System

Abstract

A system is provided.SOLUTION: A system comprising: means for integrating a plurality of query routes and accepting queries from users in a unified manner; generation and AI means for analyzing queries accepted from users and classifying topics and complexity of problems; means for generating responses to queries from a knowledgebase and providing personalized answers to users; and means for automatically executing an escalation procedure according to urgency and complexity of queries.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Modern customer support involves multiple inquiry channels (help, chat, email, and telephone), which creates common issues in responding to inquiries, such as a lack of consistency, staff shortages, difficulty in providing 24-hour support, and difficulty in sharing information. As a result, the user experience is compromised and the efficiency of support operations is reduced. The purpose of this invention is to solve these issues and achieve consistency, efficiency, automation, and cost reduction in customer support. [Means for solving the problem]

[0005] This invention first provides a means for integrating multiple inquiry routes and centrally accepting inquiries from users. Next, it provides a means for analyzing inquiries received from users using generative AI and classifying the topic and complexity of the problem. It also provides a means for generating responses to inquiries from a knowledge base and providing users with personalized answers. It also includes a means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry. This enables consistent customer support operations, compensates for labor shortages, enables 24-hour support, and reduces costs. It also includes a means for converting user voice input into text and a means for saving inquiry dialogue history and using it to respond to future inquiries.

[0006] "Contact channels" refers to the multiple methods a user uses to contact customer support, such as help, chat, email, or phone.

[0007] "Integration" means combining multiple separate inquiry routes into one and managing them centrally.

[0008] "Centralized acceptance" refers to accepting inquiries from multiple inquiry routes in one place.

[0009] "Generative AI" refers to a system or process that uses artificial intelligence technology to analyze user inquiries and generate appropriate responses.

[0010] "Analysis" means examining the inquiries received from users in detail to identify the topic and complexity of the problem.

[0011] "Classification" refers to sorting the analyzed inquiries into specific categories (e.g., technical support, account issues, payment issues, etc.).

[0012] A "knowledge base" is a database that compiles information and methods for responding to inquiries, and serves as a reference point for the generation AI when generating answers.

[0013] "Personalization" refers to providing each user with individually appropriate answers based on their past interaction history and user profile information.

[0014] "Escalation procedure" refers to the process of automatically transferring complex or urgent inquiries that cannot be handled by the generation AI to the appropriate specialist staff or higher support level.

[0015] "Converting voice input to text" refers to the process of converting voice data into text when a user makes a voice inquiry.

[0016] "Interaction History" refers to a record of past interactions between a User and Customer Support.

[0017] "Storage" means recording and storing data such as conversation history so that it is not lost.

[0018] "Customer support" refers to the general work of responding to customer inquiries and problems related to products and services. [Brief explanation of the drawings]

[0019] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0020] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0021] First, the terms used in the following description will be explained.

[0022] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0023] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0024] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0025] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0026] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0027] [First embodiment]

[0028] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0029] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0030] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0031] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0032] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0033] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0034] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0035] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0036] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0037] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0038] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0039] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0040] MODE FOR CARRYING OUT THE INVENTION

[0041] This invention relates to a customer support system that integrates multiple inquiry routes, centrally accepts inquiries from users, analyzes and classifies them using generative AI, generates personalized answers from a knowledge base, and automatically executes escalation procedures as necessary.

[0042] 1. User Interface Configuration

[0043] The server provides a web form accessible to users that integrates multiple contact channels (help, chat, email, phone) and includes an interface that allows users to select the type of inquiry they wish to make. Users can then enter the required information and submit their inquiry.

[0044] 2. Accepting and Processing User Input

[0045] The device captures the user's query entered into the web form. If the user speaks the query, it is converted into text using speech recognition technology. This input data is then sent to the server.

[0046] 3. Inquiry Analysis and Classification

[0047] The server sends the received text data to the generation AI, which performs the following steps:

[0048] 1. Extract keywords and phrases from the inquiry to identify the topic and complexity of the problem.

[0049] 2. Categorize inquiries into categories (e.g., technical support, account issues, payment issues, etc.).

[0050] 4. Generating and sending auto-replies

[0051] The server retrieves relevant information from a knowledge base based on the classified inquiry, and then the generative AI generates a personalized answer based on past interaction history and user profile information. This answer is then sent to the user via the server.

[0052] 5. Escalation Procedures

[0053] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[0054] 6. Saving inquiry history

[0055] The server stores all dialogue history in a database, which is used to respond to future inquiries and contribute to improving the response quality of the entire system.

[0056] Specific examples

[0057] 1. The user accesses the web form, enters "Account Recovery" and submits it.

[0058] 2. The device captures this input and sends it to the server.

[0059] 3. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[0060] 4. The server retrieves the account recovery instructions from the knowledge base, and the generative AI generates a personalized answer.

[0061] 5. The server sends the answer to the user.

[0062] 6. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[0063] 7. Finally, the server stores the conversation history and uses it for future responses.

[0064] In this way, the present invention makes it possible to improve the efficiency of customer support operations, realize 24-hour support, provide consistent responses, reduce costs, and also contribute to improving the user experience.

[0065] The processing flow will be explained below.

[0066] Step 1:

[0067] A user visits a support page on a website and opens a contact form. The user selects the type of inquiry and enters the inquiry details using text input or voice input. The user clicks the submit button.

[0068] Step 2:

[0069] The device captures the user's input data. In the case of text input, it stores the data as text. In the case of voice input, it uses a speech recognition API to convert the voice data into text.

[0070] Step 3:

[0071] The device sends the captured text data to the server.

[0072] Step 4:

[0073] The server passes the received text data to the generation AI, which analyzes the inquiry and extracts keywords and phrases. Based on the analysis results, it identifies the topic and complexity of the problem and classifies the inquiry into a specific category.

[0074] Step 5:

[0075] The server sends a query to the knowledge base based on the classified query content, and retrieves related information from the knowledge base.

[0076] Step 6:

[0077] The server passes the acquired information to the generation AI, which then references past interaction history and user profile information to generate a personalized response.

[0078] Step 7:

[0079] The server sends the generated answer to the user. If the user receives the answer and has additional questions or inquiries, the process continues from step 1 again.

[0080] Step 8:

[0081] The server uses the generated AI to determine the urgency and complexity of the inquiry, automatically executing escalation procedures as needed, and transferring the inquiry to a specialist staff member if the issue is complex technical.

[0082] Step 9:

[0083] The server stores the dialogue history in a database. All inquiry dialogues are recorded as logs and used to handle future inquiries and improve the knowledge base.

[0084] Step 10:

[0085] The server notifies the user of the progress of the escalation and completion of the automated response, allowing the user to see in real time how their inquiry is being handled.

[0086] Example 1

[0087] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0088] Current customer support systems often result in delayed responses and misunderstandings. In particular, when multiple inquiry channels exist, managing inquiries can become cumbersome, potentially reducing user satisfaction. Furthermore, the analysis and classification of inquiries is performed manually, making it difficult to provide a prompt response. Furthermore, the lack of personalized responses means that user needs cannot be fully met. The present invention aims to solve these problems.

[0089] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0090] In this invention, the server includes means for integrating multiple inquiry routes and centrally accepting inquiries from users, generation AI means for analyzing the inquiries received from users and classifying the topic and complexity of the problem, means for generating responses to the inquiries from a database and providing personalized answers to the users, means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry, means for generating answers by referencing the user's profile information and past dialogue history, means for analyzing and classifying received data as text, and means for automatically sending answers to the inquiries to the users. This enables centralized management of inquiries, rapid analysis and classification, and personalized responses.

[0091] "Multiple inquiry channels" refers to multiple ways that users can contact you (help, chat, email, phone, etc.).

[0092] "Centralized reception means" refers to a method of receiving and managing inquiries received from multiple inquiry routes in a single integrated system.

[0093] "Generative AI means" refers to artificial intelligence technology that analyzes the content of inquiries, extracts keywords and phrases, and classifies them.

[0094] A "database" refers to an information system that accumulates information and knowledge bases necessary for responding to inquiries and manages them in a searchable manner.

[0095] "Personalized answers" refer to responses that provide answers optimized for a specific user based on the user's individual profile information and past interaction history.

[0096] "Escalation procedures" refer to methods for automatically transferring inquiries to specific specialist staff or departments depending on the urgency and complexity of the inquiry.

[0097] "Profile information" refers to individual data such as a user's basic information, past inquiry history, and behavioral patterns.

[0098] "Interaction history" refers to data that records all past interactions between a user and a support system.

[0099] "Means for analyzing and classifying as text" refers to a technology for converting received inquiry data into text format and analyzing and classifying the content.

[0100] "Means for automatic sending" refers to the method by which the system automatically sends the generated answers to the user.

[0101] This invention is a customer support system that integrates multiple inquiry routes, centrally accepts inquiries from users, analyzes and classifies them using a generation AI, generates personalized answers from a database, and automatically executes escalation procedures as needed. This system consists of a server that provides a web form, a terminal that accepts user input, a generation AI that analyzes the content of the inquiry, and a means for sending answers to users.

[0102] The server provides a web form that users can access and centrally accept inquiries. This form integrates multiple inquiry routes, such as help, chat, email, and phone. Users enter the required information in the form and submit their inquiry.

[0103] The inquiry entered by the user into the web form is captured by the device. In the case of voice input, the device converts the voice into text using voice recognition technology (e.g., Google Cloud Speech-to-Text). The converted text data is sent to the server.

[0104] The server sends the received text data to a generation AI (e.g., OpenAI's GPT-4), which extracts keywords and phrases from the inquiry and identifies the topic and complexity of the problem. The generation AI then categorizes the inquiry into categories such as technical support, account issues, and payment issues based on the extracted keywords.

[0105] The server searches and retrieves relevant information from a database (e.g., Confluence) based on the category information obtained from the generation AI. If necessary, the generation AI references past interaction history and user profile information to generate a personalized answer. The generated answer is sent to the user via the server.

[0106] The generative AI determines the urgency and complexity of the inquiry and automatically executes escalation procedures, where the server routes the inquiry to the necessary specialist staff and notifies the user of the progress.

[0107] The server stores all interaction history in a database (e.g., MySQL). The stored interaction history is used as reference material for responding to future inquiries, contributing to improving the response quality of the entire system.

[0108] Specific examples

[0109] For example, if a user visits a web form, types in "Account Recovery," and submits it, the system will:

[0110] 1. The device captures this input and sends it to the server.

[0111] 2. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[0112] 3. The server retrieves the account recovery instructions from the database, and the generative AI generates a personalized answer.

[0113] 4. The server sends the answer to the user.

[0114] 5. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[0115] 6. Finally, the server stores the conversation history and uses it for future responses.

[0116] Example prompts to input to the generative AI model

[0117] User asks: "My account is locked. How do I recover it?"

[0118] Prompt: "What should I do if a user's account is locked? Please suggest a specific solution based on past interaction history."

[0119] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0120] Step 1:

[0121] The server provides a web form that users can access, which integrates multiple contact channels, including help, chat, email, and phone.

[0122] Input: The server receives the user's access to a web form.

[0123] Output: Displays an interface where the user can enter a query.

[0124] Specific behavior: Initializes each field of the web form and displays the options for the inquiry route.

[0125] Step 2:

[0126] The user enters the necessary information and submits the inquiry.

[0127] Input: A user enters their inquiry or contact information into a web form and clicks submit.

[0128] Output: The input query data is generated and sent to the terminal.

[0129] What it does: Captures data entered by the user and converts it into structured data, such as JSON format.

[0130] Step 3:

[0131] The device captures the user's input and sends it to the server.

[0132] Input: The query data entered by the user.

[0133] Output: The query data is sent to the server.

[0134] Specific operation: In the case of voice input, the device uses voice recognition technology to convert the voice into text, and then sends the converted text data to the server.

[0135] Step 4:

[0136] The server sends the received text data to the generation AI.

[0137] Input: Text data sent from the terminal.

[0138] Output: Text data sent to the generation AI.

[0139] Specific operation: The server checks the received data, generates an appropriate prompt sentence, and sends the text data to the generation AI.

[0140] Step 5:

[0141] The generation AI analyzes the inquiry content and extracts keywords and phrases.

[0142] Input: Text data sent to the generation AI.

[0143] Output: Extracted keywords and phrases.

[0144] How it works: Generative AI uses natural language processing techniques to analyze text data and extract important keywords and phrases.

[0145] Step 6:

[0146] The generative AI classifies inquiries into categories based on the extracted keywords.

[0147] Input: Extracted keywords or phrases.

[0148] Output: Categorized inquiry categories.

[0149] How it works: Generative AI uses rule-based techniques and machine learning models to classify text data into categories such as tech support, account issues, and payment issues.

[0150] Step 7:

[0151] The server searches and retrieves related information from the database based on the category information obtained from the generation AI.

[0152] Input: The classified inquiry category.

[0153] Output: Database results with relevant information.

[0154] Specific operations: The server executes a database query to obtain knowledge base information related to the inquiry category.

[0155] Step 8:

[0156] The generative AI references past interaction history and user profile information to generate personalized answers.

[0157] Input: Relevant information retrieved from databases, past interaction history, and user profile information.

[0158] Output: A personalized answer.

[0159] Specific behavior: Generative AI synthesizes the acquired information and generates recommended answers.

[0160] Step 9:

[0161] The server generates the answer and sends it to the user.

[0162] Input: Personalized answers from generative AI.

[0163] Output: The answer sent to the user.

[0164] What happens: The server formats the answer and sends it to the user in some form, such as email or chat.

[0165] Step 10:

[0166] The generative AI determines the urgency and complexity of the inquiry and automatically executes escalation procedures.

[0167] Input: Generative AI assessment of the urgency and complexity of the inquiry.

[0168] Output: Notification if escalation is required.

[0169] Specific behavior: The generative AI assesses the urgency and complexity of the inquiry and, if necessary, issues instructions to transfer the inquiry to specialized staff.

[0170] Step 11:

[0171] The server will notify the user of the progress of the escalation.

[0172] Input: Escalation procedure progress information.

[0173] Output: Progress notification to the user.

[0174] What it does: The server periodically checks the progress and notifies the user with status updates.

[0175] Step 12:

[0176] The server stores all interaction history in a database.

[0177] Input: Interaction data stored on the server.

[0178] Output: The saved interaction history.

[0179] Specific operation: The server records the dialogue history in a database and makes it available for use in responding to future inquiries.

[0180] (Application example 1)

[0181] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0182] Customer support on modern online shopping sites can be complicated for users, as there are multiple routes for inquiries (email, chat, phone, etc.). This means that there is a demand for consistent and prompt responses to inquiries. There is also a growing demand for customer support that supports voice input, and a system that can respond to this demand efficiently is required.

[0183] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0184] In this invention, the server includes: means for integrating multiple inquiry routes and centrally accepting inquiries from users; means for providing an interface for users to make inquiries through smart devices; speech recognition means for converting voice inquiries into text; generative AI means for analyzing inquiries accepted from users and classifying the topic and complexity of the problem; means for creating prompt sentences using a generative AI model and generating responses to the inquiries from a knowledge base to provide users with personalized answers; means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry; and means for saving the dialogue history of the inquiry and using it for responding to future inquiries. This enables fast and consistent customer support, improving the user experience and efficiently responding to inquiries.

[0185] "Multiple inquiry routes" refers to multiple ways that users can make inquiries, such as email, chat, phone, and voice input.

[0186] "Smart devices" refers to portable electronic devices with internet connectivity, such as smartphones and tablets.

[0187] "Interface" refers to the function of providing a screen and input means for users to interact with a system.

[0188] "Voice recognition means" refers to technology that converts voice into text data when a user makes a voice inquiry.

[0189] "Generative AI methods" refers to artificial intelligence technologies that analyze incoming inquiries, categorize them by topic and complexity, and generate appropriate responses.

[0190] "Generative AI model" refers to a trained AI model that generates prompts based on the content of an inquiry and creates appropriate responses.

[0191] A "prompt" is a phrase that is input into an AI model to generate an optimal response.

[0192] A "knowledge base" refers to a database that stores past inquiry data and knowledge, and is used as a source of information for generating responses.

[0193] "Personalized answers" refers to responses that are customized based on a user's individual situation and history.

[0194] "Escalation measures" refers to the function of automatically transferring an inquiry to specialized staff depending on the urgency and complexity of the inquiry.

[0195] "Dialogue history" refers to a record of past inquiries and responses with a user.

[0196] This invention relates to a customer support system that integrates multiple inquiry channels and centrally accepts inquiries from users. It is intended to improve the efficiency of customer support, primarily on online shopping sites, and enhance the user experience. The system for implementing this invention is as follows:

[0197] Overall system configuration

[0198] The system consists of three main elements: a server, a terminal, and a user.

[0199] User Interface Settings

[0200] The server provides a user-accessible web form or smart device interface that integrates multiple inquiry channels (email, chat, phone, voice input) and allows users to easily select the type of inquiry they wish to make. The user can then enter the required information and submit the inquiry.

[0201] Accepting and processing user input

[0202] When a user types or records a query via a smart device, the data is captured by the device. If the user speaks the query, the voice data is converted into text using voice recognition technology (e.g., SpeechRecognition library). The converted text data is then sent to the server.

[0203] Inquiry analysis and classification

[0204] The server sends the received text data to a generative AI (e.g., OpenAI's language model), which extracts keywords and phrases from the inquiry, identifies the topic and complexity of the problem, and classifies the inquiry into categories (e.g., technical support, order issues, payment issues, etc.).

[0205] Generate and send auto-replies

[0206] The server retrieves relevant information from a knowledge base based on the classified inquiry, and the generative AI model generates a personalized answer based on past interaction history and user profile information. This answer is then sent to the user via the server.

[0207] Escalation Procedures

[0208] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[0209] Saving inquiry history

[0210] The server stores all dialogue history in a database, which is used to respond to future inquiries and contribute to improving the response quality of the entire system.

[0211] Specific examples

[0212] The user opens the smartphone app and makes a voice inquiry saying, "Please tell me how to return an item." The device captures the voice data and converts it into text using the SpeechRecognition library. The text passed to the generation AI is analyzed and classified as "I would like to know how to return an item." The generation AI model retrieves relevant return methods from the knowledge base and generates a personalized answer saying, "To find out how to return an item, select your order history from the account page and click the return option." This answer is then sent to the user via the server.

[0213] Prompt Sentence Examples

[0214] An example prompt is:

[0215] Inquiry: How do I return the item?

[0216] Please parse this and generate an appropriate answer.

[0217] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0218] Step 1:

[0219] The user types or records their inquiry via voice via their smart device.

[0220] Input: Text or voice data of the inquiry.

[0221] Output: The captured text or audio data.

[0222] Specific operation: The user opens the smartphone app and asks by voice, "Please tell me how to return the item."

[0223] Step 2:

[0224] The device captures the data and uses voice recognition technology to convert the voice data into text.

[0225] Input: Captured audio data.

[0226] Output: The converted text data.

[0227] What happens: The device captures audio data and converts it to text using the SpeechRecognition library.

[0228] Step 3:

[0229] The terminal transmits the converted text data to the server.

[0230] Input: The converted text data.

[0231] Output: The text data sent to the server.

[0232] Specific operation: The device makes an API request to send the converted text data to the server.

[0233] Step 4:

[0234] The server sends the received text data to the generation AI.

[0235] Input: Text data sent to the server.

[0236] Output: Text data sent to the generation AI.

[0237] Specific operation: The server sends text data to the generative AI model via the API.

[0238] Step 5:

[0239] Generative AI analyzes the query, extracting keywords and phrases to identify the topic and complexity of the problem.

[0240] Input: Text data sent to the generation AI.

[0241] Output: The topic and complexity of the parsed query.

[0242] How it works: A generative AI model (e.g., OpenAI's language model) analyzes text data, extracts important keywords, and classifies them into problem categories.

[0243] Step 6:

[0244] The server retrieves relevant information from a knowledge base based on the classified query.

[0245] Input: The topic and complexity of the parsed query.

[0246] Output: Relevant information retrieved from the knowledge base.

[0247] Specific operation: The server accesses the knowledge base and obtains relevant information (e.g., how to process returns) based on the analysis results.

[0248] Step 7:

[0249] The generative AI model references past interaction history and user profile information to generate personalized answers.

[0250] Input: Relevant information retrieved from the knowledge base, and the user's profile information.

[0251] Output: A personalized answer.

[0252] How it works: The generative AI model uses the user's past inquiry history and account information to generate the most appropriate answer.

[0253] Step 8:

[0254] The server generates the answer and sends it to the user.

[0255] Input: Personalized answers from generative AI.

[0256] Output: The answer sent to the user.

[0257] What it does: The server sends a personalized answer to the user's smart device.

[0258] Step 9:

[0259] Generative AI determines the urgency and complexity of the inquiry and automatically executes escalation procedures.

[0260] Input: User's query and its analysis results.

[0261] Output: The query routed to the required specialist.

[0262] Specific operation: The generating AI determines the urgency of the inquiry, and if necessary, the server transfers the inquiry to specialized staff.

[0263] Step 10:

[0264] The server stores all interaction history in a database.

[0265] Input: User interaction history.

[0266] Output: Dialogue history stored in a database.

[0267] Specific operation: The server stores all interaction history with the user in a database and uses it to respond to future inquiries.

[0268] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0269] MODE FOR CARRYING OUT THE INVENTION

[0270] This invention relates to a customer support system that integrates multiple inquiry routes, centrally accepts inquiries from users, analyzes and classifies them using generative AI, generates personalized answers from a knowledge base, and automatically executes escalation procedures as needed. Furthermore, by combining it with an emotion engine that recognizes user emotions, the content and tone of the response can be adjusted to achieve more personalized responses.

[0271] 1. User Interface Configuration

[0272] The server provides a web form accessible to users that integrates multiple contact channels (help, chat, email, phone) and includes an interface that allows users to select the type of inquiry they wish to make. Users can then enter the required information and submit their inquiry.

[0273] 2. Accepting and Processing User Input

[0274] The device captures the user's query entered into the web form. If the user speaks the query, it is converted into text using speech recognition technology. This input data is then sent to the server.

[0275] 3. Inquiry Analysis and Classification

[0276] The server then passes the received text data to the generation AI, which analyzes the inquiry, extracts keywords and phrases, and, based on the analysis results, identifies the topic and complexity of the problem and classifies the inquiry into a specific category.

[0277] 4. Generating and sending auto-replies

[0278] The server sends the query to a knowledge base based on the classified query content, retrieves relevant information from the knowledge base, and then the generative AI references past interaction history and user profile information to generate a personalized answer. This answer is then sent to the user via the server.

[0279] 5. Emotional analysis of users using an emotion engine

[0280] The server passes the user's inquiry data to the emotion engine, which analyzes the user's emotions from text and voice data and provides the results to the generation AI. The generation AI takes this emotional data into account and appropriately adjusts the content and tone of the response.

[0281] 6. Escalation Procedures

[0282] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[0283] 7. Saving inquiry history

[0284] The server stores all dialogue history in a database. This history can be used to respond to future inquiries, contributing to improving the response quality of the entire system. In addition, user emotional data is also saved as a log, enabling more accurate and personalized responses for future inquiries.

[0285] Specific examples

[0286] 1. The user accesses the web form, enters "Account Recovery" and submits it.

[0287] 2. The device captures this input and sends it to the server.

[0288] 3. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[0289] 4. The server retrieves the account recovery instructions from the knowledge base, and the generative AI generates a personalized answer.

[0290] 5. The server sends the answer to the user.

[0291] 6. At the same time, the server passes the user's input data to the emotion engine, which analyzes the user's emotions. The generative AI takes this analysis into account and adjusts the tone and content of the response.

[0292] 7. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[0293] 8. Finally, the server stores the dialogue history and emotion data for future use.

[0294] In this way, this invention makes it possible to improve the efficiency of customer support operations, realize 24-hour support, provide consistent responses, and reduce costs. In addition, the use of an emotion engine enables responses that are more in tune with emotions, contributing to an improved user experience.

[0295] The processing flow will be explained below.

[0296] Step 1:

[0297] A user visits a website's support page and opens a contact form. They select the type of inquiry they want, choosing a category such as "Account Recovery" or "Technical Support," then enter their inquiry details using text input or voice input and click the submit button.

[0298] Step 2:

[0299] The device captures the user's input data. In the case of text input, it stores the data as text. In the case of voice input, it uses a speech recognition API to convert the voice data into text.

[0300] Step 3:

[0301] The device sends the captured text data to the server.

[0302] Step 4:

[0303] The server passes the received text data to the generation AI, which analyzes the inquiry and extracts keywords and phrases. Based on the analysis results, the AI ​​identifies the topic and complexity of the problem and classifies the inquiry into a specific category, such as technical support, account issues, or payment issues.

[0304] Step 5:

[0305] The server sends a query to the knowledge base based on the classified query content, and retrieves related information from the knowledge base.

[0306] Step 6:

[0307] The server passes the acquired information to the generation AI, which then references past interaction history and user profile information to generate a personalized response.

[0308] Step 7:

[0309] The server then sends the generated response to the user, analyzing the user's emotions using an emotion engine and adjusting the tone and content accordingly. For example, if the user is feeling frustrated, the tone of the response will be more friendly.

[0310] Step 8:

[0311] The server uses the AI ​​to determine the urgency and complexity of the inquiry, and automatically executes escalation procedures as needed. For example, if the issue is technically difficult, the inquiry will be transferred to a specialist.

[0312] Step 9:

[0313] The server stores the conversation history in a database. All conversations are logged and used to handle future inquiries and improve the knowledge base. In addition, user sentiment data is also stored.

[0314] Step 10:

[0315] The server notifies the user of the progress of the escalation and completion of the automated response, allowing the user to see in real time how their inquiry is being handled.

[0316] Specific examples

[0317] 1. The user accesses the web form, enters "account recovery" as the inquiry, and clicks submit.

[0318] 2. The device captures this input data and sends it to the server.

[0319] 3. The server passes the content to the generation AI, which then classifies "Account Recovery" into the technical support category.

[0320] 4. The server retrieves the account recovery instructions from the knowledge base, and the generative AI generates a personalized answer.

[0321] 5. The server provides the generated answer to the user, with an emotion engine analyzing the user's emotions and adjusting the tone of the answer to be more friendly.

[0322] 6. If the user's problem is not resolved, the generation AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[0323] 7. Finally, the server stores the dialogue history and emotion data for use in responding to future inquiries.

[0324] Through this flow, the present invention improves the efficiency of customer support operations, enables 24-hour support, provides consistent responses, reduces costs, and improves user experience.

[0325] Example 2

[0326] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0327] In conventional customer support systems, multiple inquiry routes were managed separately, making it difficult to provide consistent responses. Furthermore, the lack of escalation procedures and emotionally sensitive responses made it difficult to improve the user experience. Furthermore, the lack of effective storage and utilization of inquiry history made it difficult to improve support quality.

[0328] The identification process 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 a means for integrating multiple inquiry routes and centrally accepting inquiries from users, a generation AI means for analyzing inquiries accepted from users and classifying the topic and complexity of the problem, a means for generating responses to inquiries from an information base and providing users with personalized answers, a means for passing inquiry data to a sentiment analysis engine and adjusting the content and tone of the response based on the analysis results, a means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry, and a means for saving the dialogue history and sentiment data of the inquiry and using them in future inquiry responses. This enables consistent and prompt responses, responses that are sensitive to the user's sentiment, prompt escalation, and effective use of the inquiry history.

[0329] "Multiple contact channels" are ways for users to contact you through different channels, such as help, chat, email, or phone.

[0330] "Centralized reception" refers to the integrated management of information received from multiple inquiry methods in a single system.

[0331] "Analysis" refers to the process by which the generative AI understands the content of the inquiry it receives, extracts keywords and phrases, and evaluates the topic and complexity of the problem.

[0332] "Generative AI" refers to a system that uses artificial intelligence technology to understand the content of an inquiry and provide appropriate answers or classifications.

[0333] An "information base" refers to a database that compiles past inquiry data and specialized knowledge, and is used to generate appropriate answers.

[0334] "Personalized answers" refer to responses that are individually tailored to the user, taking into account their profile and past interaction history.

[0335] An "emotion analysis engine" refers to a system that analyzes emotions from user inquiry text and voice data and provides the results to the generative AI.

[0336] "Escalation Procedures" refers to the process of automatically transferring inquiries to higher-level specialist staff based on the urgency or complexity of the inquiry.

[0337] "Dialogue history" refers to a record of past communication between the user and the system, and is used as reference material for responding to future inquiries.

[0338] "Emotion data" is data that indicates the emotional state of a user extracted by an emotion analysis engine.

[0339] This invention relates to a customer support system that integrates multiple inquiry routes and centrally accepts inquiries from users. This system uses generative AI and a sentiment analysis engine to analyze inquiries, generate automatic responses, execute escalation procedures, and provide users with personalized responses.

[0340] Hardware and software used

[0341] This system uses the following hardware and software.

[0342] Server: Plays a central role in accepting and processing user queries.

[0343] Terminal: Captures user input (text and voice) and sends it to the server.

[0344] Generative AI models: Analyze incoming queries and classify the topic and complexity of the problem.

[0345] Sentiment analysis engine: Analyzes user inquiry data and recognizes emotions.

[0346] Data processing and calculation

[0347] Each function of the system is described in detail below.

[0348] User Interface Settings

[0349] The server provides a user-accessible web form that integrates multiple inquiry channels, such as help, chat, email, and phone, and includes an interface through which the user can select the type of inquiry, through which the user can enter the required information and submit the inquiry.

[0350] Receiving and processing inquiries

[0351] The user's query entered into the web form is captured by the device. If the query is voice, it is converted to text using speech recognition technology. This input data is then sent to the server.

[0352] Inquiry analysis and classification

[0353] The server then passes the received text data to the generation AI, which analyzes the query and extracts keywords and phrases. Based on the analysis results, the AI ​​identifies the topic and complexity of the problem and classifies the query into a specific category.

[0354] Generate and send auto-replies

[0355] The server sends a query to an information base based on the classified query content to retrieve relevant information. The generation AI then references past interaction history and user profile information to generate a personalized answer, which is then sent to the user via the server.

[0356] Sentiment analysis and response tailoring

[0357] The server passes the user's inquiry data to the emotion analysis engine. The emotion engine analyzes the user's emotions from the text and voice data and provides the results to the generation AI. The generation AI takes this emotional data into account and adjusts the content and tone of the response appropriately.

[0358] Escalation Procedures

[0359] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[0360] Saving inquiry history

[0361] The server stores all dialogue history in a database. This history can be used to respond to future inquiries, contributing to improving the response quality of the entire system. In addition, user emotional data is also saved as a log, enabling more accurate and personalized responses for future inquiries.

[0362] Specific examples

[0363] 1. The user accesses the web form, enters "Account Recovery" and submits it.

[0364] 2. The device captures this input and sends it to the server.

[0365] 3. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[0366] 4. The server retrieves account recovery instructions from its information base, and the generative AI generates a personalized answer.

[0367] 5. The server sends the answer to the user.

[0368] 6. At the same time, the server passes the user's input data to the emotion analysis engine, which analyzes the user's emotions. The generative AI takes this analysis into account and adjusts the tone and content of the response.

[0369] 7. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[0370] 8. Finally, the server stores the dialogue history and emotion data for future use.

[0371] In this way, this invention makes it possible to improve the efficiency of customer support operations, realize 24-hour support, provide consistent responses, and reduce costs. In addition, the use of a sentiment analysis engine enables responses that are more sensitive to emotions, contributing to an improved user experience.

[0372] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0373] Step 1:

[0374] A user accesses a web form, enters inquiry information, and presses the "Submit" button. At this time, the input is text-based information or voice data. Take the example of a user entering "Account recovery." Input: The user's inquiry. Output: The input inquiry data.

[0375] Step 2:

[0376] The device captures the user's input, and if the input is voice, converts it into text data using voice recognition technology. The converted text data is sent to the server. Input: User's raw voice or text. Output: Text data.

[0377] Step 3:

[0378] The server sends the received text data to the generation AI. The generation AI analyzes the received inquiry and extracts keywords and phrases. Input: Text data received by the server. Output: Keyword extraction and analysis results by the generation AI.

[0379] Step 4:

[0380] The server identifies the topic and complexity of the inquiry based on the analysis results from the generative AI and classifies it into an appropriate category. For example, "account recovery" is classified into the "technical support" category. Input: Analysis results from the generative AI. Output: Enquiry category classification.

[0381] Step 5:

[0382] The server queries the information base for the classified inquiry and retrieves relevant information. The data retrieved from the information base is sent to the generation AI, which generates a personalized response by referencing past interaction history and user profile information. Input: Inquiry category classification. Output: Personalized response.

[0383] Step 6:

[0384] The server sends the generated personalized response to the user, for example, "Hello, here are the steps to recover your account." Input: The generated personalized response. Output: The response sent to the user.

[0385] Step 7:

[0386] The server passes the inquiry data to the sentiment analysis engine, which identifies the user's sentiment and provides the results to the generation AI. For example, if the user enters "I'm in a hurry," the sentiment analysis engine recognizes a high level of urgency. Input: User inquiry data. Output: Sentiment analysis results.

[0387] Step 8:

[0388] The generation AI takes into account the results of the sentiment analysis and adjusts the content and tone of the response. For example, it generates a response that takes sentiment into account, such as "We will respond immediately." Input: Sentiment analysis results. Output: Adjusted response.

[0389] Step 9:

[0390] The server adjusts the content and tone of the response and then sends it to the user. Input: Adjusted response. Output: Adjusted response sent to the user.

[0391] Step 10:

[0392] The generation AI reassess the urgency and complexity of the inquiry and determines whether escalation is necessary. Urgent or highly difficult inquiries are automatically transferred to specialized staff. The server notifies the user of the progress of the escalation. Input: Reassessed urgency and complexity. Output: Transfer to specialized staff, notification to user.

[0393] Step 11:

[0394] The server stores all dialogue history and emotion data in a database. The stored data is used to respond to future inquiries and helps improve the response quality of the entire system. Input: dialogue history and emotion data. Output: stored data.

[0395] (Application example 2)

[0396] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0397] Conventional customer support systems have a lack of uniformity in responses to user inquiries, making it difficult to provide personalized responses. Furthermore, responses that do not take into account the user's feelings can lead to a decline in customer satisfaction. Furthermore, manual escalation procedures can lead to delays and errors.

[0398] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for integrating multiple inquiry routes and centrally accepting inquiries from users; generation AI means for analyzing inquiries accepted from users and classifying the topic and complexity of the problem; means for generating responses to inquiries from a knowledge base and providing users with personalized answers; emotion engine means for analyzing user emotions and adjusting the content and tone of the response; means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry; and means for saving the inquiry dialogue history and using it for future inquiry responses. This enables more efficient customer support operations, 24-hour support, consistent responses, cost reduction, and improved customer satisfaction through responses that are sensitive to emotions.

[0399] "Enquiry channels" refer to the multiple ways or channels through which users can make inquiries.

[0400] "A means of centralized reception" refers to a function that allows inquiries sent from multiple inquiry routes to be received in one place.

[0401] "Generative AI means of analyzing and classifying" refers to the generative AI's ability to analyze the content of a query and classify it based on the topic and complexity of the problem.

[0402] A "knowledge base" refers to a database that stores and organizes answers to inquiries and information.

[0403] "Means for providing personalized answers" refers to the ability to generate and provide individually customized answers based on the user's inquiry and profile information.

[0404] "Emotion engine means" refers to a system that analyzes a user's emotions and adjusts the content and tone of the response based on the results.

[0405] "Means for implementing escalation procedures" refers to the ability to automatically route inquiries to specialized staff based on the urgency and complexity of the inquiry.

[0406] "Means for saving dialogue history and using it to respond to future inquiries" refers to a function that records the content and responses of past inquiries and uses them to respond to future inquiries.

[0407] A specific system for implementing the present invention has the following configuration.

[0408] First, the server integrates multiple inquiry channels and centrally accepts user inquiries. This includes web forms, chat boxes, and voice input. When a user makes an inquiry, the data is sent to the server. The server captures this data and passes it to a generative AI model for analysis. The analysis classifies the topic and complexity of the problem.

[0409] The server then references the knowledge base and uses a generative AI model to generate a personalized answer. This answer also takes into account the user's profile information and past interaction history. The emotion engine also analyzes the emotions from the user's input data and provides the results to the generative AI model. The generative AI model uses this emotional data to appropriately adjust the content and tone of the response.

[0410] If the inquiry is urgent or complex, the server automatically executes an escalation procedure and transfers the inquiry to a specialist staff member. The user is notified of the progress. In addition, the dialogue history and emotion data are recorded in a database and used to handle future inquiries.

[0411] Hardware and software used:

[0412] Server: A cloud server that provides an API endpoint (e.g., AWS, Google Cloud)

[0413] Generative AI models: text analysis and response generation (e.g., GPT-4)

[0414] Emotion engine: Sentiment analysis (e.g. IBM Watson, Microsoft Azure Emotion API)

[0415] Device: Capturing user input (e.g. smartphone, HMD)

[0416] Examples:

[0417] For example, if a user voice-inquires within a virtual store's smartphone app, "How do I return an item?", the voice data is sent to a server and converted into text using speech recognition technology. A generative AI model analyzes the inquiry, retrieves information about the "return procedure" from a knowledge base, and generates and sends a personalized response taking into account the user's profile information. Furthermore, an emotion engine analyzes the user's emotions (e.g., frustration), and the generative AI model adjusts the tone of the response taking these emotions into account.

[0418] Example prompt sentence:

[0419] I have an item I'd like to return, how do I go about doing that?

[0420] "I'm having issues with the quality of the product I purchased. What should I do?"

[0421] In this way, real-time responses that are sensitive to emotions become possible, contributing to an improved user experience.

[0422] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0423] Step 1:

[0424] The user makes an inquiry. The user selects an appropriate input method from multiple inquiry routes (e.g., web form, chat box, voice input) and types or speaks the inquiry content. The input data is converted into text using voice recognition technology.

[0425] Input: User text or voice data

[0426] Output: Query data converted to text

[0427] Step 2:

[0428] The device captures the user's input data and sends it to the server. If there is voice input, the device processes it to convert the voice into text.

[0429] Input: User's inquiry (text or voice data)

[0430] Output: Text data sent to the server

[0431] Step 3:

[0432] The server then passes the received text data to a generative AI model for analysis, which extracts keywords and phrases from the query and categorizes them based on the topic and complexity of the problem.

[0433] Input: Text data

[0434] Output: Parsed data (topic and complexity classification results)

[0435] Step 4:

[0436] The server uses the parsed data to consult a knowledge base to retrieve relevant information, and a generative AI model takes this information, along with the user's profile information, into account to generate a personalized answer.

[0437] Input: Analytics data, knowledge base information, user profile information

[0438] Output: Personalized answer

[0439] Step 5:

[0440] The server passes the user's text data to the emotion engine for emotion analysis. The emotion engine analyzes the user's emotions from the text data and provides the results to the generative AI model.

[0441] Input: User's text data

[0442] Output: Emotion analysis results

[0443] Step 6:

[0444] The generative AI model takes into account the results of sentiment analysis and adjusts the content and tone of the response accordingly, resulting in a personalized response that takes the user's emotions into account.

[0445] Input: Sentiment analysis results, personalized answers

[0446] Output: Emotionally sensitive personalized answers

[0447] Step 7:

[0448] The server then sends the generated answer to the user, and simultaneously automatically routes the inquiry to a specialist if escalation is required, and notifies the user of the progress.

[0449] Input: Emotionally sensitive personalized answers

[0450] Output: Response sent to user, escalation action

[0451] Step 8:

[0452] The server stores all dialogue history and emotion data in a database, which is used to respond to future inquiries and contribute to improving the response quality of the entire system.

[0453] Input: Dialogue history, emotion data

[0454] Output: Historical data stored in a database

[0455] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0456] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0457] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0458] [Second embodiment]

[0459] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0460] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0461] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0462] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0463] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0464] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0465] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0466] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0467] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0468] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0469] In the smart glasses 214, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0470] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0471] MODE FOR CARRYING OUT THE INVENTION

[0472] This invention relates to a customer support system that integrates multiple inquiry routes, centrally accepts inquiries from users, analyzes and classifies them using generative AI, generates personalized answers from a knowledge base, and automatically executes escalation procedures as necessary.

[0473] 1. User Interface Configuration

[0474] The server provides a web form accessible to users that integrates multiple contact channels (help, chat, email, phone) and includes an interface that allows users to select the type of inquiry they wish to make. Users can then enter the required information and submit their inquiry.

[0475] 2. Accepting and Processing User Input

[0476] The device captures the user's query entered into the web form. If the user speaks the query, it is converted into text using speech recognition technology. This input data is then sent to the server.

[0477] 3. Inquiry Analysis and Classification

[0478] The server sends the received text data to the generation AI, which performs the following steps:

[0479] 1. Extract keywords and phrases from the inquiry to identify the topic and complexity of the problem.

[0480] 2. Categorize inquiries into categories (e.g., technical support, account issues, payment issues, etc.).

[0481] 4. Generating and sending auto-replies

[0482] The server retrieves relevant information from a knowledge base based on the classified inquiry, and then the generative AI generates a personalized answer based on past interaction history and user profile information. This answer is then sent to the user via the server.

[0483] 5. Escalation Procedures

[0484] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[0485] 6. Saving inquiry history

[0486] The server stores all dialogue history in a database, which is used to respond to future inquiries and contribute to improving the response quality of the entire system.

[0487] Specific examples

[0488] 1. The user accesses the web form, enters "Account Recovery" and submits it.

[0489] 2. The device captures this input and sends it to the server.

[0490] 3. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[0491] 4. The server retrieves the account recovery instructions from the knowledge base, and the generative AI generates a personalized answer.

[0492] 5. The server sends the answer to the user.

[0493] 6. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[0494] 7. Finally, the server stores the conversation history and uses it for future responses.

[0495] In this way, the present invention makes it possible to improve the efficiency of customer support operations, realize 24-hour support, provide consistent responses, reduce costs, and also contribute to improving the user experience.

[0496] The processing flow will be explained below.

[0497] Step 1:

[0498] A user visits a support page on a website and opens a contact form. The user selects the type of inquiry and enters the inquiry details using text input or voice input. The user clicks the submit button.

[0499] Step 2:

[0500] The device captures the user's input data. In the case of text input, it stores the data as text. In the case of voice input, it uses a speech recognition API to convert the voice data into text.

[0501] Step 3:

[0502] The device sends the captured text data to the server.

[0503] Step 4:

[0504] The server passes the received text data to the generation AI, which analyzes the inquiry and extracts keywords and phrases. Based on the analysis results, it identifies the topic and complexity of the problem and classifies the inquiry into a specific category.

[0505] Step 5:

[0506] The server sends a query to the knowledge base based on the classified query content, and retrieves related information from the knowledge base.

[0507] Step 6:

[0508] The server passes the acquired information to the generation AI, which then references past interaction history and user profile information to generate a personalized response.

[0509] Step 7:

[0510] The server sends the generated answer to the user. If the user receives the answer and has additional questions or inquiries, the process continues from step 1 again.

[0511] Step 8:

[0512] The server uses the generated AI to determine the urgency and complexity of the inquiry, automatically executing escalation procedures as needed, and transferring the inquiry to a specialist staff member if the issue is complex technical.

[0513] Step 9:

[0514] The server stores the dialogue history in a database. All inquiry dialogues are recorded as logs and used to handle future inquiries and improve the knowledge base.

[0515] Step 10:

[0516] The server notifies the user of the progress of the escalation and completion of the automated response, allowing the user to see in real time how their inquiry is being handled.

[0517] Example 1

[0518] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0519] Current customer support systems often result in delayed responses and misunderstandings. In particular, when multiple inquiry channels exist, managing inquiries can become cumbersome, potentially reducing user satisfaction. Furthermore, the analysis and classification of inquiries is performed manually, making it difficult to provide a prompt response. Furthermore, the lack of personalized responses means that user needs cannot be fully met. The present invention aims to solve these problems.

[0520] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0521] In this invention, the server includes means for integrating multiple inquiry routes and centrally accepting inquiries from users, generation AI means for analyzing the inquiries received from users and classifying the topic and complexity of the problem, means for generating responses to the inquiries from a database and providing personalized answers to the users, means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry, means for generating answers by referencing the user's profile information and past dialogue history, means for analyzing and classifying received data as text, and means for automatically sending answers to the inquiries to the users. This enables centralized management of inquiries, rapid analysis and classification, and personalized responses.

[0522] "Multiple inquiry channels" refers to multiple ways that users can contact you (help, chat, email, phone, etc.).

[0523] "Centralized reception means" refers to a method of receiving and managing inquiries received from multiple inquiry routes in a single integrated system.

[0524] "Generative AI means" refers to artificial intelligence technology that analyzes the content of inquiries, extracts keywords and phrases, and classifies them.

[0525] A "database" refers to an information system that accumulates information and knowledge bases necessary for responding to inquiries and manages them in a searchable manner.

[0526] "Personalized answers" refer to responses that provide answers optimized for a specific user based on the user's individual profile information and past interaction history.

[0527] "Escalation procedures" refer to methods for automatically transferring inquiries to specific specialist staff or departments depending on the urgency and complexity of the inquiry.

[0528] "Profile information" refers to individual data such as a user's basic information, past inquiry history, and behavioral patterns.

[0529] "Interaction history" refers to data that records all past interactions between a user and a support system.

[0530] "Means for analyzing and classifying as text" refers to a technology for converting received inquiry data into text format and analyzing and classifying the content.

[0531] "Means for automatic sending" refers to the method by which the system automatically sends the generated answers to the user.

[0532] This invention is a customer support system that integrates multiple inquiry routes, centrally accepts inquiries from users, analyzes and classifies them using a generation AI, generates personalized answers from a database, and automatically executes escalation procedures as needed. This system consists of a server that provides a web form, a terminal that accepts user input, a generation AI that analyzes the content of the inquiry, and a means for sending answers to users.

[0533] The server provides a web form that users can access and centrally accept inquiries. This form integrates multiple inquiry routes, such as help, chat, email, and phone. Users enter the required information in the form and submit their inquiry.

[0534] The inquiry entered by the user into the web form is captured by the device. In the case of voice input, the device converts the voice into text using voice recognition technology (e.g., Google Cloud Speech-to-Text). The converted text data is sent to the server.

[0535] The server sends the received text data to a generation AI (e.g., OpenAI's GPT-4), which extracts keywords and phrases from the inquiry and identifies the topic and complexity of the problem. The generation AI then categorizes the inquiry into categories such as technical support, account issues, and payment issues based on the extracted keywords.

[0536] The server searches and retrieves relevant information from a database (e.g., Confluence) based on the category information obtained from the generation AI. If necessary, the generation AI references past interaction history and user profile information to generate a personalized answer. The generated answer is sent to the user via the server.

[0537] The generative AI determines the urgency and complexity of the inquiry and automatically executes escalation procedures, where the server routes the inquiry to the necessary specialist staff and notifies the user of the progress.

[0538] The server stores all interaction history in a database (e.g., MySQL). The stored interaction history is used as reference material for responding to future inquiries, contributing to improving the response quality of the entire system.

[0539] Specific examples

[0540] For example, if a user visits a web form, types in "Account Recovery," and submits it, the system will:

[0541] 1. The device captures this input and sends it to the server.

[0542] 2. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[0543] 3. The server retrieves the account recovery instructions from the database, and the generative AI generates a personalized answer.

[0544] 4. The server sends the answer to the user.

[0545] 5. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[0546] 6. Finally, the server stores the conversation history and uses it for future responses.

[0547] Example prompts to input to the generative AI model

[0548] User asks: "My account is locked. How do I recover it?"

[0549] Prompt: "What should I do if a user's account is locked? Please suggest a specific solution based on past interaction history."

[0550] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0551] Step 1:

[0552] The server provides a web form that users can access, which integrates multiple contact channels, including help, chat, email, and phone.

[0553] Input: The server receives the user's access to a web form.

[0554] Output: Displays an interface where the user can enter a query.

[0555] Specific behavior: Initializes each field of the web form and displays the options for the inquiry route.

[0556] Step 2:

[0557] The user enters the necessary information and submits the inquiry.

[0558] Input: A user enters their inquiry or contact information into a web form and clicks submit.

[0559] Output: The input query data is generated and sent to the terminal.

[0560] What it does: Captures data entered by the user and converts it into structured data, such as JSON format.

[0561] Step 3:

[0562] The device captures the user's input and sends it to the server.

[0563] Input: The query data entered by the user.

[0564] Output: The query data is sent to the server.

[0565] Specific operation: In the case of voice input, the device uses voice recognition technology to convert the voice into text, and then sends the converted text data to the server.

[0566] Step 4:

[0567] The server sends the received text data to the generation AI.

[0568] Input: Text data sent from the terminal.

[0569] Output: Text data sent to the generation AI.

[0570] Specific operation: The server checks the received data, generates an appropriate prompt sentence, and sends the text data to the generation AI.

[0571] Step 5:

[0572] The generation AI analyzes the inquiry content and extracts keywords and phrases.

[0573] Input: Text data sent to the generation AI.

[0574] Output: Extracted keywords and phrases.

[0575] How it works: Generative AI uses natural language processing techniques to analyze text data and extract important keywords and phrases.

[0576] Step 6:

[0577] The generative AI classifies inquiries into categories based on the extracted keywords.

[0578] Input: Extracted keywords or phrases.

[0579] Output: Categorized inquiry categories.

[0580] How it works: Generative AI uses rule-based techniques and machine learning models to classify text data into categories such as tech support, account issues, and payment issues.

[0581] Step 7:

[0582] The server searches and retrieves related information from the database based on the category information obtained from the generation AI.

[0583] Input: The classified inquiry category.

[0584] Output: Database results with relevant information.

[0585] Specific operations: The server executes a database query to obtain knowledge base information related to the inquiry category.

[0586] Step 8:

[0587] The generative AI references past interaction history and user profile information to generate personalized answers.

[0588] Input: Relevant information retrieved from databases, past interaction history, and user profile information.

[0589] Output: A personalized answer.

[0590] Specific behavior: Generative AI synthesizes the acquired information and generates recommended answers.

[0591] Step 9:

[0592] The server generates the answer and sends it to the user.

[0593] Input: Personalized answers from generative AI.

[0594] Output: The answer sent to the user.

[0595] What happens: The server formats the answer and sends it to the user in some form, such as email or chat.

[0596] Step 10:

[0597] The generative AI determines the urgency and complexity of the inquiry and automatically executes escalation procedures.

[0598] Input: Generative AI assessment of the urgency and complexity of the inquiry.

[0599] Output: Notification if escalation is required.

[0600] Specific behavior: The generative AI assesses the urgency and complexity of the inquiry and, if necessary, issues instructions to transfer the inquiry to specialized staff.

[0601] Step 11:

[0602] The server will notify the user of the progress of the escalation.

[0603] Input: Escalation procedure progress information.

[0604] Output: Progress notification to the user.

[0605] What it does: The server periodically checks the progress and notifies the user with status updates.

[0606] Step 12:

[0607] The server stores all interaction history in a database.

[0608] Input: Interaction data stored on the server.

[0609] Output: The saved interaction history.

[0610] Specific operation: The server records the dialogue history in a database and makes it available for use in responding to future inquiries.

[0611] (Application example 1)

[0612] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0613] Customer support on modern online shopping sites can be complicated for users, as there are multiple routes for inquiries (email, chat, phone, etc.). This means that there is a demand for consistent and prompt responses to inquiries. There is also a growing demand for customer support that supports voice input, and a system that can respond to this demand efficiently is required.

[0614] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0615] In this invention, the server includes: means for integrating multiple inquiry routes and centrally accepting inquiries from users; means for providing an interface for users to make inquiries through smart devices; speech recognition means for converting voice inquiries into text; generative AI means for analyzing inquiries accepted from users and classifying the topic and complexity of the problem; means for creating prompt sentences using a generative AI model and generating responses to the inquiries from a knowledge base to provide users with personalized answers; means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry; and means for saving the dialogue history of the inquiry and using it for responding to future inquiries. This enables fast and consistent customer support, improving the user experience and efficiently responding to inquiries.

[0616] "Multiple inquiry routes" refers to multiple ways that users can make inquiries, such as email, chat, phone, and voice input.

[0617] "Smart devices" refers to portable electronic devices with internet connectivity, such as smartphones and tablets.

[0618] "Interface" refers to the function of providing a screen and input means for users to interact with a system.

[0619] "Voice recognition means" refers to technology that converts voice into text data when a user makes a voice inquiry.

[0620] "Generative AI methods" refers to artificial intelligence technologies that analyze incoming inquiries, categorize them by topic and complexity, and generate appropriate responses.

[0621] "Generative AI model" refers to a trained AI model that generates prompts based on the content of an inquiry and creates appropriate responses.

[0622] A "prompt" is a phrase that is input into an AI model to generate an optimal response.

[0623] A "knowledge base" refers to a database that stores past inquiry data and knowledge, and is used as a source of information for generating responses.

[0624] "Personalized answers" refers to responses that are customized based on a user's individual situation and history.

[0625] "Escalation measures" refers to the function of automatically transferring an inquiry to specialized staff depending on the urgency and complexity of the inquiry.

[0626] "Dialogue history" refers to a record of past inquiries and responses with a user.

[0627] This invention relates to a customer support system that integrates multiple inquiry channels and centrally accepts inquiries from users. It is intended to improve the efficiency of customer support, primarily on online shopping sites, and enhance the user experience. The system for implementing this invention is as follows:

[0628] Overall system configuration

[0629] The system consists of three main elements: a server, a terminal, and a user.

[0630] User Interface Settings

[0631] The server provides a user-accessible web form or smart device interface that integrates multiple inquiry channels (email, chat, phone, voice input) and allows users to easily select the type of inquiry they wish to make. The user can then enter the required information and submit the inquiry.

[0632] Accepting and processing user input

[0633] When a user types or records a query via a smart device, the data is captured by the device. If the user speaks the query, the voice data is converted into text using voice recognition technology (e.g., SpeechRecognition library). The converted text data is then sent to the server.

[0634] Inquiry analysis and classification

[0635] The server sends the received text data to a generative AI (e.g., OpenAI's language model), which extracts keywords and phrases from the inquiry, identifies the topic and complexity of the problem, and classifies the inquiry into categories (e.g., technical support, order issues, payment issues, etc.).

[0636] Generate and send auto-replies

[0637] The server retrieves relevant information from a knowledge base based on the classified inquiry, and the generative AI model generates a personalized answer based on past interaction history and user profile information. This answer is then sent to the user via the server.

[0638] Escalation Procedures

[0639] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[0640] Saving inquiry history

[0641] The server stores all dialogue history in a database, which is used to respond to future inquiries and contribute to improving the response quality of the entire system.

[0642] Specific examples

[0643] The user opens the smartphone app and makes a voice inquiry saying, "Please tell me how to return an item." The device captures the voice data and converts it into text using the SpeechRecognition library. The text passed to the generation AI is analyzed and classified as "I would like to know how to return an item." The generation AI model retrieves relevant return methods from the knowledge base and generates a personalized answer saying, "To find out how to return an item, select your order history from the account page and click the return option." This answer is then sent to the user via the server.

[0644] Prompt Sentence Examples

[0645] An example prompt is:

[0646] Inquiry: How do I return the item?

[0647] Please parse this and generate an appropriate answer.

[0648] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0649] Step 1:

[0650] The user types or records their inquiry via voice via their smart device.

[0651] Input: Text or voice data of the inquiry.

[0652] Output: The captured text or audio data.

[0653] Specific operation: The user opens the smartphone app and asks by voice, "Please tell me how to return the item."

[0654] Step 2:

[0655] The device captures the data and uses voice recognition technology to convert the voice data into text.

[0656] Input: Captured audio data.

[0657] Output: The converted text data.

[0658] What happens: The device captures audio data and converts it to text using the SpeechRecognition library.

[0659] Step 3:

[0660] The terminal transmits the converted text data to the server.

[0661] Input: The converted text data.

[0662] Output: The text data sent to the server.

[0663] Specific operation: The device makes an API request to send the converted text data to the server.

[0664] Step 4:

[0665] The server sends the received text data to the generation AI.

[0666] Input: Text data sent to the server.

[0667] Output: Text data sent to the generation AI.

[0668] Specific operation: The server sends text data to the generative AI model via the API.

[0669] Step 5:

[0670] Generative AI analyzes the query, extracting keywords and phrases to identify the topic and complexity of the problem.

[0671] Input: Text data sent to the generation AI.

[0672] Output: The topic and complexity of the parsed query.

[0673] How it works: A generative AI model (e.g., OpenAI's language model) analyzes text data, extracts important keywords, and classifies them into problem categories.

[0674] Step 6:

[0675] The server retrieves relevant information from a knowledge base based on the classified query.

[0676] Input: The topic and complexity of the parsed query.

[0677] Output: Relevant information retrieved from the knowledge base.

[0678] Specific operation: The server accesses the knowledge base and obtains relevant information (e.g., how to process returns) based on the analysis results.

[0679] Step 7:

[0680] The generative AI model references past interaction history and user profile information to generate personalized answers.

[0681] Input: Relevant information retrieved from the knowledge base, and the user's profile information.

[0682] Output: A personalized answer.

[0683] How it works: The generative AI model uses the user's past inquiry history and account information to generate the most appropriate answer.

[0684] Step 8:

[0685] The server generates the answer and sends it to the user.

[0686] Input: Personalized answers from generative AI.

[0687] Output: The answer sent to the user.

[0688] What it does: The server sends a personalized answer to the user's smart device.

[0689] Step 9:

[0690] Generative AI determines the urgency and complexity of the inquiry and automatically executes escalation procedures.

[0691] Input: User's query and its analysis results.

[0692] Output: The query routed to the required specialist.

[0693] Specific operation: The generating AI determines the urgency of the inquiry, and if necessary, the server transfers the inquiry to specialized staff.

[0694] Step 10:

[0695] The server stores all interaction history in a database.

[0696] Input: User interaction history.

[0697] Output: Dialogue history stored in a database.

[0698] Specific operation: The server stores all interaction history with the user in a database and uses it to respond to future inquiries.

[0699] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0700] MODE FOR CARRYING OUT THE INVENTION

[0701] This invention relates to a customer support system that integrates multiple inquiry routes, centrally accepts inquiries from users, analyzes and classifies them using generative AI, generates personalized answers from a knowledge base, and automatically executes escalation procedures as needed. Furthermore, by combining it with an emotion engine that recognizes user emotions, the content and tone of the response can be adjusted to achieve more personalized responses.

[0702] 1. User Interface Configuration

[0703] The server provides a web form accessible to users that integrates multiple contact channels (help, chat, email, phone) and includes an interface that allows users to select the type of inquiry they wish to make. Users can then enter the required information and submit their inquiry.

[0704] 2. Accepting and Processing User Input

[0705] The device captures the user's query entered into the web form. If the user speaks the query, it is converted into text using speech recognition technology. This input data is then sent to the server.

[0706] 3. Inquiry Analysis and Classification

[0707] The server then passes the received text data to the generation AI, which analyzes the inquiry, extracts keywords and phrases, and, based on the analysis results, identifies the topic and complexity of the problem and classifies the inquiry into a specific category.

[0708] 4. Generating and sending auto-replies

[0709] The server sends the query to a knowledge base based on the classified query content, retrieves relevant information from the knowledge base, and then the generative AI references past interaction history and user profile information to generate a personalized answer. This answer is then sent to the user via the server.

[0710] 5. Emotional analysis of users using an emotion engine

[0711] The server passes the user's inquiry data to the emotion engine, which analyzes the user's emotions from text and voice data and provides the results to the generation AI. The generation AI takes this emotional data into account and appropriately adjusts the content and tone of the response.

[0712] 6. Escalation Procedures

[0713] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[0714] 7. Saving inquiry history

[0715] The server stores all dialogue history in a database. This history can be used to respond to future inquiries, contributing to improving the response quality of the entire system. In addition, user emotional data is also saved as a log, enabling more accurate and personalized responses for future inquiries.

[0716] Specific examples

[0717] 1. The user accesses the web form, enters "Account Recovery" and submits it.

[0718] 2. The device captures this input and sends it to the server.

[0719] 3. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[0720] 4. The server retrieves the account recovery instructions from the knowledge base, and the generative AI generates a personalized answer.

[0721] 5. The server sends the answer to the user.

[0722] 6. At the same time, the server passes the user's input data to the emotion engine, which analyzes the user's emotions. The generative AI takes this analysis into account and adjusts the tone and content of the response.

[0723] 7. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[0724] 8. Finally, the server stores the dialogue history and emotion data for future use.

[0725] In this way, this invention makes it possible to improve the efficiency of customer support operations, realize 24-hour support, provide consistent responses, and reduce costs. In addition, the use of an emotion engine enables responses that are more in tune with emotions, contributing to an improved user experience.

[0726] The processing flow will be explained below.

[0727] Step 1:

[0728] A user visits a website's support page and opens a contact form. They select the type of inquiry they want, choosing a category such as "Account Recovery" or "Technical Support," then enter their inquiry details using text input or voice input and click the submit button.

[0729] Step 2:

[0730] The device captures the user's input data. In the case of text input, it stores the data as text. In the case of voice input, it uses a speech recognition API to convert the voice data into text.

[0731] Step 3:

[0732] The device sends the captured text data to the server.

[0733] Step 4:

[0734] The server passes the received text data to the generation AI, which analyzes the inquiry and extracts keywords and phrases. Based on the analysis results, the AI ​​identifies the topic and complexity of the problem and classifies the inquiry into a specific category, such as technical support, account issues, or payment issues.

[0735] Step 5:

[0736] The server sends a query to the knowledge base based on the classified query content, and retrieves related information from the knowledge base.

[0737] Step 6:

[0738] The server passes the acquired information to the generation AI, which then references past interaction history and user profile information to generate a personalized response.

[0739] Step 7:

[0740] The server then sends the generated response to the user, analyzing the user's emotions using an emotion engine and adjusting the tone and content accordingly. For example, if the user is feeling frustrated, the tone of the response will be more friendly.

[0741] Step 8:

[0742] The server uses the AI ​​to determine the urgency and complexity of the inquiry, and automatically executes escalation procedures as needed. For example, if the issue is technically difficult, the inquiry will be transferred to a specialist.

[0743] Step 9:

[0744] The server stores the conversation history in a database. All conversations are logged and used to handle future inquiries and improve the knowledge base. In addition, user sentiment data is also stored.

[0745] Step 10:

[0746] The server notifies the user of the progress of the escalation and completion of the automated response, allowing the user to see in real time how their inquiry is being handled.

[0747] Specific examples

[0748] 1. The user accesses the web form, enters "account recovery" as the inquiry, and clicks submit.

[0749] 2. The device captures this input data and sends it to the server.

[0750] 3. The server passes the content to the generation AI, which then classifies "Account Recovery" into the technical support category.

[0751] 4. The server retrieves the account recovery instructions from the knowledge base, and the generative AI generates a personalized answer.

[0752] 5. The server provides the generated answer to the user, with an emotion engine analyzing the user's emotions and adjusting the tone of the answer to be more friendly.

[0753] 6. If the user's problem is not resolved, the generation AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[0754] 7. Finally, the server stores the dialogue history and emotion data for use in responding to future inquiries.

[0755] Through this flow, the present invention improves the efficiency of customer support operations, enables 24-hour support, provides consistent responses, reduces costs, and improves user experience.

[0756] Example 2

[0757] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0758] In conventional customer support systems, multiple inquiry routes were managed separately, making it difficult to provide consistent responses. Furthermore, the lack of escalation procedures and emotionally sensitive responses made it difficult to improve the user experience. Furthermore, the lack of effective storage and utilization of inquiry history made it difficult to improve support quality.

[0759] The identification process 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 a means for integrating multiple inquiry routes and centrally accepting inquiries from users, a generation AI means for analyzing inquiries accepted from users and classifying the topic and complexity of the problem, a means for generating responses to inquiries from an information base and providing users with personalized answers, a means for passing inquiry data to a sentiment analysis engine and adjusting the content and tone of the response based on the analysis results, a means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry, and a means for saving the dialogue history and sentiment data of the inquiry and using them in future inquiry responses. This enables consistent and prompt responses, responses that are sensitive to the user's sentiment, prompt escalation, and effective use of the inquiry history.

[0760] "Multiple contact channels" are ways for users to contact you through different channels, such as help, chat, email, or phone.

[0761] "Centralized reception" refers to the integrated management of information received from multiple inquiry methods in a single system.

[0762] "Analysis" refers to the process by which the generative AI understands the content of the inquiry it receives, extracts keywords and phrases, and evaluates the topic and complexity of the problem.

[0763] "Generative AI" refers to a system that uses artificial intelligence technology to understand the content of an inquiry and provide appropriate answers or classifications.

[0764] An "information base" refers to a database that compiles past inquiry data and specialized knowledge, and is used to generate appropriate answers.

[0765] "Personalized answers" refer to responses that are individually tailored to the user, taking into account their profile and past interaction history.

[0766] An "emotion analysis engine" refers to a system that analyzes emotions from user inquiry text and voice data and provides the results to the generative AI.

[0767] "Escalation Procedures" refers to the process of automatically transferring inquiries to higher-level specialist staff based on the urgency or complexity of the inquiry.

[0768] "Dialogue history" refers to a record of past communication between the user and the system, and is used as reference material for responding to future inquiries.

[0769] "Emotion data" is data that indicates the emotional state of a user extracted by an emotion analysis engine.

[0770] This invention relates to a customer support system that integrates multiple inquiry routes and centrally accepts inquiries from users. This system uses generative AI and a sentiment analysis engine to analyze inquiries, generate automatic responses, execute escalation procedures, and provide users with personalized responses.

[0771] Hardware and software used

[0772] This system uses the following hardware and software.

[0773] Server: Plays a central role in accepting and processing user queries.

[0774] Terminal: Captures user input (text and voice) and sends it to the server.

[0775] Generative AI models: Analyze incoming queries and classify the topic and complexity of the problem.

[0776] Sentiment analysis engine: Analyzes user inquiry data and recognizes emotions.

[0777] Data processing and calculation

[0778] Each function of the system is described in detail below.

[0779] User Interface Settings

[0780] The server provides a user-accessible web form that integrates multiple inquiry channels, such as help, chat, email, and phone, and includes an interface through which the user can select the type of inquiry, through which the user can enter the required information and submit the inquiry.

[0781] Receiving and processing inquiries

[0782] The user's query entered into the web form is captured by the device. If the query is voice, it is converted to text using speech recognition technology. This input data is then sent to the server.

[0783] Inquiry analysis and classification

[0784] The server then passes the received text data to the generation AI, which analyzes the query and extracts keywords and phrases. Based on the analysis results, the AI ​​identifies the topic and complexity of the problem and classifies the query into a specific category.

[0785] Generate and send auto-replies

[0786] The server sends a query to an information base based on the classified query content to retrieve relevant information. The generation AI then references past interaction history and user profile information to generate a personalized answer, which is then sent to the user via the server.

[0787] Sentiment analysis and response tailoring

[0788] The server passes the user's inquiry data to the emotion analysis engine. The emotion engine analyzes the user's emotions from the text and voice data and provides the results to the generation AI. The generation AI takes this emotional data into account and adjusts the content and tone of the response appropriately.

[0789] Escalation Procedures

[0790] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[0791] Saving inquiry history

[0792] The server stores all dialogue history in a database. This history can be used to respond to future inquiries, contributing to improving the response quality of the entire system. In addition, user emotional data is also saved as a log, enabling more accurate and personalized responses for future inquiries.

[0793] Specific examples

[0794] 1. The user accesses the web form, enters "Account Recovery" and submits it.

[0795] 2. The device captures this input and sends it to the server.

[0796] 3. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[0797] 4. The server retrieves account recovery instructions from its information base, and the generative AI generates a personalized answer.

[0798] 5. The server sends the answer to the user.

[0799] 6. At the same time, the server passes the user's input data to the emotion analysis engine, which analyzes the user's emotions. The generative AI takes this analysis into account and adjusts the tone and content of the response.

[0800] 7. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[0801] 8. Finally, the server stores the dialogue history and emotion data for future use.

[0802] In this way, this invention makes it possible to improve the efficiency of customer support operations, realize 24-hour support, provide consistent responses, and reduce costs. In addition, the use of a sentiment analysis engine enables responses that are more sensitive to emotions, contributing to an improved user experience.

[0803] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0804] Step 1:

[0805] A user accesses a web form, enters inquiry information, and presses the "Submit" button. At this time, the input is text-based information or voice data. Take the example of a user entering "Account recovery." Input: The user's inquiry. Output: The input inquiry data.

[0806] Step 2:

[0807] The device captures the user's input, and if the input is voice, converts it into text data using voice recognition technology. The converted text data is sent to the server. Input: User's raw voice or text. Output: Text data.

[0808] Step 3:

[0809] The server sends the received text data to the generation AI. The generation AI analyzes the received inquiry and extracts keywords and phrases. Input: Text data received by the server. Output: Keyword extraction and analysis results by the generation AI.

[0810] Step 4:

[0811] The server identifies the topic and complexity of the inquiry based on the analysis results from the generative AI and classifies it into an appropriate category. For example, "account recovery" is classified into the "technical support" category. Input: Analysis results from the generative AI. Output: Enquiry category classification.

[0812] Step 5:

[0813] The server queries the information base for the classified inquiry and retrieves relevant information. The data retrieved from the information base is sent to the generation AI, which generates a personalized response by referencing past interaction history and user profile information. Input: Inquiry category classification. Output: Personalized response.

[0814] Step 6:

[0815] The server sends the generated personalized response to the user, for example, "Hello, here are the steps to recover your account." Input: The generated personalized response. Output: The response sent to the user.

[0816] Step 7:

[0817] The server passes the inquiry data to the sentiment analysis engine, which identifies the user's sentiment and provides the results to the generation AI. For example, if the user enters "I'm in a hurry," the sentiment analysis engine recognizes a high level of urgency. Input: User inquiry data. Output: Sentiment analysis results.

[0818] Step 8:

[0819] The generation AI takes into account the results of the sentiment analysis and adjusts the content and tone of the response. For example, it generates a response that takes sentiment into account, such as "We will respond immediately." Input: Sentiment analysis results. Output: Adjusted response.

[0820] Step 9:

[0821] The server adjusts the content and tone of the response and then sends it to the user. Input: Adjusted response. Output: Adjusted response sent to the user.

[0822] Step 10:

[0823] The generation AI reassess the urgency and complexity of the inquiry and determines whether escalation is necessary. Urgent or highly difficult inquiries are automatically transferred to specialized staff. The server notifies the user of the progress of the escalation. Input: Reassessed urgency and complexity. Output: Transfer to specialized staff, notification to user.

[0824] Step 11:

[0825] The server stores all dialogue history and emotion data in a database. The stored data is used to respond to future inquiries and helps improve the response quality of the entire system. Input: dialogue history and emotion data. Output: stored data.

[0826] (Application example 2)

[0827] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0828] Conventional customer support systems have a lack of uniformity in responses to user inquiries, making it difficult to provide personalized responses. Furthermore, responses that do not take into account the user's feelings can lead to a decline in customer satisfaction. Furthermore, manual escalation procedures can lead to delays and errors.

[0829] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for integrating multiple inquiry routes and centrally accepting inquiries from users; generation AI means for analyzing inquiries accepted from users and classifying the topic and complexity of the problem; means for generating responses to inquiries from a knowledge base and providing users with personalized answers; emotion engine means for analyzing user emotions and adjusting the content and tone of the response; means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry; and means for saving the inquiry dialogue history and using it for future inquiry responses. This enables more efficient customer support operations, 24-hour support, consistent responses, cost reduction, and improved customer satisfaction through responses that are sensitive to emotions.

[0830] "Enquiry channels" refer to the multiple ways or channels through which users can make inquiries.

[0831] "A means of centralized reception" refers to a function that allows inquiries sent from multiple inquiry routes to be received in one place.

[0832] "Generative AI means of analyzing and classifying" refers to the generative AI's ability to analyze the content of a query and classify it based on the topic and complexity of the problem.

[0833] A "knowledge base" refers to a database that stores and organizes answers to inquiries and information.

[0834] "Means for providing personalized answers" refers to the ability to generate and provide individually customized answers based on the user's inquiry and profile information.

[0835] "Emotion engine means" refers to a system that analyzes a user's emotions and adjusts the content and tone of the response based on the results.

[0836] "Means for implementing escalation procedures" refers to the ability to automatically route inquiries to specialized staff based on the urgency and complexity of the inquiry.

[0837] "Means for saving dialogue history and using it to respond to future inquiries" refers to a function that records the content and responses of past inquiries and uses them to respond to future inquiries.

[0838] A specific system for implementing the present invention has the following configuration.

[0839] First, the server integrates multiple inquiry channels and centrally accepts user inquiries. This includes web forms, chat boxes, and voice input. When a user makes an inquiry, the data is sent to the server. The server captures this data and passes it to a generative AI model for analysis. The analysis classifies the topic and complexity of the problem.

[0840] The server then references the knowledge base and uses a generative AI model to generate a personalized answer. This answer also takes into account the user's profile information and past interaction history. The emotion engine also analyzes the emotions from the user's input data and provides the results to the generative AI model. The generative AI model uses this emotional data to appropriately adjust the content and tone of the response.

[0841] If the inquiry is urgent or complex, the server automatically executes an escalation procedure and transfers the inquiry to a specialist staff member. The user is notified of the progress. In addition, the dialogue history and emotion data are recorded in a database and used to handle future inquiries.

[0842] Hardware and software used:

[0843] Server: A cloud server that provides an API endpoint (e.g., AWS, Google Cloud)

[0844] Generative AI models: text analysis and response generation (e.g., GPT-4)

[0845] Emotion engine: Sentiment analysis (e.g. IBM Watson, Microsoft Azure Emotion API)

[0846] Device: Capturing user input (e.g. smartphone, HMD)

[0847] Examples:

[0848] For example, if a user voice-inquires within a virtual store's smartphone app, "How do I return an item?", the voice data is sent to a server and converted into text using speech recognition technology. A generative AI model analyzes the inquiry, retrieves information about the "return procedure" from a knowledge base, and generates and sends a personalized response taking into account the user's profile information. Furthermore, an emotion engine analyzes the user's emotions (e.g., frustration), and the generative AI model adjusts the tone of the response taking these emotions into account.

[0849] Example prompt sentence:

[0850] I have an item I'd like to return, how do I go about doing that?

[0851] "I'm having issues with the quality of the product I purchased. What should I do?"

[0852] In this way, real-time responses that are sensitive to emotions become possible, contributing to an improved user experience.

[0853] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0854] Step 1:

[0855] The user makes an inquiry. The user selects an appropriate input method from multiple inquiry routes (e.g., web form, chat box, voice input) and types or speaks the inquiry content. The input data is converted into text using voice recognition technology.

[0856] Input: User text or voice data

[0857] Output: Query data converted to text

[0858] Step 2:

[0859] The device captures the user's input data and sends it to the server. If there is voice input, the device processes it to convert the voice into text.

[0860] Input: User's inquiry (text or voice data)

[0861] Output: Text data sent to the server

[0862] Step 3:

[0863] The server then passes the received text data to a generative AI model for analysis, which extracts keywords and phrases from the query and categorizes them based on the topic and complexity of the problem.

[0864] Input: Text data

[0865] Output: Parsed data (topic and complexity classification results)

[0866] Step 4:

[0867] The server uses the parsed data to consult a knowledge base to retrieve relevant information, and a generative AI model takes this information, along with the user's profile information, into account to generate a personalized answer.

[0868] Input: Analytics data, knowledge base information, user profile information

[0869] Output: Personalized answer

[0870] Step 5:

[0871] The server passes the user's text data to the emotion engine for emotion analysis. The emotion engine analyzes the user's emotions from the text data and provides the results to the generative AI model.

[0872] Input: User's text data

[0873] Output: Emotion analysis results

[0874] Step 6:

[0875] The generative AI model takes into account the results of sentiment analysis and adjusts the content and tone of the response accordingly, resulting in a personalized response that takes the user's emotions into account.

[0876] Input: Sentiment analysis results, personalized answers

[0877] Output: Emotionally sensitive personalized answers

[0878] Step 7:

[0879] The server then sends the generated answer to the user, and simultaneously automatically routes the inquiry to a specialist if escalation is required, and notifies the user of the progress.

[0880] Input: Emotionally sensitive personalized answers

[0881] Output: Response sent to user, escalation action

[0882] Step 8:

[0883] The server stores all dialogue history and emotion data in a database, which is used to respond to future inquiries and contribute to improving the response quality of the entire system.

[0884] Input: Dialogue history, emotion data

[0885] Output: Historical data stored in a database

[0886] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0887] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0888] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0889] [Third embodiment]

[0890] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0891] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0892] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0893] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0894] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0895] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0896] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0897] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0898] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0899] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0900] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0901] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0902] MODE FOR CARRYING OUT THE INVENTION

[0903] This invention relates to a customer support system that integrates multiple inquiry routes, centrally accepts inquiries from users, analyzes and classifies them using generative AI, generates personalized answers from a knowledge base, and automatically executes escalation procedures as necessary.

[0904] 1. User Interface Configuration

[0905] The server provides a web form accessible to users that integrates multiple contact channels (help, chat, email, phone) and includes an interface that allows users to select the type of inquiry they wish to make. Users can then enter the required information and submit their inquiry.

[0906] 2. Accepting and Processing User Input

[0907] The device captures the user's query entered into the web form. If the user speaks the query, it is converted into text using speech recognition technology. This input data is then sent to the server.

[0908] 3. Inquiry Analysis and Classification

[0909] The server sends the received text data to the generation AI, which performs the following steps:

[0910] 1. Extract keywords and phrases from the inquiry to identify the topic and complexity of the problem.

[0911] 2. Categorize inquiries into categories (e.g., technical support, account issues, payment issues, etc.).

[0912] 4. Generating and sending auto-replies

[0913] The server retrieves relevant information from a knowledge base based on the classified inquiry, and then the generative AI generates a personalized answer based on past interaction history and user profile information. This answer is then sent to the user via the server.

[0914] 5. Escalation Procedures

[0915] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[0916] 6. Saving inquiry history

[0917] The server stores all dialogue history in a database, which is used to respond to future inquiries and contribute to improving the response quality of the entire system.

[0918] Specific examples

[0919] 1. The user accesses the web form, enters "Account Recovery" and submits it.

[0920] 2. The device captures this input and sends it to the server.

[0921] 3. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[0922] 4. The server retrieves the account recovery instructions from the knowledge base, and the generative AI generates a personalized answer.

[0923] 5. The server sends the answer to the user.

[0924] 6. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[0925] 7. Finally, the server stores the conversation history and uses it for future responses.

[0926] In this way, the present invention makes it possible to improve the efficiency of customer support operations, realize 24-hour support, provide consistent responses, reduce costs, and also contribute to improving the user experience.

[0927] The processing flow will be explained below.

[0928] Step 1:

[0929] A user visits a support page on a website and opens a contact form. The user selects the type of inquiry and enters the inquiry details using text input or voice input. The user clicks the submit button.

[0930] Step 2:

[0931] The device captures the user's input data. In the case of text input, it stores the data as text. In the case of voice input, it uses a speech recognition API to convert the voice data into text.

[0932] Step 3:

[0933] The device sends the captured text data to the server.

[0934] Step 4:

[0935] The server passes the received text data to the generation AI, which analyzes the inquiry and extracts keywords and phrases. Based on the analysis results, it identifies the topic and complexity of the problem and classifies the inquiry into a specific category.

[0936] Step 5:

[0937] The server sends a query to the knowledge base based on the classified query content, and retrieves related information from the knowledge base.

[0938] Step 6:

[0939] The server passes the acquired information to the generation AI, which then references past interaction history and user profile information to generate a personalized response.

[0940] Step 7:

[0941] The server sends the generated answer to the user. If the user receives the answer and has additional questions or inquiries, the process continues from step 1 again.

[0942] Step 8:

[0943] The server uses the generated AI to determine the urgency and complexity of the inquiry, automatically executing escalation procedures as needed, and transferring the inquiry to a specialist staff member if the issue is complex technical.

[0944] Step 9:

[0945] The server stores the dialogue history in a database. All inquiry dialogues are recorded as logs and used to handle future inquiries and improve the knowledge base.

[0946] Step 10:

[0947] The server notifies the user of the progress of the escalation and completion of the automated response, allowing the user to see in real time how their inquiry is being handled.

[0948] Example 1

[0949] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0950] Current customer support systems often result in delayed responses and misunderstandings. In particular, when multiple inquiry channels exist, managing inquiries can become cumbersome, potentially reducing user satisfaction. Furthermore, the analysis and classification of inquiries is performed manually, making it difficult to provide a prompt response. Furthermore, the lack of personalized responses means that user needs cannot be fully met. The present invention aims to solve these problems.

[0951] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0952] In this invention, the server includes means for integrating multiple inquiry routes and centrally accepting inquiries from users, generation AI means for analyzing the inquiries received from users and classifying the topic and complexity of the problem, means for generating responses to the inquiries from a database and providing personalized answers to the users, means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry, means for generating answers by referencing the user's profile information and past dialogue history, means for analyzing and classifying received data as text, and means for automatically sending answers to the inquiries to the users. This enables centralized management of inquiries, rapid analysis and classification, and personalized responses.

[0953] "Multiple inquiry channels" refers to multiple ways that users can contact you (help, chat, email, phone, etc.).

[0954] "Centralized reception means" refers to a method of receiving and managing inquiries received from multiple inquiry routes in a single integrated system.

[0955] "Generative AI means" refers to artificial intelligence technology that analyzes the content of inquiries, extracts keywords and phrases, and classifies them.

[0956] A "database" refers to an information system that accumulates information and knowledge bases necessary for responding to inquiries and manages them in a searchable manner.

[0957] "Personalized answers" refer to responses that provide answers optimized for a specific user based on the user's individual profile information and past interaction history.

[0958] "Escalation procedures" refer to methods for automatically transferring inquiries to specific specialist staff or departments depending on the urgency and complexity of the inquiry.

[0959] "Profile information" refers to individual data such as a user's basic information, past inquiry history, and behavioral patterns.

[0960] "Interaction history" refers to data that records all past interactions between a user and a support system.

[0961] "Means for analyzing and classifying as text" refers to a technology for converting received inquiry data into text format and analyzing and classifying the content.

[0962] "Means for automatic sending" refers to the method by which the system automatically sends the generated answers to the user.

[0963] This invention is a customer support system that integrates multiple inquiry routes, centrally accepts inquiries from users, analyzes and classifies them using a generation AI, generates personalized answers from a database, and automatically executes escalation procedures as needed. This system consists of a server that provides a web form, a terminal that accepts user input, a generation AI that analyzes the content of the inquiry, and a means for sending answers to users.

[0964] The server provides a web form that users can access and centrally accept inquiries. This form integrates multiple inquiry routes, such as help, chat, email, and phone. Users enter the required information in the form and submit their inquiry.

[0965] The inquiry entered by the user into the web form is captured by the device. In the case of voice input, the device converts the voice into text using voice recognition technology (e.g., Google Cloud Speech-to-Text). The converted text data is sent to the server.

[0966] The server sends the received text data to a generation AI (e.g., OpenAI's GPT-4), which extracts keywords and phrases from the inquiry and identifies the topic and complexity of the problem. The generation AI then categorizes the inquiry into categories such as technical support, account issues, and payment issues based on the extracted keywords.

[0967] The server searches and retrieves relevant information from a database (e.g., Confluence) based on the category information obtained from the generation AI. If necessary, the generation AI references past interaction history and user profile information to generate a personalized answer. The generated answer is sent to the user via the server.

[0968] The generative AI determines the urgency and complexity of the inquiry and automatically executes escalation procedures, where the server routes the inquiry to the necessary specialist staff and notifies the user of the progress.

[0969] The server stores all interaction history in a database (e.g., MySQL). The stored interaction history is used as reference material for responding to future inquiries, contributing to improving the response quality of the entire system.

[0970] Specific examples

[0971] For example, if a user visits a web form, types in "Account Recovery," and submits it, the system will:

[0972] 1. The device captures this input and sends it to the server.

[0973] 2. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[0974] 3. The server retrieves the account recovery instructions from the database, and the generative AI generates a personalized answer.

[0975] 4. The server sends the answer to the user.

[0976] 5. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[0977] 6. Finally, the server stores the conversation history and uses it for future responses.

[0978] Example prompts to input to the generative AI model

[0979] User asks: "My account is locked. How do I recover it?"

[0980] Prompt: "What should I do if a user's account is locked? Please suggest a specific solution based on past interaction history."

[0981] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0982] Step 1:

[0983] The server provides a web form that users can access, which integrates multiple contact channels, including help, chat, email, and phone.

[0984] Input: The server receives the user's access to a web form.

[0985] Output: Displays an interface where the user can enter a query.

[0986] Specific behavior: Initializes each field of the web form and displays the options for the inquiry route.

[0987] Step 2:

[0988] The user enters the necessary information and submits the inquiry.

[0989] Input: A user enters their inquiry or contact information into a web form and clicks submit.

[0990] Output: The input query data is generated and sent to the terminal.

[0991] What it does: Captures data entered by the user and converts it into structured data, such as JSON format.

[0992] Step 3:

[0993] The device captures the user's input and sends it to the server.

[0994] Input: The query data entered by the user.

[0995] Output: The query data is sent to the server.

[0996] Specific operation: In the case of voice input, the device uses voice recognition technology to convert the voice into text, and then sends the converted text data to the server.

[0997] Step 4:

[0998] The server sends the received text data to the generation AI.

[0999] Input: Text data sent from the terminal.

[1000] Output: Text data sent to the generation AI.

[1001] Specific operation: The server checks the received data, generates an appropriate prompt sentence, and sends the text data to the generation AI.

[1002] Step 5:

[1003] The generation AI analyzes the inquiry content and extracts keywords and phrases.

[1004] Input: Text data sent to the generation AI.

[1005] Output: Extracted keywords and phrases.

[1006] How it works: Generative AI uses natural language processing techniques to analyze text data and extract important keywords and phrases.

[1007] Step 6:

[1008] The generative AI classifies inquiries into categories based on the extracted keywords.

[1009] Input: Extracted keywords or phrases.

[1010] Output: Categorized inquiry categories.

[1011] How it works: Generative AI uses rule-based techniques and machine learning models to classify text data into categories such as tech support, account issues, and payment issues.

[1012] Step 7:

[1013] The server searches and retrieves related information from the database based on the category information obtained from the generation AI.

[1014] Input: The classified inquiry category.

[1015] Output: Database results with relevant information.

[1016] Specific operations: The server executes a database query to obtain knowledge base information related to the inquiry category.

[1017] Step 8:

[1018] The generative AI references past interaction history and user profile information to generate personalized answers.

[1019] Input: Relevant information retrieved from databases, past interaction history, and user profile information.

[1020] Output: A personalized answer.

[1021] Specific behavior: Generative AI synthesizes the acquired information and generates recommended answers.

[1022] Step 9:

[1023] The server generates the answer and sends it to the user.

[1024] Input: Personalized answers from generative AI.

[1025] Output: The answer sent to the user.

[1026] What happens: The server formats the answer and sends it to the user in some form, such as email or chat.

[1027] Step 10:

[1028] The generative AI determines the urgency and complexity of the inquiry and automatically executes escalation procedures.

[1029] Input: Generative AI assessment of the urgency and complexity of the inquiry.

[1030] Output: Notification if escalation is required.

[1031] Specific behavior: The generative AI assesses the urgency and complexity of the inquiry and, if necessary, issues instructions to transfer the inquiry to specialized staff.

[1032] Step 11:

[1033] The server will notify the user of the progress of the escalation.

[1034] Input: Escalation procedure progress information.

[1035] Output: Progress notification to the user.

[1036] What it does: The server periodically checks the progress and notifies the user with status updates.

[1037] Step 12:

[1038] The server stores all interaction history in a database.

[1039] Input: Interaction data stored on the server.

[1040] Output: The saved interaction history.

[1041] Specific operation: The server records the dialogue history in a database and makes it available for use in responding to future inquiries.

[1042] (Application example 1)

[1043] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1044] Customer support on modern online shopping sites can be complicated for users, as there are multiple routes for inquiries (email, chat, phone, etc.). This means that there is a demand for consistent and prompt responses to inquiries. There is also a growing demand for customer support that supports voice input, and a system that can respond to this demand efficiently is required.

[1045] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1046] In this invention, the server includes: means for integrating multiple inquiry routes and centrally accepting inquiries from users; means for providing an interface for users to make inquiries through smart devices; speech recognition means for converting voice inquiries into text; generative AI means for analyzing inquiries accepted from users and classifying the topic and complexity of the problem; means for creating prompt sentences using a generative AI model and generating responses to the inquiries from a knowledge base to provide users with personalized answers; means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry; and means for saving the dialogue history of the inquiry and using it for responding to future inquiries. This enables fast and consistent customer support, improving the user experience and efficiently responding to inquiries.

[1047] "Multiple inquiry routes" refers to multiple ways that users can make inquiries, such as email, chat, phone, and voice input.

[1048] "Smart devices" refers to portable electronic devices with internet connectivity, such as smartphones and tablets.

[1049] "Interface" refers to the function of providing a screen and input means for users to interact with a system.

[1050] "Voice recognition means" refers to technology that converts voice into text data when a user makes a voice inquiry.

[1051] "Generative AI methods" refers to artificial intelligence technologies that analyze incoming inquiries, categorize them by topic and complexity, and generate appropriate responses.

[1052] "Generative AI model" refers to a trained AI model that generates prompts based on the content of an inquiry and creates appropriate responses.

[1053] A "prompt" is a phrase that is input into an AI model to generate an optimal response.

[1054] A "knowledge base" refers to a database that stores past inquiry data and knowledge, and is used as a source of information for generating responses.

[1055] "Personalized answers" refers to responses that are customized based on a user's individual situation and history.

[1056] "Escalation measures" refers to the function of automatically transferring an inquiry to specialized staff depending on the urgency and complexity of the inquiry.

[1057] "Dialogue history" refers to a record of past inquiries and responses with a user.

[1058] This invention relates to a customer support system that integrates multiple inquiry channels and centrally accepts inquiries from users. It is intended to improve the efficiency of customer support, primarily on online shopping sites, and enhance the user experience. The system for implementing this invention is as follows:

[1059] Overall system configuration

[1060] The system consists of three main elements: a server, a terminal, and a user.

[1061] User Interface Settings

[1062] The server provides a user-accessible web form or smart device interface that integrates multiple inquiry channels (email, chat, phone, voice input) and allows users to easily select the type of inquiry they wish to make. The user can then enter the required information and submit the inquiry.

[1063] Accepting and processing user input

[1064] When a user types or records a query via a smart device, the data is captured by the device. If the user speaks the query, the voice data is converted into text using voice recognition technology (e.g., SpeechRecognition library). The converted text data is then sent to the server.

[1065] Inquiry analysis and classification

[1066] The server sends the received text data to a generative AI (e.g., OpenAI's language model), which extracts keywords and phrases from the inquiry, identifies the topic and complexity of the problem, and classifies the inquiry into categories (e.g., technical support, order issues, payment issues, etc.).

[1067] Generate and send auto-replies

[1068] The server retrieves relevant information from a knowledge base based on the classified inquiry, and the generative AI model generates a personalized answer based on past interaction history and user profile information. This answer is then sent to the user via the server.

[1069] Escalation Procedures

[1070] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[1071] Saving inquiry history

[1072] The server stores all dialogue history in a database, which is used to respond to future inquiries and contribute to improving the response quality of the entire system.

[1073] Specific examples

[1074] The user opens the smartphone app and makes a voice inquiry saying, "Please tell me how to return an item." The device captures the voice data and converts it into text using the SpeechRecognition library. The text passed to the generation AI is analyzed and classified as "I would like to know how to return an item." The generation AI model retrieves relevant return methods from the knowledge base and generates a personalized answer saying, "To find out how to return an item, select your order history from the account page and click the return option." This answer is then sent to the user via the server.

[1075] Prompt Sentence Examples

[1076] An example prompt is:

[1077] Inquiry: How do I return the item?

[1078] Please parse this and generate an appropriate answer.

[1079] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1080] Step 1:

[1081] The user types or records their inquiry via voice via their smart device.

[1082] Input: Text or voice data of the inquiry.

[1083] Output: The captured text or audio data.

[1084] Specific operation: The user opens the smartphone app and asks by voice, "Please tell me how to return the item."

[1085] Step 2:

[1086] The device captures the data and uses voice recognition technology to convert the voice data into text.

[1087] Input: Captured audio data.

[1088] Output: The converted text data.

[1089] What happens: The device captures audio data and converts it to text using the SpeechRecognition library.

[1090] Step 3:

[1091] The terminal transmits the converted text data to the server.

[1092] Input: The converted text data.

[1093] Output: The text data sent to the server.

[1094] Specific operation: The device makes an API request to send the converted text data to the server.

[1095] Step 4:

[1096] The server sends the received text data to the generation AI.

[1097] Input: Text data sent to the server.

[1098] Output: Text data sent to the generation AI.

[1099] Specific operation: The server sends text data to the generative AI model via the API.

[1100] Step 5:

[1101] Generative AI analyzes the query, extracting keywords and phrases to identify the topic and complexity of the problem.

[1102] Input: Text data sent to the generation AI.

[1103] Output: The topic and complexity of the parsed query.

[1104] How it works: A generative AI model (e.g., OpenAI's language model) analyzes text data, extracts important keywords, and classifies them into problem categories.

[1105] Step 6:

[1106] The server retrieves relevant information from a knowledge base based on the classified query.

[1107] Input: The topic and complexity of the parsed query.

[1108] Output: Relevant information retrieved from the knowledge base.

[1109] Specific operation: The server accesses the knowledge base and obtains relevant information (e.g., how to process returns) based on the analysis results.

[1110] Step 7:

[1111] The generative AI model references past interaction history and user profile information to generate personalized answers.

[1112] Input: Relevant information retrieved from the knowledge base, and the user's profile information.

[1113] Output: A personalized answer.

[1114] How it works: The generative AI model uses the user's past inquiry history and account information to generate the most appropriate answer.

[1115] Step 8:

[1116] The server generates the answer and sends it to the user.

[1117] Input: Personalized answers from generative AI.

[1118] Output: The answer sent to the user.

[1119] What it does: The server sends a personalized answer to the user's smart device.

[1120] Step 9:

[1121] Generative AI determines the urgency and complexity of the inquiry and automatically executes escalation procedures.

[1122] Input: User's query and its analysis results.

[1123] Output: The query routed to the required specialist.

[1124] Specific operation: The generating AI determines the urgency of the inquiry, and if necessary, the server transfers the inquiry to specialized staff.

[1125] Step 10:

[1126] The server stores all interaction history in a database.

[1127] Input: User interaction history.

[1128] Output: Dialogue history stored in a database.

[1129] Specific operation: The server stores all interaction history with the user in a database and uses it to respond to future inquiries.

[1130] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1131] MODE FOR CARRYING OUT THE INVENTION

[1132] This invention relates to a customer support system that integrates multiple inquiry routes, centrally accepts inquiries from users, analyzes and classifies them using generative AI, generates personalized answers from a knowledge base, and automatically executes escalation procedures as needed. Furthermore, by combining it with an emotion engine that recognizes user emotions, the content and tone of the response can be adjusted to achieve more personalized responses.

[1133] 1. User Interface Configuration

[1134] The server provides a web form accessible to users that integrates multiple contact channels (help, chat, email, phone) and includes an interface that allows users to select the type of inquiry they wish to make. Users can then enter the required information and submit their inquiry.

[1135] 2. Accepting and Processing User Input

[1136] The device captures the user's query entered into the web form. If the user speaks the query, it is converted into text using speech recognition technology. This input data is then sent to the server.

[1137] 3. Inquiry Analysis and Classification

[1138] The server then passes the received text data to the generation AI, which analyzes the inquiry, extracts keywords and phrases, and, based on the analysis results, identifies the topic and complexity of the problem and classifies the inquiry into a specific category.

[1139] 4. Generating and sending auto-replies

[1140] The server sends the query to a knowledge base based on the classified query content, retrieves relevant information from the knowledge base, and then the generative AI references past interaction history and user profile information to generate a personalized answer. This answer is then sent to the user via the server.

[1141] 5. Emotional analysis of users using an emotion engine

[1142] The server passes the user's inquiry data to the emotion engine, which analyzes the user's emotions from text and voice data and provides the results to the generation AI. The generation AI takes this emotional data into account and appropriately adjusts the content and tone of the response.

[1143] 6. Escalation Procedures

[1144] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[1145] 7. Saving inquiry history

[1146] The server stores all dialogue history in a database. This history can be used to respond to future inquiries, contributing to improving the response quality of the entire system. In addition, user emotional data is also saved as a log, enabling more accurate and personalized responses for future inquiries.

[1147] Specific examples

[1148] 1. The user accesses the web form, enters "Account Recovery" and submits it.

[1149] 2. The device captures this input and sends it to the server.

[1150] 3. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[1151] 4. The server retrieves the account recovery instructions from the knowledge base, and the generative AI generates a personalized answer.

[1152] 5. The server sends the answer to the user.

[1153] 6. At the same time, the server passes the user's input data to the emotion engine, which analyzes the user's emotions. The generative AI takes this analysis into account and adjusts the tone and content of the response.

[1154] 7. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[1155] 8. Finally, the server stores the dialogue history and emotion data for future use.

[1156] In this way, this invention makes it possible to improve the efficiency of customer support operations, realize 24-hour support, provide consistent responses, and reduce costs. In addition, the use of an emotion engine enables responses that are more in tune with emotions, contributing to an improved user experience.

[1157] The processing flow will be explained below.

[1158] Step 1:

[1159] A user visits a website's support page and opens a contact form. They select the type of inquiry they want, choosing a category such as "Account Recovery" or "Technical Support," then enter their inquiry details using text input or voice input and click the submit button.

[1160] Step 2:

[1161] The device captures the user's input data. In the case of text input, it stores the data as text. In the case of voice input, it uses a speech recognition API to convert the voice data into text.

[1162] Step 3:

[1163] The device sends the captured text data to the server.

[1164] Step 4:

[1165] The server passes the received text data to the generation AI, which analyzes the inquiry and extracts keywords and phrases. Based on the analysis results, the AI ​​identifies the topic and complexity of the problem and classifies the inquiry into a specific category, such as technical support, account issues, or payment issues.

[1166] Step 5:

[1167] The server sends a query to the knowledge base based on the classified query content, and retrieves related information from the knowledge base.

[1168] Step 6:

[1169] The server passes the acquired information to the generation AI, which then references past interaction history and user profile information to generate a personalized response.

[1170] Step 7:

[1171] The server then sends the generated response to the user, analyzing the user's emotions using an emotion engine and adjusting the tone and content accordingly. For example, if the user is feeling frustrated, the tone of the response will be more friendly.

[1172] Step 8:

[1173] The server uses the AI ​​to determine the urgency and complexity of the inquiry, and automatically executes escalation procedures as needed. For example, if the issue is technically difficult, the inquiry will be transferred to a specialist.

[1174] Step 9:

[1175] The server stores the conversation history in a database. All conversations are logged and used to handle future inquiries and improve the knowledge base. In addition, user sentiment data is also stored.

[1176] Step 10:

[1177] The server notifies the user of the progress of the escalation and completion of the automated response, allowing the user to see in real time how their inquiry is being handled.

[1178] Specific examples

[1179] 1. The user accesses the web form, enters "account recovery" as the inquiry, and clicks submit.

[1180] 2. The device captures this input data and sends it to the server.

[1181] 3. The server passes the content to the generation AI, which then classifies "Account Recovery" into the technical support category.

[1182] 4. The server retrieves the account recovery instructions from the knowledge base, and the generative AI generates a personalized answer.

[1183] 5. The server provides the generated answer to the user, with an emotion engine analyzing the user's emotions and adjusting the tone of the answer to be more friendly.

[1184] 6. If the user's problem is not resolved, the generation AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[1185] 7. Finally, the server stores the dialogue history and emotion data for use in responding to future inquiries.

[1186] Through this flow, the present invention improves the efficiency of customer support operations, enables 24-hour support, provides consistent responses, reduces costs, and improves user experience.

[1187] Example 2

[1188] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1189] In conventional customer support systems, multiple inquiry routes were managed separately, making it difficult to provide consistent responses. Furthermore, the lack of escalation procedures and emotionally sensitive responses made it difficult to improve the user experience. Furthermore, the lack of effective storage and utilization of inquiry history made it difficult to improve support quality.

[1190] The identification process 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 a means for integrating multiple inquiry routes and centrally accepting inquiries from users, a generation AI means for analyzing inquiries accepted from users and classifying the topic and complexity of the problem, a means for generating responses to inquiries from an information base and providing users with personalized answers, a means for passing inquiry data to a sentiment analysis engine and adjusting the content and tone of the response based on the analysis results, a means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry, and a means for saving the dialogue history and sentiment data of the inquiry and using them in future inquiry responses. This enables consistent and prompt responses, responses that are sensitive to the user's sentiment, prompt escalation, and effective use of the inquiry history.

[1191] "Multiple contact channels" are ways for users to contact you through different channels, such as help, chat, email, or phone.

[1192] "Centralized reception" refers to the integrated management of information received from multiple inquiry methods in a single system.

[1193] "Analysis" refers to the process by which the generative AI understands the content of the inquiry it receives, extracts keywords and phrases, and evaluates the topic and complexity of the problem.

[1194] "Generative AI" refers to a system that uses artificial intelligence technology to understand the content of an inquiry and provide appropriate answers or classifications.

[1195] An "information base" refers to a database that compiles past inquiry data and specialized knowledge, and is used to generate appropriate answers.

[1196] "Personalized answers" refer to responses that are individually tailored to the user, taking into account their profile and past interaction history.

[1197] An "emotion analysis engine" refers to a system that analyzes emotions from user inquiry text and voice data and provides the results to the generative AI.

[1198] "Escalation Procedures" refers to the process of automatically transferring inquiries to higher-level specialist staff based on the urgency or complexity of the inquiry.

[1199] "Dialogue history" refers to a record of past communication between the user and the system, and is used as reference material for responding to future inquiries.

[1200] "Emotion data" is data that indicates the emotional state of a user extracted by an emotion analysis engine.

[1201] This invention relates to a customer support system that integrates multiple inquiry routes and centrally accepts inquiries from users. This system uses generative AI and a sentiment analysis engine to analyze inquiries, generate automatic responses, execute escalation procedures, and provide users with personalized responses.

[1202] Hardware and software used

[1203] This system uses the following hardware and software.

[1204] Server: Plays a central role in accepting and processing user queries.

[1205] Terminal: Captures user input (text and voice) and sends it to the server.

[1206] Generative AI models: Analyze incoming queries and classify the topic and complexity of the problem.

[1207] Sentiment analysis engine: Analyzes user inquiry data and recognizes emotions.

[1208] Data processing and calculation

[1209] Each function of the system is described in detail below.

[1210] User Interface Settings

[1211] The server provides a user-accessible web form that integrates multiple inquiry channels, such as help, chat, email, and phone, and includes an interface through which the user can select the type of inquiry, through which the user can enter the required information and submit the inquiry.

[1212] Receiving and processing inquiries

[1213] The user's query entered into the web form is captured by the device. If the query is voice, it is converted to text using speech recognition technology. This input data is then sent to the server.

[1214] Inquiry analysis and classification

[1215] The server then passes the received text data to the generation AI, which analyzes the query and extracts keywords and phrases. Based on the analysis results, the AI ​​identifies the topic and complexity of the problem and classifies the query into a specific category.

[1216] Generate and send auto-replies

[1217] The server sends a query to an information base based on the classified query content to retrieve relevant information. The generation AI then references past interaction history and user profile information to generate a personalized answer, which is then sent to the user via the server.

[1218] Sentiment analysis and response tailoring

[1219] The server passes the user's inquiry data to the emotion analysis engine. The emotion engine analyzes the user's emotions from the text and voice data and provides the results to the generation AI. The generation AI takes this emotional data into account and adjusts the content and tone of the response appropriately.

[1220] Escalation Procedures

[1221] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[1222] Saving inquiry history

[1223] The server stores all dialogue history in a database. This history can be used to respond to future inquiries, contributing to improving the response quality of the entire system. In addition, user emotional data is also saved as a log, enabling more accurate and personalized responses for future inquiries.

[1224] Specific examples

[1225] 1. The user accesses the web form, enters "Account Recovery" and submits it.

[1226] 2. The device captures this input and sends it to the server.

[1227] 3. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[1228] 4. The server retrieves account recovery instructions from its information base, and the generative AI generates a personalized answer.

[1229] 5. The server sends the answer to the user.

[1230] 6. At the same time, the server passes the user's input data to the emotion analysis engine, which analyzes the user's emotions. The generative AI takes this analysis into account and adjusts the tone and content of the response.

[1231] 7. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[1232] 8. Finally, the server stores the dialogue history and emotion data for future use.

[1233] In this way, this invention makes it possible to improve the efficiency of customer support operations, realize 24-hour support, provide consistent responses, and reduce costs. In addition, the use of a sentiment analysis engine enables responses that are more sensitive to emotions, contributing to an improved user experience.

[1234] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1235] Step 1:

[1236] A user accesses a web form, enters inquiry information, and presses the "Submit" button. At this time, the input is text-based information or voice data. Take the example of a user entering "Account recovery." Input: The user's inquiry. Output: The input inquiry data.

[1237] Step 2:

[1238] The device captures the user's input, and if the input is voice, converts it into text data using voice recognition technology. The converted text data is sent to the server. Input: User's raw voice or text. Output: Text data.

[1239] Step 3:

[1240] The server sends the received text data to the generation AI. The generation AI analyzes the received inquiry and extracts keywords and phrases. Input: Text data received by the server. Output: Keyword extraction and analysis results by the generation AI.

[1241] Step 4:

[1242] The server identifies the topic and complexity of the inquiry based on the analysis results from the generative AI and classifies it into an appropriate category. For example, "account recovery" is classified into the "technical support" category. Input: Analysis results from the generative AI. Output: Enquiry category classification.

[1243] Step 5:

[1244] The server queries the information base for the classified inquiry and retrieves relevant information. The data retrieved from the information base is sent to the generation AI, which generates a personalized response by referencing past interaction history and user profile information. Input: Inquiry category classification. Output: Personalized response.

[1245] Step 6:

[1246] The server sends the generated personalized response to the user, for example, "Hello, here are the steps to recover your account." Input: The generated personalized response. Output: The response sent to the user.

[1247] Step 7:

[1248] The server passes the inquiry data to the sentiment analysis engine, which identifies the user's sentiment and provides the results to the generation AI. For example, if the user enters "I'm in a hurry," the sentiment analysis engine recognizes a high level of urgency. Input: User inquiry data. Output: Sentiment analysis results.

[1249] Step 8:

[1250] The generation AI takes into account the results of the sentiment analysis and adjusts the content and tone of the response. For example, it generates a response that takes sentiment into account, such as "We will respond immediately." Input: Sentiment analysis results. Output: Adjusted response.

[1251] Step 9:

[1252] The server adjusts the content and tone of the response and then sends it to the user. Input: Adjusted response. Output: Adjusted response sent to the user.

[1253] Step 10:

[1254] The generation AI reassess the urgency and complexity of the inquiry and determines whether escalation is necessary. Urgent or highly difficult inquiries are automatically transferred to specialized staff. The server notifies the user of the progress of the escalation. Input: Reassessed urgency and complexity. Output: Transfer to specialized staff, notification to user.

[1255] Step 11:

[1256] The server stores all dialogue history and emotion data in a database. The stored data is used to respond to future inquiries and helps improve the response quality of the entire system. Input: dialogue history and emotion data. Output: stored data.

[1257] (Application example 2)

[1258] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1259] Conventional customer support systems have a lack of uniformity in responses to user inquiries, making it difficult to provide personalized responses. Furthermore, responses that do not take into account the user's feelings can lead to a decline in customer satisfaction. Furthermore, manual escalation procedures can lead to delays and errors.

[1260] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for integrating multiple inquiry routes and centrally accepting inquiries from users; generation AI means for analyzing inquiries accepted from users and classifying the topic and complexity of the problem; means for generating responses to inquiries from a knowledge base and providing users with personalized answers; emotion engine means for analyzing user emotions and adjusting the content and tone of the response; means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry; and means for saving the inquiry dialogue history and using it for future inquiry responses. This enables more efficient customer support operations, 24-hour support, consistent responses, cost reduction, and improved customer satisfaction through responses that are sensitive to emotions.

[1261] "Enquiry channels" refer to the multiple ways or channels through which users can make inquiries.

[1262] "A means of centralized reception" refers to a function that allows inquiries sent from multiple inquiry routes to be received in one place.

[1263] "Generative AI means of analyzing and classifying" refers to the generative AI's ability to analyze the content of a query and classify it based on the topic and complexity of the problem.

[1264] A "knowledge base" refers to a database that stores and organizes answers to inquiries and information.

[1265] "Means for providing personalized answers" refers to the ability to generate and provide individually customized answers based on the user's inquiry and profile information.

[1266] "Emotion engine means" refers to a system that analyzes a user's emotions and adjusts the content and tone of the response based on the results.

[1267] "Means for implementing escalation procedures" refers to the ability to automatically route inquiries to specialized staff based on the urgency and complexity of the inquiry.

[1268] "Means for saving dialogue history and using it to respond to future inquiries" refers to a function that records the content and responses of past inquiries and uses them to respond to future inquiries.

[1269] A specific system for implementing the present invention has the following configuration.

[1270] First, the server integrates multiple inquiry channels and centrally accepts user inquiries. This includes web forms, chat boxes, and voice input. When a user makes an inquiry, the data is sent to the server. The server captures this data and passes it to a generative AI model for analysis. The analysis classifies the topic and complexity of the problem.

[1271] The server then references the knowledge base and uses a generative AI model to generate a personalized answer. This answer also takes into account the user's profile information and past interaction history. The emotion engine also analyzes the emotions from the user's input data and provides the results to the generative AI model. The generative AI model uses this emotional data to appropriately adjust the content and tone of the response.

[1272] If the inquiry is urgent or complex, the server automatically executes an escalation procedure and transfers the inquiry to a specialist staff member. The user is notified of the progress. In addition, the dialogue history and emotion data are recorded in a database and used to handle future inquiries.

[1273] Hardware and software used:

[1274] Server: A cloud server that provides an API endpoint (e.g., AWS, Google Cloud)

[1275] Generative AI models: text analysis and response generation (e.g., GPT-4)

[1276] Emotion engine: Sentiment analysis (e.g. IBM Watson, Microsoft Azure Emotion API)

[1277] Device: Capturing user input (e.g. smartphone, HMD)

[1278] Examples:

[1279] For example, if a user voice-inquires within a virtual store's smartphone app, "How do I return an item?", the voice data is sent to a server and converted into text using speech recognition technology. A generative AI model analyzes the inquiry, retrieves information about the "return procedure" from a knowledge base, and generates and sends a personalized response taking into account the user's profile information. Furthermore, an emotion engine analyzes the user's emotions (e.g., frustration), and the generative AI model adjusts the tone of the response taking these emotions into account.

[1280] Example prompt sentence:

[1281] I have an item I'd like to return, how do I go about doing that?

[1282] "I'm having issues with the quality of the product I purchased. What should I do?"

[1283] In this way, real-time responses that are sensitive to emotions become possible, contributing to an improved user experience.

[1284] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1285] Step 1:

[1286] The user makes an inquiry. The user selects an appropriate input method from multiple inquiry routes (e.g., web form, chat box, voice input) and types or speaks the inquiry content. The input data is converted into text using voice recognition technology.

[1287] Input: User text or voice data

[1288] Output: Query data converted to text

[1289] Step 2:

[1290] The device captures the user's input data and sends it to the server. If there is voice input, the device processes it to convert the voice into text.

[1291] Input: User's inquiry (text or voice data)

[1292] Output: Text data sent to the server

[1293] Step 3:

[1294] The server then passes the received text data to a generative AI model for analysis, which extracts keywords and phrases from the query and categorizes them based on the topic and complexity of the problem.

[1295] Input: Text data

[1296] Output: Parsed data (topic and complexity classification results)

[1297] Step 4:

[1298] The server uses the parsed data to consult a knowledge base to retrieve relevant information, and a generative AI model takes this information, along with the user's profile information, into account to generate a personalized answer.

[1299] Input: Analytics data, knowledge base information, user profile information

[1300] Output: Personalized answer

[1301] Step 5:

[1302] The server passes the user's text data to the emotion engine for emotion analysis. The emotion engine analyzes the user's emotions from the text data and provides the results to the generative AI model.

[1303] Input: User's text data

[1304] Output: Emotion analysis results

[1305] Step 6:

[1306] The generative AI model takes into account the results of sentiment analysis and adjusts the content and tone of the response accordingly, resulting in a personalized response that takes the user's emotions into account.

[1307] Input: Sentiment analysis results, personalized answers

[1308] Output: Emotionally sensitive personalized answers

[1309] Step 7:

[1310] The server then sends the generated answer to the user, and simultaneously automatically routes the inquiry to a specialist if escalation is required, and notifies the user of the progress.

[1311] Input: Emotionally sensitive personalized answers

[1312] Output: Response sent to user, escalation action

[1313] Step 8:

[1314] The server stores all dialogue history and emotion data in a database, which is used to respond to future inquiries and contribute to improving the response quality of the entire system.

[1315] Input: Dialogue history, emotion data

[1316] Output: Historical data stored in a database

[1317] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1318] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1319] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1320] [Fourth embodiment]

[1321] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1322] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1323] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1324] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1325] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1326] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1327] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1328] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1329] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1330] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1331] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1332] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1333] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1334] MODE FOR CARRYING OUT THE INVENTION

[1335] This invention relates to a customer support system that integrates multiple inquiry routes, centrally accepts inquiries from users, analyzes and classifies them using generative AI, generates personalized answers from a knowledge base, and automatically executes escalation procedures as necessary.

[1336] 1. User Interface Configuration

[1337] The server provides a web form accessible to users that integrates multiple contact channels (help, chat, email, phone) and includes an interface that allows users to select the type of inquiry they wish to make. Users can then enter the required information and submit their inquiry.

[1338] 2. Accepting and Processing User Input

[1339] The device captures the user's query entered into the web form. If the user speaks the query, it is converted into text using speech recognition technology. This input data is then sent to the server.

[1340] 3. Inquiry Analysis and Classification

[1341] The server sends the received text data to the generation AI, which performs the following steps:

[1342] 1. Extract keywords and phrases from the inquiry to identify the topic and complexity of the problem.

[1343] 2. Categorize inquiries into categories (e.g., technical support, account issues, payment issues, etc.).

[1344] 4. Generating and sending auto-replies

[1345] The server retrieves relevant information from a knowledge base based on the classified inquiry, and then the generative AI generates a personalized answer based on past interaction history and user profile information. This answer is then sent to the user via the server.

[1346] 5. Escalation Procedures

[1347] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[1348] 6. Saving inquiry history

[1349] The server stores all dialogue history in a database, which is used to respond to future inquiries and contribute to improving the response quality of the entire system.

[1350] Specific examples

[1351] 1. The user accesses the web form, enters "Account Recovery" and submits it.

[1352] 2. The device captures this input and sends it to the server.

[1353] 3. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[1354] 4. The server retrieves the account recovery instructions from the knowledge base, and the generative AI generates a personalized answer.

[1355] 5. The server sends the answer to the user.

[1356] 6. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[1357] 7. Finally, the server stores the conversation history and uses it for future responses.

[1358] In this way, the present invention makes it possible to improve the efficiency of customer support operations, realize 24-hour support, provide consistent responses, reduce costs, and also contribute to improving the user experience.

[1359] The processing flow will be explained below.

[1360] Step 1:

[1361] A user visits a support page on a website and opens a contact form. The user selects the type of inquiry and enters the inquiry details using text input or voice input. The user clicks the submit button.

[1362] Step 2:

[1363] The device captures the user's input data. In the case of text input, it stores the data as text. In the case of voice input, it uses a speech recognition API to convert the voice data into text.

[1364] Step 3:

[1365] The device sends the captured text data to the server.

[1366] Step 4:

[1367] The server passes the received text data to the generation AI, which analyzes the inquiry and extracts keywords and phrases. Based on the analysis results, it identifies the topic and complexity of the problem and classifies the inquiry into a specific category.

[1368] Step 5:

[1369] The server sends a query to the knowledge base based on the classified query content, and retrieves related information from the knowledge base.

[1370] Step 6:

[1371] The server passes the acquired information to the generation AI, which then references past interaction history and user profile information to generate a personalized response.

[1372] Step 7:

[1373] The server sends the generated answer to the user. If the user receives the answer and has additional questions or inquiries, the process continues from step 1 again.

[1374] Step 8:

[1375] The server uses the generated AI to determine the urgency and complexity of the inquiry, automatically executing escalation procedures as needed, and transferring the inquiry to a specialist staff member if the issue is complex technical.

[1376] Step 9:

[1377] The server stores the dialogue history in a database. All inquiry dialogues are recorded as logs and used to handle future inquiries and improve the knowledge base.

[1378] Step 10:

[1379] The server notifies the user of the progress of the escalation and completion of the automated response, allowing the user to see in real time how their inquiry is being handled.

[1380] Example 1

[1381] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1382] Current customer support systems often result in delayed responses and misunderstandings. In particular, when multiple inquiry channels exist, managing inquiries can become cumbersome, potentially reducing user satisfaction. Furthermore, the analysis and classification of inquiries is performed manually, making it difficult to provide a prompt response. Furthermore, the lack of personalized responses means that user needs cannot be fully met. The present invention aims to solve these problems.

[1383] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1384] In this invention, the server includes means for integrating multiple inquiry routes and centrally accepting inquiries from users, generation AI means for analyzing the inquiries received from users and classifying the topic and complexity of the problem, means for generating responses to the inquiries from a database and providing personalized answers to the users, means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry, means for generating answers by referencing the user's profile information and past dialogue history, means for analyzing and classifying received data as text, and means for automatically sending answers to the inquiries to the users. This enables centralized management of inquiries, rapid analysis and classification, and personalized responses.

[1385] "Multiple inquiry channels" refers to multiple ways that users can contact you (help, chat, email, phone, etc.).

[1386] "Centralized reception means" refers to a method of receiving and managing inquiries received from multiple inquiry routes in a single integrated system.

[1387] "Generative AI means" refers to artificial intelligence technology that analyzes the content of inquiries, extracts keywords and phrases, and classifies them.

[1388] A "database" refers to an information system that accumulates information and knowledge bases necessary for responding to inquiries and manages them in a searchable manner.

[1389] "Personalized answers" refer to responses that provide answers optimized for a specific user based on the user's individual profile information and past interaction history.

[1390] "Escalation procedures" refer to methods for automatically transferring inquiries to specific specialist staff or departments depending on the urgency and complexity of the inquiry.

[1391] "Profile information" refers to individual data such as a user's basic information, past inquiry history, and behavioral patterns.

[1392] "Interaction history" refers to data that records all past interactions between a user and a support system.

[1393] "Means for analyzing and classifying as text" refers to a technology for converting received inquiry data into text format and analyzing and classifying the content.

[1394] "Means for automatic sending" refers to the method by which the system automatically sends the generated answers to the user.

[1395] This invention is a customer support system that integrates multiple inquiry routes, centrally accepts inquiries from users, analyzes and classifies them using a generation AI, generates personalized answers from a database, and automatically executes escalation procedures as needed. This system consists of a server that provides a web form, a terminal that accepts user input, a generation AI that analyzes the content of the inquiry, and a means for sending answers to users.

[1396] The server provides a web form that users can access and centrally accept inquiries. This form integrates multiple inquiry routes, such as help, chat, email, and phone. Users enter the required information in the form and submit their inquiry.

[1397] The inquiry entered by the user into the web form is captured by the device. In the case of voice input, the device converts the voice into text using voice recognition technology (e.g., Google Cloud Speech-to-Text). The converted text data is sent to the server.

[1398] The server sends the received text data to a generation AI (e.g., OpenAI's GPT-4), which extracts keywords and phrases from the inquiry and identifies the topic and complexity of the problem. The generation AI then categorizes the inquiry into categories such as technical support, account issues, and payment issues based on the extracted keywords.

[1399] The server searches and retrieves relevant information from a database (e.g., Confluence) based on the category information obtained from the generation AI. If necessary, the generation AI references past interaction history and user profile information to generate a personalized answer. The generated answer is sent to the user via the server.

[1400] The generative AI determines the urgency and complexity of the inquiry and automatically executes escalation procedures, where the server routes the inquiry to the necessary specialist staff and notifies the user of the progress.

[1401] The server stores all interaction history in a database (e.g., MySQL). The stored interaction history is used as reference material for responding to future inquiries, contributing to improving the response quality of the entire system.

[1402] Specific examples

[1403] For example, if a user visits a web form, types in "Account Recovery," and submits it, the system will:

[1404] 1. The device captures this input and sends it to the server.

[1405] 2. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[1406] 3. The server retrieves the account recovery instructions from the database, and the generative AI generates a personalized answer.

[1407] 4. The server sends the answer to the user.

[1408] 5. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[1409] 6. Finally, the server stores the conversation history and uses it for future responses.

[1410] Example prompts to input to the generative AI model

[1411] User asks: "My account is locked. How do I recover it?"

[1412] Prompt: "What should I do if a user's account is locked? Please suggest a specific solution based on past interaction history."

[1413] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1414] Step 1:

[1415] The server provides a web form that users can access, which integrates multiple contact channels, including help, chat, email, and phone.

[1416] Input: The server receives the user's access to a web form.

[1417] Output: Displays an interface where the user can enter a query.

[1418] Specific behavior: Initializes each field of the web form and displays the options for the inquiry route.

[1419] Step 2:

[1420] The user enters the necessary information and submits the inquiry.

[1421] Input: A user enters their inquiry or contact information into a web form and clicks submit.

[1422] Output: The input query data is generated and sent to the terminal.

[1423] What it does: Captures data entered by the user and converts it into structured data, such as JSON format.

[1424] Step 3:

[1425] The device captures the user's input and sends it to the server.

[1426] Input: The query data entered by the user.

[1427] Output: The query data is sent to the server.

[1428] Specific operation: In the case of voice input, the device uses voice recognition technology to convert the voice into text, and then sends the converted text data to the server.

[1429] Step 4:

[1430] The server sends the received text data to the generation AI.

[1431] Input: Text data sent from the terminal.

[1432] Output: Text data sent to the generation AI.

[1433] Specific operation: The server checks the received data, generates an appropriate prompt sentence, and sends the text data to the generation AI.

[1434] Step 5:

[1435] The generation AI analyzes the inquiry content and extracts keywords and phrases.

[1436] Input: Text data sent to the generation AI.

[1437] Output: Extracted keywords and phrases.

[1438] How it works: Generative AI uses natural language processing techniques to analyze text data and extract important keywords and phrases.

[1439] Step 6:

[1440] The generative AI classifies inquiries into categories based on the extracted keywords.

[1441] Input: Extracted keywords or phrases.

[1442] Output: Categorized inquiry categories.

[1443] How it works: Generative AI uses rule-based techniques and machine learning models to classify text data into categories such as tech support, account issues, and payment issues.

[1444] Step 7:

[1445] The server searches and retrieves related information from the database based on the category information obtained from the generation AI.

[1446] Input: The classified inquiry category.

[1447] Output: Database results with relevant information.

[1448] Specific operations: The server executes a database query to obtain knowledge base information related to the inquiry category.

[1449] Step 8:

[1450] The generative AI references past interaction history and user profile information to generate personalized answers.

[1451] Input: Relevant information retrieved from databases, past interaction history, and user profile information.

[1452] Output: A personalized answer.

[1453] Specific behavior: Generative AI synthesizes the acquired information and generates recommended answers.

[1454] Step 9:

[1455] The server generates the answer and sends it to the user.

[1456] Input: Personalized answers from generative AI.

[1457] Output: The answer sent to the user.

[1458] What happens: The server formats the answer and sends it to the user in some form, such as email or chat.

[1459] Step 10:

[1460] The generative AI determines the urgency and complexity of the inquiry and automatically executes escalation procedures.

[1461] Input: Generative AI assessment of the urgency and complexity of the inquiry.

[1462] Output: Notification if escalation is required.

[1463] Specific behavior: The generative AI assesses the urgency and complexity of the inquiry and, if necessary, issues instructions to transfer the inquiry to specialized staff.

[1464] Step 11:

[1465] The server will notify the user of the progress of the escalation.

[1466] Input: Escalation procedure progress information.

[1467] Output: Progress notification to the user.

[1468] What it does: The server periodically checks the progress and notifies the user with status updates.

[1469] Step 12:

[1470] The server stores all interaction history in a database.

[1471] Input: Interaction data stored on the server.

[1472] Output: The saved interaction history.

[1473] Specific operation: The server records the dialogue history in a database and makes it available for use in responding to future inquiries.

[1474] (Application example 1)

[1475] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1476] Customer support on modern online shopping sites can be complicated for users, as there are multiple routes for inquiries (email, chat, phone, etc.). This means that there is a demand for consistent and prompt responses to inquiries. There is also a growing demand for customer support that supports voice input, and a system that can respond to this demand efficiently is required.

[1477] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1478] In this invention, the server includes: means for integrating multiple inquiry routes and centrally accepting inquiries from users; means for providing an interface for users to make inquiries through smart devices; speech recognition means for converting voice inquiries into text; generative AI means for analyzing inquiries accepted from users and classifying the topic and complexity of the problem; means for creating prompt sentences using a generative AI model and generating responses to the inquiries from a knowledge base to provide users with personalized answers; means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry; and means for saving the dialogue history of the inquiry and using it for responding to future inquiries. This enables fast and consistent customer support, improving the user experience and efficiently responding to inquiries.

[1479] "Multiple inquiry routes" refers to multiple ways that users can make inquiries, such as email, chat, phone, and voice input.

[1480] "Smart devices" refers to portable electronic devices with internet connectivity, such as smartphones and tablets.

[1481] "Interface" refers to the function of providing a screen and input means for users to interact with a system.

[1482] "Voice recognition means" refers to technology that converts voice into text data when a user makes a voice inquiry.

[1483] "Generative AI methods" refers to artificial intelligence technologies that analyze incoming inquiries, categorize them by topic and complexity, and generate appropriate responses.

[1484] "Generative AI model" refers to a trained AI model that generates prompts based on the content of an inquiry and creates appropriate responses.

[1485] A "prompt" is a phrase that is input into an AI model to generate an optimal response.

[1486] A "knowledge base" refers to a database that stores past inquiry data and knowledge, and is used as a source of information for generating responses.

[1487] "Personalized answers" refers to responses that are customized based on a user's individual situation and history.

[1488] "Escalation measures" refers to the function of automatically transferring an inquiry to specialized staff depending on the urgency and complexity of the inquiry.

[1489] "Dialogue history" refers to a record of past inquiries and responses with a user.

[1490] This invention relates to a customer support system that integrates multiple inquiry channels and centrally accepts inquiries from users. It is intended to improve the efficiency of customer support, primarily on online shopping sites, and enhance the user experience. The system for implementing this invention is as follows:

[1491] Overall system configuration

[1492] The system consists of three main elements: a server, a terminal, and a user.

[1493] User Interface Settings

[1494] The server provides a user-accessible web form or smart device interface that integrates multiple inquiry channels (email, chat, phone, voice input) and allows users to easily select the type of inquiry they wish to make. The user can then enter the required information and submit the inquiry.

[1495] Accepting and processing user input

[1496] When a user types or records a query via a smart device, the data is captured by the device. If the user speaks the query, the voice data is converted into text using voice recognition technology (e.g., SpeechRecognition library). The converted text data is then sent to the server.

[1497] Inquiry analysis and classification

[1498] The server sends the received text data to a generative AI (e.g., OpenAI's language model), which extracts keywords and phrases from the inquiry, identifies the topic and complexity of the problem, and classifies the inquiry into categories (e.g., technical support, order issues, payment issues, etc.).

[1499] Generate and send auto-replies

[1500] The server retrieves relevant information from a knowledge base based on the classified inquiry, and the generative AI model generates a personalized answer based on past interaction history and user profile information. This answer is then sent to the user via the server.

[1501] Escalation Procedures

[1502] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[1503] Saving inquiry history

[1504] The server stores all dialogue history in a database, which is used to respond to future inquiries and contribute to improving the response quality of the entire system.

[1505] Specific examples

[1506] The user opens the smartphone app and makes a voice inquiry saying, "Please tell me how to return an item." The device captures the voice data and converts it into text using the SpeechRecognition library. The text passed to the generation AI is analyzed and classified as "I would like to know how to return an item." The generation AI model retrieves relevant return methods from the knowledge base and generates a personalized answer saying, "To find out how to return an item, select your order history from the account page and click the return option." This answer is then sent to the user via the server.

[1507] Prompt Sentence Examples

[1508] An example prompt is:

[1509] Inquiry: How do I return the item?

[1510] Please parse this and generate an appropriate answer.

[1511] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1512] Step 1:

[1513] The user types or records their inquiry via voice via their smart device.

[1514] Input: Text or voice data of the inquiry.

[1515] Output: The captured text or audio data.

[1516] Specific operation: The user opens the smartphone app and asks by voice, "Please tell me how to return the item."

[1517] Step 2:

[1518] The device captures the data and uses voice recognition technology to convert the voice data into text.

[1519] Input: Captured audio data.

[1520] Output: The converted text data.

[1521] What happens: The device captures audio data and converts it to text using the SpeechRecognition library.

[1522] Step 3:

[1523] The terminal transmits the converted text data to the server.

[1524] Input: The converted text data.

[1525] Output: The text data sent to the server.

[1526] Specific operation: The device makes an API request to send the converted text data to the server.

[1527] Step 4:

[1528] The server sends the received text data to the generation AI.

[1529] Input: Text data sent to the server.

[1530] Output: Text data sent to the generation AI.

[1531] Specific operation: The server sends text data to the generative AI model via the API.

[1532] Step 5:

[1533] Generative AI analyzes the query, extracting keywords and phrases to identify the topic and complexity of the problem.

[1534] Input: Text data sent to the generation AI.

[1535] Output: The topic and complexity of the parsed query.

[1536] How it works: A generative AI model (e.g., OpenAI's language model) analyzes text data, extracts important keywords, and classifies them into problem categories.

[1537] Step 6:

[1538] The server retrieves relevant information from a knowledge base based on the classified query.

[1539] Input: The topic and complexity of the parsed query.

[1540] Output: Relevant information retrieved from the knowledge base.

[1541] Specific operation: The server accesses the knowledge base and obtains relevant information (e.g., how to process returns) based on the analysis results.

[1542] Step 7:

[1543] The generative AI model references past interaction history and user profile information to generate personalized answers.

[1544] Input: Relevant information retrieved from the knowledge base, and the user's profile information.

[1545] Output: A personalized answer.

[1546] How it works: The generative AI model uses the user's past inquiry history and account information to generate the most appropriate answer.

[1547] Step 8:

[1548] The server generates the answer and sends it to the user.

[1549] Input: Personalized answers from generative AI.

[1550] Output: The answer sent to the user.

[1551] What it does: The server sends a personalized answer to the user's smart device.

[1552] Step 9:

[1553] Generative AI determines the urgency and complexity of the inquiry and automatically executes escalation procedures.

[1554] Input: User's query and its analysis results.

[1555] Output: The query routed to the required specialist.

[1556] Specific operation: The generating AI determines the urgency of the inquiry, and if necessary, the server transfers the inquiry to specialized staff.

[1557] Step 10:

[1558] The server stores all interaction history in a database.

[1559] Input: User interaction history.

[1560] Output: Dialogue history stored in a database.

[1561] Specific operation: The server stores all interaction history with the user in a database and uses it to respond to future inquiries.

[1562] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1563] MODE FOR CARRYING OUT THE INVENTION

[1564] This invention relates to a customer support system that integrates multiple inquiry routes, centrally accepts inquiries from users, analyzes and classifies them using generative AI, generates personalized answers from a knowledge base, and automatically executes escalation procedures as needed. Furthermore, by combining it with an emotion engine that recognizes user emotions, the content and tone of the response can be adjusted to achieve more personalized responses.

[1565] 1. User Interface Configuration

[1566] The server provides a web form accessible to users that integrates multiple contact channels (help, chat, email, phone) and includes an interface that allows users to select the type of inquiry they wish to make. Users can then enter the required information and submit their inquiry.

[1567] 2. Accepting and Processing User Input

[1568] The device captures the user's query entered into the web form. If the user speaks the query, it is converted into text using speech recognition technology. This input data is then sent to the server.

[1569] 3. Inquiry Analysis and Classification

[1570] The server then passes the received text data to the generation AI, which analyzes the inquiry, extracts keywords and phrases, and, based on the analysis results, identifies the topic and complexity of the problem and classifies the inquiry into a specific category.

[1571] 4. Generating and sending auto-replies

[1572] The server sends the query to a knowledge base based on the classified query content, retrieves relevant information from the knowledge base, and then the generative AI references past interaction history and user profile information to generate a personalized answer. This answer is then sent to the user via the server.

[1573] 5. Emotional analysis of users using an emotion engine

[1574] The server passes the user's inquiry data to the emotion engine, which analyzes the user's emotions from text and voice data and provides the results to the generation AI. The generation AI takes this emotional data into account and appropriately adjusts the content and tone of the response.

[1575] 6. Escalation Procedures

[1576] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[1577] 7. Saving inquiry history

[1578] The server stores all dialogue history in a database. This history can be used to respond to future inquiries, contributing to improving the response quality of the entire system. In addition, user emotional data is also saved as a log, enabling more accurate and personalized responses for future inquiries.

[1579] Specific examples

[1580] 1. The user accesses the web form, enters "Account Recovery" and submits it.

[1581] 2. The device captures this input and sends it to the server.

[1582] 3. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[1583] 4. The server retrieves the account recovery instructions from the knowledge base, and the generative AI generates a personalized answer.

[1584] 5. The server sends the answer to the user.

[1585] 6. At the same time, the server passes the user's input data to the emotion engine, which analyzes the user's emotions. The generative AI takes this analysis into account and adjusts the tone and content of the response.

[1586] 7. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[1587] 8. Finally, the server stores the dialogue history and emotion data for future use.

[1588] In this way, this invention makes it possible to improve the efficiency of customer support operations, realize 24-hour support, provide consistent responses, and reduce costs. In addition, the use of an emotion engine enables responses that are more in tune with emotions, contributing to an improved user experience.

[1589] The processing flow will be explained below.

[1590] Step 1:

[1591] A user visits a website's support page and opens a contact form. They select the type of inquiry they want, choosing a category such as "Account Recovery" or "Technical Support," then enter their inquiry details using text input or voice input and click the submit button.

[1592] Step 2:

[1593] The device captures the user's input data. In the case of text input, it stores the data as text. In the case of voice input, it uses a speech recognition API to convert the voice data into text.

[1594] Step 3:

[1595] The device sends the captured text data to the server.

[1596] Step 4:

[1597] The server passes the received text data to the generation AI, which analyzes the inquiry and extracts keywords and phrases. Based on the analysis results, the AI ​​identifies the topic and complexity of the problem and classifies the inquiry into a specific category, such as technical support, account issues, or payment issues.

[1598] Step 5:

[1599] The server sends a query to the knowledge base based on the classified query content, and retrieves related information from the knowledge base.

[1600] Step 6:

[1601] The server passes the acquired information to the generation AI, which then references past interaction history and user profile information to generate a personalized response.

[1602] Step 7:

[1603] The server then sends the generated response to the user, analyzing the user's emotions using an emotion engine and adjusting the tone and content accordingly. For example, if the user is feeling frustrated, the tone of the response will be more friendly.

[1604] Step 8:

[1605] The server uses the AI ​​to determine the urgency and complexity of the inquiry, and automatically executes escalation procedures as needed. For example, if the issue is technically difficult, the inquiry will be transferred to a specialist.

[1606] Step 9:

[1607] The server stores the conversation history in a database. All conversations are logged and used to handle future inquiries and improve the knowledge base. In addition, user sentiment data is also stored.

[1608] Step 10:

[1609] The server notifies the user of the progress of the escalation and completion of the automated response, allowing the user to see in real time how their inquiry is being handled.

[1610] Specific examples

[1611] 1. The user accesses the web form, enters "account recovery" as the inquiry, and clicks submit.

[1612] 2. The device captures this input data and sends it to the server.

[1613] 3. The server passes the content to the generation AI, which then classifies "Account Recovery" into the technical support category.

[1614] 4. The server retrieves the account recovery instructions from the knowledge base, and the generative AI generates a personalized answer.

[1615] 5. The server provides the generated answer to the user, with an emotion engine analyzing the user's emotions and adjusting the tone of the answer to be more friendly.

[1616] 6. If the user's problem is not resolved, the generation AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[1617] 7. Finally, the server stores the dialogue history and emotion data for use in responding to future inquiries.

[1618] Through this flow, the present invention improves the efficiency of customer support operations, enables 24-hour support, provides consistent responses, reduces costs, and improves user experience.

[1619] Example 2

[1620] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1621] In conventional customer support systems, multiple inquiry routes were managed separately, making it difficult to provide consistent responses. Furthermore, the lack of escalation procedures and emotionally sensitive responses made it difficult to improve the user experience. Furthermore, the lack of effective storage and utilization of inquiry history made it difficult to improve support quality.

[1622] The identification process 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 a means for integrating multiple inquiry routes and centrally accepting inquiries from users, a generation AI means for analyzing inquiries accepted from users and classifying the topic and complexity of the problem, a means for generating responses to inquiries from an information base and providing users with personalized answers, a means for passing inquiry data to a sentiment analysis engine and adjusting the content and tone of the response based on the analysis results, a means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry, and a means for saving the dialogue history and sentiment data of the inquiry and using them in future inquiry responses. This enables consistent and prompt responses, responses that are sensitive to the user's sentiment, prompt escalation, and effective use of the inquiry history.

[1623] "Multiple contact channels" are ways for users to contact you through different channels, such as help, chat, email, or phone.

[1624] "Centralized reception" refers to the integrated management of information received from multiple inquiry methods in a single system.

[1625] "Analysis" refers to the process by which the generative AI understands the content of the inquiry it receives, extracts keywords and phrases, and evaluates the topic and complexity of the problem.

[1626] "Generative AI" refers to a system that uses artificial intelligence technology to understand the content of an inquiry and provide appropriate answers or classifications.

[1627] An "information base" refers to a database that compiles past inquiry data and specialized knowledge, and is used to generate appropriate answers.

[1628] "Personalized answers" refer to responses that are individually tailored to the user, taking into account their profile and past interaction history.

[1629] An "emotion analysis engine" refers to a system that analyzes emotions from user inquiry text and voice data and provides the results to the generative AI.

[1630] "Escalation Procedures" refers to the process of automatically transferring inquiries to higher-level specialist staff based on the urgency or complexity of the inquiry.

[1631] "Dialogue history" refers to a record of past communication between the user and the system, and is used as reference material for responding to future inquiries.

[1632] "Emotion data" is data that indicates the emotional state of a user extracted by an emotion analysis engine.

[1633] This invention relates to a customer support system that integrates multiple inquiry routes and centrally accepts inquiries from users. This system uses generative AI and a sentiment analysis engine to analyze inquiries, generate automatic responses, execute escalation procedures, and provide users with personalized responses.

[1634] Hardware and software used

[1635] This system uses the following hardware and software.

[1636] Server: Plays a central role in accepting and processing user queries.

[1637] Terminal: Captures user input (text and voice) and sends it to the server.

[1638] Generative AI models: Analyze incoming queries and classify the topic and complexity of the problem.

[1639] Sentiment analysis engine: Analyzes user inquiry data and recognizes emotions.

[1640] Data processing and calculation

[1641] Each function of the system is described in detail below.

[1642] User Interface Settings

[1643] The server provides a user-accessible web form that integrates multiple inquiry channels, such as help, chat, email, and phone, and includes an interface through which the user can select the type of inquiry, through which the user can enter the required information and submit the inquiry.

[1644] Receiving and processing inquiries

[1645] The user's query entered into the web form is captured by the device. If the query is voice, it is converted to text using speech recognition technology. This input data is then sent to the server.

[1646] Inquiry analysis and classification

[1647] The server then passes the received text data to the generation AI, which analyzes the query and extracts keywords and phrases. Based on the analysis results, the AI ​​identifies the topic and complexity of the problem and classifies the query into a specific category.

[1648] Generate and send auto-replies

[1649] The server sends a query to an information base based on the classified query content to retrieve relevant information. The generation AI then references past interaction history and user profile information to generate a personalized answer, which is then sent to the user via the server.

[1650] Sentiment analysis and response tailoring

[1651] The server passes the user's inquiry data to the emotion analysis engine. The emotion engine analyzes the user's emotions from the text and voice data and provides the results to the generation AI. The generation AI takes this emotional data into account and adjusts the content and tone of the response appropriately.

[1652] Escalation Procedures

[1653] The AI ​​generator determines the urgency and complexity of the inquiry and automatically executes escalation procedures, transferring the inquiry to the necessary specialist staff, while the server notifies the user of the progress of the escalation.

[1654] Saving inquiry history

[1655] The server stores all dialogue history in a database. This history can be used to respond to future inquiries, contributing to improving the response quality of the entire system. In addition, user emotional data is also saved as a log, enabling more accurate and personalized responses for future inquiries.

[1656] Specific examples

[1657] 1. The user accesses the web form, enters "Account Recovery" and submits it.

[1658] 2. The device captures this input and sends it to the server.

[1659] 3. The server passes the content to the generation AI, which classifies "Account Recovery" into the technical support category.

[1660] 4. The server retrieves account recovery instructions from its information base, and the generative AI generates a personalized answer.

[1661] 5. The server sends the answer to the user.

[1662] 6. At the same time, the server passes the user's input data to the emotion analysis engine, which analyzes the user's emotions. The generative AI takes this analysis into account and adjusts the tone and content of the response.

[1663] 7. If the problem is not resolved, the generative AI automatically escalates the issue and the server transfers the inquiry to a specialist.

[1664] 8. Finally, the server stores the dialogue history and emotion data for future use.

[1665] In this way, this invention makes it possible to improve the efficiency of customer support operations, realize 24-hour support, provide consistent responses, and reduce costs. In addition, the use of a sentiment analysis engine enables responses that are more sensitive to emotions, contributing to an improved user experience.

[1666] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1667] Step 1:

[1668] A user accesses a web form, enters inquiry information, and presses the "Submit" button. At this time, the input is text-based information or voice data. Take the example of a user entering "Account recovery." Input: The user's inquiry. Output: The input inquiry data.

[1669] Step 2:

[1670] The device captures the user's input, and if the input is voice, converts it into text data using voice recognition technology. The converted text data is sent to the server. Input: User's raw voice or text. Output: Text data.

[1671] Step 3:

[1672] The server sends the received text data to the generation AI. The generation AI analyzes the received inquiry and extracts keywords and phrases. Input: Text data received by the server. Output: Keyword extraction and analysis results by the generation AI.

[1673] Step 4:

[1674] The server identifies the topic and complexity of the inquiry based on the analysis results from the generative AI and classifies it into an appropriate category. For example, "account recovery" is classified into the "technical support" category. Input: Analysis results from the generative AI. Output: Enquiry category classification.

[1675] Step 5:

[1676] The server queries the information base for the classified inquiry and retrieves relevant information. The data retrieved from the information base is sent to the generation AI, which generates a personalized response by referencing past interaction history and user profile information. Input: Inquiry category classification. Output: Personalized response.

[1677] Step 6:

[1678] The server sends the generated personalized response to the user, for example, "Hello, here are the steps to recover your account." Input: The generated personalized response. Output: The response sent to the user.

[1679] Step 7:

[1680] The server passes the inquiry data to the sentiment analysis engine, which identifies the user's sentiment and provides the results to the generation AI. For example, if the user enters "I'm in a hurry," the sentiment analysis engine recognizes a high level of urgency. Input: User inquiry data. Output: Sentiment analysis results.

[1681] Step 8:

[1682] The generation AI takes into account the results of the sentiment analysis and adjusts the content and tone of the response. For example, it generates a response that takes sentiment into account, such as "We will respond immediately." Input: Sentiment analysis results. Output: Adjusted response.

[1683] Step 9:

[1684] The server adjusts the content and tone of the response and then sends it to the user. Input: Adjusted response. Output: Adjusted response sent to the user.

[1685] Step 10:

[1686] The generation AI reassess the urgency and complexity of the inquiry and determines whether escalation is necessary. Urgent or highly difficult inquiries are automatically transferred to specialized staff. The server notifies the user of the progress of the escalation. Input: Reassessed urgency and complexity. Output: Transfer to specialized staff, notification to user.

[1687] Step 11:

[1688] The server stores all dialogue history and emotion data in a database. The stored data is used to respond to future inquiries and helps improve the response quality of the entire system. Input: dialogue history and emotion data. Output: stored data.

[1689] (Application example 2)

[1690] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1691] Conventional customer support systems have a lack of uniformity in responses to user inquiries, making it difficult to provide personalized responses. Furthermore, responses that do not take into account the user's feelings can lead to a decline in customer satisfaction. Furthermore, manual escalation procedures can lead to delays and errors.

[1692] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for integrating multiple inquiry routes and centrally accepting inquiries from users; generation AI means for analyzing inquiries accepted from users and classifying the topic and complexity of the problem; means for generating responses to inquiries from a knowledge base and providing users with personalized answers; emotion engine means for analyzing user emotions and adjusting the content and tone of the response; means for automatically executing escalation procedures depending on the urgency and complexity of the inquiry; and means for saving the inquiry dialogue history and using it for future inquiry responses. This enables more efficient customer support operations, 24-hour support, consistent responses, cost reduction, and improved customer satisfaction through responses that are sensitive to emotions.

[1693] "Enquiry channels" refer to the multiple ways or channels through which users can make inquiries.

[1694] "A means of centralized reception" refers to a function that allows inquiries sent from multiple inquiry routes to be received in one place.

[1695] "Generative AI means of analyzing and classifying" refers to the generative AI's ability to analyze the content of a query and classify it based on the topic and complexity of the problem.

[1696] A "knowledge base" refers to a database that stores and organizes answers to inquiries and information.

[1697] "Means for providing personalized answers" refers to the ability to generate and provide individually customized answers based on the user's inquiry and profile information.

[1698] "Emotion engine means" refers to a system that analyzes a user's emotions and adjusts the content and tone of the response based on the results.

[1699] "Means for implementing escalation procedures" refers to the ability to automatically route inquiries to specialized staff based on the urgency and complexity of the inquiry.

[1700] "Means for saving dialogue history and using it to respond to future inquiries" refers to a function that records the content and responses of past inquiries and uses them to respond to future inquiries.

[1701] A specific system for implementing the present invention has the following configuration.

[1702] First, the server integrates multiple inquiry channels and centrally accepts user inquiries. This includes web forms, chat boxes, and voice input. When a user makes an inquiry, the data is sent to the server. The server captures this data and passes it to a generative AI model for analysis. The analysis classifies the topic and complexity of the problem.

[1703] The server then references the knowledge base and uses a generative AI model to generate a personalized answer. This answer also takes into account the user's profile information and past interaction history. The emotion engine also analyzes the emotions from the user's input data and provides the results to the generative AI model. The generative AI model uses this emotional data to appropriately adjust the content and tone of the response.

[1704] If the inquiry is urgent or complex, the server automatically executes an escalation procedure and transfers the inquiry to a specialist staff member. The user is notified of the progress. In addition, the dialogue history and emotion data are recorded in a database and used to handle future inquiries.

[1705] Hardware and software used:

[1706] Server: A cloud server that provides an API endpoint (e.g., AWS, Google Cloud)

[1707] Generative AI models: text analysis and response generation (e.g., GPT-4)

[1708] Emotion engine: Sentiment analysis (e.g. IBM Watson, Microsoft Azure Emotion API)

[1709] Device: Capturing user input (e.g. smartphone, HMD)

[1710] Examples:

[1711] For example, if a user voice-inquires within a virtual store's smartphone app, "How do I return an item?", the voice data is sent to a server and converted into text using speech recognition technology. A generative AI model analyzes the inquiry, retrieves information about the "return procedure" from a knowledge base, and generates and sends a personalized response taking into account the user's profile information. Furthermore, an emotion engine analyzes the user's emotions (e.g., frustration), and the generative AI model adjusts the tone of the response taking these emotions into account.

[1712] Example prompt sentence:

[1713] I have an item I'd like to return, how do I go about doing that?

[1714] "I'm having issues with the quality of the product I purchased. What should I do?"

[1715] In this way, real-time responses that are sensitive to emotions become possible, contributing to an improved user experience.

[1716] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1717] Step 1:

[1718] The user makes an inquiry. The user selects an appropriate input method from multiple inquiry routes (e.g., web form, chat box, voice input) and types or speaks the inquiry content. The input data is converted into text using voice recognition technology.

[1719] Input: User text or voice data

[1720] Output: Query data converted to text

[1721] Step 2:

[1722] The device captures the user's input data and sends it to the server. If there is voice input, the device processes it to convert the voice into text.

[1723] Input: User's inquiry (text or voice data)

[1724] Output: Text data sent to the server

[1725] Step 3:

[1726] The server then passes the received text data to a generative AI model for analysis, which extracts keywords and phrases from the query and categorizes them based on the topic and complexity of the problem.

[1727] Input: Text data

[1728] Output: Parsed data (topic and complexity classification results)

[1729] Step 4:

[1730] The server uses the parsed data to consult a knowledge base to retrieve relevant information, and a generative AI model takes this information, along with the user's profile information, into account to generate a personalized answer.

[1731] Input: Analytics data, knowledge base information, user profile information

[1732] Output: Personalized answer

[1733] Step 5:

[1734] The server passes the user's text data to the emotion engine for emotion analysis. The emotion engine analyzes the user's emotions from the text data and provides the results to the generative AI model.

[1735] Input: User's text data

[1736] Output: Emotion analysis results

[1737] Step 6:

[1738] The generative AI model takes into account the results of sentiment analysis and adjusts the content and tone of the response accordingly, resulting in a personalized response that takes the user's emotions into account.

[1739] Input: Sentiment analysis results, personalized answers

[1740] Output: Emotionally sensitive personalized answers

[1741] Step 7:

[1742] The server then sends the generated answer to the user, and simultaneously automatically routes the inquiry to a specialist if escalation is required, and notifies the user of the progress.

[1743] Input: Emotionally sensitive personalized answers

[1744] Output: Response sent to user, escalation action

[1745] Step 8:

[1746] The server stores all dialogue history and emotion data in a database, which is used to respond to future inquiries and contribute to improving the response quality of the entire system.

[1747] Input: Dialogue history, emotion data

[1748] Output: Historical data stored in a database

[1749] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1750] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1751] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1752] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1753] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1754] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1755] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1756] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1757] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1758] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1759] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1760] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1761] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1762] 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.

[1763] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1764] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1765] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1766] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1767] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1768] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1769] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1770] The following is further disclosed regarding the above embodiment.

[1771] (Claim 1)

[1772] A means to integrate multiple inquiry routes and receive inquiries from users in a centralized manner,

[1773] A generative AI method for analyzing user inquiries and classifying the topic and complexity of the problem;

[1774] A means for generating responses to inquiries from a knowledge base to provide users with personalized answers;

[1775] A system that includes a means to automatically implement escalation procedures depending on the urgency and complexity of the inquiry.

[1776] (Claim 2)

[1777] 10. The system of claim 1, further comprising means for converting voice input from a user into text.

[1778] (Claim 3)

[1779] 10. The system of claim 1, further comprising means for storing a dialogue history of the inquiry for use in handling future inquiries.

[1780] "Example 1"

[1781] (Claim 1)

[1782] A means to integrate multiple inquiry routes and receive inquiries from users in a centralized manner,

[1783] A generative AI method for analyzing user inquiries and classifying the topic and complexity of the problem;

[1784] a means for generating responses to inquiries from a database and providing personalized answers to users;

[1785] A means of automatically implementing escalation procedures depending on the urgency and complexity of the inquiry;

[1786] A means for generating an answer by referencing the user's profile information and past interaction history;

[1787] means for parsing and classifying the received data as text;

[1788] A means of automatically sending responses to inquiries to users;

[1789] A system including:

[1790] (Claim 2)

[1791] 10. The system of claim 1, further comprising means for converting voice input from a user into text.

[1792] (Claim 3)

[1793] 10. The system of claim 1, further comprising means for storing a dialogue history of the inquiry for use in handling future inquiries.

[1794] "Application Example 1"

[1795] (Claim 1)

[1796] A means to integrate multiple inquiry routes and receive inquiries from users in a centralized manner,

[1797] A generative AI method for analyzing user inquiries and classifying the topic and complexity of the problem;

[1798] A means for generating responses to inquiries from a knowledge base to provide users with personalized answers;

[1799] A means of automatically implementing escalation procedures depending on the urgency and complexity of the inquiry;

[1800] means for providing an interface for users to make inquiries via smart devices;

[1801] a speech recognition means for converting a voice query into text;

[1802] A system that includes a means for generating prompts using a generative AI model to generate optimal responses for a user.

[1803] (Claim 2)

[1804] 10. The system of claim 1, further comprising means for converting voice input from a user into text.

[1805] (Claim 3)

[1806] 10. The system of claim 1, further comprising means for storing a dialogue history of the inquiry for use in handling future inquiries.

[1807] "Example 2: Combining Emotion Engines"

[1808] (Claim 1)

[1809] A means to integrate multiple inquiry routes and receive inquiries from users in a centralized manner,

[1810] A generative AI method for analyzing user inquiries and classifying the topic and complexity of the problem;

[1811] means for generating responses to inquiries from an information base and providing personalized answers to users;

[1812] A means of passing inquiry data to a sentiment analysis engine and adjusting the content and tone of responses based on the analysis results;

[1813] A means of automatically implementing escalation procedures depending on the urgency and complexity of the inquiry;

[1814] The system includes a means for storing the dialogue history and emotion data of inquiries and using them to respond to future inquiries.

[1815] (Claim 2)

[1816] 10. The system of claim 1, further comprising means for converting voice input from a user into text.

[1817] (Claim 3)

[1818] 10. The system of claim 1, further comprising means for notifying a user of the progress of the escalation.

[1819] "Application example 2 when combining emotion engines"

[1820] (Claim 1)

[1821] A means to integrate multiple inquiry routes and receive inquiries from users in a centralized manner,

[1822] A generative AI method for analyzing user inquiries and classifying the topic and complexity of the problem;

[1823] A means for generating responses to inquiries from a knowledge base to provide users with personalized answers;

[1824] an emotion engine means for analyzing a user's emotion and adjusting the content and tone of a response;

[1825] A means of automatically implementing escalation procedures depending on the urgency and complexity of the inquiry;

[1826] A means for storing the history of inquiries and using it to respond to future inquiries;

[1827] A system including:

[1828] (Claim 2)

[1829] 10. The system of claim 1, further comprising means for converting voice input from a user into text.

[1830] (Claim 3)

[1831] 10. The system of claim 1, further comprising means for using the user's emotional data to adjust the content and tone of a response generated based on the query content. [Explanation of symbols]

[1832] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means to integrate multiple inquiry routes and receive inquiries from users in a centralized manner, A generative AI method for analyzing user inquiries and classifying the topic and complexity of the problem; A means for generating responses to inquiries from a knowledge base to provide users with personalized answers; A system that includes a means to automatically implement escalation procedures depending on the urgency and complexity of the inquiry.

2. The system of claim 1 further comprising means for converting voice input from a user into text.

3. The system of claim 1 further comprising means for storing a dialogue history of the inquiry for use in handling future inquiries.

Images