system

Abstract

A system is provided.SOLUTION: A system comprising: means for pre-authoring a FAQ database and troubleshooting guides; means for training a generative AI model to generate appropriate answers to user questions; means for providing a user interface for a user to input questions; means for passing user-input questions to the generative AI model to generate answers; and means for returning the generated answers to the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Conventional customer support systems often leave customers waiting for long periods of time, preventing them from receiving prompt and appropriate support. Furthermore, conventional systems require high training and education costs, and are limited in their scope and quality. The present invention aims to utilize generative AI to provide prompt and appropriate answers to customer questions and problems, thereby improving the efficiency of customer support and customer satisfaction. [Means for solving the problem]

[0005] The present invention is a system that includes a means for creating an FAQ database and a troubleshooting guide in advance, a means for training a generative AI model to generate appropriate answers to user questions, a means for providing a user interface for the user to input questions, a means for passing the questions input by the user to the generative AI model to generate answers, and a means for returning the generated answers to the user.The system also includes a means for collecting feedback from the user and adjusting and improving the generative AI model.Furthermore, the generative AI model uses the FAQ database and the troubleshooting guide as a training data set.

[0006] The "FAQ database" is a database that organizes and stores frequently asked questions and their answers.

[0007] A "Troubleshooting Guide" is a guide that systematically summarizes problems that users may encounter and their solutions.

[0008] A "generative AI model" is an artificial intelligence model that uses natural language generation technology to generate appropriate responses to user input.

[0009] A "user interface" is a screen or application that allows a user to enter questions or inquiries into a system.

[0010] A "feedback collection method" is a mechanism for obtaining ratings and comments from users and using them to improve the system and AI model.

[0011] A "training dataset" is a collection of data used to train a generative AI model.

[0012] "Appropriate answer generation" refers to the process of using a generative AI model to create accurate and useful answers to users' questions.

[0013] A "support system" is a system that has a set of functions and means for accepting user inquiries and providing solutions. [Brief explanation of the drawings]

[0014] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14]FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0015] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

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

[0017] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0018] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0019] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0020] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0022] [First embodiment]

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

[0024] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0025] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0026] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0027] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0028] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0029] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

[0031] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0032] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0033] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0034] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0035] This invention is a customer support system that utilizes a generative AI model. This system creates an FAQ database and a troubleshooting guide in advance, and then trains the generative AI model based on them to generate appropriate answers to user questions.

[0036] First, the server creates a customer support FAQ database and troubleshooting guide, which are built based on past inquiry data and expert knowledge, covering frequently asked questions and solutions to problems.

[0037] The server then uses this FAQ database and troubleshooting guide to train a generative AI model designed to generate appropriate responses to natural language questions from users. The model uses automated learning algorithms to learn patterns from the dataset and generate optimal answers based on that knowledge.

[0038] When a user uses a device to input a question, the device sends the question to a server, which passes the question to a generative AI model, which then generates the optimal response to the question. The generated response is then sent back to the device via the server and displayed to the user.

[0039] For example, if a user asks, "How do I return an item?", the device sends this question to the server. The server passes this question to a generative AI model, which then generates a response such as, "To find out how to return an item, select the item you want to return from your order history and follow the return procedure from there. For detailed instructions, please see the link below." The server then sends this response back to the device, allowing the user to view the answer on the screen.

[0040] Furthermore, the server has the ability to collect user feedback and adjust and improve the generative AI model based on it. Based on the feedback provided by the user, the model's response accuracy can be periodically improved. When a user gives feedback such as "This answer was very helpful," the data is stored on the server and used for the next model training.

[0041] In this way, the customer support system of the present invention realizes quick and appropriate customer responses and improves customer satisfaction. Furthermore, by improving the response accuracy of the AI, it is possible to reduce training and education costs and provide efficient customer support.

[0042] The processing flow will be explained below.

[0043] Step 1:

[0044] The server creates a FAQ database and troubleshooting guide based on experts and past support data, organizing and storing frequently asked questions and their solutions.

[0045] Step 2:

[0046] The server trains a generative AI model based on the FAQ database and troubleshooting guide it creates. The model learns from these datasets to generate appropriate responses to natural language input from users.

[0047] Step 3:

[0048] The server provides a user interface (UI) for users to enter questions, which can be implemented as a web page or a mobile application.

[0049] Step 4:

[0050] A user accesses the customer support system through the UI and inputs a question, such as "How do I return a product?"

[0051] Step 5:

[0052] The device sends the user-entered question to the server as an HTTP request, which is sent in an appropriate format (e.g., JSON).

[0053] Step 6:

[0054] The server inputs the received question into a generative AI model, which analyzes the question and generates the best answer based on pre-trained data.

[0055] Step 7:

[0056] The server takes the generated answer and formats it in a way that is easy for the user to understand, for example as text with links and additional information.

[0057] Step 8:

[0058] The server returns a formatted response to the terminal, which then displays the response on the user's terminal.

[0059] Step 9:

[0060] The user checks the answer displayed on the terminal, and if the question is not resolved, the user can enter the question again.

[0061] Step 10:

[0062] Users provide feedback on the answers they receive from the system, such as rating the answer as "this answer was helpful" or "this answer was insufficient."

[0063] Step 11:

[0064] The server stores the collected feedback from users in a database, which is used for the next model training.

[0065] Step 12:

[0066] The server retrains the generative AI model with newly collected feedback data to improve the accuracy and coverage of answers, thus maintaining a cycle of system improvement.

[0067] Each step in the embodiment of the present invention is specifically performed in this manner to provide prompt and appropriate customer support to users.

[0068] Example 1

[0069] 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."

[0070] Existing customer support systems face challenges in providing prompt and appropriate responses to user inquiries. Furthermore, they lack a means to continuously improve the system based on user feedback, making it difficult to improve response accuracy. Furthermore, they are unable to provide consistent answers to the diverse problems users face, potentially leading to a decline in customer satisfaction.

[0071] 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.

[0072] In this invention, the server includes a means for creating an FAQ database and a troubleshooting guide in advance, a means for training a generative AI model to generate appropriate answers to user questions, a means for providing a user interface for users to input questions, and a means for collecting feedback and adjusting and improving the generative AI model based on the feedback, thereby enabling prompt and appropriate customer support, improving customer satisfaction, and achieving continuous improvement of the system.

[0073] An "FAQ database" is a database that is built based on past inquiry data and expert knowledge and centrally manages answers to frequently asked inquiries.

[0074] A "troubleshooting guide" is a systematic guide that describes in detail how to solve problems and issues that users may encounter.

[0075] A "generative AI model" is an artificial intelligence model trained to generate appropriate answers to user questions using natural language processing techniques.

[0076] A "user interface" is an interactive screen or operating means through which a user accesses a system and inputs questions.

[0077] "Feedback" refers to the evaluations and opinions that users provide regarding the system's responses, and this information is used to improve the system.

[0078] A "training dataset" is a collection of data used to train a generative AI model, collected from FAQ databases and troubleshooting guides.

[0079] A "terminal" is a device, such as a computer or smartphone, that a user uses to access the system.

[0080] The "server" is a central processing unit that receives questions from users and passes them to a generative AI model to generate answers.

[0081] This invention relates to a customer support system that utilizes a generative AI model. This system creates an FAQ database and a troubleshooting guide in advance, and then trains a generative AI model based on them to generate appropriate answers to user questions.

[0082] First, the server uses software and data management tools to create an FAQ database and troubleshooting guide for customer support. Specific hardware includes a database server and a high-performance computer, and software includes a database management system such as MySQL (registered trademark) or PostgreSQL. This database and guide are built based on past inquiry data and expert knowledge, and cover frequently asked questions and solutions to problems.

[0083] The server then uses the FAQ database and troubleshooting guide to train a generative AI model. The generative AI model is created using libraries such as TENSORFLOW (registered trademark) and PyTorch. It is designed to use an automatic learning algorithm to generate appropriate responses to natural language questions from users. Specifically, question and answer pairs are extracted from the FAQ database and troubleshooting guide and provided to the generative AI model as a training dataset.

[0084] When a user inputs a question using a device, the question is sent from the device to a server. The device is assumed to be a PC or smartphone, and a dedicated user interface is provided for it. For example, if a user asks, "How do I return a product?", the question is sent from the device to the server. The server passes the received question to a generative AI model, and the model generates an optimal response. The generated response is then sent back to the device via the server and displayed to the user.

[0085] An example of a generated response might be, "To find out how to return an item, please select the item you wish to return from your order history and follow the return procedure from there. For detailed instructions, please see the link below." This response responds quickly and accurately to the information the user is seeking.

[0086] Furthermore, the server has the ability to collect user feedback and adjust and improve the generative AI model based on that feedback. When a user provides feedback such as "This answer was very helpful," that data is stored on the server and used for the next model training. Collecting this feedback and retraining the model improves the response accuracy of the generative AI model and improves the performance of the system as a whole.

[0087] Examples of prompts include:

[0088] Please tell me how to return the product

[0089] "I haven't received the product I ordered, what should I do?"

[0090] Please tell me how to change my payment method.

[0091] The system aims to provide efficient and consistent customer support and improve customer satisfaction. In addition, the improved accuracy of AI responses will reduce training and education costs, resulting in more efficient support.

[0092] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0093] Step 1: Creating a FAQ database and troubleshooting guide

[0094] The server collects past inquiry data and expert knowledge, and uses them to create an FAQ database and troubleshooting guide. Specifically, it uses a database management system (e.g., MySQL, PostgreSQL) to categorize, organize, and store the collected data.

[0095] Input: past inquiry data, expert knowledge

[0096] Output: FAQ database, troubleshooting guide

[0097] Step 2: Training the generative AI model

[0098] The server trains the generative AI model using data from the FAQ database and troubleshooting guides. Specifically, it prepares a training dataset and builds and trains the model using libraries such as TensorFlow and PyTorch.

[0099] Input: FAQ database, troubleshooting guide

[0100] Output: A trained generative AI model

[0101] Step 3: Accepting user questions

[0102] A user inputs a question using a terminal, which then sends the question to a server. Specifically, a question is input in natural language through a user interface and then sent to the server via a network.

[0103] Input: User question (natural language)

[0104] Output: Question data (received on the server side)

[0105] Step 4: Question processing and response generation

[0106] The server passes the received question to a generative AI model, which then generates an appropriate response to the question. Specifically, the question data is input into the AI ​​model, and the model generates the optimal response based on the training data.

[0107] Input: Question data

[0108] Output: The generated response

[0109] Step 5: Returning a response

[0110] The server sends the generated response back to the terminal, which then displays the received response to the user. Specifically, the response data is sent to the terminal via the network, and the user can check the response on the screen.

[0111] Input: Generated response

[0112] Output: The response that is displayed to the user

[0113] Step 6: Gather feedback

[0114] The user inputs the feedback they provide on the terminal, and the terminal transmits the feedback to the server. Specifically, the user inputs their thoughts and evaluations on the response, which are then stored on the server.

[0115] Input: User feedback

[0116] Output: Feedback data stored on the server

[0117] Step 7: Adjust and improve the model

[0118] The server periodically retrains the generative AI model based on the collected feedback to improve response accuracy. Specifically, it adds the feedback data to the training data and retrains the generative AI model.

[0119] Input: Feedback data

[0120] Output: An improved generative AI model

[0121] (Application example 1)

[0122] 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."

[0123] Conventional customer support systems often struggle to respond to user inquiries quickly and accurately. They also lack the functionality to efficiently collect user feedback and improve the accuracy of their models. As a result, problems such as low user satisfaction and inefficient support operations arise.

[0124] 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.

[0125] In this invention, the server includes means for creating an FAQ database and a troubleshooting guide in advance, means for training a generative AI model to generate appropriate answers to user questions, means for providing a user interface for users to input questions, means for passing questions input by users to the generative AI model to generate answers, means for returning the generated answers to the users, and means for collecting user feedback and improving the accuracy of the generative AI model, thereby enabling faster responses to user inquiries, providing accurate responses, and improving the model based on the feedback.

[0126] An "FAQ database" is an information resource that systematically organizes frequently asked questions and their answers.

[0127] A "troubleshooting guide" is a guide that specifically describes problems that users often encounter and how to solve them.

[0128] A "generative AI model" is an artificial intelligence model that generates appropriate answers to questions posed in natural language by users.

[0129] "User interface" refers to the screen and operating methods that users use to input information into a system.

[0130] "Feedback" refers to opinions and ratings provided by users about the quality and usability of an answer.

[0131] The present invention is a customer support system that uses a FAQ database and troubleshooting guides to respond to user questions. This system is intended for use particularly in online shopping sites, and generates real-time responses to user questions via smartphones.

[0132] The server first creates a FAQ database and troubleshooting guide, then uses them to train a generative AI model. The generative AI model uses an automated learning algorithm, such as OpenAI's GPT-4, to learn patterns from the dataset. It also collects user feedback and uses that feedback to improve the model's accuracy.

[0133] When a user inputs a question from their smartphone, the device sends the question to a server. The server passes the received question to a generative AI model, which generates an appropriate answer. The generated answer is then sent back to the smartphone via the server and displayed to the user. This allows the user to receive a quick and accurate answer.

[0134] As a specific use case, if a user asks, "How do I return an item?", the device sends this question to the server. The server passes this question to the generative AI model, which then generates a response such as, "Select the item you want to return from your order history and complete the return procedure. Click here for details." The server then sends this response back to the device, allowing the user to view the answer on their smartphone screen.

[0135] User feedback is collected in the form of, for example, "This answer was very helpful." This feedback is then stored on the server and used for the next training. This allows the generative AI model to regularly improve its response accuracy, enabling the provision of higher quality customer support.

[0136] The specific hardware used includes smartphones as user devices, high-performance computers as servers, and GPT-4 as a generative AI model, while the software used includes various APIs, cloud services, and database management systems.

[0137] Example prompt sentence:

[0138] When a user asks, "How do I return an item?", respond with: "Select the item you want to return from your order history and follow the return process. Click this link for more information."

[0139] As a result, this customer support system can provide efficient and satisfying service and significantly reduce the support work load of companies.

[0140] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0141] Step 1:

[0142] The user inputs a question through the smartphone's user interface, for example, "How do I return a product?"

[0143] Input: User question

[0144] Output: Data of the questions entered on the smartphone

[0145] Step 2:

[0146] The device converts the questions entered by the user into data format and sends it to the server. This data is structured in JSON format or similar.

[0147] Input: Questions typed into the user's smartphone

[0148] Output: Structured question data (e.g., JSON format)

[0149] Step 3:

[0150] The server passes the received question to the generative AI model. Specifically, it converts the question data into a format that the generative AI model can process and inputs it into the model.

[0151] Input: Structured question data

[0152] Output: Input data prepared for a generative AI model

[0153] Step 4:

[0154] A generative AI model (e.g., GPT-4) analyzes the provided question data and generates the optimal answer by referencing a FAQ database and troubleshooting guide.

[0155] Input: Input data for the generative AI model

[0156] Output: The generated answer

[0157] Step 5:

[0158] The server receives the answers from the generative AI model and converts them into a data format to send back to the user.

[0159] Input: Answer from generative AI model

[0160] Output: Structured response data

[0161] Step 6:

[0162] The terminal receives the answer data from the server and displays it on the screen, allowing the user to check the answer to the question.

[0163] Input: Response data from the server

[0164] Output: Answer displayed on smartphone screen

[0165] Step 7:

[0166] The user can input feedback about the displayed answer and send it to the server via the terminal. For example, the user can give feedback such as "This answer was very helpful."

[0167] Input: User feedback

[0168] Output: Feedback data sent from the device to the server

[0169] Step 8:

[0170] The server collects and stores user feedback to add to the training dataset of the generative AI model, which will improve the model's accuracy the next time it is trained.

[0171] Input: Feedback data from the device

[0172] Output: Feedback added to the training dataset

[0173] 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.

[0174] This invention is a customer support system that utilizes a generative AI model and further combines it with an emotion engine that recognizes user emotions. This system creates a FAQ database and troubleshooting guide in advance, and then trains the generative AI model based on these to generate appropriate answers to user questions. The emotion engine also analyzes user emotions and adjusts the generative AI model's responses to provide more appropriate customer support.

[0175] First, the server creates a customer support FAQ database and troubleshooting guide, which are built based on past inquiry data and expert knowledge, covering frequently asked questions and solutions to problems.

[0176] The server then uses this FAQ database and troubleshooting guide to train a generative AI model, which learns from these datasets to generate appropriate responses to natural language input from users.

[0177] Furthermore, the server provides a user interface (UI) for users to input their questions, which can be implemented as a web page or a mobile application and easily accessible to users.

[0178] The new system incorporates an emotion engine, which can analyze user emotions and apply responses. When a user uses a device to input a question, the device sends the question to a server. The server then passes the received question to the emotion engine, which analyzes the user's emotions. The analyzed emotion data is then passed to a generative AI model, which then generates an appropriate response based on that data.

[0179] For example, if a user asks, "How do I return an item?", the device sends this question to the server. At the same time, the emotion engine analyzes the tone and context of the user's speech and, if it determines that the user is in a hurry, the generative AI model generates a quick and specific response such as, "The return process is simple. Please follow the steps below immediately."

[0180] The server then formats the generated answer and sends it back to the user, who can review the answer displayed on their device and provide feedback if needed, such as "This answer was very helpful." This feedback is stored on the server and used to retrain the generative AI model and emotion engine.

[0181] In this way, the customer support system of the present invention combines a generative AI model and an emotion engine to provide fast and appropriate customer service and improve customer satisfaction. Furthermore, by recognizing the user's emotions and adjusting responses accordingly, it is possible to provide more personalized support. The above is a specific embodiment for implementing the present invention.

[0182] The processing flow will be explained below.

[0183] Step 1:

[0184] The server uses experts and past support data to create a FAQ database and troubleshooting guides that organize and store frequently asked questions and their solutions.

[0185] Step 2:

[0186] The server uses the created FAQ database and troubleshooting guide to train a generative AI model, which is then able to learn patterns from this data and generate appropriate responses to users' natural language input.

[0187] Step 3:

[0188] The server provides a user interface (UI) for users to enter their questions, implemented as a web page or mobile application, making it easy for users to access and use.

[0189] Step 4:

[0190] The user uses a terminal to access the provided UI and input a question into the customer support system, for example, "How do I return a product?"

[0191] Step 5:

[0192] The device sends the user-entered question to the server as an HTTP request, which is sent in an appropriate format (e.g., JSON).

[0193] Step 6:

[0194] The server passes the received question to the emotion engine, which analyzes the emotion from the user's input text and passes the analysis results to the generative AI model.

[0195] Step 7:

[0196] The emotion engine analyzes emotional data from user input, for example classifying emotions from user text as "positive," "negative," or "neutral." This data is then used to adjust the generative AI model's response.

[0197] Step 8:

[0198] The server inputs the emotion data obtained from the emotion engine into the generative AI model, which then references this data to generate a response that corresponds to the user's emotion.

[0199] Step 9:

[0200] The server takes the generated answer and formats it in a format that is easy for the user to understand, for example as text with links and additional information.

[0201] Step 10:

[0202] The server returns a formatted response to the terminal, which then displays the response on the user's terminal.

[0203] Step 11:

[0204] The user checks the answer displayed on the terminal, and if the answer is not what they expected, they can enter the question again.

[0205] Step 12:

[0206] Users provide feedback on the answers they receive from the system, such as rating the answer as "this answer was helpful" or "this answer was insufficient."

[0207] Step 13:

[0208] The server stores the collected feedback from users in a database, which is used for the next retraining of the model and emotion engine.

[0209] Step 14:

[0210] The server uses newly collected feedback data to retrain the generative AI model and emotion engine to improve the accuracy and coverage of answers, maintaining a cycle of system improvement.

[0211] Through this series of steps, the present invention combines a generative AI model and an emotion engine to provide fast and appropriate customer support, thereby increasing user satisfaction and realizing efficient support.

[0212] Example 2

[0213] 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."

[0214] Conventional customer support systems have difficulty in providing appropriate answers to user questions quickly, and are unable to generate responses that take into account the user's emotions. Therefore, in order to improve user satisfaction, it is necessary to provide personalized support that takes into account the user's emotions.

[0215] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0216] In this invention, the server includes means for creating an FAQ database and a troubleshooting guide, means for training a generative AI model to generate appropriate answers to user questions, means for providing a user interface for users to input questions, means for passing the questions input by the users to the generative AI model and a sentiment analysis engine to generate answers, and means for returning the generated answers to the users, thereby enabling the provision of prompt and appropriate answers that take into consideration the user's emotions.

[0217] An "FAQ database" is a collection of information that organizes and stores answers to frequently asked questions from users.

[0218] A "troubleshooting guide" is a manual that systematically describes solutions to problems and issues that users may encounter.

[0219] A "generative AI model" is an artificial intelligence model trained to generate appropriate responses to natural language input from a user.

[0220] A "user interface" is an interactive screen or device that allows a user to enter questions into a system and receive answers from the system.

[0221] An "emotion analysis engine" is a technology that analyzes language data entered by a user and estimates the user's emotional state.

[0222] "Feedback" refers to a user inputting an evaluation or opinion on the answer received from the system.

[0223] "Server" means the computer system that runs the entire customer support system, storing and processing data, training, generating responses, etc.

[0224] This invention is a customer support system that utilizes a generative AI model and further combines it with a sentiment analysis engine that recognizes user emotions. This system creates a FAQ database and troubleshooting guide in advance, and then trains the generative AI model based on these to generate appropriate answers to user questions. The sentiment analysis engine also analyzes user emotions and adjusts the generative AI model's responses to provide more appropriate customer support.

[0225] Hardware and software used

[0226] Server: A platform that creates FAQ databases and troubleshooting guides for customer support, and runs generative AI models and sentiment analysis engines.

[0227] Device: The device (e.g., computer, smartphone) used by the user to enter questions and exchange information with the server.

[0228] Generative AI model: An AI model that generates optimal answers to user questions. Examples include OpenAI's GPT-3 (registered trademark) and GPT-4.

[0229] Sentiment analysis engine: An engine for analyzing user emotions and adjusting the generated response. Example: IBM's Watson® Tone Analyzer.

[0230] Specific examples

[0231] 1. Creating a FAQ database and troubleshooting guide

[0232] The server creates a FAQ database and troubleshooting guide for customer support. This database is built based on past inquiry data and expert knowledge. For example, a question about "how to return a product" may include specific instructions such as "bring proof of purchase and go to the store to complete the procedure."

[0233] 2. Training the generative AI model

[0234] The server uses the FAQ database and troubleshooting guide to train the generative AI model, for example, to respond appropriately to inquiries containing the keyword "returns."

[0235] 3. Providing a User Interface (UI)

[0236] The server provides a UI for users to enter their questions. This UI can be implemented as a web page or a mobile application. Once users log in, it provides a chat box and a search bar to make it easy for them to enter their questions.

[0237] 4. User Question Processing

[0238] The user uses the terminal to input a specific question, for example, "How do I return a product?", and the terminal transmits the user's question to the server in real time.

[0239] 5. Emotion Analysis

[0240] The server passes the received question to a sentiment analysis engine, which analyzes the content and tone of the question. The sentiment analysis engine infers the user's emotions, such as whether they are angry, confused, or in a hurry, from the context and wording. For example, if the question contains the phrase "I want to return it quickly," the sentiment analysis engine will determine that the user is in a hurry.

[0241] 6. Response generation using generative AI models

[0242] The generative AI model generates appropriate responses based on the emotional data provided by the sentiment analysis engine. For example, if it determines that the user is in a hurry, it generates a quick and specific response such as, "The return process is simple. Please follow the steps below immediately."

[0243] 7. Returning Responses

[0244] The server formats the generated response and sends it back to the user. The response created by the generative AI model is formatted into HTML or JSON format and sent to the terminal as an HTTP response.

[0245] 8. User Acknowledgment and Feedback

[0246] The user checks the answer displayed on the device. If the answer is helpful, they can enter feedback such as "This answer was very helpful." A feedback form is provided on the device so that users can easily submit their opinions.

[0247] 9. Save feedback and retrain

[0248] The server stores user feedback and uses it to retrain the generative AI model and sentiment analysis engine. The feedback data is stored in a database and used for the next training run, allowing the system to become more sophisticated over time and better meet user expectations.

[0249] Prompt Sentence Examples

[0250] Please tell me about the return procedure.

[0251] "My item hasn't arrived, what should I do?"

[0252] Please tell me about the product warranty period.

[0253] This allows the customer support system to respond quickly and appropriately to user questions, provide personalized service that takes user emotions into consideration, and increase customer satisfaction through continuous improvement through feedback.

[0254] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0255] Step 1:

[0256] The server creates an FAQ database and troubleshooting guide. The server collects past inquiry data and expert knowledge, and stores the data constructed based on that in the database. For example, if there are many inquiries about how to return a product, the server organizes and stores the specific procedures for those inquiries in the database.

[0257] Input: Historical inquiry data, expert knowledge

[0258] Output: FAQ database, troubleshooting guide

[0259] Step 2:

[0260] The server trains the generative AI model using the FAQ database and troubleshooting guide. The server collects information from both databases and trains the generative AI model to improve its ability to generate appropriate answers to user questions. For example, the server repeatedly trains a generative AI model (e.g., GPT-4) on data related to the keyword "returns."

[0261] Input: FAQ database, troubleshooting guide

[0262] Output: A trained generative AI model

[0263] Step 3:

[0264] The server provides a user interface (UI) for users to enter questions. This UI is implemented as a web page or mobile application, and allows users to easily enter questions. Specifically, it includes a chat box and a search bar.

[0265] Input: Web page and mobile application design information

[0266] Output: The implemented user interface

[0267] Step 4:

[0268] The user uses the terminal to enter a specific question, for example, "How do I return an item?" This question is sent to the server through the UI.

[0269] Input: User question

[0270] Output: Question data sent to the server

[0271] Step 5:

[0272] The server receives the question from the device and passes it to the emotion analysis engine. The emotion analysis engine analyzes the tone and context of the question to estimate the user's emotional state. For example, a question containing the phrase "I want to return it quickly" indicates a sense of urgency.

[0273] Input: User question data

[0274] Output: Emotion data

[0275] Step 6:

[0276] The generative AI model receives the emotion data and generates an appropriate response based on it. For example, if it determines that the user is in a hurry, it will generate a quick and specific response such as, "The return process is easy. Please follow the steps below immediately."

[0277] Input: Emotion data

[0278] Output: Response data

[0279] Step 7:

[0280] The server formats the generated response data and returns it to the user device. For example, it formats the response created by the generative AI model into HTML or JSON format and sends it to the device as an HTTP response.

[0281] Input: Response data

[0282] Output: Response sent back to the user's terminal

[0283] Step 8:

[0284] The user checks the answer displayed on the terminal. If the problem is solved and the answer is helpful, the user enters feedback such as "This answer was very helpful." This feedback is then sent back to the server.

[0285] Input: User feedback

[0286] Output: Feedback data sent to the server

[0287] Step 9:

[0288] The server stores the feedback data and uses it to retrain the generative AI model and sentiment analysis engine. The feedback data is stored in a database and used for the next training session.

[0289] Input: Feedback data

[0290] Output: Improved generative AI models and sentiment analysis engines

[0291] These steps enable the customer support system to respond quickly and appropriately to user questions, provide personalized service that takes user sentiment into account, and increase customer satisfaction through continuous improvement through feedback.

[0292] (Application example 2)

[0293] 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."

[0294] Conventional customer support systems are required to respond to user questions quickly and appropriately, but they lack the technology to respond in a way that takes the user's emotions into consideration. This increases the likelihood of users feeling stressed or dissatisfied. In addition, there are limited ways to improve the system based on feedback, making it difficult to improve the quality of support.

[0295] The identification processing by the identification 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 creating an FAQ database and a troubleshooting guide in advance, means for training a generative AI model to generate appropriate answers to user questions, means for providing a user interface for the user to input questions, means for passing questions input by the user to the generative AI model to generate answers, means for analyzing the user's emotions using a sentiment analysis engine, means for adjusting the response of the generative AI model based on the results of the sentiment analysis, and means for returning the generated answers to the user. This makes it possible to provide quick and appropriate answers to user questions and to provide personalized responses that take the user's emotions into consideration.

[0296] The "FAQ Database" is a database of frequently asked questions and their answers, built based on past inquiry data and troubleshooting guides.

[0297] A "Troubleshooting Guide" is a collection of procedures or guidelines that provide solutions to specific problems or malfunctions.

[0298] A "generative AI model" is a model trained using machine learning techniques to generate appropriate responses to a user's natural language input.

[0299] A "user interface" is an interface that allows users to access the system and input questions, and is provided as a web page or mobile application.

[0300] An "emotion analysis engine" is a technology that analyzes and identifies emotions from user comments and input content.

[0301] A "feedback system" is a mechanism for collecting evaluations and opinions from users and using them to improve and adjust the system.

[0302] The following procedures and system configuration are required to realize the invention.

[0303] System Configuration

[0304] The system of this invention mainly consists of a server, a terminal, and a user. The server has an FAQ database and a troubleshooting guide, and includes a generative AI model and a sentiment analysis engine. The terminal provides a user interface for users to input questions.

[0305] Program processing

[0306] Creating a FAQ database and troubleshooting guide

[0307] The server creates a FAQ database and troubleshooting guide based on past inquiry data and expert knowledge, covering frequently asked questions and solutions to problems.

[0308] Training generative AI models

[0309] The server uses the FAQ database and troubleshooting guide to train a generative AI model, which is then used to generate appropriate responses to natural language input from users.

[0310] Providing a user interface

[0311] The server provides a user interface implemented as a web page or mobile application, allowing users to easily input questions via their devices.

[0312] Parsing questions and generating answers

[0313] When a user inputs a question using a terminal, the question is sent to the server. The server passes the question to a sentiment analysis engine, which analyzes the user's sentiment. For example, if a user asks, "How do I return a product?", the sentiment analysis engine analyzes the tone and context of the user's speech and determines that the user is in a hurry.

[0314] Regulating responses based on emotions

[0315] The analyzed emotion data is passed to a generative AI model, which then generates an appropriate response based on that data. For rushed users, a quick and specific response such as "Returns are easy, just follow the steps below" is generated.

[0316] Returning answers and collecting feedback

[0317] The server formats the generated answer and sends it back to the user via the device. The user can review the answer displayed on the device and provide feedback if necessary. For example, they could say, "This answer was very helpful." This feedback is stored on the server and used to retrain the generative AI model and sentiment analysis engine.

[0318] Hardware and software used

[0319] The system uses the following hardware and software:

[0320] Server hardware: FAQ database, troubleshooting guide, generative AI model, sentiment analysis engine

[0321] Natural Language Toolkit (NLTK): A natural language processing library

[0322] TensorFlow / Keras: Training and inference of machine learning models

[0323] Emotion Recognition API: External service for emotion analysis

[0324] Flask: Building a server-side API

[0325] React Native: Frontend for smartphone applications

[0326] Specific examples

[0327] When a user asks "How do I return an item?" the system behaves as follows:

[0328] Example prompt sentence:

[0329] User Question: How do I return an item? Sentiment: I'm in a hurry

[0330] Based on this prompt, the generative AI model responds with, "The return process is simple. Please follow the steps below immediately." This allows for a fast, appropriate, and personalized response to the user's question.

[0331] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0332] Step 1:

[0333] A user opens a customer support application using a terminal and enters a question. The entered question is sent to the server by the terminal. The input of this step is the user's question text, and the output is the transmission of the question data to the server.

[0334] Step 2:

[0335] The server sends the received question to a sentiment analysis engine to analyze the user's sentiment. The sentiment analysis engine uses natural language processing technology (such as NLTK) to analyze the tone and context of the speech and identify the user's sentiment (e.g., hurry, irritated, calm). The input of this step is the question data, and the output is the user's sentiment data.

[0336] Step 3:

[0337] The emotion data analyzed by the emotion analysis engine is passed to the generative AI model by the server. The generative AI model generates the optimal response to the user's question based on the training dataset (FAQ database and troubleshooting guide). The input of this step is the question data and emotion data, and the output is the provisional response data.

[0338] Step 4:

[0339] The tentative responses generated by the generative AI model are adjusted based on emotional data to be formatted to suit the user's emotions. For example, if the user is determined to be in a hurry, the response will be adjusted to be brief and quick. The inputs of this step are tentative response data and emotional data, and the output is the final response data.

[0340] Step 5:

[0341] The final response data is sent back from the server to the terminal and displayed to the user. The user checks the displayed answer and provides feedback if necessary. The input of this step is the final response data, and the output is the response displayed on the user terminal and user feedback.

[0342] Step 6:

[0343] User feedback is sent from the device to the server, which collects and stores it. The collected feedback is used to retrain the generative AI model and sentiment analysis engine. The input of this step is user feedback, and the output is the storage of feedback data and model improvements.

[0344] These processing steps enable the system to generate fast, relevant answers to user questions, personalize responses by taking into account user sentiment, and incorporate feedback to continuously improve the overall system quality and accuracy.

[0345] 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.

[0346] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0347] 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.

[0348] [Second embodiment]

[0349] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0350] 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.

[0351] 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).

[0352] 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.

[0353] 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.

[0354] 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).

[0355] 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.

[0356] 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.

[0357] 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.

[0358] 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.

[0359] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0360] 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."

[0361] This invention is a customer support system that utilizes a generative AI model. This system creates an FAQ database and a troubleshooting guide in advance, and then trains the generative AI model based on them to generate appropriate answers to user questions.

[0362] First, the server creates a customer support FAQ database and troubleshooting guide, which are built based on past inquiry data and expert knowledge, covering frequently asked questions and solutions to problems.

[0363] The server then uses this FAQ database and troubleshooting guide to train a generative AI model designed to generate appropriate responses to natural language questions from users. The model uses automated learning algorithms to learn patterns from the dataset and generate optimal answers based on that knowledge.

[0364] When a user uses a device to input a question, the device sends the question to a server, which passes the question to a generative AI model, which then generates the optimal response to the question. The generated response is then sent back to the device via the server and displayed to the user.

[0365] For example, if a user asks, "How do I return an item?", the device sends this question to the server. The server passes this question to a generative AI model, which then generates a response such as, "To find out how to return an item, select the item you want to return from your order history and follow the return procedure from there. For detailed instructions, please see the link below." The server then sends this response back to the device, allowing the user to view the answer on the screen.

[0366] Furthermore, the server has the ability to collect user feedback and adjust and improve the generative AI model based on it. Based on the feedback provided by the user, the model's response accuracy can be periodically improved. When a user gives feedback such as "This answer was very helpful," the data is stored on the server and used for the next model training.

[0367] In this way, the customer support system of the present invention realizes quick and appropriate customer responses and improves customer satisfaction. Furthermore, by improving the response accuracy of the AI, it is possible to reduce training and education costs and provide efficient customer support.

[0368] The processing flow will be explained below.

[0369] Step 1:

[0370] The server creates a FAQ database and troubleshooting guide based on experts and past support data, organizing and storing frequently asked questions and their solutions.

[0371] Step 2:

[0372] The server trains a generative AI model based on the FAQ database and troubleshooting guide it creates. The model learns from these datasets to generate appropriate responses to natural language input from users.

[0373] Step 3:

[0374] The server provides a user interface (UI) for users to enter questions, which can be implemented as a web page or a mobile application.

[0375] Step 4:

[0376] A user accesses the customer support system through the UI and inputs a question, such as "How do I return a product?"

[0377] Step 5:

[0378] The device sends the user-entered question to the server as an HTTP request, which is sent in an appropriate format (e.g., JSON).

[0379] Step 6:

[0380] The server inputs the received question into a generative AI model, which analyzes the question and generates the best answer based on pre-trained data.

[0381] Step 7:

[0382] The server takes the generated answer and formats it in a way that is easy for the user to understand, for example as text with links and additional information.

[0383] Step 8:

[0384] The server returns a formatted response to the terminal, which then displays the response on the user's terminal.

[0385] Step 9:

[0386] The user checks the answer displayed on the terminal, and if the question is not resolved, the user can enter the question again.

[0387] Step 10:

[0388] Users provide feedback on the answers they receive from the system, such as rating the answer as "this answer was helpful" or "this answer was insufficient."

[0389] Step 11:

[0390] The server stores the collected feedback from users in a database, which is used for the next model training.

[0391] Step 12:

[0392] The server retrains the generative AI model with newly collected feedback data to improve the accuracy and coverage of answers, thus maintaining a cycle of system improvement.

[0393] Each step in the embodiment of the present invention is specifically performed in this manner to provide prompt and appropriate customer support to users.

[0394] Example 1

[0395] 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."

[0396] Existing customer support systems face challenges in providing prompt and appropriate responses to user inquiries. Furthermore, they lack a means to continuously improve the system based on user feedback, making it difficult to improve response accuracy. Furthermore, they are unable to provide consistent answers to the diverse problems users face, potentially leading to a decline in customer satisfaction.

[0397] 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.

[0398] In this invention, the server includes a means for creating an FAQ database and a troubleshooting guide in advance, a means for training a generative AI model to generate appropriate answers to user questions, a means for providing a user interface for users to input questions, and a means for collecting feedback and adjusting and improving the generative AI model based on the feedback, thereby enabling prompt and appropriate customer support, improving customer satisfaction, and achieving continuous improvement of the system.

[0399] An "FAQ database" is a database that is built based on past inquiry data and expert knowledge and centrally manages answers to frequently asked inquiries.

[0400] A "troubleshooting guide" is a systematic guide that describes in detail how to solve problems and issues that users may encounter.

[0401] A "generative AI model" is an artificial intelligence model trained to generate appropriate answers to user questions using natural language processing techniques.

[0402] A "user interface" is an interactive screen or operating means through which a user accesses a system and inputs questions.

[0403] "Feedback" refers to the evaluations and opinions that users provide regarding the system's responses, and this information is used to improve the system.

[0404] A "training dataset" is a collection of data used to train a generative AI model, collected from FAQ databases and troubleshooting guides.

[0405] A "terminal" is a device, such as a computer or smartphone, that a user uses to access the system.

[0406] The "server" is a central processing unit that receives questions from users and passes them to a generative AI model to generate answers.

[0407] This invention relates to a customer support system that utilizes a generative AI model. This system creates an FAQ database and a troubleshooting guide in advance, and then trains a generative AI model based on them to generate appropriate answers to user questions.

[0408] First, the server uses software and data management tools to create an FAQ database and troubleshooting guide for customer support. Specific hardware includes a database server or high-performance computer, and software includes a database management system such as MySQL or PostgreSQL. This database and guide are built based on past inquiry data and expert knowledge, and cover frequently asked questions and solutions to problems.

[0409] The server then uses the FAQ database and troubleshooting guide to train a generative AI model. The generative AI model is created using libraries such as TensorFlow and PyTorch. It is designed to use an automatic learning algorithm to generate appropriate responses to natural language questions from users. Specifically, question and answer pairs are extracted from the FAQ database and troubleshooting guide and provided to the generative AI model as a training dataset.

[0410] When a user inputs a question using a device, the question is sent from the device to a server. The device is assumed to be a PC or smartphone, and a dedicated user interface is provided for it. For example, if a user asks, "How do I return a product?", the question is sent from the device to the server. The server passes the received question to a generative AI model, and the model generates an optimal response. The generated response is then sent back to the device via the server and displayed to the user.

[0411] An example of a generated response might be, "To find out how to return an item, please select the item you wish to return from your order history and follow the return procedure from there. For detailed instructions, please see the link below." This response responds quickly and accurately to the information the user is seeking.

[0412] Furthermore, the server has the ability to collect user feedback and adjust and improve the generative AI model based on that feedback. When a user provides feedback such as "This answer was very helpful," that data is stored on the server and used for the next model training. Collecting this feedback and retraining the model improves the response accuracy of the generative AI model and improves the performance of the system as a whole.

[0413] Examples of prompts include:

[0414] Please tell me how to return the product

[0415] "I haven't received the product I ordered, what should I do?"

[0416] Please tell me how to change my payment method.

[0417] The system aims to provide efficient and consistent customer support and improve customer satisfaction. In addition, the improved accuracy of AI responses will reduce training and education costs, resulting in more efficient support.

[0418] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0419] Step 1: Creating a FAQ database and troubleshooting guide

[0420] The server collects past inquiry data and expert knowledge, and uses them to create an FAQ database and troubleshooting guide. Specifically, it uses a database management system (e.g., MySQL, PostgreSQL) to categorize, organize, and store the collected data.

[0421] Input: past inquiry data, expert knowledge

[0422] Output: FAQ database, troubleshooting guide

[0423] Step 2: Training the generative AI model

[0424] The server trains the generative AI model using data from the FAQ database and troubleshooting guides. Specifically, it prepares a training dataset and builds and trains the model using libraries such as TensorFlow and PyTorch.

[0425] Input: FAQ database, troubleshooting guide

[0426] Output: A trained generative AI model

[0427] Step 3: Accepting user questions

[0428] A user inputs a question using a terminal, which then sends the question to a server. Specifically, a question is input in natural language through a user interface and then sent to the server via a network.

[0429] Input: User question (natural language)

[0430] Output: Question data (received on the server side)

[0431] Step 4: Question processing and response generation

[0432] The server passes the received question to a generative AI model, which then generates an appropriate response to the question. Specifically, the question data is input into the AI ​​model, and the model generates the optimal response based on the training data.

[0433] Input: Question data

[0434] Output: The generated response

[0435] Step 5: Returning a response

[0436] The server sends the generated response back to the terminal, which then displays the received response to the user. Specifically, the response data is sent to the terminal via the network, and the user can check the response on the screen.

[0437] Input: Generated response

[0438] Output: The response that is displayed to the user

[0439] Step 6: Gather feedback

[0440] The user inputs the feedback they provide on the terminal, and the terminal transmits the feedback to the server. Specifically, the user inputs their thoughts and evaluations on the response, which are then stored on the server.

[0441] Input: User feedback

[0442] Output: Feedback data stored on the server

[0443] Step 7: Adjust and improve the model

[0444] The server periodically retrains the generative AI model based on the collected feedback to improve response accuracy. Specifically, it adds the feedback data to the training data and retrains the generative AI model.

[0445] Input: Feedback data

[0446] Output: An improved generative AI model

[0447] (Application example 1)

[0448] 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."

[0449] Conventional customer support systems often struggle to respond to user inquiries quickly and accurately. They also lack the functionality to efficiently collect user feedback and improve the accuracy of their models. As a result, problems such as low user satisfaction and inefficient support operations arise.

[0450] 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.

[0451] In this invention, the server includes means for creating an FAQ database and a troubleshooting guide in advance, means for training a generative AI model to generate appropriate answers to user questions, means for providing a user interface for users to input questions, means for passing questions input by users to the generative AI model to generate answers, means for returning the generated answers to the users, and means for collecting user feedback and improving the accuracy of the generative AI model, thereby enabling faster responses to user inquiries, providing accurate responses, and improving the model based on the feedback.

[0452] An "FAQ database" is an information resource that systematically organizes frequently asked questions and their answers.

[0453] A "troubleshooting guide" is a guide that specifically describes problems that users often encounter and how to solve them.

[0454] A "generative AI model" is an artificial intelligence model that generates appropriate answers to questions posed in natural language by users.

[0455] "User interface" refers to the screen and operating methods that users use to input information into a system.

[0456] "Feedback" refers to opinions and ratings provided by users about the quality and usability of an answer.

[0457] The present invention is a customer support system that uses a FAQ database and troubleshooting guides to respond to user questions. This system is intended for use particularly in online shopping sites, and generates real-time responses to user questions via smartphones.

[0458] The server first creates a database of FAQs and troubleshooting guides, then uses them to train a generative AI model, which uses an automated learning algorithm, such as OpenAI's GPT-4, to learn patterns from the dataset and collects user feedback, which it then uses to improve the model's accuracy.

[0459] When a user inputs a question from their smartphone, the device sends the question to a server. The server passes the received question to a generative AI model, which generates an appropriate answer. The generated answer is then sent back to the smartphone via the server and displayed to the user. This allows the user to receive a quick and accurate answer.

[0460] As a specific use case, if a user asks, "How do I return an item?", the device sends this question to the server. The server passes this question to the generative AI model, which then generates a response such as, "Select the item you want to return from your order history and complete the return procedure. Click here for details." The server then sends this response back to the device, allowing the user to view the answer on their smartphone screen.

[0461] User feedback is collected in the form of, for example, "This answer was very helpful." This feedback is then stored on the server and used for the next training. This allows the generative AI model to regularly improve its response accuracy, enabling the provision of higher quality customer support.

[0462] The specific hardware used includes smartphones as user devices, high-performance computers as servers, and GPT-4 as a generative AI model, while the software used includes various APIs, cloud services, and database management systems.

[0463] Example prompt sentence:

[0464] When a user asks, "How do I return an item?", respond with: "Select the item you want to return from your order history and follow the return process. Click this link for more information."

[0465] As a result, this customer support system can provide efficient and satisfying service and significantly reduce the support work load of companies.

[0466] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0467] Step 1:

[0468] The user inputs a question through the smartphone's user interface, for example, "How do I return a product?"

[0469] Input: User question

[0470] Output: Data of the questions entered on the smartphone

[0471] Step 2:

[0472] The device converts the questions entered by the user into data format and sends it to the server. This data is structured in JSON format or similar.

[0473] Input: Questions typed into the user's smartphone

[0474] Output: Structured question data (e.g., JSON format)

[0475] Step 3:

[0476] The server passes the received question to the generative AI model. Specifically, it converts the question data into a format that the generative AI model can process and inputs it into the model.

[0477] Input: Structured question data

[0478] Output: Input data prepared for a generative AI model

[0479] Step 4:

[0480] A generative AI model (e.g., GPT-4) analyzes the provided question data and generates the optimal answer by referencing a FAQ database and troubleshooting guide.

[0481] Input: Input data for the generative AI model

[0482] Output: The generated answer

[0483] Step 5:

[0484] The server receives the answers from the generative AI model and converts them into a data format to send back to the user.

[0485] Input: Answer from generative AI model

[0486] Output: Structured response data

[0487] Step 6:

[0488] The terminal receives the answer data from the server and displays it on the screen, allowing the user to check the answer to the question.

[0489] Input: Response data from the server

[0490] Output: Answer displayed on smartphone screen

[0491] Step 7:

[0492] The user can input feedback about the displayed answer and send it to the server via the terminal. For example, the user can give feedback such as "This answer was very helpful."

[0493] Input: User feedback

[0494] Output: Feedback data sent from the device to the server

[0495] Step 8:

[0496] The server collects and stores user feedback to add to the training dataset of the generative AI model, which will improve the model's accuracy the next time it is trained.

[0497] Input: Feedback data from the device

[0498] Output: Feedback added to the training dataset

[0499] 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.

[0500] This invention is a customer support system that utilizes a generative AI model and further combines it with an emotion engine that recognizes user emotions. This system creates a FAQ database and troubleshooting guide in advance, and then trains the generative AI model based on these to generate appropriate answers to user questions. The emotion engine also analyzes user emotions and adjusts the generative AI model's responses to provide more appropriate customer support.

[0501] First, the server creates a customer support FAQ database and troubleshooting guide, which are built based on past inquiry data and expert knowledge, covering frequently asked questions and solutions to problems.

[0502] The server then uses this FAQ database and troubleshooting guide to train a generative AI model, which learns from these datasets to generate appropriate responses to natural language input from users.

[0503] Furthermore, the server provides a user interface (UI) for users to input their questions, which can be implemented as a web page or a mobile application and easily accessible to users.

[0504] The new system incorporates an emotion engine, which can analyze user emotions and apply responses. When a user uses a device to input a question, the device sends the question to a server. The server then passes the received question to the emotion engine, which analyzes the user's emotions. The analyzed emotion data is then passed to a generative AI model, which then generates an appropriate response based on that data.

[0505] For example, if a user asks, "How do I return an item?", the device sends this question to the server. At the same time, the emotion engine analyzes the tone and context of the user's speech and, if it determines that the user is in a hurry, the generative AI model generates a quick and specific response such as, "The return process is simple. Please follow the steps below immediately."

[0506] The server then formats the generated answer and sends it back to the user, who can review the answer displayed on their device and provide feedback if needed, such as "This answer was very helpful." This feedback is stored on the server and used to retrain the generative AI model and emotion engine.

[0507] In this way, the customer support system of the present invention combines a generative AI model and an emotion engine to provide fast and appropriate customer service and improve customer satisfaction. Furthermore, by recognizing the user's emotions and adjusting responses accordingly, it is possible to provide more personalized support. The above is a specific embodiment for implementing the present invention.

[0508] The processing flow will be explained below.

[0509] Step 1:

[0510] The server uses experts and past support data to create a FAQ database and troubleshooting guides that organize and store frequently asked questions and their solutions.

[0511] Step 2:

[0512] The server uses the created FAQ database and troubleshooting guide to train a generative AI model, which is then able to learn patterns from this data and generate appropriate responses to users' natural language input.

[0513] Step 3:

[0514] The server provides a user interface (UI) for users to enter their questions, implemented as a web page or mobile application, making it easy for users to access and use.

[0515] Step 4:

[0516] The user uses a terminal to access the provided UI and input a question into the customer support system, for example, "How do I return a product?"

[0517] Step 5:

[0518] The device sends the user-entered question to the server as an HTTP request, which is sent in an appropriate format (e.g., JSON).

[0519] Step 6:

[0520] The server passes the received question to the emotion engine, which analyzes the emotion from the user's input text and passes the analysis results to the generative AI model.

[0521] Step 7:

[0522] The emotion engine analyzes emotional data from user input, for example classifying emotions from user text as "positive," "negative," or "neutral." This data is then used to adjust the generative AI model's response.

[0523] Step 8:

[0524] The server inputs the emotion data obtained from the emotion engine into the generative AI model, which then references this data to generate a response that corresponds to the user's emotion.

[0525] Step 9:

[0526] The server takes the generated answer and formats it in a format that is easy for the user to understand, for example as text with links and additional information.

[0527] Step 10:

[0528] The server returns a formatted response to the terminal, which then displays the response on the user's terminal.

[0529] Step 11:

[0530] The user checks the answer displayed on the terminal, and if the answer is not what they expected, they can enter the question again.

[0531] Step 12:

[0532] Users provide feedback on the answers they receive from the system, such as rating the answer as "this answer was helpful" or "this answer was insufficient."

[0533] Step 13:

[0534] The server stores the collected feedback from users in a database, which is used for the next retraining of the model and emotion engine.

[0535] Step 14:

[0536] The server uses newly collected feedback data to retrain the generative AI model and emotion engine to improve the accuracy and coverage of answers, maintaining a cycle of system improvement.

[0537] Through this series of steps, the present invention combines a generative AI model and an emotion engine to provide fast and appropriate customer support, thereby increasing user satisfaction and realizing efficient support.

[0538] Example 2

[0539] 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."

[0540] Conventional customer support systems have difficulty in providing appropriate answers to user questions quickly, and are unable to generate responses that take into account the user's emotions. Therefore, in order to improve user satisfaction, it is necessary to provide personalized support that takes into account the user's emotions.

[0541] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0542] In this invention, the server includes means for creating an FAQ database and a troubleshooting guide, means for training a generative AI model to generate appropriate answers to user questions, means for providing a user interface for users to input questions, means for passing the questions input by the users to the generative AI model and a sentiment analysis engine to generate answers, and means for returning the generated answers to the users, thereby enabling the provision of prompt and appropriate answers that take into consideration the user's emotions.

[0543] An "FAQ database" is a collection of information that organizes and stores answers to frequently asked questions from users.

[0544] A "troubleshooting guide" is a manual that systematically describes solutions to problems and issues that users may encounter.

[0545] A "generative AI model" is an artificial intelligence model trained to generate appropriate responses to natural language input from a user.

[0546] A "user interface" is an interactive screen or device that allows a user to enter questions into a system and receive answers from the system.

[0547] An "emotion analysis engine" is a technology that analyzes language data entered by a user and estimates the user's emotional state.

[0548] "Feedback" refers to a user inputting an evaluation or opinion on the answer received from the system.

[0549] "Server" means the computer system that runs the entire customer support system, storing and processing data, training, generating responses, etc.

[0550] This invention is a customer support system that utilizes a generative AI model and further combines it with a sentiment analysis engine that recognizes user emotions. This system creates a FAQ database and troubleshooting guide in advance, and then trains the generative AI model based on these to generate appropriate answers to user questions. The sentiment analysis engine also analyzes user emotions and adjusts the generative AI model's responses to provide more appropriate customer support.

[0551] Hardware and software used

[0552] Server: A platform that creates FAQ databases and troubleshooting guides for customer support, and runs generative AI models and sentiment analysis engines.

[0553] Device: The device (e.g., computer, smartphone) used by the user to enter questions and exchange information with the server.

[0554] Generative AI model: An AI model that generates optimal answers to user questions. Examples include OpenAI's GPT-3 and GPT-4.

[0555] Sentiment analysis engine: An engine for analyzing user sentiment and adjusting the generated response. Example: IBM's Watson Tone Analyzer.

[0556] Specific examples

[0557] 1. Creating a FAQ database and troubleshooting guide

[0558] The server creates a FAQ database and troubleshooting guide for customer support. This database is built based on past inquiry data and expert knowledge. For example, a question about "how to return a product" may include specific instructions such as "bring proof of purchase and go to the store to complete the procedure."

[0559] 2. Training the generative AI model

[0560] The server uses the FAQ database and troubleshooting guide to train the generative AI model, for example, to respond appropriately to inquiries containing the keyword "returns."

[0561] 3. Providing a User Interface (UI)

[0562] The server provides a UI for users to enter their questions. This UI can be implemented as a web page or a mobile application. Once users log in, it provides a chat box and a search bar to make it easy for them to enter their questions.

[0563] 4. User Question Processing

[0564] The user uses the terminal to input a specific question, for example, "How do I return a product?", and the terminal transmits the user's question to the server in real time.

[0565] 5. Emotion Analysis

[0566] The server passes the received question to a sentiment analysis engine, which analyzes the content and tone of the question. The sentiment analysis engine infers the user's emotions, such as whether they are angry, confused, or in a hurry, from the context and wording. For example, if the question contains the phrase "I want to return it quickly," the sentiment analysis engine will determine that the user is in a hurry.

[0567] 6. Response generation using generative AI models

[0568] The generative AI model generates appropriate responses based on the emotional data provided by the sentiment analysis engine. For example, if it determines that the user is in a hurry, it generates a quick and specific response such as, "The return process is simple. Please follow the steps below immediately."

[0569] 7. Returning Responses

[0570] The server formats the generated response and sends it back to the user. The response created by the generative AI model is formatted into HTML or JSON format and sent to the terminal as an HTTP response.

[0571] 8. User Acknowledgment and Feedback

[0572] The user checks the answer displayed on the device. If the answer is helpful, they can enter feedback such as "This answer was very helpful." A feedback form is provided on the device so that users can easily submit their opinions.

[0573] 9. Save feedback and retrain

[0574] The server stores user feedback and uses it to retrain the generative AI model and sentiment analysis engine. The feedback data is stored in a database and used for the next training run, allowing the system to become more sophisticated over time and better meet user expectations.

[0575] Prompt Sentence Examples

[0576] Please tell me about the return procedure.

[0577] "My item hasn't arrived, what should I do?"

[0578] Please tell me about the product warranty period.

[0579] This allows the customer support system to respond quickly and appropriately to user questions, provide personalized service that takes user emotions into consideration, and increase customer satisfaction through continuous improvement through feedback.

[0580] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0581] Step 1:

[0582] The server creates an FAQ database and troubleshooting guide. The server collects past inquiry data and expert knowledge, and stores the data constructed based on that in the database. For example, if there are many inquiries about how to return a product, the server organizes and stores the specific procedures for those inquiries in the database.

[0583] Input: Historical inquiry data, expert knowledge

[0584] Output: FAQ database, troubleshooting guide

[0585] Step 2:

[0586] The server trains the generative AI model using the FAQ database and troubleshooting guide. The server collects information from both databases and trains the generative AI model to improve its ability to generate appropriate answers to user questions. For example, the server repeatedly trains a generative AI model (e.g., GPT-4) on data related to the keyword "returns."

[0587] Input: FAQ database, troubleshooting guide

[0588] Output: A trained generative AI model

[0589] Step 3:

[0590] The server provides a user interface (UI) for users to enter questions. This UI is implemented as a web page or mobile application, and allows users to easily enter questions. Specifically, it includes a chat box and a search bar.

[0591] Input: Web page and mobile application design information

[0592] Output: The implemented user interface

[0593] Step 4:

[0594] The user uses the terminal to enter a specific question, for example, "How do I return an item?" This question is sent to the server through the UI.

[0595] Input: User question

[0596] Output: Question data sent to the server

[0597] Step 5:

[0598] The server receives the question from the device and passes it to the emotion analysis engine. The emotion analysis engine analyzes the tone and context of the question to estimate the user's emotional state. For example, a question containing the phrase "I want to return it quickly" indicates a sense of urgency.

[0599] Input: User question data

[0600] Output: Emotion data

[0601] Step 6:

[0602] The generative AI model receives the emotion data and generates an appropriate response based on it. For example, if it determines that the user is in a hurry, it will generate a quick and specific response such as, "The return process is easy. Please follow the steps below immediately."

[0603] Input: Emotion data

[0604] Output: Response data

[0605] Step 7:

[0606] The server formats the generated response data and returns it to the user device. For example, it formats the response created by the generative AI model into HTML or JSON format and sends it to the device as an HTTP response.

[0607] Input: Response data

[0608] Output: Response sent back to the user's terminal

[0609] Step 8:

[0610] The user checks the answer displayed on the terminal. If the problem is solved and the answer is helpful, the user enters feedback such as "This answer was very helpful." This feedback is then sent back to the server.

[0611] Input: User feedback

[0612] Output: Feedback data sent to the server

[0613] Step 9:

[0614] The server stores the feedback data and uses it to retrain the generative AI model and sentiment analysis engine. The feedback data is stored in a database and used for the next training session.

[0615] Input: Feedback data

[0616] Output: Improved generative AI models and sentiment analysis engines

[0617] These steps enable the customer support system to respond quickly and appropriately to user questions, provide personalized service that takes user sentiment into account, and increase customer satisfaction through continuous improvement through feedback.

[0618] (Application example 2)

[0619] 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."

[0620] Conventional customer support systems are required to respond to user questions quickly and appropriately, but they lack the technology to respond in a way that takes the user's emotions into consideration. This increases the likelihood of users feeling stressed or dissatisfied. In addition, there are limited ways to improve the system based on feedback, making it difficult to improve the quality of support.

[0621] The identification processing by the identification 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 creating an FAQ database and a troubleshooting guide in advance, means for training a generative AI model to generate appropriate answers to user questions, means for providing a user interface for the user to input questions, means for passing questions input by the user to the generative AI model to generate answers, means for analyzing the user's emotions using a sentiment analysis engine, means for adjusting the response of the generative AI model based on the results of the sentiment analysis, and means for returning the generated answers to the user. This makes it possible to provide quick and appropriate answers to user questions and to provide personalized responses that take the user's emotions into consideration.

[0622] The "FAQ Database" is a database of frequently asked questions and their answers, built based on past inquiry data and troubleshooting guides.

[0623] A "Troubleshooting Guide" is a collection of procedures or guidelines that provide solutions to specific problems or malfunctions.

[0624] A "generative AI model" is a model trained using machine learning techniques to generate appropriate responses to a user's natural language input.

[0625] A "user interface" is an interface that allows users to access the system and input questions, and is provided as a web page or mobile application.

[0626] An "emotion analysis engine" is a technology that analyzes and identifies emotions from user comments and input content.

[0627] A "feedback system" is a mechanism for collecting evaluations and opinions from users and using them to improve and adjust the system.

[0628] The following procedures and system configuration are required to realize the invention.

[0629] System Configuration

[0630] The system of this invention mainly consists of a server, a terminal, and a user. The server has an FAQ database and a troubleshooting guide, and includes a generative AI model and a sentiment analysis engine. The terminal provides a user interface for users to input questions.

[0631] Program processing

[0632] Creating a FAQ database and troubleshooting guide

[0633] The server creates a FAQ database and troubleshooting guide based on past inquiry data and expert knowledge, covering frequently asked questions and solutions to problems.

[0634] Training generative AI models

[0635] The server uses the FAQ database and troubleshooting guide to train a generative AI model, which is then used to generate appropriate responses to natural language input from users.

[0636] Providing a user interface

[0637] The server provides a user interface implemented as a web page or mobile application, allowing users to easily input questions via their devices.

[0638] Parsing questions and generating answers

[0639] When a user inputs a question using a terminal, the question is sent to the server. The server passes the question to a sentiment analysis engine, which analyzes the user's sentiment. For example, if a user asks, "How do I return a product?", the sentiment analysis engine analyzes the tone and context of the user's speech and determines that the user is in a hurry.

[0640] Regulating responses based on emotions

[0641] The analyzed emotion data is passed to a generative AI model, which then generates an appropriate response based on that data. For rushed users, a quick and specific response such as "Returns are easy, just follow the steps below" is generated.

[0642] Returning answers and collecting feedback

[0643] The server formats the generated answer and sends it back to the user via the device. The user can review the answer displayed on the device and provide feedback if necessary. For example, they could say, "This answer was very helpful." This feedback is stored on the server and used to retrain the generative AI model and sentiment analysis engine.

[0644] Hardware and software used

[0645] The system uses the following hardware and software:

[0646] Server hardware: FAQ database, troubleshooting guide, generative AI model, sentiment analysis engine

[0647] Natural Language Toolkit (NLTK): A natural language processing library

[0648] TensorFlow / Keras: Training and inference of machine learning models

[0649] Emotion Recognition API: External service for emotion analysis

[0650] Flask: Building a server-side API

[0651] React Native: Frontend for smartphone applications

[0652] Specific examples

[0653] When a user asks "How do I return an item?" the system behaves as follows:

[0654] Example prompt sentence:

[0655] User Question: How do I return an item? Sentiment: I'm in a hurry

[0656] Based on this prompt, the generative AI model responds with, "The return process is simple. Please follow the steps below immediately." This allows for a fast, appropriate, and personalized response to the user's question.

[0657] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0658] Step 1:

[0659] A user opens a customer support application using a terminal and enters a question. The entered question is sent to the server by the terminal. The input of this step is the user's question text, and the output is the transmission of the question data to the server.

[0660] Step 2:

[0661] The server sends the received question to a sentiment analysis engine to analyze the user's sentiment. The sentiment analysis engine uses natural language processing technology (such as NLTK) to analyze the tone and context of the speech and identify the user's sentiment (e.g., hurry, irritated, calm). The input of this step is the question data, and the output is the user's sentiment data.

[0662] Step 3:

[0663] The emotion data analyzed by the emotion analysis engine is passed to the generative AI model by the server. The generative AI model generates the optimal response to the user's question based on the training dataset (FAQ database and troubleshooting guide). The input of this step is the question data and emotion data, and the output is the provisional response data.

[0664] Step 4:

[0665] The tentative responses generated by the generative AI model are adjusted based on emotional data to be formatted to suit the user's emotions. For example, if the user is determined to be in a hurry, the response will be adjusted to be brief and quick. The inputs of this step are tentative response data and emotional data, and the output is the final response data.

[0666] Step 5:

[0667] The final response data is sent back from the server to the terminal and displayed to the user. The user checks the displayed answer and provides feedback if necessary. The input of this step is the final response data, and the output is the response displayed on the user terminal and user feedback.

[0668] Step 6:

[0669] User feedback is sent from the device to the server, which collects and stores it. The collected feedback is used to retrain the generative AI model and sentiment analysis engine. The input of this step is user feedback, and the output is the storage of feedback data and model improvements.

[0670] These processing steps enable the system to generate fast, relevant answers to user questions, personalize responses by taking into account user sentiment, and incorporate feedback to continuously improve the overall system quality and accuracy.

[0671] 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.

[0672] 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.

[0673] 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.

[0674] [Third embodiment]

[0675] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0676] 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.

[0677] 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).

[0678] 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.

[0679] 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.

[0680] 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).

[0681] 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.

[0682] 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.

[0683] 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.

[0684] 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.

[0685] 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.

[0686] 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."

[0687] This invention is a customer support system that utilizes a generative AI model. This system creates an FAQ database and a troubleshooting guide in advance, and then trains the generative AI model based on them to generate appropriate answers to user questions.

[0688] First, the server creates a customer support FAQ database and troubleshooting guide, which are built based on past inquiry data and expert knowledge, covering frequently asked questions and solutions to problems.

[0689] The server then uses this FAQ database and troubleshooting guide to train a generative AI model designed to generate appropriate responses to natural language questions from users. The model uses automated learning algorithms to learn patterns from the dataset and generate optimal answers based on that knowledge.

[0690] When a user uses a device to input a question, the device sends the question to a server, which passes the question to a generative AI model, which then generates the optimal response to the question. The generated response is then sent back to the device via the server and displayed to the user.

[0691] For example, if a user asks, "How do I return an item?", the device sends this question to the server. The server passes this question to a generative AI model, which then generates a response such as, "To find out how to return an item, select the item you want to return from your order history and follow the return procedure from there. For detailed instructions, please see the link below." The server then sends this response back to the device, allowing the user to view the answer on the screen.

[0692] Furthermore, the server has the ability to collect user feedback and adjust and improve the generative AI model based on it. Based on the feedback provided by the user, the model's response accuracy can be periodically improved. When a user gives feedback such as "This answer was very helpful," the data is stored on the server and used for the next model training.

[0693] In this way, the customer support system of the present invention realizes quick and appropriate customer responses and improves customer satisfaction. Furthermore, by improving the response accuracy of the AI, it is possible to reduce training and education costs and provide efficient customer support.

[0694] The processing flow will be explained below.

[0695] Step 1:

[0696] The server creates a FAQ database and troubleshooting guide based on experts and past support data, organizing and storing frequently asked questions and their solutions.

[0697] Step 2:

[0698] The server trains a generative AI model based on the FAQ database and troubleshooting guide it creates. The model learns from these datasets to generate appropriate responses to natural language input from users.

[0699] Step 3:

[0700] The server provides a user interface (UI) for users to enter questions, which can be implemented as a web page or a mobile application.

[0701] Step 4:

[0702] A user accesses the customer support system through the UI and inputs a question, such as "How do I return a product?"

[0703] Step 5:

[0704] The device sends the user-entered question to the server as an HTTP request, which is sent in an appropriate format (e.g., JSON).

[0705] Step 6:

[0706] The server inputs the received question into a generative AI model, which analyzes the question and generates the best answer based on pre-trained data.

[0707] Step 7:

[0708] The server takes the generated answer and formats it in a way that is easy for the user to understand, for example as text with links and additional information.

[0709] Step 8:

[0710] The server returns a formatted response to the terminal, which then displays the response on the user's terminal.

[0711] Step 9:

[0712] The user checks the answer displayed on the terminal, and if the question is not resolved, the user can enter the question again.

[0713] Step 10:

[0714] Users provide feedback on the answers they receive from the system, such as rating the answer as "this answer was helpful" or "this answer was insufficient."

[0715] Step 11:

[0716] The server stores the collected feedback from users in a database, which is used for the next model training.

[0717] Step 12:

[0718] The server retrains the generative AI model with newly collected feedback data to improve the accuracy and coverage of answers, thus maintaining a cycle of system improvement.

[0719] Each step in the embodiment of the present invention is specifically performed in this manner to provide prompt and appropriate customer support to users.

[0720] Example 1

[0721] 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."

[0722] Existing customer support systems face challenges in providing prompt and appropriate responses to user inquiries. Furthermore, they lack a means to continuously improve the system based on user feedback, making it difficult to improve response accuracy. Furthermore, they are unable to provide consistent answers to the diverse problems users face, potentially leading to a decline in customer satisfaction.

[0723] 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.

[0724] In this invention, the server includes a means for creating an FAQ database and a troubleshooting guide in advance, a means for training a generative AI model to generate appropriate answers to user questions, a means for providing a user interface for users to input questions, and a means for collecting feedback and adjusting and improving the generative AI model based on the feedback, thereby enabling prompt and appropriate customer support, improving customer satisfaction, and achieving continuous improvement of the system.

[0725] An "FAQ database" is a database that is built based on past inquiry data and expert knowledge and centrally manages answers to frequently asked inquiries.

[0726] A "troubleshooting guide" is a systematic guide that describes in detail how to solve problems and issues that users may encounter.

[0727] A "generative AI model" is an artificial intelligence model trained to generate appropriate answers to user questions using natural language processing techniques.

[0728] A "user interface" is an interactive screen or operating means through which a user accesses a system and inputs questions.

[0729] "Feedback" refers to the evaluations and opinions that users provide regarding the system's responses, and this information is used to improve the system.

[0730] A "training dataset" is a collection of data used to train a generative AI model, collected from FAQ databases and troubleshooting guides.

[0731] A "terminal" is a device, such as a computer or smartphone, that a user uses to access the system.

[0732] The "server" is a central processing unit that receives questions from users and passes them to a generative AI model to generate answers.

[0733] This invention relates to a customer support system that utilizes a generative AI model. This system creates an FAQ database and a troubleshooting guide in advance, and then trains a generative AI model based on them to generate appropriate answers to user questions.

[0734] First, the server uses software and data management tools to create an FAQ database and troubleshooting guide for customer support. Specific hardware includes a database server or high-performance computer, and software includes a database management system such as MySQL or PostgreSQL. This database and guide are built based on past inquiry data and expert knowledge, and cover frequently asked questions and solutions to problems.

[0735] The server then uses the FAQ database and troubleshooting guide to train a generative AI model. The generative AI model is created using libraries such as TensorFlow and PyTorch. It is designed to use an automatic learning algorithm to generate appropriate responses to natural language questions from users. Specifically, question and answer pairs are extracted from the FAQ database and troubleshooting guide and provided to the generative AI model as a training dataset.

[0736] When a user inputs a question using a device, the question is sent from the device to a server. The device is assumed to be a PC or smartphone, and a dedicated user interface is provided for it. For example, if a user asks, "How do I return a product?", the question is sent from the device to the server. The server passes the received question to a generative AI model, and the model generates an optimal response. The generated response is then sent back to the device via the server and displayed to the user.

[0737] An example of a generated response might be, "To find out how to return an item, please select the item you wish to return from your order history and follow the return procedure from there. For detailed instructions, please see the link below." This response responds quickly and accurately to the information the user is seeking.

[0738] Furthermore, the server has the ability to collect user feedback and adjust and improve the generative AI model based on that feedback. When a user provides feedback such as "This answer was very helpful," that data is stored on the server and used for the next model training. Collecting this feedback and retraining the model improves the response accuracy of the generative AI model and improves the performance of the system as a whole.

[0739] Examples of prompts include:

[0740] Please tell me how to return the product

[0741] "I haven't received the product I ordered, what should I do?"

[0742] Please tell me how to change my payment method.

[0743] The system aims to provide efficient and consistent customer support and improve customer satisfaction. In addition, the improved accuracy of AI responses will reduce training and education costs, resulting in more efficient support.

[0744] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0745] Step 1: Creating a FAQ database and troubleshooting guide

[0746] The server collects past inquiry data and expert knowledge, and uses them to create an FAQ database and troubleshooting guide. Specifically, it uses a database management system (e.g., MySQL, PostgreSQL) to categorize, organize, and store the collected data.

[0747] Input: past inquiry data, expert knowledge

[0748] Output: FAQ database, troubleshooting guide

[0749] Step 2: Training the generative AI model

[0750] The server trains the generative AI model using data from the FAQ database and troubleshooting guides. Specifically, it prepares a training dataset and builds and trains the model using libraries such as TensorFlow and PyTorch.

[0751] Input: FAQ database, troubleshooting guide

[0752] Output: A trained generative AI model

[0753] Step 3: Accepting user questions

[0754] A user inputs a question using a terminal, which then sends the question to a server. Specifically, a question is input in natural language through a user interface and then sent to the server via a network.

[0755] Input: User question (natural language)

[0756] Output: Question data (received on the server side)

[0757] Step 4: Question processing and response generation

[0758] The server passes the received question to a generative AI model, which then generates an appropriate response to the question. Specifically, the question data is input into the AI ​​model, and the model generates the optimal response based on the training data.

[0759] Input: Question data

[0760] Output: The generated response

[0761] Step 5: Returning a response

[0762] The server sends the generated response back to the terminal, which then displays the received response to the user. Specifically, the response data is sent to the terminal via the network, and the user can check the response on the screen.

[0763] Input: Generated response

[0764] Output: The response that is displayed to the user

[0765] Step 6: Gather feedback

[0766] The user inputs the feedback they provide on the terminal, and the terminal transmits the feedback to the server. Specifically, the user inputs their thoughts and evaluations on the response, which are then stored on the server.

[0767] Input: User feedback

[0768] Output: Feedback data stored on the server

[0769] Step 7: Adjust and improve the model

[0770] The server periodically retrains the generative AI model based on the collected feedback to improve response accuracy. Specifically, it adds the feedback data to the training data and retrains the generative AI model.

[0771] Input: Feedback data

[0772] Output: An improved generative AI model

[0773] (Application example 1)

[0774] 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."

[0775] Conventional customer support systems often struggle to respond to user inquiries quickly and accurately. They also lack the functionality to efficiently collect user feedback and improve the accuracy of their models. As a result, problems such as low user satisfaction and inefficient support operations arise.

[0776] 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.

[0777] In this invention, the server includes means for creating an FAQ database and a troubleshooting guide in advance, means for training a generative AI model to generate appropriate answers to user questions, means for providing a user interface for users to input questions, means for passing questions input by users to the generative AI model to generate answers, means for returning the generated answers to the users, and means for collecting user feedback and improving the accuracy of the generative AI model, thereby enabling faster responses to user inquiries, providing accurate responses, and improving the model based on the feedback.

[0778] An "FAQ database" is an information resource that systematically organizes frequently asked questions and their answers.

[0779] A "troubleshooting guide" is a guide that specifically describes problems that users often encounter and how to solve them.

[0780] A "generative AI model" is an artificial intelligence model that generates appropriate answers to questions posed in natural language by users.

[0781] "User interface" refers to the screen and operating methods that users use to input information into a system.

[0782] "Feedback" refers to opinions and ratings provided by users about the quality and usability of an answer.

[0783] The present invention is a customer support system that uses a FAQ database and troubleshooting guides to respond to user questions. This system is intended for use particularly in online shopping sites, and generates real-time responses to user questions via smartphones.

[0784] The server first creates a database of FAQs and troubleshooting guides, then uses them to train a generative AI model, which uses an automated learning algorithm, such as OpenAI's GPT-4, to learn patterns from the dataset and collects user feedback, which it then uses to improve the model's accuracy.

[0785] When a user inputs a question from their smartphone, the device sends the question to a server. The server passes the received question to a generative AI model, which generates an appropriate answer. The generated answer is then sent back to the smartphone via the server and displayed to the user. This allows the user to receive a quick and accurate answer.

[0786] As a specific use case, if a user asks, "How do I return an item?", the device sends this question to the server. The server passes this question to the generative AI model, which then generates a response such as, "Select the item you want to return from your order history and complete the return procedure. Click here for details." The server then sends this response back to the device, allowing the user to view the answer on their smartphone screen.

[0787] User feedback is collected in the form of, for example, "This answer was very helpful." This feedback is then stored on the server and used for the next training. This allows the generative AI model to regularly improve its response accuracy, enabling the provision of higher quality customer support.

[0788] The specific hardware used includes smartphones as user devices, high-performance computers as servers, and GPT-4 as a generative AI model, while the software used includes various APIs, cloud services, and database management systems.

[0789] Example prompt sentence:

[0790] When a user asks, "How do I return an item?", respond with: "Select the item you want to return from your order history and follow the return process. Click this link for more information."

[0791] As a result, this customer support system can provide efficient and satisfying service and significantly reduce the support work load of companies.

[0792] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0793] Step 1:

[0794] The user inputs a question through the smartphone's user interface, for example, "How do I return a product?"

[0795] Input: User question

[0796] Output: Data of the questions entered on the smartphone

[0797] Step 2:

[0798] The device converts the questions entered by the user into data format and sends it to the server. This data is structured in JSON format or similar.

[0799] Input: Questions typed into the user's smartphone

[0800] Output: Structured question data (e.g., JSON format)

[0801] Step 3:

[0802] The server passes the received question to the generative AI model. Specifically, it converts the question data into a format that the generative AI model can process and inputs it into the model.

[0803] Input: Structured question data

[0804] Output: Input data prepared for a generative AI model

[0805] Step 4:

[0806] A generative AI model (e.g., GPT-4) analyzes the provided question data and generates the optimal answer by referencing a FAQ database and troubleshooting guide.

[0807] Input: Input data for the generative AI model

[0808] Output: The generated answer

[0809] Step 5:

[0810] The server receives the answers from the generative AI model and converts them into a data format to send back to the user.

[0811] Input: Answer from generative AI model

[0812] Output: Structured response data

[0813] Step 6:

[0814] The terminal receives the answer data from the server and displays it on the screen, allowing the user to check the answer to the question.

[0815] Input: Response data from the server

[0816] Output: Answer displayed on smartphone screen

[0817] Step 7:

[0818] The user can input feedback about the displayed answer and send it to the server via the terminal. For example, the user can give feedback such as "This answer was very helpful."

[0819] Input: User feedback

[0820] Output: Feedback data sent from the device to the server

[0821] Step 8:

[0822] The server collects and stores user feedback to add to the training dataset of the generative AI model, which will improve the model's accuracy the next time it is trained.

[0823] Input: Feedback data from the device

[0824] Output: Feedback added to the training dataset

[0825] 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.

[0826] This invention is a customer support system that utilizes a generative AI model and further combines it with an emotion engine that recognizes user emotions. This system creates a FAQ database and troubleshooting guide in advance, and then trains the generative AI model based on these to generate appropriate answers to user questions. The emotion engine also analyzes user emotions and adjusts the generative AI model's responses to provide more appropriate customer support.

[0827] First, the server creates a customer support FAQ database and troubleshooting guide, which are built based on past inquiry data and expert knowledge, covering frequently asked questions and solutions to problems.

[0828] The server then uses this FAQ database and troubleshooting guide to train a generative AI model, which learns from these datasets to generate appropriate responses to natural language input from users.

[0829] Furthermore, the server provides a user interface (UI) for users to input their questions, which can be implemented as a web page or a mobile application and easily accessible to users.

[0830] The new system incorporates an emotion engine, which can analyze user emotions and apply responses. When a user uses a device to input a question, the device sends the question to a server. The server then passes the received question to the emotion engine, which analyzes the user's emotions. The analyzed emotion data is then passed to a generative AI model, which then generates an appropriate response based on that data.

[0831] For example, if a user asks, "How do I return an item?", the device sends this question to the server. At the same time, the emotion engine analyzes the tone and context of the user's speech and, if it determines that the user is in a hurry, the generative AI model generates a quick and specific response such as, "The return process is simple. Please follow the steps below immediately."

[0832] The server then formats the generated answer and sends it back to the user, who can review the answer displayed on their device and provide feedback if needed, such as "This answer was very helpful." This feedback is stored on the server and used to retrain the generative AI model and emotion engine.

[0833] In this way, the customer support system of the present invention combines a generative AI model and an emotion engine to provide fast and appropriate customer service and improve customer satisfaction. Furthermore, by recognizing the user's emotions and adjusting responses accordingly, it is possible to provide more personalized support. The above is a specific embodiment for implementing the present invention.

[0834] The processing flow will be explained below.

[0835] Step 1:

[0836] The server uses experts and past support data to create a FAQ database and troubleshooting guides that organize and store frequently asked questions and their solutions.

[0837] Step 2:

[0838] The server uses the created FAQ database and troubleshooting guide to train a generative AI model, which is then able to learn patterns from this data and generate appropriate responses to users' natural language input.

[0839] Step 3:

[0840] The server provides a user interface (UI) for users to enter their questions, implemented as a web page or mobile application, making it easy for users to access and use.

[0841] Step 4:

[0842] The user uses a terminal to access the provided UI and input a question into the customer support system, for example, "How do I return a product?"

[0843] Step 5:

[0844] The device sends the user-entered question to the server as an HTTP request, which is sent in an appropriate format (e.g., JSON).

[0845] Step 6:

[0846] The server passes the received question to the emotion engine, which analyzes the emotion from the user's input text and passes the analysis results to the generative AI model.

[0847] Step 7:

[0848] The emotion engine analyzes emotional data from user input, for example classifying emotions from user text as "positive," "negative," or "neutral." This data is then used to adjust the generative AI model's response.

[0849] Step 8:

[0850] The server inputs the emotion data obtained from the emotion engine into the generative AI model, which then references this data to generate a response that corresponds to the user's emotion.

[0851] Step 9:

[0852] The server takes the generated answer and formats it in a format that is easy for the user to understand, for example as text with links and additional information.

[0853] Step 10:

[0854] The server returns a formatted response to the terminal, which then displays the response on the user's terminal.

[0855] Step 11:

[0856] The user checks the answer displayed on the terminal, and if the answer is not what they expected, they can enter the question again.

[0857] Step 12:

[0858] Users provide feedback on the answers they receive from the system, such as rating the answer as "this answer was helpful" or "this answer was insufficient."

[0859] Step 13:

[0860] The server stores the collected feedback from users in a database, which is used for the next retraining of the model and emotion engine.

[0861] Step 14:

[0862] The server uses newly collected feedback data to retrain the generative AI model and emotion engine to improve the accuracy and coverage of answers, maintaining a cycle of system improvement.

[0863] Through this series of steps, the present invention combines a generative AI model and an emotion engine to provide fast and appropriate customer support, thereby increasing user satisfaction and realizing efficient support.

[0864] Example 2

[0865] 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."

[0866] Conventional customer support systems have difficulty in providing appropriate answers to user questions quickly, and are unable to generate responses that take into account the user's emotions. Therefore, in order to improve user satisfaction, it is necessary to provide personalized support that takes into account the user's emotions.

[0867] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0868] In this invention, the server includes means for creating an FAQ database and a troubleshooting guide, means for training a generative AI model to generate appropriate answers to user questions, means for providing a user interface for users to input questions, means for passing the questions input by the users to the generative AI model and a sentiment analysis engine to generate answers, and means for returning the generated answers to the users, thereby enabling the provision of prompt and appropriate answers that take into consideration the user's emotions.

[0869] An "FAQ database" is a collection of information that organizes and stores answers to frequently asked questions from users.

[0870] A "troubleshooting guide" is a manual that systematically describes solutions to problems and issues that users may encounter.

[0871] A "generative AI model" is an artificial intelligence model trained to generate appropriate responses to natural language input from a user.

[0872] A "user interface" is an interactive screen or device that allows a user to enter questions into a system and receive answers from the system.

[0873] An "emotion analysis engine" is a technology that analyzes language data entered by a user and estimates the user's emotional state.

[0874] "Feedback" refers to a user inputting an evaluation or opinion on the answer received from the system.

[0875] "Server" means the computer system that runs the entire customer support system, storing and processing data, training, generating responses, etc.

[0876] This invention is a customer support system that utilizes a generative AI model and further combines it with a sentiment analysis engine that recognizes user emotions. This system creates a FAQ database and troubleshooting guide in advance, and then trains the generative AI model based on these to generate appropriate answers to user questions. The sentiment analysis engine also analyzes user emotions and adjusts the generative AI model's responses to provide more appropriate customer support.

[0877] Hardware and software used

[0878] Server: A platform that creates FAQ databases and troubleshooting guides for customer support, and runs generative AI models and sentiment analysis engines.

[0879] Device: The device (e.g., computer, smartphone) used by the user to enter questions and exchange information with the server.

[0880] Generative AI model: An AI model that generates optimal answers to user questions. Examples include OpenAI's GPT-3 and GPT-4.

[0881] Sentiment analysis engine: An engine for analyzing user sentiment and adjusting the generated response. Example: IBM's Watson Tone Analyzer.

[0882] Specific examples

[0883] 1. Creating a FAQ database and troubleshooting guide

[0884] The server creates a FAQ database and troubleshooting guide for customer support. This database is built based on past inquiry data and expert knowledge. For example, a question about "how to return a product" may include specific instructions such as "bring proof of purchase and go to the store to complete the procedure."

[0885] 2. Training the generative AI model

[0886] The server uses the FAQ database and troubleshooting guide to train the generative AI model, for example, to respond appropriately to inquiries containing the keyword "returns."

[0887] 3. Providing a User Interface (UI)

[0888] The server provides a UI for users to enter their questions. This UI can be implemented as a web page or a mobile application. Once users log in, it provides a chat box and a search bar to make it easy for them to enter their questions.

[0889] 4. User Question Processing

[0890] The user uses the terminal to input a specific question, for example, "How do I return a product?", and the terminal transmits the user's question to the server in real time.

[0891] 5. Emotion Analysis

[0892] The server passes the received question to a sentiment analysis engine, which analyzes the content and tone of the question. The sentiment analysis engine infers the user's emotions, such as whether they are angry, confused, or in a hurry, from the context and wording. For example, if the question contains the phrase "I want to return it quickly," the sentiment analysis engine will determine that the user is in a hurry.

[0893] 6. Response generation using generative AI models

[0894] The generative AI model generates appropriate responses based on the emotional data provided by the sentiment analysis engine. For example, if it determines that the user is in a hurry, it generates a quick and specific response such as, "The return process is simple. Please follow the steps below immediately."

[0895] 7. Returning Responses

[0896] The server formats the generated response and sends it back to the user. The response created by the generative AI model is formatted into HTML or JSON format and sent to the terminal as an HTTP response.

[0897] 8. User Acknowledgment and Feedback

[0898] The user checks the answer displayed on the device. If the answer is helpful, they can enter feedback such as "This answer was very helpful." A feedback form is provided on the device so that users can easily submit their opinions.

[0899] 9. Save feedback and retrain

[0900] The server stores user feedback and uses it to retrain the generative AI model and sentiment analysis engine. The feedback data is stored in a database and used for the next training run, allowing the system to become more sophisticated over time and better meet user expectations.

[0901] Prompt Sentence Examples

[0902] Please tell me about the return procedure.

[0903] "My item hasn't arrived, what should I do?"

[0904] Please tell me about the product warranty period.

[0905] This allows the customer support system to respond quickly and appropriately to user questions, provide personalized service that takes user emotions into consideration, and increase customer satisfaction through continuous improvement through feedback.

[0906] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0907] Step 1:

[0908] The server creates an FAQ database and troubleshooting guide. The server collects past inquiry data and expert knowledge, and stores the data constructed based on that in the database. For example, if there are many inquiries about how to return a product, the server organizes and stores the specific procedures for those inquiries in the database.

[0909] Input: Historical inquiry data, expert knowledge

[0910] Output: FAQ database, troubleshooting guide

[0911] Step 2:

[0912] The server trains the generative AI model using the FAQ database and troubleshooting guide. The server collects information from both databases and trains the generative AI model to improve its ability to generate appropriate answers to user questions. For example, the server repeatedly trains a generative AI model (e.g., GPT-4) on data related to the keyword "returns."

[0913] Input: FAQ database, troubleshooting guide

[0914] Output: A trained generative AI model

[0915] Step 3:

[0916] The server provides a user interface (UI) for users to enter questions. This UI is implemented as a web page or mobile application, and allows users to easily enter questions. Specifically, it includes a chat box and a search bar.

[0917] Input: Web page and mobile application design information

[0918] Output: The implemented user interface

[0919] Step 4:

[0920] The user uses the terminal to enter a specific question, for example, "How do I return an item?" This question is sent to the server through the UI.

[0921] Input: User question

[0922] Output: Question data sent to the server

[0923] Step 5:

[0924] The server receives the question from the device and passes it to the emotion analysis engine. The emotion analysis engine analyzes the tone and context of the question to estimate the user's emotional state. For example, a question containing the phrase "I want to return it quickly" indicates a sense of urgency.

[0925] Input: User question data

[0926] Output: Emotion data

[0927] Step 6:

[0928] The generative AI model receives the emotion data and generates an appropriate response based on it. For example, if it determines that the user is in a hurry, it will generate a quick and specific response such as, "The return process is easy. Please follow the steps below immediately."

[0929] Input: Emotion data

[0930] Output: Response data

[0931] Step 7:

[0932] The server formats the generated response data and returns it to the user device. For example, it formats the response created by the generative AI model into HTML or JSON format and sends it to the device as an HTTP response.

[0933] Input: Response data

[0934] Output: Response sent back to the user's terminal

[0935] Step 8:

[0936] The user checks the answer displayed on the terminal. If the problem is solved and the answer is helpful, the user enters feedback such as "This answer was very helpful." This feedback is then sent back to the server.

[0937] Input: User feedback

[0938] Output: Feedback data sent to the server

[0939] Step 9:

[0940] The server stores the feedback data and uses it to retrain the generative AI model and sentiment analysis engine. The feedback data is stored in a database and used for the next training session.

[0941] Input: Feedback data

[0942] Output: Improved generative AI models and sentiment analysis engines

[0943] These steps enable the customer support system to respond quickly and appropriately to user questions, provide personalized service that takes user sentiment into account, and increase customer satisfaction through continuous improvement through feedback.

[0944] (Application example 2)

[0945] 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."

[0946] Conventional customer support systems are required to respond to user questions quickly and appropriately, but they lack the technology to respond in a way that takes the user's emotions into consideration. This increases the likelihood of users feeling stressed or dissatisfied. In addition, there are limited ways to improve the system based on feedback, making it difficult to improve the quality of support.

[0947] The identification processing by the identification 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 creating an FAQ database and a troubleshooting guide in advance, means for training a generative AI model to generate appropriate answers to user questions, means for providing a user interface for the user to input questions, means for passing questions input by the user to the generative AI model to generate answers, means for analyzing the user's emotions using a sentiment analysis engine, means for adjusting the response of the generative AI model based on the results of the sentiment analysis, and means for returning the generated answers to the user. This makes it possible to provide quick and appropriate answers to user questions and to provide personalized responses that take the user's emotions into consideration.

[0948] The "FAQ Database" is a database of frequently asked questions and their answers, built based on past inquiry data and troubleshooting guides.

[0949] A "Troubleshooting Guide" is a collection of procedures or guidelines that provide solutions to specific problems or malfunctions.

[0950] A "generative AI model" is a model trained using machine learning techniques to generate appropriate responses to a user's natural language input.

[0951] A "user interface" is an interface that allows users to access the system and input questions, and is provided as a web page or mobile application.

[0952] An "emotion analysis engine" is a technology that analyzes and identifies emotions from user comments and input content.

[0953] A "feedback system" is a mechanism for collecting evaluations and opinions from users and using them to improve and adjust the system.

[0954] The following procedures and system configuration are required to realize the invention.

[0955] System Configuration

[0956] The system of this invention mainly consists of a server, a terminal, and a user. The server has an FAQ database and a troubleshooting guide, and includes a generative AI model and a sentiment analysis engine. The terminal provides a user interface for users to input questions.

[0957] Program processing

[0958] Creating a FAQ database and troubleshooting guide

[0959] The server creates a FAQ database and troubleshooting guide based on past inquiry data and expert knowledge, covering frequently asked questions and solutions to problems.

[0960] Training generative AI models

[0961] The server uses the FAQ database and troubleshooting guide to train a generative AI model, which is then used to generate appropriate responses to natural language input from users.

[0962] Providing a user interface

[0963] The server provides a user interface implemented as a web page or mobile application, allowing users to easily input questions via their devices.

[0964] Parsing questions and generating answers

[0965] When a user inputs a question using a terminal, the question is sent to the server. The server passes the question to a sentiment analysis engine, which analyzes the user's sentiment. For example, if a user asks, "How do I return a product?", the sentiment analysis engine analyzes the tone and context of the user's speech and determines that the user is in a hurry.

[0966] Regulating responses based on emotions

[0967] The analyzed emotion data is passed to a generative AI model, which then generates an appropriate response based on that data. For rushed users, a quick and specific response such as "Returns are easy, just follow the steps below" is generated.

[0968] Returning answers and collecting feedback

[0969] The server formats the generated answer and sends it back to the user via the device. The user can review the answer displayed on the device and provide feedback if necessary. For example, they could say, "This answer was very helpful." This feedback is stored on the server and used to retrain the generative AI model and sentiment analysis engine.

[0970] Hardware and software used

[0971] The system uses the following hardware and software:

[0972] Server hardware: FAQ database, troubleshooting guide, generative AI model, sentiment analysis engine

[0973] Natural Language Toolkit (NLTK): A natural language processing library

[0974] TensorFlow / Keras: Training and inference of machine learning models

[0975] Emotion Recognition API: External service for emotion analysis

[0976] Flask: Building a server-side API

[0977] React Native: Frontend for smartphone applications

[0978] Specific examples

[0979] When a user asks "How do I return an item?" the system behaves as follows:

[0980] Example prompt sentence:

[0981] User Question: How do I return an item? Sentiment: I'm in a hurry

[0982] Based on this prompt, the generative AI model responds with, "The return process is simple. Please follow the steps below immediately." This allows for a fast, appropriate, and personalized response to the user's question.

[0983] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0984] Step 1:

[0985] A user opens a customer support application using a terminal and enters a question. The entered question is sent to the server by the terminal. The input of this step is the user's question text, and the output is the transmission of the question data to the server.

[0986] Step 2:

[0987] The server sends the received question to a sentiment analysis engine to analyze the user's sentiment. The sentiment analysis engine uses natural language processing technology (such as NLTK) to analyze the tone and context of the speech and identify the user's sentiment (e.g., hurry, irritated, calm). The input of this step is the question data, and the output is the user's sentiment data.

[0988] Step 3:

[0989] The emotion data analyzed by the emotion analysis engine is passed to the generative AI model by the server. The generative AI model generates the optimal response to the user's question based on the training dataset (FAQ database and troubleshooting guide). The input of this step is the question data and emotion data, and the output is the provisional response data.

[0990] Step 4:

[0991] The tentative responses generated by the generative AI model are adjusted based on emotional data to be formatted to suit the user's emotions. For example, if the user is determined to be in a hurry, the response will be adjusted to be brief and quick. The inputs of this step are tentative response data and emotional data, and the output is the final response data.

[0992] Step 5:

[0993] The final response data is sent back from the server to the terminal and displayed to the user. The user checks the displayed answer and provides feedback if necessary. The input of this step is the final response data, and the output is the response displayed on the user terminal and user feedback.

[0994] Step 6:

[0995] User feedback is sent from the device to the server, which collects and stores it. The collected feedback is used to retrain the generative AI model and sentiment analysis engine. The input of this step is user feedback, and the output is the storage of feedback data and model improvements.

[0996] These processing steps enable the system to generate fast, relevant answers to user questions, personalize responses by taking into account user sentiment, and incorporate feedback to continuously improve the overall system quality and accuracy.

[0997] 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.

[0998] 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.

[0999] 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.

[1000] [Fourth embodiment]

[1001] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1002] 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.

[1003] 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).

[1004] 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.

[1005] 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.

[1006] 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).

[1007] 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.

[1008] 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.

[1009] 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.

[1010] 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.

[1011] 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.

[1012] 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.

[1013] 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."

[1014] This invention is a customer support system that utilizes a generative AI model. This system creates an FAQ database and a troubleshooting guide in advance, and then trains the generative AI model based on them to generate appropriate answers to user questions.

[1015] First, the server creates a customer support FAQ database and troubleshooting guide, which are built based on past inquiry data and expert knowledge, covering frequently asked questions and solutions to problems.

[1016] The server then uses this FAQ database and troubleshooting guide to train a generative AI model designed to generate appropriate responses to natural language questions from users. The model uses automated learning algorithms to learn patterns from the dataset and generate optimal answers based on that knowledge.

[1017] When a user uses a device to input a question, the device sends the question to a server, which passes the question to a generative AI model, which then generates the optimal response to the question. The generated response is then sent back to the device via the server and displayed to the user.

[1018] For example, if a user asks, "How do I return an item?", the device sends this question to the server. The server passes this question to a generative AI model, which then generates a response such as, "To find out how to return an item, select the item you want to return from your order history and follow the return procedure from there. For detailed instructions, please see the link below." The server then sends this response back to the device, allowing the user to view the answer on the screen.

[1019] Furthermore, the server has the ability to collect user feedback and adjust and improve the generative AI model based on it. Based on the feedback provided by the user, the model's response accuracy can be periodically improved. When a user gives feedback such as "This answer was very helpful," the data is stored on the server and used for the next model training.

[1020] In this way, the customer support system of the present invention realizes quick and appropriate customer responses and improves customer satisfaction. Furthermore, by improving the response accuracy of the AI, it is possible to reduce training and education costs and provide efficient customer support.

[1021] The processing flow will be explained below.

[1022] Step 1:

[1023] The server creates a FAQ database and troubleshooting guide based on experts and past support data, organizing and storing frequently asked questions and their solutions.

[1024] Step 2:

[1025] The server trains a generative AI model based on the FAQ database and troubleshooting guide it creates. The model learns from these datasets to generate appropriate responses to natural language input from users.

[1026] Step 3:

[1027] The server provides a user interface (UI) for users to enter questions, which can be implemented as a web page or a mobile application.

[1028] Step 4:

[1029] A user accesses the customer support system through the UI and inputs a question, such as "How do I return a product?"

[1030] Step 5:

[1031] The device sends the user-entered question to the server as an HTTP request, which is sent in an appropriate format (e.g., JSON).

[1032] Step 6:

[1033] The server inputs the received question into a generative AI model, which analyzes the question and generates the best answer based on pre-trained data.

[1034] Step 7:

[1035] The server takes the generated answer and formats it in a way that is easy for the user to understand, for example as text with links and additional information.

[1036] Step 8:

[1037] The server returns a formatted response to the terminal, which then displays the response on the user's terminal.

[1038] Step 9:

[1039] The user checks the answer displayed on the terminal, and if the question is not resolved, the user can enter the question again.

[1040] Step 10:

[1041] Users provide feedback on the answers they receive from the system, such as rating the answer as "this answer was helpful" or "this answer was insufficient."

[1042] Step 11:

[1043] The server stores the collected feedback from users in a database, which is used for the next model training.

[1044] Step 12:

[1045] The server retrains the generative AI model with newly collected feedback data to improve the accuracy and coverage of answers, thus maintaining a cycle of system improvement.

[1046] Each step in the embodiment of the present invention is specifically performed in this manner to provide prompt and appropriate customer support to users.

[1047] Example 1

[1048] 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."

[1049] Existing customer support systems face challenges in providing prompt and appropriate responses to user inquiries. Furthermore, they lack a means to continuously improve the system based on user feedback, making it difficult to improve response accuracy. Furthermore, they are unable to provide consistent answers to the diverse problems users face, potentially leading to a decline in customer satisfaction.

[1050] 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.

[1051] In this invention, the server includes a means for creating an FAQ database and a troubleshooting guide in advance, a means for training a generative AI model to generate appropriate answers to user questions, a means for providing a user interface for users to input questions, and a means for collecting feedback and adjusting and improving the generative AI model based on the feedback, thereby enabling prompt and appropriate customer support, improving customer satisfaction, and achieving continuous improvement of the system.

[1052] An "FAQ database" is a database that is built based on past inquiry data and expert knowledge and centrally manages answers to frequently asked inquiries.

[1053] A "troubleshooting guide" is a systematic guide that describes in detail how to solve problems and issues that users may encounter.

[1054] A "generative AI model" is an artificial intelligence model trained to generate appropriate answers to user questions using natural language processing techniques.

[1055] A "user interface" is an interactive screen or operating means through which a user accesses a system and inputs questions.

[1056] "Feedback" refers to the evaluations and opinions that users provide regarding the system's responses, and this information is used to improve the system.

[1057] A "training dataset" is a collection of data used to train a generative AI model, collected from FAQ databases and troubleshooting guides.

[1058] A "terminal" is a device, such as a computer or smartphone, that a user uses to access the system.

[1059] The "server" is a central processing unit that receives questions from users and passes them to a generative AI model to generate answers.

[1060] This invention relates to a customer support system that utilizes a generative AI model. This system creates an FAQ database and a troubleshooting guide in advance, and then trains a generative AI model based on them to generate appropriate answers to user questions.

[1061] First, the server uses software and data management tools to create an FAQ database and troubleshooting guide for customer support. Specific hardware includes a database server or high-performance computer, and software includes a database management system such as MySQL or PostgreSQL. This database and guide are built based on past inquiry data and expert knowledge, and cover frequently asked questions and solutions to problems.

[1062] The server then uses the FAQ database and troubleshooting guide to train a generative AI model. The generative AI model is created using libraries such as TensorFlow and PyTorch. It is designed to use an automatic learning algorithm to generate appropriate responses to natural language questions from users. Specifically, question and answer pairs are extracted from the FAQ database and troubleshooting guide and provided to the generative AI model as a training dataset.

[1063] When a user inputs a question using a device, the question is sent from the device to a server. The device is assumed to be a PC or smartphone, and a dedicated user interface is provided for it. For example, if a user asks, "How do I return a product?", the question is sent from the device to the server. The server passes the received question to a generative AI model, and the model generates an optimal response. The generated response is then sent back to the device via the server and displayed to the user.

[1064] An example of a generated response might be, "To find out how to return an item, please select the item you wish to return from your order history and follow the return procedure from there. For detailed instructions, please see the link below." This response responds quickly and accurately to the information the user is seeking.

[1065] Furthermore, the server has the ability to collect user feedback and adjust and improve the generative AI model based on that feedback. When a user provides feedback such as "This answer was very helpful," that data is stored on the server and used for the next model training. Collecting this feedback and retraining the model improves the response accuracy of the generative AI model and improves the performance of the system as a whole.

[1066] Examples of prompts include:

[1067] Please tell me how to return the product

[1068] "I haven't received the product I ordered, what should I do?"

[1069] Please tell me how to change my payment method.

[1070] The system aims to provide efficient and consistent customer support and improve customer satisfaction. In addition, the improved accuracy of AI responses will reduce training and education costs, resulting in more efficient support.

[1071] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1072] Step 1: Creating a FAQ database and troubleshooting guide

[1073] The server collects past inquiry data and expert knowledge, and uses them to create an FAQ database and troubleshooting guide. Specifically, it uses a database management system (e.g., MySQL, PostgreSQL) to categorize, organize, and store the collected data.

[1074] Input: past inquiry data, expert knowledge

[1075] Output: FAQ database, troubleshooting guide

[1076] Step 2: Training the generative AI model

[1077] The server trains the generative AI model using data from the FAQ database and troubleshooting guides. Specifically, it prepares a training dataset and builds and trains the model using libraries such as TensorFlow and PyTorch.

[1078] Input: FAQ database, troubleshooting guide

[1079] Output: A trained generative AI model

[1080] Step 3: Accepting user questions

[1081] A user inputs a question using a terminal, which then sends the question to a server. Specifically, a question is input in natural language through a user interface and then sent to the server via a network.

[1082] Input: User question (natural language)

[1083] Output: Question data (received on the server side)

[1084] Step 4: Question processing and response generation

[1085] The server passes the received question to a generative AI model, which then generates an appropriate response to the question. Specifically, the question data is input into the AI ​​model, and the model generates the optimal response based on the training data.

[1086] Input: Question data

[1087] Output: The generated response

[1088] Step 5: Returning a response

[1089] The server sends the generated response back to the terminal, which then displays the received response to the user. Specifically, the response data is sent to the terminal via the network, and the user can check the response on the screen.

[1090] Input: Generated response

[1091] Output: The response that is displayed to the user

[1092] Step 6: Gather feedback

[1093] The user inputs the feedback they provide on the terminal, and the terminal transmits the feedback to the server. Specifically, the user inputs their thoughts and evaluations on the response, which are then stored on the server.

[1094] Input: User feedback

[1095] Output: Feedback data stored on the server

[1096] Step 7: Adjust and improve the model

[1097] The server periodically retrains the generative AI model based on the collected feedback to improve response accuracy. Specifically, it adds the feedback data to the training data and retrains the generative AI model.

[1098] Input: Feedback data

[1099] Output: An improved generative AI model

[1100] (Application example 1)

[1101] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1102] Conventional customer support systems often struggle to respond to user inquiries quickly and accurately. They also lack the functionality to efficiently collect user feedback and improve the accuracy of their models. As a result, problems such as low user satisfaction and inefficient support operations arise.

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

[1104] In this invention, the server includes means for creating an FAQ database and a troubleshooting guide in advance, means for training a generative AI model to generate appropriate answers to user questions, means for providing a user interface for users to input questions, means for passing questions input by users to the generative AI model to generate answers, means for returning the generated answers to the users, and means for collecting user feedback and improving the accuracy of the generative AI model, thereby enabling faster responses to user inquiries, providing accurate responses, and improving the model based on the feedback.

[1105] An "FAQ database" is an information resource that systematically organizes frequently asked questions and their answers.

[1106] A "troubleshooting guide" is a guide that specifically describes problems that users often encounter and how to solve them.

[1107] A "generative AI model" is an artificial intelligence model that generates appropriate answers to questions posed in natural language by users.

[1108] "User interface" refers to the screen and operating methods that users use to input information into a system.

[1109] "Feedback" refers to opinions and ratings provided by users about the quality and usability of an answer.

[1110] The present invention is a customer support system that uses a FAQ database and troubleshooting guides to respond to user questions. This system is intended for use particularly in online shopping sites, and generates real-time responses to user questions via smartphones.

[1111] The server first creates a database of FAQs and troubleshooting guides, then uses them to train a generative AI model, which uses an automated learning algorithm, such as OpenAI's GPT-4, to learn patterns from the dataset and collects user feedback, which it then uses to improve the model's accuracy.

[1112] When a user inputs a question from their smartphone, the device sends the question to a server. The server passes the received question to a generative AI model, which generates an appropriate answer. The generated answer is then sent back to the smartphone via the server and displayed to the user. This allows the user to receive a quick and accurate answer.

[1113] As a specific use case, if a user asks, "How do I return an item?", the device sends this question to the server. The server passes this question to the generative AI model, which then generates a response such as, "Select the item you want to return from your order history and complete the return procedure. Click here for details." The server then sends this response back to the device, allowing the user to view the answer on their smartphone screen.

[1114] User feedback is collected in the form of, for example, "This answer was very helpful." This feedback is then stored on the server and used for the next training. This allows the generative AI model to regularly improve its response accuracy, enabling the provision of higher quality customer support.

[1115] The specific hardware used includes smartphones as user devices, high-performance computers as servers, and GPT-4 as a generative AI model, while the software used includes various APIs, cloud services, and database management systems.

[1116] Example prompt sentence:

[1117] When a user asks, "How do I return an item?", respond with: "Select the item you want to return from your order history and follow the return process. Click this link for more information."

[1118] As a result, this customer support system can provide efficient and satisfying service and significantly reduce the support work load of companies.

[1119] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1120] Step 1:

[1121] The user inputs a question through the smartphone's user interface, for example, "How do I return a product?"

[1122] Input: User question

[1123] Output: Data of the questions entered on the smartphone

[1124] Step 2:

[1125] The device converts the questions entered by the user into data format and sends it to the server. This data is structured in JSON format or similar.

[1126] Input: Questions typed into the user's smartphone

[1127] Output: Structured question data (e.g., JSON format)

[1128] Step 3:

[1129] The server passes the received question to the generative AI model. Specifically, it converts the question data into a format that the generative AI model can process and inputs it into the model.

[1130] Input: Structured question data

[1131] Output: Input data prepared for a generative AI model

[1132] Step 4:

[1133] A generative AI model (e.g., GPT-4) analyzes the provided question data and generates the optimal answer by referencing a FAQ database and troubleshooting guide.

[1134] Input: Input data for the generative AI model

[1135] Output: The generated answer

[1136] Step 5:

[1137] The server receives the answers from the generative AI model and converts them into a data format to send back to the user.

[1138] Input: Answer from generative AI model

[1139] Output: Structured response data

[1140] Step 6:

[1141] The terminal receives the answer data from the server and displays it on the screen, allowing the user to check the answer to the question.

[1142] Input: Response data from the server

[1143] Output: Answer displayed on smartphone screen

[1144] Step 7:

[1145] The user can input feedback about the displayed answer and send it to the server via the terminal. For example, the user can give feedback such as "This answer was very helpful."

[1146] Input: User feedback

[1147] Output: Feedback data sent from the device to the server

[1148] Step 8:

[1149] The server collects and stores user feedback to add to the training dataset of the generative AI model, which will improve the model's accuracy the next time it is trained.

[1150] Input: Feedback data from the device

[1151] Output: Feedback added to the training dataset

[1152] 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.

[1153] This invention is a customer support system that utilizes a generative AI model and further combines it with an emotion engine that recognizes user emotions. This system creates a FAQ database and troubleshooting guide in advance, and then trains the generative AI model based on these to generate appropriate answers to user questions. The emotion engine also analyzes user emotions and adjusts the generative AI model's responses to provide more appropriate customer support.

[1154] First, the server creates a customer support FAQ database and troubleshooting guide, which are built based on past inquiry data and expert knowledge, covering frequently asked questions and solutions to problems.

[1155] The server then uses this FAQ database and troubleshooting guide to train a generative AI model, which learns from these datasets to generate appropriate responses to natural language input from users.

[1156] Furthermore, the server provides a user interface (UI) for users to input their questions, which can be implemented as a web page or a mobile application and easily accessible to users.

[1157] The new system incorporates an emotion engine, which can analyze user emotions and apply responses. When a user uses a device to input a question, the device sends the question to a server. The server then passes the received question to the emotion engine, which analyzes the user's emotions. The analyzed emotion data is then passed to a generative AI model, which then generates an appropriate response based on that data.

[1158] For example, if a user asks, "How do I return an item?", the device sends this question to the server. At the same time, the emotion engine analyzes the tone and context of the user's speech and, if it determines that the user is in a hurry, the generative AI model generates a quick and specific response such as, "The return process is simple. Please follow the steps below immediately."

[1159] The server then formats the generated answer and sends it back to the user, who can review the answer displayed on their device and provide feedback if needed, such as "This answer was very helpful." This feedback is stored on the server and used to retrain the generative AI model and emotion engine.

[1160] In this way, the customer support system of the present invention combines a generative AI model and an emotion engine to provide fast and appropriate customer service and improve customer satisfaction. Furthermore, by recognizing the user's emotions and adjusting responses accordingly, it is possible to provide more personalized support. The above is a specific embodiment for implementing the present invention.

[1161] The processing flow will be explained below.

[1162] Step 1:

[1163] The server uses experts and past support data to create a FAQ database and troubleshooting guides that organize and store frequently asked questions and their solutions.

[1164] Step 2:

[1165] The server uses the created FAQ database and troubleshooting guide to train a generative AI model, which is then able to learn patterns from this data and generate appropriate responses to users' natural language input.

[1166] Step 3:

[1167] The server provides a user interface (UI) for users to enter their questions, implemented as a web page or mobile application, making it easy for users to access and use.

[1168] Step 4:

[1169] The user uses a terminal to access the provided UI and input a question into the customer support system, for example, "How do I return a product?"

[1170] Step 5:

[1171] The device sends the user-entered question to the server as an HTTP request, which is sent in an appropriate format (e.g., JSON).

[1172] Step 6:

[1173] The server passes the received question to the emotion engine, which analyzes the emotion from the user's input text and passes the analysis results to the generative AI model.

[1174] Step 7:

[1175] The emotion engine analyzes emotional data from user input, for example classifying emotions from user text as "positive," "negative," or "neutral." This data is then used to adjust the generative AI model's response.

[1176] Step 8:

[1177] The server inputs the emotion data obtained from the emotion engine into the generative AI model, which then references this data to generate a response that corresponds to the user's emotion.

[1178] Step 9:

[1179] The server takes the generated answer and formats it in a format that is easy for the user to understand, for example as text with links and additional information.

[1180] Step 10:

[1181] The server returns a formatted response to the terminal, which then displays the response on the user's terminal.

[1182] Step 11:

[1183] The user checks the answer displayed on the terminal, and if the answer is not what they expected, they can enter the question again.

[1184] Step 12:

[1185] Users provide feedback on the answers they receive from the system, such as rating the answer as "this answer was helpful" or "this answer was insufficient."

[1186] Step 13:

[1187] The server stores the collected feedback from users in a database, which is used for the next retraining of the model and emotion engine.

[1188] Step 14:

[1189] The server uses newly collected feedback data to retrain the generative AI model and emotion engine to improve the accuracy and coverage of answers, maintaining a cycle of system improvement.

[1190] Through this series of steps, the present invention combines a generative AI model and an emotion engine to provide fast and appropriate customer support, thereby increasing user satisfaction and realizing efficient support.

[1191] Example 2

[1192] 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."

[1193] Conventional customer support systems have difficulty in providing appropriate answers to user questions quickly, and are unable to generate responses that take into account the user's emotions. Therefore, in order to improve user satisfaction, it is necessary to provide personalized support that takes into account the user's emotions.

[1194] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1195] In this invention, the server includes means for creating an FAQ database and a troubleshooting guide, means for training a generative AI model to generate appropriate answers to user questions, means for providing a user interface for users to input questions, means for passing the questions input by the users to the generative AI model and a sentiment analysis engine to generate answers, and means for returning the generated answers to the users, thereby enabling the provision of prompt and appropriate answers that take into consideration the user's emotions.

[1196] An "FAQ database" is a collection of information that organizes and stores answers to frequently asked questions from users.

[1197] A "troubleshooting guide" is a manual that systematically describes solutions to problems and issues that users may encounter.

[1198] A "generative AI model" is an artificial intelligence model trained to generate appropriate responses to natural language input from a user.

[1199] A "user interface" is an interactive screen or device that allows a user to enter questions into a system and receive answers from the system.

[1200] An "emotion analysis engine" is a technology that analyzes language data entered by a user and estimates the user's emotional state.

[1201] "Feedback" refers to a user inputting an evaluation or opinion on the answer received from the system.

[1202] "Server" means the computer system that runs the entire customer support system, storing and processing data, training, generating responses, etc.

[1203] This invention is a customer support system that utilizes a generative AI model and further combines it with a sentiment analysis engine that recognizes user emotions. This system creates a FAQ database and troubleshooting guide in advance, and then trains the generative AI model based on these to generate appropriate answers to user questions. The sentiment analysis engine also analyzes user emotions and adjusts the generative AI model's responses to provide more appropriate customer support.

[1204] Hardware and software used

[1205] Server: A platform that creates FAQ databases and troubleshooting guides for customer support, and runs generative AI models and sentiment analysis engines.

[1206] Device: The device (e.g., computer, smartphone) used by the user to enter questions and exchange information with the server.

[1207] Generative AI model: An AI model that generates optimal answers to user questions. Examples include OpenAI's GPT-3 and GPT-4.

[1208] Sentiment analysis engine: An engine for analyzing user sentiment and adjusting the generated response. Example: IBM's Watson Tone Analyzer.

[1209] Specific examples

[1210] 1. Creating a FAQ database and troubleshooting guide

[1211] The server creates a FAQ database and troubleshooting guide for customer support. This database is built based on past inquiry data and expert knowledge. For example, a question about "how to return a product" may include specific instructions such as "bring proof of purchase and go to the store to complete the procedure."

[1212] 2. Training the generative AI model

[1213] The server uses the FAQ database and troubleshooting guide to train the generative AI model, for example, to respond appropriately to inquiries containing the keyword "returns."

[1214] 3. Providing a User Interface (UI)

[1215] The server provides a UI for users to enter their questions. This UI can be implemented as a web page or a mobile application. Once users log in, it provides a chat box and a search bar to make it easy for them to enter their questions.

[1216] 4. User Question Processing

[1217] The user uses the terminal to input a specific question, for example, "How do I return a product?", and the terminal transmits the user's question to the server in real time.

[1218] 5. Emotion Analysis

[1219] The server passes the received question to a sentiment analysis engine, which analyzes the content and tone of the question. The sentiment analysis engine infers the user's emotions, such as whether they are angry, confused, or in a hurry, from the context and wording. For example, if the question contains the phrase "I want to return it quickly," the sentiment analysis engine will determine that the user is in a hurry.

[1220] 6. Response generation using generative AI models

[1221] The generative AI model generates appropriate responses based on the emotional data provided by the sentiment analysis engine. For example, if it determines that the user is in a hurry, it generates a quick and specific response such as, "The return process is simple. Please follow the steps below immediately."

[1222] 7. Returning Responses

[1223] The server formats the generated response and sends it back to the user. The response created by the generative AI model is formatted into HTML or JSON format and sent to the terminal as an HTTP response.

[1224] 8. User Acknowledgment and Feedback

[1225] The user checks the answer displayed on the device. If the answer is helpful, they can enter feedback such as "This answer was very helpful." A feedback form is provided on the device so that users can easily submit their opinions.

[1226] 9. Save feedback and retrain

[1227] The server stores user feedback and uses it to retrain the generative AI model and sentiment analysis engine. The feedback data is stored in a database and used for the next training run, allowing the system to become more sophisticated over time and better meet user expectations.

[1228] Prompt Sentence Examples

[1229] Please tell me about the return procedure.

[1230] "My item hasn't arrived, what should I do?"

[1231] Please tell me about the product warranty period.

[1232] This allows the customer support system to respond quickly and appropriately to user questions, provide personalized service that takes user emotions into consideration, and increase customer satisfaction through continuous improvement through feedback.

[1233] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1234] Step 1:

[1235] The server creates an FAQ database and troubleshooting guide. The server collects past inquiry data and expert knowledge, and stores the data constructed based on that in the database. For example, if there are many inquiries about how to return a product, the server organizes and stores the specific procedures for those inquiries in the database.

[1236] Input: Historical inquiry data, expert knowledge

[1237] Output: FAQ database, troubleshooting guide

[1238] Step 2:

[1239] The server trains the generative AI model using the FAQ database and troubleshooting guide. The server collects information from both databases and trains the generative AI model to improve its ability to generate appropriate answers to user questions. For example, the server repeatedly trains a generative AI model (e.g., GPT-4) on data related to the keyword "returns."

[1240] Input: FAQ database, troubleshooting guide

[1241] Output: A trained generative AI model

[1242] Step 3:

[1243] The server provides a user interface (UI) for users to enter questions. This UI is implemented as a web page or mobile application, and allows users to easily enter questions. Specifically, it includes a chat box and a search bar.

[1244] Input: Web page and mobile application design information

[1245] Output: The implemented user interface

[1246] Step 4:

[1247] The user uses the terminal to enter a specific question, for example, "How do I return an item?" This question is sent to the server through the UI.

[1248] Input: User question

[1249] Output: Question data sent to the server

[1250] Step 5:

[1251] The server receives the question from the device and passes it to the emotion analysis engine. The emotion analysis engine analyzes the tone and context of the question to estimate the user's emotional state. For example, a question containing the phrase "I want to return it quickly" indicates a sense of urgency.

[1252] Input: User question data

[1253] Output: Emotion data

[1254] Step 6:

[1255] The generative AI model receives the emotion data and generates an appropriate response based on it. For example, if it determines that the user is in a hurry, it will generate a quick and specific response such as, "The return process is easy. Please follow the steps below immediately."

[1256] Input: Emotion data

[1257] Output: Response data

[1258] Step 7:

[1259] The server formats the generated response data and returns it to the user device. For example, it formats the response created by the generative AI model into HTML or JSON format and sends it to the device as an HTTP response.

[1260] Input: Response data

[1261] Output: Response sent back to the user's terminal

[1262] Step 8:

[1263] The user checks the answer displayed on the terminal. If the problem is solved and the answer is helpful, the user enters feedback such as "This answer was very helpful." This feedback is then sent back to the server.

[1264] Input: User feedback

[1265] Output: Feedback data sent to the server

[1266] Step 9:

[1267] The server stores the feedback data and uses it to retrain the generative AI model and sentiment analysis engine. The feedback data is stored in a database and used for the next training session.

[1268] Input: Feedback data

[1269] Output: Improved generative AI models and sentiment analysis engines

[1270] These steps enable the customer support system to respond quickly and appropriately to user questions, provide personalized service that takes user sentiment into account, and increase customer satisfaction through continuous improvement through feedback.

[1271] (Application example 2)

[1272] 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."

[1273] Conventional customer support systems are required to respond to user questions quickly and appropriately, but they lack the technology to respond in a way that takes the user's emotions into consideration. This increases the likelihood of users feeling stressed or dissatisfied. In addition, there are limited ways to improve the system based on feedback, making it difficult to improve the quality of support.

[1274] The identification processing by the identification 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 creating an FAQ database and a troubleshooting guide in advance, means for training a generative AI model to generate appropriate answers to user questions, means for providing a user interface for the user to input questions, means for passing questions input by the user to the generative AI model to generate answers, means for analyzing the user's emotions using a sentiment analysis engine, means for adjusting the response of the generative AI model based on the results of the sentiment analysis, and means for returning the generated answers to the user. This makes it possible to provide quick and appropriate answers to user questions and to provide personalized responses that take the user's emotions into consideration.

[1275] The "FAQ Database" is a database of frequently asked questions and their answers, built based on past inquiry data and troubleshooting guides.

[1276] A "Troubleshooting Guide" is a collection of procedures or guidelines that provide solutions to specific problems or malfunctions.

[1277] A "generative AI model" is a model trained using machine learning techniques to generate appropriate responses to a user's natural language input.

[1278] A "user interface" is an interface that allows users to access the system and input questions, and is provided as a web page or mobile application.

[1279] An "emotion analysis engine" is a technology that analyzes and identifies emotions from user comments and input content.

[1280] A "feedback system" is a mechanism for collecting evaluations and opinions from users and using them to improve and adjust the system.

[1281] The following procedures and system configuration are required to realize the invention.

[1282] System Configuration

[1283] The system of this invention mainly consists of a server, a terminal, and a user. The server has an FAQ database and a troubleshooting guide, and includes a generative AI model and a sentiment analysis engine. The terminal provides a user interface for users to input questions.

[1284] Program processing

[1285] Creating a FAQ database and troubleshooting guide

[1286] The server creates a FAQ database and troubleshooting guide based on past inquiry data and expert knowledge, covering frequently asked questions and solutions to problems.

[1287] Training generative AI models

[1288] The server uses the FAQ database and troubleshooting guide to train a generative AI model, which is then used to generate appropriate responses to natural language input from users.

[1289] Providing a user interface

[1290] The server provides a user interface implemented as a web page or mobile application, allowing users to easily input questions via their devices.

[1291] Parsing questions and generating answers

[1292] When a user inputs a question using a terminal, the question is sent to the server. The server passes the question to a sentiment analysis engine, which analyzes the user's sentiment. For example, if a user asks, "How do I return a product?", the sentiment analysis engine analyzes the tone and context of the user's speech and determines that the user is in a hurry.

[1293] Regulating responses based on emotions

[1294] The analyzed emotion data is passed to a generative AI model, which then generates an appropriate response based on that data. For rushed users, a quick and specific response such as "Returns are easy, just follow the steps below" is generated.

[1295] Returning answers and collecting feedback

[1296] The server formats the generated answer and sends it back to the user via the device. The user can review the answer displayed on the device and provide feedback if necessary. For example, they could say, "This answer was very helpful." This feedback is stored on the server and used to retrain the generative AI model and sentiment analysis engine.

[1297] Hardware and software used

[1298] The system uses the following hardware and software:

[1299] Server hardware: FAQ database, troubleshooting guide, generative AI model, sentiment analysis engine

[1300] Natural Language Toolkit (NLTK): A natural language processing library

[1301] TensorFlow / Keras: Training and inference of machine learning models

[1302] Emotion Recognition API: External service for emotion analysis

[1303] Flask: Building a server-side API

[1304] React Native: Frontend for smartphone applications

[1305] Specific examples

[1306] When a user asks "How do I return an item?" the system behaves as follows:

[1307] Example prompt sentence:

[1308] User Question: How do I return an item? Sentiment: I'm in a hurry

[1309] Based on this prompt, the generative AI model responds with, "The return process is simple. Please follow the steps below immediately." This allows for a fast, appropriate, and personalized response to the user's question.

[1310] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1311] Step 1:

[1312] A user opens a customer support application using a terminal and enters a question. The entered question is sent to the server by the terminal. The input of this step is the user's question text, and the output is the transmission of the question data to the server.

[1313] Step 2:

[1314] The server sends the received question to a sentiment analysis engine to analyze the user's sentiment. The sentiment analysis engine uses natural language processing technology (such as NLTK) to analyze the tone and context of the speech and identify the user's sentiment (e.g., hurry, irritated, calm). The input of this step is the question data, and the output is the user's sentiment data.

[1315] Step 3:

[1316] The emotion data analyzed by the emotion analysis engine is passed to the generative AI model by the server. The generative AI model generates the optimal response to the user's question based on the training dataset (FAQ database and troubleshooting guide). The input of this step is the question data and emotion data, and the output is the provisional response data.

[1317] Step 4:

[1318] The tentative responses generated by the generative AI model are adjusted based on emotional data to be formatted to suit the user's emotions. For example, if the user is determined to be in a hurry, the response will be adjusted to be brief and quick. The inputs of this step are tentative response data and emotional data, and the output is the final response data.

[1319] Step 5:

[1320] The final response data is sent back from the server to the terminal and displayed to the user. The user checks the displayed answer and provides feedback if necessary. The input of this step is the final response data, and the output is the response displayed on the user terminal and user feedback.

[1321] Step 6:

[1322] User feedback is sent from the device to the server, which collects and stores it. The collected feedback is used to retrain the generative AI model and sentiment analysis engine. The input of this step is user feedback, and the output is the storage of feedback data and model improvements.

[1323] These processing steps enable the system to generate fast, relevant answers to user questions, personalize responses by taking into account user sentiment, and incorporate feedback to continuously improve the overall system quality and accuracy.

[1324] 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.

[1325] 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.

[1326] 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.

[1327] 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.

[1328] 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.

[1329] 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.

[1330] 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).

[1331] 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.

[1332] 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."

[1333] 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.

[1334] 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).

[1335] 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.

[1336] 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.

[1337] 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.

[1338] 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.

[1339] 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.

[1340] 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.

[1341] 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.

[1342] 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.

[1343] 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.

[1344] 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.

[1345] The following is further disclosed regarding the above embodiment.

[1346] (Claim 1)

[1347] A means of pre-creating a FAQ database and troubleshooting guide;

[1348] a means for training a generative AI model to generate appropriate answers to user questions;

[1349] means for providing a user interface for a user to input a question;

[1350] A means for passing a question input by a user to a generative AI model to generate an answer;

[1351] means for returning the generated answer to the user;

[1352] A system including:

[1353] (Claim 2)

[1354] 10. The system of claim 1, further comprising means for collecting user feedback to adjust and improve the generative AI model.

[1355] (Claim 3)

[1356] 10. The system of claim 1, wherein the generative AI model uses an FAQ database and a troubleshooting guide as a training data set.

[1357] "Example 1"

[1358] (Claim 1)

[1359] A means of pre-creating a FAQ database and troubleshooting guide;

[1360] a means for training a generative AI model to generate appropriate answers to user questions;

[1361] means for providing a user interface for a user to input a question;

[1362] A means for passing a question input by a user to a generative AI model to generate an answer;

[1363] means for returning the generated answer to the user;

[1364] A means to collect feedback and use it to adjust and improve the generative AI model;

[1365] A system including:

[1366] (Claim 2)

[1367] 10. The system of claim 1, wherein the generative AI model uses an FAQ database and a troubleshooting guide as a training data set.

[1368] (Claim 3)

[1369] 2. The system of claim 1, further comprising means for a user to input a question using a terminal and for the terminal to transmit the question to a server.

[1370] "Application Example 1"

[1371] (Claim 1)

[1372] A means of pre-creating a FAQ database and troubleshooting guide;

[1373] a means for training a generative AI model to generate appropriate answers to user questions;

[1374] means for providing a user interface for a user to input a question;

[1375] A means for passing a question input by a user to a generative AI model to generate an answer;

[1376] means for returning the generated answer to the user;

[1377] A means to collect user feedback and improve the accuracy of the generative AI model; and

[1378] A system including:

[1379] (Claim 2)

[1380] The system described in claim 1, characterized in that the system is installed on a smartphone and has the function of the generative AI model responding in real time to questions entered by the user.

[1381] (Claim 3)

[1382] 10. The system of claim 1, wherein the generative AI model uses an FAQ database and a troubleshooting guide as a training data set and generates optimal answers based on the training.

[1383] "Example 2: Combining Emotion Engines"

[1384] (Claim 1)

[1385] A means of creating a FAQ database and troubleshooting guide;

[1386] a means for training a generative AI model to generate appropriate answers to user questions;

[1387] means for providing a user interface for a user to input a question;

[1388] A means of generating answers by passing questions entered by users to a generative AI model and a sentiment analysis engine;

[1389] means for returning the generated answer to the user;

[1390] A system including:

[1391] (Claim 2)

[1392] 10. The system of claim 1, further comprising means for analyzing a user's emotion and adjusting the response of the generative AI model based on the emotion.

[1393] (Claim 3)

[1394] 10. The system of claim 1, further comprising means for collecting user feedback to adjust and improve the generative AI model and sentiment analysis engine.

[1395] (Claim 4)

[1396] 10. The system of claim 1, wherein the generative AI model uses an FAQ database and a troubleshooting guide as a training data set.

[1397] "Application example 2 when combining emotion engines"

[1398] (Claim 1)

[1399] A means of pre-creating a FAQ database and troubleshooting guide;

[1400] a means for training a generative AI model to generate appropriate answers to user questions;

[1401] means for providing a user interface for a user to input a question;

[1402] A means for passing a question input by a user to a generative AI model to generate an answer;

[1403] means for analyzing user emotions using a sentiment analysis engine;

[1404] a means for adjusting the response of the generative AI model based on the results of the sentiment analysis;

[1405] means for returning the generated answer to the user;

[1406] A system including:

[1407] (Claim 2)

[1408] 10. The system of claim 1, further comprising means for collecting user feedback to adjust and improve the generative AI model and sentiment analysis engine.

[1409] (Claim 3)

[1410] 10. The system of claim 1, wherein the generative AI model uses an FAQ database and a troubleshooting guide as a training data set, and the sentiment analysis engine analyzes user sentiment. [Explanation of symbols]

[1411] 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 of pre-creating a FAQ database and troubleshooting guide; a means for training a generative AI model to generate appropriate answers to user questions; means for providing a user interface for a user to input a question; A means for passing a question input by a user to a generative AI model to generate an answer; means for returning the generated answer to the user; A system including:

2. 10. The system of claim 1, further comprising means for collecting user feedback to adjust and improve the generative AI model.

3. 10. The system of claim 1, wherein the generative AI model uses an FAQ database and a troubleshooting guide as a training data set.

Images