escalation" is the act of passing on a difficult problem to a higher-ranking or specialized person.

[1710] A "support team" is a group of specialists responsible for handling customer inquiries and resolving problems.

[1711] "Recognizing emotions" means analyzing customer inquiries and identifying the types of emotions they contain.

[1712] "Adjusting" means appropriately changing the information and responses provided to match the perceived emotions.

[1713] A "system" is a mechanism that provides a certain function through the coordinated action of multiple components or means.

[1714] The system for implementing this invention analyzes customer inquiries, recognizes their emotions, and generates and adjusts appropriate responses based on those emotions in a content distribution service using smartphones. The specific operation of the system will be described below.

[1715] First, the user submits an inquiry using a smartphone app. The user's inquiry is in the format of, for example, "Please tell me about the latest TV dramas." The server receives this inquiry and analyzes it using a natural language processing (NLP) library. During this process, SpaCy is used to extract keywords from the inquiry and identify the subject of the inquiry.

[1716] Next, the server uses VADER Sentiment to perform sentiment analysis on the inquiry and recognizes the user's sentiment as either "positive," "negative," or "neutral." Based on this sentiment data and extracted keywords, the server searches the SQLite database and retrieves the relevant response. The retrieved response is then adjusted by the sentiment engine based on the user's sentiment. For example, if the user is "positive," the response will be written in a friendly style.

[1717] The adjusted response is sent from the server to the user's smartphone and displayed on the app screen. If a suitable response does not exist in the database, or if the user's sentiment is very strong and negative, the server escalates the inquiry to the support team. During this escalation process, the support team is notified of the detailed inquiry and the results of the sentiment analysis.

[1718] Hardware and software to be used

[1719] Hardware: Smartphone (iOS or Android)

[1720] Software: Flask (Python framework), SpaCy (natural language processing library), VADER Sentiment (sentiment analysis library), SQLite (database)

[1721] Specific example

[1722] For example, if a user sends an inquiry asking "Please tell me the latest TV dramas," the server will process it as follows:

[1723] 1. Receive user inquiries.

[1724] 2. Analyze the inquiries and extract the keywords "latest" and "drama".

[1725] 3. Conduct sentiment analysis and recognize the user's emotions as "neutral."

[1726] 4. Search the database and obtain the answer, "The latest drama is XX."

[1727] 5. Adjust the response based on the user's emotions and generate an answer in the format "The latest drama is XX. (User's emotion: Neutral)".

[1728] 6. Send the generated response to the user and display it on the smartphone app screen.

[1729] Example of a prompt

[1730] 1. Inquiry: "Please tell me about the latest TV dramas."

[1731] 2. Answer: "The latest drama is XX. (User sentiment: Neutral)"

[1732] This makes it possible to provide automated responses to user inquiries that take emotions into consideration.

[1733] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1734] Step 1:

[1735] Users submit their inquiries through a smartphone app. The inquiry content is provided in text format. At this point, users ask specific questions such as, "Please tell me about the latest TV dramas." The output is the text data of the inquiry.

[1736] Step 2:

[1737] The server processes the received query. First, it passes the query to a natural language processing (NLP) library. The SpaCy library is used for this purpose. The input is the query text received in step 1, and the output is a list of keywords. Specifically, keywords such as "latest" and "drama" are extracted from the query.

[1738] Step 3:

[1739] The server performs sentiment analysis on the extracted keywords. The VADER Sentiment library is used for this purpose. The input consists of the keywords extracted in step 2 and the original query. The output is a sentiment classification (positive, negative, or neutral). For example, the query "Please tell me the latest TV dramas" would be classified as "neutral."

[1740] Step 4:

[1741] The server searches the database based on the extracted keywords and sentiment analysis results. It executes queries against the SQLite database to retrieve the relevant answers. The inputs are the keywords from step 2 and the sentiment classification results from step 3. The output is the corresponding answer text. Specific answers such as "The latest drama is XX" can be obtained.

[1742] Step 5:

[1743] The server adjusts the acquired responses based on the sentiment analysis results. This process involves changing the writing style and adding supplementary information according to the sentiment. The input is the response text acquired in step 4 and the sentiment classification result obtained in step 3. The output is the adjusted response text. For example, the response might reflect a "neutral" sentiment recognition, resulting in "The latest drama is XX. (User's sentiment: Neutral)."

[1744] Step 6:

[1745] The server sends the adjusted response to the user. To do this, it uses the Flask framework to return the response as an HTTP response. The input is the response text adjusted in step 5, and the output is the response displayed on the user's smartphone.

[1746] Step 7:

[1747] If a suitable answer cannot be retrieved from the database, or if the sentiment analysis indicates very strong negative emotions, the server escalates the inquiry to the support team. The input consists of the sentiment analysis results obtained in step 3 and the database search results that were deemed irrelevant in step 4. The output is a notification to the support team, which includes the inquiry details and the sentiment analysis results.

[1748] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1749] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1750] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1751] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1752] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1753] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1754] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1755] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1756] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1757] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[1758] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[1759] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[1760] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[1761] 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.

[1762] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[1763] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[1764] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[1765] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[1766] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[1767] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[1768] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[1769] The following is further disclosed regarding the embodiments described above.

[1770] (Claim 1)

[1771] A means of receiving customer inquiries,

[1772] A means for analyzing the aforementioned query content using natural language processing,

[1773] A means for searching a database based on the analyzed query content,

[1774] A means for generating an appropriate answer from the aforementioned search results,

[1775] A means for sending the generated response to the customer,

[1776] A means of escalating a customer inquiry to the support team if a suitable answer cannot be found,

[1777] A system that includes this.

[1778] (Claim 2)

[1779] The system according to claim 1, characterized in that the natural language processing extracts keywords from the query content and determines the category of the query based on these keywords.

[1780] (Claim 3)

[1781] The system according to claim 1, characterized in that the escalation means includes a notification means for providing detailed information of the inquiry to the support team.

[1782] "Example 1"

[1783] (Claim 1)

[1784] A means of receiving inquiries from users,

[1785] A means for analyzing the aforementioned query content using natural language processing,

[1786] A means for searching a database based on the analyzed query content,

[1787] A means for generating an appropriate answer from the aforementioned search results,

[1788] Means for sending the generated response to the user,

[1789] A means for escalating a user's inquiry to the support team if a suitable answer cannot be found,

[1790] A system that includes this.

[1791] (Claim 2)

[1792] The system according to claim 1, characterized in that the natural language processing extracts keywords from the query content and determines the category of the query based on these keywords.

[1793] (Claim 3)

[1794] The system according to claim 1, characterized in that the escalation means includes a notification means for providing detailed information of the inquiry to the support team.

[1795] "Application Example 1"

[1796] (Claim 1)

[1797] A means of receiving customer inquiries,

[1798] A means for analyzing the aforementioned query content using natural language processing,

[1799] A means for searching a database based on the analyzed query content,

[1800] A means for generating an appropriate answer from the aforementioned search results,

[1801] A means for sending the generated response to the customer,

[1802] A means of escalating a customer inquiry to the support team if a suitable answer cannot be found,

[1803] In the aforementioned escalation, a means for notifying the detailed information of the inquiry,

[1804] A means of easily inputting customer inquiries using a smartphone,

[1805] A means of using a generative AI model to automatically extract keywords from an inquiry and generate an answer based on them,

[1806] A means of displaying the generated answer as a prompt,

[1807] A system that includes this.

[1808] (Claim 2)

[1809] The system according to claim 1, characterized in that the natural language processing extracts keywords from the query content and determines the category of the query based on these keywords.

[1810] (Claim 3)

[1811] The system according to claim 1, characterized in that the escalation means includes a notification means for providing detailed information of the inquiry to the support team.

[1812] "Example 2 of combining an emotion engine"

[1813] (Claim 1)

[1814] A means of receiving customer inquiries,

[1815] A means for analyzing the aforementioned query content using natural language processing,

[1816] A means for searching a database based on the analyzed query content,

[1817] A means for generating an appropriate answer from the aforementioned search results,

[1818] A means for sending the generated response to the customer,

[1819] A means of escalating a customer inquiry to the support team if a suitable answer cannot be found,

[1820] A means for analyzing the customer's emotions contained in the aforementioned inquiry,

[1821] Means for adjusting responses based on the analyzed customer emotions,

[1822] A system that includes this.

[1823] (Claim 2)

[1824] The system according to claim 1, characterized in that the natural language processing extracts keywords from the query content and determines the category of the query based on these keywords.

[1825] (Claim 3)

[1826] The system according to claim 1, characterized in that the escalation means includes a notification means for providing the support team with detailed information about the inquiry and the results of customer sentiment recognition.

[1827] "Application example 2 when combining with an emotional engine"

[1828] (Claim 1)

[1829] A means of receiving customer inquiries,

[1830] A means for analyzing the aforementioned query content using natural language processing,

[1831] A means for searching a database based on the analyzed query content,

[1832] A means for generating an appropriate answer from the aforementioned search results,

[1833] A means for sending the generated response to the customer,

[1834] A means of escalating a customer inquiry to the support team if a suitable answer cannot be found,

[1835] A means of recognizing the customer's emotions regarding the content of the inquiry,

[1836] Means for adjusting the response generated based on the recognized emotion,

[1837] A system that includes this.

[1838] (Claim 2)

[1839] The system according to claim 1, characterized in that the natural language processing extracts keywords from the query content and determines the category of the query based on these keywords.

[1840] (Claim 3)

[1841] The system according to claim 1, characterized in that the escalation means includes a notification means for providing detailed information of the inquiry to the support team. [Explanation of Symbols]

[1842] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of receiving customer inquiries, A means for analyzing the aforementioned query content using natural language processing, A means for searching a database based on the analyzed query content, A means for generating an appropriate answer from the aforementioned search results, A means for sending the generated response to the customer, A means of escalating a customer inquiry to the support team if a suitable answer cannot be found, A system that includes this.

2. The system according to claim 1, characterized in that the natural language processing extracts keywords from the query content and determines the category of the query based on these keywords.

3. The system according to claim 1, characterized in that the escalation means includes a notification means for providing detailed information of the inquiry to the support team.

Images