System

Abstract

A system is provided.SOLUTION: A means for collecting case data in the past, a means for preprocessing the collected case data, a means for learning a generated AI model using the preprocessed data, a means for providing an interface through which a user can input a query, and a means for responding to the query of the user by: A system comprising: means for generating an optimal answer using the generative AI model; means for providing the generated answer to a user; means for collecting feedback from the user; means for retraining the generative AI model using the collected feedback; means for automatically generating an instructional manual; and means for providing the generated instructional manual.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] In customer service and call center operations, sufficient training and information sharing are necessary for employees to respond appropriately and quickly. However, it is difficult for all employees to receive uniform training in their daily work, which results in inconsistencies in the quality of responses. In addition, training new employees takes time and effort, and there are large differences in the responses of veteran employees. Furthermore, in order to improve the quality of responses to complaints while streamlining operations, it is necessary to automate appropriate response methods and efficiently train employees. A system that solves these issues is needed. [Means for solving the problem]

[0005] The present invention provides a means for collecting and preprocessing past case data, thereby cleansing and structuring the data. It also provides a means for training a generative AI model using the preprocessed data. This training enables past response patterns to be modeled and optimal responses to user inquiries to be generated. It also includes a means for providing an interface through which users can input inquiries and for providing the generated responses to the user. It also provides a means for collecting feedback from users and retraining the generative AI model based on that feedback. It also includes a means for automatically generating and providing training manuals based on the collected data and model output. This system enables employees to respond to complaints quickly and appropriately, and is also useful for training new employees and improving the skills of current employees, thereby improving the efficiency of customer service and call center operations and improving the quality of responses.

[0006] 1. Collection

[0007] "Collection" is the act of systematically and methodically gathering specific information or data.

[0008] 2. Pretreatment

[0009] "Preprocessing" is the process of data cleansing and structuring collected data to convert it into a form suitable for analysis and learning.

[0010] 3. Generative AI Models

[0011] A "generative AI model" is an artificial intelligence model that is trained using machine learning algorithms to make predictions or suggestions for specific problems.

[0012] 4. Interface

[0013] "Interface" refers to the user-friendly screens and input forms that allow users to interact with the system.

[0014] 5. Inquiries

[0015] "Query" means a question or request that a User enters into the System seeking specific information.

[0016] 6. Answer

[0017] "Response" refers to the answer or solution provided by the system to a user's inquiry.

[0018] 7. Feedback

[0019] "Feedback" refers to information that informs users of the appropriateness and areas for improvement of the answers they have provided.

[0020] 8. Retraining

[0021] "Retraining" is the process of retraining an existing AI model using new data and feedback.

[0022] 9. Educational Manual

[0023] An "educational manual" is an instructional material or guideline aimed at acquiring a specific job or skill.

[0024] 10. Offer

[0025] "Providing" refers to the act of supplying specific information or services to others. [Brief explanation of the drawings]

[0026] [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

[0027] 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.

[0028] First, the terms used in the following description will be explained.

[0029] 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).

[0030] 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.

[0031] 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.

[0032] 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.

[0033] 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."

[0034] [First embodiment]

[0035] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0036] 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.

[0037] 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).

[0038] 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.

[0039] 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.

[0040] 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.

[0041] 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.

[0042] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0043] 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.

[0044] 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.

[0045] 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.

[0046] 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."

[0047] MODE FOR CARRYING OUT THE INVENTION

[0048] An embodiment of the present invention will be described below.

[0049] This system is designed to improve the efficiency and quality of customer service and complaint handling in call center operations. It collects and preprocesses past case data, generates and trains AI models, provides a user interface, handles user inquiries, collects feedback and retrains, and automatically generates training manuals.

[0050] Data collection and preprocessing

[0051] First, the server collects past case data collected from call centers and customer support operations. For example, it obtains historical data on responses to complaints such as "the product arrived damaged." Next, the server preprocesses this data. In preprocessing, unnecessary data is cleansed and important data is extracted and structured. In particular, text data is tokenized, and important keywords are extracted.

[0052] AI model generation and training

[0053] Using the preprocessed data, the server builds and trains a generative AI model. This is the process of using machine learning algorithms to train the model and learn from past response patterns. For example, it can learn the appropriate way to respond to complaints and predict the optimal response.

[0054] Providing a user interface

[0055] Next, the server designs and implements a user interface, which is provided via a terminal. This interface is designed to allow users to easily access the system and input inquiries. For example, a customer service representative can ask the system, "What should we do if a customer's order is delayed?"

[0056] Responding to user inquiries

[0057] When a user inputs a query into the system, the server uses the generative AI model to generate the optimal response to that query. For example, the server suggests "how to respond to a delayed order" by suggesting "sending an apology email and offering a discount on the next order." The device then displays this response to the user.

[0058] Gathering feedback and retraining

[0059] Users can provide feedback on the appropriateness of the suggested answers, which the server collects and uses to retrain the AI ​​model, a process that improves the model's accuracy and allows it to better respond to future inquiries.

[0060] Automatic generation of training manuals

[0061] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated and provided to the terminal in PDF format.

[0062] Through the above process, this system will improve the quality of customer service and call center operations and increase operational efficiency.

[0063] The processing flow will be explained below.

[0064] Step 1: Collecting interaction data

[0065] The server collects past case data from call centers and customer support systems, specifically data including complaint details, response methods, and customer feedback.

[0066] Step 2: Preprocessing the data

[0067] The server preprocesses the collected case data, for example by cleansing unnecessary information and tokenizing and structuring important text information.

[0068] Step 3: Create a dataset

[0069] The server splits the preprocessed data into a training set and a test set, typically 80% for training and 20% for testing.

[0070] Step 4: Building and training an AI model

[0071] The server uses a machine learning algorithm to build a generative AI model and trains it using a training set. Specifically, it learns pairs of complaint details and response methods and has the model recognize patterns.

[0072] Step 5: Evaluate and tune the model

[0073] The server evaluates the trained AI model using the test set, and if the accuracy is low, adjusts the hyperparameters to optimize the model.

[0074] Step 6: Providing a User Interface

[0075] The server designs an interface that allows users to input queries and displays it on the terminal. This interface allows queries to be made by filling out a form or operating buttons.

[0076] Step 7: Enter your inquiry

[0077] A user inputs a specific complaint or inquiry into the system. For example, a user may inquire, "What should I do if my product delivery is delayed?"

[0078] Step 8: Generate an answer with the AI ​​model

[0079] The server receives the user's inquiry and uses a generative AI model to generate the optimal response, such as suggesting specific actions such as sending an apology email and offering a coupon for free shipping on the next purchase.

[0080] Step 9: Provide your answers

[0081] The device displays the generated answers from the server to the user, who then reviews the suggested answers and takes action.

[0082] Step 10: Provide feedback

[0083] The user provides feedback on the appropriateness of the provided answer, for example, rating whether the proposed action was effective or not.

[0084] Step 11: Gather feedback

[0085] The server collects feedback from users, which is used to improve the model.

[0086] Step 12: Retraining the AI ​​model

[0087] The server retrains the generative AI model based on the collected feedback, improving its accuracy and making it better able to respond to the next inquiry.

[0088] Step 13: Automatic generation of training manuals

[0089] The server automatically generates training manuals using the collected data and learning results. For example, a "manual for dealing with product defects" is created.

[0090] Step 14: Provide training manuals

[0091] The server displays the generated training manual on the terminal and makes it available to new and current employees, for example, by making it available for download in PDF format.

[0092] In this way, this system improves the quality of complaint handling and increases operational efficiency.

[0093] Example 1

[0094] 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."

[0095] In traditional customer service and call center operations, there is a demand for improved efficiency and quality in handling complaints. However, there are insufficient methods for effectively utilizing past case data, resulting in issues with the consistency and accuracy of responses. Furthermore, there is a lack of advanced functionality, such as automatic model updating based on appropriate feedback and automatic generation of training materials for new employees.

[0096] 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.

[0097] In this invention, the server includes means for collecting past case data, means for preprocessing the collected case data, means for training a machine learning model using the preprocessed data, means for providing an interface through which a user can input an inquiry, means for generating an optimal answer to the user's inquiry using the machine learning model, means for providing the generated answer to the user, means for collecting feedback from the user, means for retraining the machine learning model using the collected feedback, means for automatically generating training materials, and means for providing the generated training materials. This enables improved efficiency and quality in complaint handling, and also improves the consistency and accuracy of responses. Furthermore, automatic model updating and automatic generation of training materials are realized by utilizing appropriate feedback.

[0098] "Past case data" refers to records of past complaint and inquiry handling in call centers and customer service operations.

[0099] "Preprocessing" refers to the process of removing unnecessary information from collected data and extracting and formatting important information.

[0100] A "machine learning model" is a model that uses an algorithm to learn from collected and preprocessed data and predict the optimal response to future inquiries.

[0101] "Interface" refers to the operation screen and input form that users use to access the system and enter inquiries.

[0102] A "generative AI model" refers to an artificial intelligence model built using machine learning algorithms based on collected and pre-processed data.

[0103] "Optimal response" refers to the most appropriate response method generated based on past response cases and learning results, based on the user's inquiry.

[0104] "Feedback" refers to a user's rating or comment on the appropriateness of a proposed answer.

[0105] "Retraining" refers to the process of retraining an existing machine learning model based on collected feedback to improve its accuracy and performance.

[0106] "Educational materials" refers to guides and manuals on how to handle complaints and customer service, intended to educate new employees and improve the skills of current employees.

[0107] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the accompanying drawings. The present invention aims to improve the efficiency of customer service and complaint handling in call center operations, thereby improving the quality of responses.

[0108] First, the server collects past case data collected from call centers and customer support operations. For example, it obtains historical data on complaints such as "the product arrived damaged." This data is obtained from a database. Next, the server preprocesses this data. In preprocessing, unnecessary data such as dates and personal information is cleansed, and important data such as the type of complaint, how it was handled, and the results are extracted and structured. Furthermore, the text data is tokenized, and important keywords are extracted. Specifically, keywords such as "product," "damaged," and "arrived" are extracted.

[0109] Next, using the preprocessed data, the server uses a machine learning algorithm (e.g., using TensorFlow) to build and train a generative AI model. In this process, the server learns past complaint handling patterns and predicts the optimal response method. For example, it learns how to respond to "damaged product" cases and provides the most appropriate relevant information.

[0110] The server then designs and implements a user interface, which is provided to users through their terminals. This interface is provided through a web browser or dedicated application, and is designed to allow users to easily access the system and input inquiries. For example, a customer service representative can inquire about "what to do if a customer's order is delayed."

[0111] When a user inputs a query into the system, the server uses the generative AI model to generate the optimal response to that query. For example, the server suggests "how to respond to a delayed order" by suggesting "sending an apology email and offering a discount on the next order." The device then displays this response to the user.

[0112] Additionally, users can provide feedback on the appropriateness of the suggested answers. The server collects this feedback and retrains the AI ​​model. This process improves the model's accuracy and allows it to respond more appropriately to future inquiries.

[0113] Finally, the server automatically generates training materials based on the collected data and learning results. These materials are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated and provided to the terminal in PDF format.

[0114] As a specific example, the server collects complaint processing history data for "the product arrived damaged," tokenizes it, and extracts keywords such as "product," "damaged," and "arrived." Using the preprocessed data, a generative AI model is built using a machine learning algorithm to learn the optimal response method for "damaged product" cases. A user interface is designed and provided via a terminal, allowing the user to ask the system "how to respond if an order is late." The server generates "send an apology email and offer a discount on the next order," which is then displayed on the terminal. Based on user feedback, the server retrains the model and generates and provides a "manual for responding to product defects" in PDF format.

[0115] An example of a prompt for a generative AI model is:

[0116] Please suggest the best way to handle a complaint based on the complaint handling history below.

[0117] Complaint: The product arrived damaged.

[0118] Response history:

[0119] 1. Send new products to customers immediately

[0120] 2. Requesting a return of damaged goods at no cost to the customer

[0121] 3. Offer a discount coupon for your next order

[0122] Please suggest the best way to respond.

[0123] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0124] Step 1:

[0125] The server retrieves past case data collected from call centers and customer support operations from a database. The input is the database and its query, and the output is case data including the complaint details, response methods, results, etc. Specifically, the server uses an SQL query to extract the relevant data from the database.

[0126] Step 2:

[0127] Remove unnecessary information from the case data acquired by the server (data cleansing). The input is the case data acquired in step 1, and the output is the cleansed data. Specifically, it removes unnecessary fields such as dates and personal information.

[0128] Step 3:

[0129] The server structures the data. The input is the cleansed data from step 2, and the output is structured data. Specifically, the server reclassifies the data into fields such as complaint type, response method, and outcome.

[0130] Step 4:

[0131] The server tokenizes the text data and extracts important keywords. The input is the structured data from step 3, and the output is a list of keywords. Specifically, the server uses a natural language processing tool to tokenize the text and extract keywords such as "product," "damaged," and "arrived."

[0132] Step 5:

[0133] The server uses a machine learning algorithm (e.g., TensorFlow) to build a generative AI model. The input is the keyword list extracted in step 4 and the preprocessed data, and the output is the generated AI model. Specifically, the server inputs the data into the algorithm and trains the model.

[0134] Step 6:

[0135] The server stores the learning results and designs and implements the user interface. The input is the AI ​​model generated in step 5, and the output is the completed user interface. Specifically, the code to provide the interface through a web browser or dedicated application is written and deployed.

[0136] Step 7:

[0137] The terminal allows access to the system through a user interface. The input is a request to access the user interface, and the output is the display of the interface. Specifically, the terminal displays a login screen or a search screen.

[0138] Step 8:

[0139] A user inputs an inquiry into the system. The input is the inquiry entered by the user, and the output is the content being sent to the system. A specific operation is when the user inputs "What to do if an order is late."

[0140] Step 9:

[0141] The server uses the generative AI model to generate the optimal answer to the user's inquiry. The input is the inquiry entered in step 8, and the output is the generated answer. Specifically, the server inputs the inquiry into the model and generates an answer such as "send an apology email and offer a discount on the next order."

[0142] Step 10:

[0143] The terminal displays the generated answer to the user. The input is the answer generated in step 9, and the output is the display of the answer to the user. Specifically, the terminal displays the answer on the screen.

[0144] Step 11:

[0145] The user provides feedback on the appropriateness of the proposed answer. The input is the user's feedback, and the output is the feedback information being sent to the system. The specific behavior is that the user provides feedback as "appropriate" or "inappropriate."

[0146] Step 12:

[0147] The server collects feedback and retrains the AI ​​model. The input is user feedback and existing model data, and the output is the retrained AI model. Specifically, the server reflects the feedback in the model and retrains it.

[0148] Step 13:

[0149] The server automatically generates training materials based on the collected data and learning results. The input is the retrained AI model and learning results, and the output is the generated training manual. Specifically, the server generates content such as a "manual for dealing with product defects" in PDF format.

[0150] Step 14:

[0151] The terminal provides the generated educational materials to the user. The input is the generated educational manual, and the output is the provision of the manual. Specifically, the terminal provides a download link so that the user can download the manual.

[0152] (Application example 1)

[0153] 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."

[0154] In virtual stores, customer service requires quick and appropriate responses to complaints and inquiries, but conventional systems often make response processes cumbersome and result in inconsistent quality. Furthermore, there is a lack of feedback collection and learning processes to improve response accuracy, making continuous improvement difficult.

[0155] 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.

[0156] In this invention, the server includes means for collecting past case data, means for preprocessing the collected case data, means for training a generative AI model using the preprocessed data, means for providing an interface through which a user can input an inquiry, means for generating an optimal answer to the user's inquiry using the generative AI model, means for providing the generated answer to the user, means for collecting feedback from the user, means for retraining the generative AI model using the collected feedback, means for automatically generating a training manual, means for providing the generated training manual, means for responding to inquiries using a smart device, and means for accepting inquiries and providing answers in real time, thereby enabling improved quality and efficiency of customer service.

[0157] "Past case data" refers to data that includes the details of responses to complaints and inquiries recorded in the past in customer service operations.

[0158] "Means of collection" refers to equipment and software used to store data on past cases on a server, etc.

[0159] "Preprocessing means" refers to techniques and methods for cleansing collected data and extracting and structuring important information.

[0160] A "generative AI model" refers to an artificial intelligence model that is built using machine learning algorithms and has learned past response patterns.

[0161] "User Interface" means the screen or operating environment through which a user accesses the system and enters inquiries.

[0162] "Means for generating optimal answers" refers to technologies and functions that use generative AI models to output optimal answers to user inquiries.

[0163] "Feedback collection means" refers to any equipment or software used to collect user opinions or ratings regarding the appropriateness of responses.

[0164] "Retraining methods" are techniques or methods for retraining a generative AI model based on collected feedback to improve the model's accuracy.

[0165] "Means for automatically generating training manuals" refers to technologies and methods for automatically creating manuals for improving customer service skills based on case data and learning results.

[0166] "Smart devices" refer to portable electronic devices that can connect to the Internet, such as smartphones and head-mounted displays.

[0167] "Real-time means" refers to technologies and functions that instantly generate and provide answers to user inquiries.

[0168] MODE FOR CARRYING OUT THE INVENTION

[0169] A system for realizing the present invention in a virtual store is configured using the following hardware and software.

[0170] Hardware and software used

[0171] Hardware

[0172] Server: Used to process, store, and learn from large amounts of data.

[0173] Smart devices: Smartphones, head-mounted displays, etc., are the interface through which users input inquiries and receive responses.

[0174] software

[0175] TensorFlow: A machine learning library used to train generative AI models.

[0176] Django: Web framework, provides the user interface.

[0177] NLTK: A natural language processing library used for data preprocessing.

[0178] MySQL: Database management system used to store and manage data.

[0179] System configuration and operation

[0180] Data collection and preprocessing

[0181] The server collects past case data from call centers and customer service operations. The collected data is tokenized using NLTK to extract important keywords and phrases. The data is then cleansed of unnecessary data and stored in a structured format in a database.

[0182] Training generative AI models

[0183] Using the preprocessed data, the server uses TensorFlow to build and train a generative AI model, a process that allows the model to learn from past response patterns and predict the best response.

[0184] Providing a user interface

[0185] Using the Django framework, the server designs and implements the user interface. This interface, which runs on smart devices, is designed to allow users to easily input inquiries. For example, users can input "What should I do if my order is delayed" into the smartphone application.

[0186] Responding to user inquiries

[0187] When a user inputs a query, the server uses a generative AI model to generate the optimal answer. This answer is displayed on the smart device and provided to the user. For example, when asked how to respond to a delayed order, the server suggests sending an apology email and offering a discount on the next order.

[0188] Gathering feedback and retraining

[0189] Users can provide feedback on the appropriateness of the suggested answers, and the server collects this feedback to retrain the generative AI model. This process improves the model's accuracy, enabling it to better respond to future inquiries.

[0190] Automatic generation of training manuals

[0191] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are provided in PDF format and are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated.

[0192] Examples of specific examples and prompts

[0193] Example 1: When a customer asks, "What should we do if my order is late?" the system responds, "We will send you an apology email and offer you a discount on your next order."

[0194] Examples of prompts:

[0195] "What is the best way to handle a complaint?"

[0196] "What do you do if a customer inquires about a delayed order?"

[0197] Please tell me what to do if the product arrives damaged.

[0198] This makes it possible for this system to improve the quality and efficiency of customer service.

[0199] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0200] Step 1:

[0201] The server collects past case data from customer support history. The input is past support history data, such as complaints about "the product arrived damaged." The server saves this data in MySQL, a database management system. The output is a structured case database.

[0202] Step 2:

[0203] The server preprocesses the collected case data. The input is the stored case data, which is tokenized using NLTK to remove unnecessary information. Specific operations include cleaning the text data, extracting keywords, and structuring. The output is the preprocessed case data.

[0204] Step 3:

[0205] The server uses the preprocessed data to build and train a generative AI model. The input is the preprocessed corresponding case data, and TensorFlow is used to train the machine learning model. Specific operations include data splitting, model selection, training, and evaluation. The output is a trained generative AI model.

[0206] Step 4:

[0207] The server provides the user interface. The input is the interface code built using Django, allowing users to easily enter queries into the system. Specific operations include form design, deployment, and access management. The output is the interface available to users.

[0208] Step 5:

[0209] A user inputs a query from a terminal. The input is a query text by the user, such as "What should I do if my order is late?" The terminal sends this input to the server. The output is the query data sent to the server.

[0210] Step 6:

[0211] The server uses a generative AI model to generate an optimal answer to a user's query. The input is the user's query text, and the generative AI model predicts the optimal answer. Specific operations include performing model inference. The output is the generated answer text.

[0212] Step 7:

[0213] The terminal displays the answer received from the server to the user. The input is the answer text sent by the server, which is displayed in the user interface. Specific operations include adjusting the display layout and notifying the user. The output is the answer that the user can view.

[0214] Step 8:

[0215] The user provides feedback on the appropriateness of the proposed answer. The input is the user's feedback text, and the terminal sends this feedback to the server. The output is the feedback data to the server.

[0216] Step 9:

[0217] The server retrains the generative AI model based on the collected feedback. The input is user feedback data and collected case data, which are used to retrain the model. Specific operations include integrating the feedback data and running the retraining process. The output is an updated generative AI model.

[0218] Step 10:

[0219] The server automatically generates training manuals. The input is a trained or updated generative AI model and corresponding case data, and the training manual is created based on this. Specific operations include selecting a template, writing the content, and executing the automatic generation process. The output is a completed training manual.

[0220] Step 11:

[0221] The server provides the generated training manual. The input is the completed training manual, which is provided to the terminal in PDF format or other format. Specific operations include file generation and distribution. The output is a usable training manual.

[0222] 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.

[0223] MODE FOR CARRYING OUT THE INVENTION

[0224] An embodiment of the present invention will be described below.

[0225] This system is designed to improve the quality and efficiency of customer service and complaint handling in call center operations. By combining the process of collecting and preprocessing past case data, generating and training an AI model, and combining it with an emotion engine that recognizes user emotions, it becomes possible to provide more appropriate and emotionally sensitive responses.

[0226] Data collection and preprocessing

[0227] First, the server collects past case data from call centers and customer support systems. Specifically, it acquires data including the details of complaints, how they were handled, and customer feedback. Next, the server preprocesses this data. In preprocessing, unnecessary data is cleansed and important text information is tokenized and structured.

[0228] AI model generation and training

[0229] Using the preprocessed data, the server builds and trains a generative AI model. This is the process of using machine learning algorithms to train the model and learn from past response patterns. For example, it can learn the appropriate way to respond to complaints and predict the optimal response.

[0230] Introducing the Emotion Engine

[0231] Next, the server introduces an emotion engine, which analyzes emotions from the user's text input or voice commands and records the emotional state as data. This allows the generative AI model to reflect appropriate emotional responses to the user's inquiries.

[0232] Providing a user interface

[0233] The server designs an interface that allows users to input inquiries and provides it through a terminal. This interface is designed so that users can easily access the system and input inquiries. For example, a customer service representative can inquire about "what to do if a customer's order is delayed."

[0234] User inquiries and sentiment analysis

[0235] When a user inputs a query into the system, the server uses a generative AI model to generate the optimal response to that query. At the same time, the emotion engine analyzes the user's emotions and optimizes the response based on that emotional data. For example, if the user expresses anger, the system will provide a more polite response that includes an apology.

[0236] Providing answers

[0237] The terminal displays the generated answer from the server to the user, and the user can take action based on the suggested answer. For example, the suggested action for "what to do if an order is delayed" might be "send an apology email and offer a coupon for free shipping on the next order."

[0238] Gathering feedback and retraining

[0239] Users can provide feedback on the appropriateness of the suggested answers, which the server collects and uses to retrain the AI ​​model, a process that improves the model's accuracy and makes it better able to respond to future inquiries.

[0240] Automatic generation of training manuals

[0241] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated and provided to the terminal in PDF format.

[0242] Through the above process, this system will improve the quality and efficiency of customer service and call center operations. The introduction of an emotion engine will enable responses that take user emotions into consideration, which is expected to improve customer satisfaction.

[0243] The processing flow will be explained below.

[0244] Step 1: Collect data on past cases

[0245] The server collects past case data from call centers and customer support systems. This process includes information such as the nature of the complaint, how it was handled, and the customer's reaction. For example, data on how to handle a delayed delivery of a product can be obtained.

[0246] Step 2: Preprocessing the data

[0247] The server preprocesses the collected case data. During preprocessing, the data is cleansed, unnecessary information is removed, and necessary information is extracted and structured. Specifically, text data is tokenized and important keywords are extracted.

[0248] Step 3: Create a dataset

[0249] The server splits the preprocessed data into training and testing sets, for example, 80% of the total data as the training set and 20% as the testing set.

[0250] Step 4: Building and training an AI model

[0251] The server uses a machine learning algorithm to build a generative AI model and trains it using a training set. Specifically, it learns pairs of complaint details and response methods and has the model recognize patterns.

[0252] Step 5: Evaluate and tune the model

[0253] The server uses the test set to evaluate the trained AI model, for example, assessing the model's accuracy, suitability, and speed of response, and adjusting any inaccuracies. It also optimizes hyperparameters as necessary.

[0254] Step 6: Implementing the Emotion Engine

[0255] The server will install an emotion engine and add the ability to analyze emotions from the user's text input or voice commands. The emotion engine will recognize the user's emotions and record the data. For example, if it detects that the user is "angry," it will save this emotion data as the analysis result.

[0256] Step 7: Providing a User Interface

[0257] The server designs the user interface and provides it through the terminal. The interface includes forms and input fields, allowing users to easily enter inquiries. For example, a form is provided for entering "What to do if the product arrives damaged."

[0258] Step 8: Query entry and sentiment analysis

[0259] The user inputs a query through the interface, for example, "The product arrived damaged." At the same time, the emotion engine analyzes the user's emotions and detects "anger."

[0260] Step 9: Generate an answer with the AI ​​model

[0261] The server receives the query and uses a generative AI model to generate the optimal answer. Based on data from the emotion engine, the server provides an answer that takes the user's emotions into consideration. For example, it may suggest sending an email with an apology and applying a discount on the next order.

[0262] Step 10: Provide your answers

[0263] The device displays the generated answer from the server to the user. The user checks the suggested answer and takes action. For example, the device displays specific measures such as "send an apology email and immediately ship a replacement product" as a response to the question "What to do if the product arrives damaged."

[0264] Step 11: Enter your feedback

[0265] The user provides feedback on the appropriateness of the provided answer, for example, by rating whether the proposed response was effective.

[0266] Step 12: Gather feedback

[0267] The server collects feedback from users, including the appropriateness of the response and areas for improvement. For example, specific opinions such as "the response was a little slow" are collected.

[0268] Step 13: Retraining the AI ​​model

[0269] The server retrains the generative AI model based on the collected feedback and emotion data, improving its accuracy and response capabilities. For example, it can learn how to respond quickly to anger.

[0270] Step 14: Automatic generation of training manuals

[0271] The server automatically generates training manuals using the collected data and learning results. For example, a "manual for dealing with product defects" is generated.

[0272] Step 15: Provide training manuals

[0273] The server displays the generated training manual on the terminal and makes it available to new and current employees, for example, by providing it as a downloadable PDF file.

[0274] Through these steps, the system will improve the quality and efficiency of customer service and call center operations. The introduction of an emotion engine will enable responses that take user emotions into consideration, which is expected to improve customer satisfaction.

[0275] Example 2

[0276] 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."

[0277] In recent years, there has been a demand for improved quality and efficiency in customer service and call center operations when it comes to handling complaints. However, conventional systems have struggled to respond in a way that takes user emotions into consideration, limiting the extent to which they can improve customer satisfaction. Furthermore, they lack the ability to automatically generate appropriate response manuals, making the efficiency of employee training an issue.

[0278] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting past case data, means for preprocessing the collected case data, means for training a generative AI model using the preprocessed data, means for providing an interface through which a user can input an inquiry, means for using an emotion engine to analyze the user's emotions, means for generating an optimal answer to the user's inquiry using the generative AI model, means for providing the generated answer to the user, means for collecting feedback from the user, means for retraining the generative AI model using the collected feedback, means for automatically generating a training manual, and means for providing the generated training manual. This enables improved quality of complaint handling and business efficiency. Furthermore, the introduction of an emotion engine enables responses that take user emotions into consideration, which is expected to improve customer satisfaction. Furthermore, the automatic generation of training manuals allows for efficient training of new employees and skill development of current employees.

[0279] "Past case data" refers to data collected from customer response history, including the content of complaints, response methods, and feedback from customers.

[0280] "Preprocessing" refers to the process of cleansing unnecessary noise from collected data and tokenizing important text information to convert it into structured data.

[0281] A "generative AI model" is a model that is constructed using machine learning algorithms with preprocessed data to predict the appropriate way to handle a complaint.

[0282] An "interface" is a tool or platform through which a user can enter a query and interact with a system, such as a web form or a mobile app.

[0283] An "emotion engine" is a system that analyzes emotions from a user's text input or voice commands, and has the ability to record the user's emotional state as data.

[0284] An "Answer" is an optimal response to a user's query, created using a generative AI model and emotion engine.

[0285] "Feedback" refers to opinions and ratings provided by users regarding the appropriateness of proposed answers.

[0286] "Retraining" is the process of using collected feedback to readjust the parameters of a generative AI model to improve the model's accuracy.

[0287] The "training manual" is a document that is automatically generated based on collected data and the learning results of the generating AI model, and is used to train new employees and improve the skills of current employees.

[0288] One embodiment of the present invention is described in detail below. This system is designed to improve the quality and efficiency of customer service and complaint handling in call center operations. By combining the process of collecting and preprocessing past case data, generating and training a generative AI model, and incorporating an emotion engine that recognizes user emotions, more appropriate and emotion-sensitive responses are possible.

[0289] First, the server collects past case data from call centers and customer support systems. Specifically, it obtains data including the details of complaints, how they were handled, and customer feedback. Data is collected from a database and extracted using the Python "pandas" library. Next, the server preprocesses this data. During preprocessing, unnecessary data is cleansed, and important text information is tokenized using the "NLTK" library and converted into structured data.

[0290] Using the preprocessed data, the server builds and trains a generative AI model. TensorFlow and PyTorch are used to build the model, and the model is trained using a training dataset. Specifically, the model learns to predict the appropriate response method using paired data of the complaint content and the response method. For example, the following prompt sentence is used:

[0291] "How should you respond if a customer complains about the following? Complaint: ____"

[0292] The server then deploys an emotion engine, which uses IBM Watson and Microsoft Azure Emotion API to analyze emotions from the user's text input and voice commands. This engine records the user's emotional state as data and reflects it in the generative AI model.

[0293] The server designs an interface that allows users to enter inquiries and provides it via the terminal. The interface is provided as a web-based form or mobile app, and is designed so that users can easily access the system and enter inquiries. For example, a customer service representative can inquire about "what to do if a customer's order is delayed" through the system.

[0294] When a user inputs an inquiry into the system, the server uses a generative AI model to generate the optimal response to that inquiry. At the same time, an emotion engine analyzes the user's emotions and optimizes the response based on that emotional data. For example, if the user expresses anger, the system will provide a response that includes a more polite apology. As a specific example, when asked what to do if an order is delayed, the system will suggest responses such as "send an apology email and offer a coupon for free shipping on the next order."

[0295] The device displays the answer generated by the server to the user, who can then act on the suggested answer. The user can also provide feedback on whether the suggested answer is appropriate. The server collects the feedback and uses it to retrain the generative AI model to improve the quality of the answer. This improves the model's accuracy and makes it better able to respond to future inquiries.

[0296] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals can be used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" can be provided in PDF format.

[0297] The above is an embodiment of the system of the present invention, which improves the quality and efficiency of customer service and complaint handling in call center operations. The introduction of an emotion engine enables responses that take user emotions into consideration, which is expected to improve customer satisfaction. Furthermore, the automatic generation of training manuals allows for efficient training of new employees and skill development for current employees.

[0298] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0299] Step 1: Data collection

[0300] The server collects past case data from call centers and customer support systems. This includes data on complaints, response methods, and customer feedback. Specifically, it extracts data from a database using SQL queries. The input is past case data, and the output is the extracted raw data.

[0301] Step 2: Data Preprocessing

[0302] The server preprocesses the collected data. Specifically, it cleanses unnecessary data using the Python "pandas" library and tokenizes text information using the "NLTK" library to convert it into structured data. The input is raw data, and the output is cleansed and tokenized structured data.

[0303] Step 3: Generate an AI model

[0304] The server uses the preprocessed data to build a generative AI model. It uses the machine learning library "TensorFlow" or "PyTorch" to define a neural network model. For example, it uses pairs of complaints and response methods as input data. The input is structured training data, and the output is an initialized generative AI model.

[0305] Step 4: Training the AI ​​model

[0306] The server trains the generative AI model. Specifically, it uses the preprocessed data to apply a learning algorithm (e.g., batch learning, setting the number of epochs, etc.) to improve the accuracy of the model. The input is the training data and the initialized AI model, and the output is the trained generative AI model.

[0307] Step 5: Implementing the Emotion Engine

[0308] The server installs an emotion engine and analyzes emotions from the user's text input or voice commands. IBM Watson and Microsoft Azure Emotion API are used as emotion engines. The input is text or voice data from the user, and the output is analyzed emotional data.

[0309] Step 6: Providing a User Interface

[0310] The server designs a query input interface and provides it through the terminal, such as a web form or a mobile app. The input is the query from the user, and the output is the query data sent to the system.

[0311] Step 7: Receiving and analyzing inquiries

[0312] A user inputs a query into the system. The server receives the query and uses a generative AI model to generate the optimal answer to that query. Specifically, a prompt such as "What should we do if a customer's order is delayed?" is used. The input is the user's query, and the output is the generated answer.

[0313] Step 8: Analyze sentiment data and optimize responses

[0314] The server uses an emotion engine to analyze the user's emotions and optimizes the answer of the generative AI model based on the emotion data. The input is the query and the generated answer, and the output is an optimized answer that takes emotions into consideration.

[0315] Step 9: Provide your answers

[0316] The device displays the generated answer from the server to the user, specifically on a web page or in the chat screen of the app. The input is the optimized answer, and the output is what is displayed to the user.

[0317] Step 10: Gather feedback

[0318] The user provides feedback on the appropriateness of the response. The server collects this feedback and stores it in a database. The input is the user feedback and the output is the collected feedback data.

[0319] Step 11: Retraining the AI ​​model

[0320] The server uses the collected feedback to retrain the generative AI model. Specifically, it uses the feedback data as new training data and readjusts the model's parameters. The inputs are the feedback data and the existing generative AI model, and the output is the retrained generative AI model.

[0321] Step 12: Automatic generation of training manuals

[0322] The server automatically generates training manuals based on the training data and model results. Specifically, it uses a Python script to analyze past cases and their effectiveness, and generates training manuals. The input is the training data and the model's learning results, and the output is the generated training manual (e.g., in PDF format).

[0323] Step 13: Provide training manuals

[0324] The terminal provides the generated training manual to the user. Specifically, it is provided as a download link or email attachment. The input is the generated training manual, and the output is the manual provided to the user.

[0325] (Application example 2)

[0326] 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."

[0327] When it comes to customer service for autonomous vehicles, there are cases where responses to user complaints or problems are delayed or inappropriate. In such cases, measures that take user feelings into consideration are required, and improving customer satisfaction is a challenge.

[0328] 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 collecting past response case data, means for preprocessing the collected response case data, means for training a generative AI model using the preprocessed data, means for providing an interface through which a user can input a query, means including an emotion engine for analyzing the user's emotions, means for generating an optimal answer to the user's query using the generative AI model, means for optimizing the generated answer based on user emotion analysis, means for providing the optimized answer to the user, means for collecting feedback from the user, means for retraining the generative AI model using the collected feedback, means for automatically generating a training manual, and means for providing the generated training manual. This enables user complaints and problems regarding autonomous vehicles to be handled quickly and appropriately, thereby improving customer satisfaction.

[0329] "Past case data" refers to data that includes details of past complaints and troubles, the methods used to deal with them, and feedback from customers.

[0330] "Preprocessing" is the process of removing unnecessary data from collected data and structuring important text information.

[0331] A "generative AI model" is a model trained using machine learning algorithms and used to predict optimal responses based on past response patterns.

[0332] A "user interface" is an interface through which a user inputs queries into a system, and supports communication between the user and the system.

[0333] The "emotion engine" is a system that analyzes emotions from the user's text and voice input and records that emotional state as data.

[0334] "Feedback" means opinions or evaluations provided by users regarding the appropriateness or quality of a system's response.

[0335] "Retraining" is the process of retraining a generative AI model using newly collected feedback to improve the model's accuracy and responsiveness.

[0336] An "education manual" is a document automatically generated based on collected data and learning results, and is used to train new employees and improve the skills of current employees.

[0337] The present invention is a system for quickly and appropriately responding to user complaints and problems with autonomous vehicles. This system collects and analyzes past case data and provides optimal countermeasures that take user emotions into consideration. An embodiment of the present invention is described in detail below.

[0338] Data collection and preprocessing

[0339] First, the server collects past case data from the autonomous vehicle's operation management system and customer response system. This includes complaint details, response methods, and customer feedback. Next, the server preprocesses this data, removing unnecessary data and tokenizing and structuring useful text information.

[0340] AI model generation and training

[0341] The server uses the preprocessed data to build a generative AI model and trains it using machine learning algorithms, allowing it to learn past response patterns and predict optimal responses, such as the appropriate way to respond to complaints.

[0342] Introducing the Emotion Engine

[0343] The server then uses an emotion engine to analyze the user's text input. The emotion engine uses natural language processing tools such as TextBlob to analyze the emotion into positive, negative, and neutral, and records the emotional state as data. This allows the AI ​​model to reflect appropriate emotional responses to the user's inquiry.

[0344] User Interface

[0345] The server designs and provides an interface through which users can input inquiries via terminals, allowing users to easily report complaints and problems.

[0346] User inquiries and sentiment analysis

[0347] When a user inputs a query into the system, the server uses a generative AI model to generate the optimal response for that query. At the same time, the emotion engine analyzes the user's emotions and optimizes the response based on that emotional data. For example, if the user expresses anger, the server will provide a more polite response that includes an apology.

[0348] Examples:

[0349] User input: "My car is stuck in the middle of the road! I'm so annoyed. What can you do?"

[0350] Emotion analysis result: "negative"

[0351] Example AI response: "We apologize for the inconvenience. Our technical team is working quickly to resolve the issue. We'll offer you a 20% discount coupon for your next visit."

[0352] Providing answers

[0353] The device displays the generated answer from the server to the user, who then acts on the suggested answer.

[0354] Gathering feedback and retraining

[0355] Users can provide feedback on the appropriateness of the proposed answers, which the server collects and uses to retrain the generative AI model, a process that improves the model's accuracy and makes it better able to respond to future inquiries.

[0356] Automatic generation of training manuals

[0357] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are used to train new employees and improve the skills of current employees. For example, a "vehicle malfunction response manual" is generated and provided to the terminal in PDF format.

[0358] In this way, the system of the present invention improves the quality of customer service in self-driving vehicles and is expected to improve customer satisfaction by enabling emotionally sensitive service.

[0359] Example prompt sentence:

[0360] Prompt: "User's complaint: My car got stuck in the middle of the road! I'm so upset. What can you do about it? Emotion: negative Suggest the best solution:"

[0361] This embodiment makes it possible to specifically implement the present invention.

[0362] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0363] Step 1:

[0364] The server collects past case data from the autonomous vehicle's operation management system and customer response system. The collected data includes the details of the complaint, how it was handled, and customer feedback. The input is raw data obtained from the operation management system database, and the output is sent to pre-processing through the collection step.

[0365] Step 2:

[0366] The server preprocesses the collected data. During this process, unnecessary data is cleansed and useful text information is tokenized and structured. Processes such as text normalization, stop word removal, and tokenization are performed. The input is the collected raw data, and the output is cleansed and structured data.

[0367] Step 3:

[0368] The server uses the preprocessed data to generate and train a generative AI model. The model is trained using machine learning algorithms (e.g., deep learning) to learn appropriate countermeasures based on past response patterns. The input is the preprocessed data, and the output is a trained generative AI model.

[0369] Step 4:

[0370] The server uses an emotion engine to analyze the user's emotions. In this process, the text entered by the user is analyzed to determine whether the emotion is positive, negative, or neutral, and the data is recorded. The input is the text input from the user, and the output is the analyzed emotion data.

[0371] Step 5:

[0372] The user uses an interface to input inquiries and complaints. The server designs this interface to make it easy for the user to input inquiries. The input is the complaint or inquiry content from the user, and the output is the inquiry data that is sent to the server.

[0373] Step 6:

[0374] The server uses a generative AI model to generate the optimal answer to a user's inquiry. The answer is optimized based on the user's emotional data analyzed by the emotion engine. For example, if the emotion included anger, a particularly polite response would be suggested. The input is the user's inquiry data and emotional data, and the output is an optimized answer.

[0375] Step 7:

[0376] The terminal displays the generated answer to the user, who then reviews the proposed answer and takes action. The input is the optimized answer sent by the server, and the output is the generated answer that is displayed to the user.

[0377] Step 8:

[0378] Users provide feedback on the appropriateness of the proposed answers, which the server collects and uses to retrain the generative AI model. The input is the user feedback data, and the output is the retraining data.

[0379] Step 9:

[0380] The server automatically generates a training manual based on the collected data and learning results. This manual is used to train new employees and improve the skills of current employees. The inputs are the collected data and learning results, and the output is a training manual (e.g., a PDF document).

[0381] In this way, each step works together to create a system that can quickly and appropriately handle complaints and problems with self-driving vehicles.

[0382] 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.

[0383] 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.

[0384] 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.

[0385] [Second embodiment]

[0386] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0387] 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.

[0388] 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).

[0389] 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.

[0390] 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.

[0391] 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).

[0392] 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.

[0393] 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.

[0394] 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.

[0395] 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.

[0396] 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.

[0397] 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."

[0398] MODE FOR CARRYING OUT THE INVENTION

[0399] An embodiment of the present invention will be described below.

[0400] This system is designed to improve the efficiency and quality of customer service and complaint handling in call center operations. It collects and preprocesses past case data, generates and trains AI models, provides a user interface, handles user inquiries, collects feedback and retrains, and automatically generates training manuals.

[0401] Data collection and preprocessing

[0402] First, the server collects past case data collected from call centers and customer support operations. For example, it obtains historical data on responses to complaints such as "the product arrived damaged." Next, the server preprocesses this data. In preprocessing, unnecessary data is cleansed and important data is extracted and structured. In particular, text data is tokenized, and important keywords are extracted.

[0403] AI model generation and training

[0404] Using the preprocessed data, the server builds and trains a generative AI model. This is the process of using machine learning algorithms to train the model and learn from past response patterns. For example, it can learn the appropriate way to respond to complaints and predict the optimal response.

[0405] Providing a user interface

[0406] Next, the server designs and implements a user interface, which is provided via a terminal. This interface is designed to allow users to easily access the system and input inquiries. For example, a customer service representative can ask the system, "What should we do if a customer's order is delayed?"

[0407] Responding to user inquiries

[0408] When a user inputs a query into the system, the server uses the generative AI model to generate the optimal response to that query. For example, the server suggests "how to respond to a delayed order" by suggesting "sending an apology email and offering a discount on the next order." The device then displays this response to the user.

[0409] Gathering feedback and retraining

[0410] Users can provide feedback on the appropriateness of the suggested answers, which the server collects and uses to retrain the AI ​​model, a process that improves the model's accuracy and allows it to better respond to future inquiries.

[0411] Automatic generation of training manuals

[0412] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated and provided to the terminal in PDF format.

[0413] Through the above process, this system will improve the quality of customer service and call center operations and increase operational efficiency.

[0414] The processing flow will be explained below.

[0415] Step 1: Collecting interaction data

[0416] The server collects past case data from call centers and customer support systems, specifically data including complaint details, response methods, and customer feedback.

[0417] Step 2: Preprocessing the data

[0418] The server preprocesses the collected case data, for example by cleansing unnecessary information and tokenizing and structuring important text information.

[0419] Step 3: Create a dataset

[0420] The server splits the preprocessed data into a training set and a test set, typically 80% for training and 20% for testing.

[0421] Step 4: Building and training an AI model

[0422] The server uses a machine learning algorithm to build a generative AI model and trains it using a training set. Specifically, it learns pairs of complaint details and response methods and has the model recognize patterns.

[0423] Step 5: Evaluate and tune the model

[0424] The server evaluates the trained AI model using the test set, and if the accuracy is low, adjusts the hyperparameters to optimize the model.

[0425] Step 6: Providing a User Interface

[0426] The server designs an interface that allows users to input queries and displays it on the terminal. This interface allows queries to be made by filling out a form or operating buttons.

[0427] Step 7: Enter your inquiry

[0428] A user inputs a specific complaint or inquiry into the system. For example, a user may inquire, "What should I do if my product delivery is delayed?"

[0429] Step 8: Generate an answer with the AI ​​model

[0430] The server receives the user's inquiry and uses a generative AI model to generate the optimal response, such as suggesting specific actions such as sending an apology email and offering a coupon for free shipping on the next purchase.

[0431] Step 9: Provide your answers

[0432] The device displays the generated answers from the server to the user, who then reviews the suggested answers and takes action.

[0433] Step 10: Provide feedback

[0434] The user provides feedback on the appropriateness of the provided answer, for example, rating whether the proposed action was effective or not.

[0435] Step 11: Gather feedback

[0436] The server collects feedback from users, which is used to improve the model.

[0437] Step 12: Retraining the AI ​​model

[0438] The server retrains the generative AI model based on the collected feedback, improving its accuracy and making it better able to respond to the next inquiry.

[0439] Step 13: Automatic generation of training manuals

[0440] The server automatically generates training manuals using the collected data and learning results. For example, a "manual for dealing with product defects" is created.

[0441] Step 14: Provide training manuals

[0442] The server displays the generated training manual on the terminal and makes it available to new and current employees, for example, by making it available for download in PDF format.

[0443] In this way, this system improves the quality of complaint handling and increases operational efficiency.

[0444] Example 1

[0445] 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."

[0446] In traditional customer service and call center operations, there is a demand for improved efficiency and quality in handling complaints. However, there are insufficient methods for effectively utilizing past case data, resulting in issues with the consistency and accuracy of responses. Furthermore, there is a lack of advanced functionality, such as automatic model updating based on appropriate feedback and automatic generation of training materials for new employees.

[0447] 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.

[0448] In this invention, the server includes means for collecting past case data, means for preprocessing the collected case data, means for training a machine learning model using the preprocessed data, means for providing an interface through which a user can input an inquiry, means for generating an optimal answer to the user's inquiry using the machine learning model, means for providing the generated answer to the user, means for collecting feedback from the user, means for retraining the machine learning model using the collected feedback, means for automatically generating training materials, and means for providing the generated training materials. This enables improved efficiency and quality in complaint handling, and also improves the consistency and accuracy of responses. Furthermore, automatic model updating and automatic generation of training materials are realized by utilizing appropriate feedback.

[0449] "Past case data" refers to records of past complaint and inquiry handling in call centers and customer service operations.

[0450] "Preprocessing" refers to the process of removing unnecessary information from collected data and extracting and formatting important information.

[0451] A "machine learning model" is a model that uses an algorithm to learn from collected and preprocessed data and predict the optimal response to future inquiries.

[0452] "Interface" refers to the operation screen and input form that users use to access the system and enter inquiries.

[0453] A "generative AI model" refers to an artificial intelligence model built using machine learning algorithms based on collected and pre-processed data.

[0454] "Optimal response" refers to the most appropriate response method generated based on past response cases and learning results, based on the user's inquiry.

[0455] "Feedback" refers to a user's rating or comment on the appropriateness of a proposed answer.

[0456] "Retraining" refers to the process of retraining an existing machine learning model based on collected feedback to improve its accuracy and performance.

[0457] "Educational materials" refers to guides and manuals on how to handle complaints and customer service, intended to educate new employees and improve the skills of current employees.

[0458] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the accompanying drawings. The present invention aims to improve the efficiency of customer service and complaint handling in call center operations, thereby improving the quality of responses.

[0459] First, the server collects past case data collected from call centers and customer support operations. For example, it obtains historical data on complaints such as "the product arrived damaged." This data is obtained from a database. Next, the server preprocesses this data. In preprocessing, unnecessary data such as dates and personal information is cleansed, and important data such as the type of complaint, how it was handled, and the results are extracted and structured. Furthermore, the text data is tokenized, and important keywords are extracted. Specifically, keywords such as "product," "damaged," and "arrived" are extracted.

[0460] Next, using the preprocessed data, the server uses a machine learning algorithm (e.g., using TensorFlow) to build and train a generative AI model. In this process, the server learns past complaint handling patterns and predicts the optimal response method. For example, it learns how to respond to "damaged product" cases and provides the most appropriate relevant information.

[0461] The server then designs and implements a user interface, which is provided to users through their terminals. This interface is provided through a web browser or dedicated application, and is designed to allow users to easily access the system and input inquiries. For example, a customer service representative can inquire about "what to do if a customer's order is delayed."

[0462] When a user inputs a query into the system, the server uses the generative AI model to generate the optimal response to that query. For example, the server suggests "how to respond to a delayed order" by suggesting "sending an apology email and offering a discount on the next order." The device then displays this response to the user.

[0463] Additionally, users can provide feedback on the appropriateness of the suggested answers. The server collects this feedback and retrains the AI ​​model. This process improves the model's accuracy and allows it to respond more appropriately to future inquiries.

[0464] Finally, the server automatically generates training materials based on the collected data and learning results. These materials are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated and provided to the terminal in PDF format.

[0465] As a specific example, the server collects complaint processing history data for "the product arrived damaged," tokenizes it, and extracts keywords such as "product," "damaged," and "arrived." Using the preprocessed data, a generative AI model is built using a machine learning algorithm to learn the optimal response method for "damaged product" cases. A user interface is designed and provided via a terminal, allowing the user to ask the system "how to respond if an order is late." The server generates "send an apology email and offer a discount on the next order," which is then displayed on the terminal. Based on user feedback, the server retrains the model and generates and provides a "manual for responding to product defects" in PDF format.

[0466] An example of a prompt for a generative AI model is:

[0467] Please suggest the best way to handle a complaint based on the complaint handling history below.

[0468] Complaint: The product arrived damaged.

[0469] Response history:

[0470] 1. Send new products to customers immediately

[0471] 2. Requesting a return of damaged goods at no cost to the customer

[0472] 3. Offer a discount coupon for your next order

[0473] Please suggest the best way to respond.

[0474] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0475] Step 1:

[0476] The server retrieves past case data collected from call centers and customer support operations from a database. The input is the database and its query, and the output is case data including the complaint details, response methods, results, etc. Specifically, the server uses an SQL query to extract the relevant data from the database.

[0477] Step 2:

[0478] Remove unnecessary information from the case data acquired by the server (data cleansing). The input is the case data acquired in step 1, and the output is the cleansed data. Specifically, it removes unnecessary fields such as dates and personal information.

[0479] Step 3:

[0480] The server structures the data. The input is the cleansed data from step 2, and the output is structured data. Specifically, the server reclassifies the data into fields such as complaint type, response method, and outcome.

[0481] Step 4:

[0482] The server tokenizes the text data and extracts important keywords. The input is the structured data from step 3, and the output is a list of keywords. Specifically, the server uses a natural language processing tool to tokenize the text and extract keywords such as "product," "damaged," and "arrived."

[0483] Step 5:

[0484] The server uses a machine learning algorithm (e.g., TensorFlow) to build a generative AI model. The input is the keyword list extracted in step 4 and the preprocessed data, and the output is the generated AI model. Specifically, the server inputs the data into the algorithm and trains the model.

[0485] Step 6:

[0486] The server stores the learning results and designs and implements the user interface. The input is the AI ​​model generated in step 5, and the output is the completed user interface. Specifically, the code to provide the interface through a web browser or dedicated application is written and deployed.

[0487] Step 7:

[0488] The terminal allows access to the system through a user interface. The input is a request to access the user interface, and the output is the display of the interface. Specifically, the terminal displays a login screen or a search screen.

[0489] Step 8:

[0490] A user inputs an inquiry into the system. The input is the inquiry entered by the user, and the output is the content being sent to the system. A specific operation is when the user inputs "What to do if an order is late."

[0491] Step 9:

[0492] The server uses the generative AI model to generate the optimal answer to the user's inquiry. The input is the inquiry entered in step 8, and the output is the generated answer. Specifically, the server inputs the inquiry into the model and generates an answer such as "send an apology email and offer a discount on the next order."

[0493] Step 10:

[0494] The terminal displays the generated answer to the user. The input is the answer generated in step 9, and the output is the display of the answer to the user. Specifically, the terminal displays the answer on the screen.

[0495] Step 11:

[0496] The user provides feedback on the appropriateness of the proposed answer. The input is the user's feedback, and the output is the feedback information being sent to the system. The specific behavior is that the user provides feedback as "appropriate" or "inappropriate."

[0497] Step 12:

[0498] The server collects feedback and retrains the AI ​​model. The input is user feedback and existing model data, and the output is the retrained AI model. Specifically, the server reflects the feedback in the model and retrains it.

[0499] Step 13:

[0500] The server automatically generates training materials based on the collected data and learning results. The input is the retrained AI model and learning results, and the output is the generated training manual. Specifically, the server generates content such as a "manual for dealing with product defects" in PDF format.

[0501] Step 14:

[0502] The terminal provides the generated educational materials to the user. The input is the generated educational manual, and the output is the provision of the manual. Specifically, the terminal provides a download link so that the user can download the manual.

[0503] (Application example 1)

[0504] 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."

[0505] In virtual stores, customer service requires quick and appropriate responses to complaints and inquiries, but conventional systems often make response processes cumbersome and result in inconsistent quality. Furthermore, there is a lack of feedback collection and learning processes to improve response accuracy, making continuous improvement difficult.

[0506] 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.

[0507] In this invention, the server includes means for collecting past case data, means for preprocessing the collected case data, means for training a generative AI model using the preprocessed data, means for providing an interface through which a user can input an inquiry, means for generating an optimal answer to the user's inquiry using the generative AI model, means for providing the generated answer to the user, means for collecting feedback from the user, means for retraining the generative AI model using the collected feedback, means for automatically generating a training manual, means for providing the generated training manual, means for responding to inquiries using a smart device, and means for accepting inquiries and providing answers in real time, thereby enabling improved quality and efficiency of customer service.

[0508] "Past case data" refers to data that includes the details of responses to complaints and inquiries recorded in the past in customer service operations.

[0509] "Means of collection" refers to equipment and software used to store data on past cases on a server, etc.

[0510] "Preprocessing means" refers to techniques and methods for cleansing collected data and extracting and structuring important information.

[0511] A "generative AI model" refers to an artificial intelligence model that is built using machine learning algorithms and has learned past response patterns.

[0512] "User Interface" means the screen or operating environment through which a user accesses the system and enters inquiries.

[0513] "Means for generating optimal answers" refers to technologies and functions that use generative AI models to output optimal answers to user inquiries.

[0514] "Feedback collection means" refers to any equipment or software used to collect user opinions or ratings regarding the appropriateness of responses.

[0515] "Retraining methods" are techniques or methods for retraining a generative AI model based on collected feedback to improve the model's accuracy.

[0516] "Means for automatically generating training manuals" refers to technologies and methods for automatically creating manuals for improving customer service skills based on case data and learning results.

[0517] "Smart devices" refer to portable electronic devices that can connect to the Internet, such as smartphones and head-mounted displays.

[0518] "Real-time means" refers to technologies and functions that instantly generate and provide answers to user inquiries.

[0519] MODE FOR CARRYING OUT THE INVENTION

[0520] A system for realizing the present invention in a virtual store is configured using the following hardware and software.

[0521] Hardware and software used

[0522] Hardware

[0523] Server: Used to process, store, and learn from large amounts of data.

[0524] Smart devices: Smartphones, head-mounted displays, etc., are the interface through which users input inquiries and receive responses.

[0525] software

[0526] TensorFlow: A machine learning library used to train generative AI models.

[0527] Django: Web framework, provides the user interface.

[0528] NLTK: A natural language processing library used for data preprocessing.

[0529] MySQL: Database management system used to store and manage data.

[0530] System configuration and operation

[0531] Data collection and preprocessing

[0532] The server collects past case data from call centers and customer service operations. The collected data is tokenized using NLTK to extract important keywords and phrases. The data is then cleansed of unnecessary data and stored in a structured format in a database.

[0533] Training generative AI models

[0534] Using the preprocessed data, the server uses TensorFlow to build and train a generative AI model, a process that allows the model to learn from past response patterns and predict the best response.

[0535] Providing a user interface

[0536] Using the Django framework, the server designs and implements the user interface. This interface, which runs on smart devices, is designed to allow users to easily input inquiries. For example, users can input "What should I do if my order is delayed" into the smartphone application.

[0537] Responding to user inquiries

[0538] When a user inputs a query, the server uses a generative AI model to generate the optimal answer. This answer is displayed on the smart device and provided to the user. For example, when asked how to respond to a delayed order, the server suggests sending an apology email and offering a discount on the next order.

[0539] Gathering feedback and retraining

[0540] Users can provide feedback on the appropriateness of the suggested answers, and the server collects this feedback to retrain the generative AI model. This process improves the model's accuracy, enabling it to better respond to future inquiries.

[0541] Automatic generation of training manuals

[0542] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are provided in PDF format and are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated.

[0543] Examples of specific examples and prompts

[0544] Example 1: When a customer asks, "What should we do if my order is late?" the system responds, "We will send you an apology email and offer you a discount on your next order."

[0545] Examples of prompts:

[0546] "What is the best way to handle a complaint?"

[0547] "What do you do if a customer inquires about a delayed order?"

[0548] Please tell me what to do if the product arrives damaged.

[0549] This makes it possible for this system to improve the quality and efficiency of customer service.

[0550] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0551] Step 1:

[0552] The server collects past case data from customer support history. The input is past support history data, such as complaints about "the product arrived damaged." The server saves this data in MySQL, a database management system. The output is a structured case database.

[0553] Step 2:

[0554] The server preprocesses the collected case data. The input is the stored case data, which is tokenized using NLTK to remove unnecessary information. Specific operations include cleaning the text data, extracting keywords, and structuring. The output is the preprocessed case data.

[0555] Step 3:

[0556] The server uses the preprocessed data to build and train a generative AI model. The input is the preprocessed corresponding case data, and TensorFlow is used to train the machine learning model. Specific operations include data splitting, model selection, training, and evaluation. The output is a trained generative AI model.

[0557] Step 4:

[0558] The server provides the user interface. The input is the interface code built using Django, allowing users to easily enter queries into the system. Specific operations include form design, deployment, and access management. The output is the interface available to users.

[0559] Step 5:

[0560] A user inputs a query from a terminal. The input is a query text by the user, such as "What should I do if my order is late?" The terminal sends this input to the server. The output is the query data sent to the server.

[0561] Step 6:

[0562] The server uses a generative AI model to generate an optimal answer to a user's query. The input is the user's query text, and the generative AI model predicts the optimal answer. Specific operations include performing model inference. The output is the generated answer text.

[0563] Step 7:

[0564] The terminal displays the answer received from the server to the user. The input is the answer text sent by the server, which is displayed in the user interface. Specific operations include adjusting the display layout and notifying the user. The output is the answer that the user can view.

[0565] Step 8:

[0566] The user provides feedback on the appropriateness of the proposed answer. The input is the user's feedback text, and the terminal sends this feedback to the server. The output is the feedback data to the server.

[0567] Step 9:

[0568] The server retrains the generative AI model based on the collected feedback. The input is user feedback data and collected case data, which are used to retrain the model. Specific operations include integrating the feedback data and running the retraining process. The output is an updated generative AI model.

[0569] Step 10:

[0570] The server automatically generates training manuals. The input is a trained or updated generative AI model and corresponding case data, and the training manual is created based on this. Specific operations include selecting a template, writing the content, and executing the automatic generation process. The output is a completed training manual.

[0571] Step 11:

[0572] The server provides the generated training manual. The input is the completed training manual, which is provided to the terminal in PDF format or other format. Specific operations include file generation and distribution. The output is a usable training manual.

[0573] 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.

[0574] MODE FOR CARRYING OUT THE INVENTION

[0575] An embodiment of the present invention will be described below.

[0576] This system is designed to improve the quality and efficiency of customer service and complaint handling in call center operations. By combining the process of collecting and preprocessing past case data, generating and training an AI model, and combining it with an emotion engine that recognizes user emotions, it becomes possible to provide more appropriate and emotionally sensitive responses.

[0577] Data collection and preprocessing

[0578] First, the server collects past case data from call centers and customer support systems. Specifically, it acquires data including the details of complaints, how they were handled, and customer feedback. Next, the server preprocesses this data. In preprocessing, unnecessary data is cleansed and important text information is tokenized and structured.

[0579] AI model generation and training

[0580] Using the preprocessed data, the server builds and trains a generative AI model. This is the process of using machine learning algorithms to train the model and learn from past response patterns. For example, it can learn the appropriate way to respond to complaints and predict the optimal response.

[0581] Introducing the Emotion Engine

[0582] Next, the server introduces an emotion engine, which analyzes emotions from the user's text input or voice commands and records the emotional state as data. This allows the generative AI model to reflect appropriate emotional responses to the user's inquiries.

[0583] Providing a user interface

[0584] The server designs an interface that allows users to input inquiries and provides it through a terminal. This interface is designed so that users can easily access the system and input inquiries. For example, a customer service representative can inquire about "what to do if a customer's order is delayed."

[0585] User inquiries and sentiment analysis

[0586] When a user inputs a query into the system, the server uses a generative AI model to generate the optimal response to that query. At the same time, the emotion engine analyzes the user's emotions and optimizes the response based on that emotional data. For example, if the user expresses anger, the system will provide a more polite response that includes an apology.

[0587] Providing answers

[0588] The terminal displays the generated answer from the server to the user, and the user can take action based on the suggested answer. For example, the suggested action for "what to do if an order is delayed" might be "send an apology email and offer a coupon for free shipping on the next order."

[0589] Gathering feedback and retraining

[0590] Users can provide feedback on the appropriateness of the suggested answers, which the server collects and uses to retrain the AI ​​model, a process that improves the model's accuracy and makes it better able to respond to future inquiries.

[0591] Automatic generation of training manuals

[0592] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated and provided to the terminal in PDF format.

[0593] Through the above process, this system will improve the quality and efficiency of customer service and call center operations. The introduction of an emotion engine will enable responses that take user emotions into consideration, which is expected to improve customer satisfaction.

[0594] The processing flow will be explained below.

[0595] Step 1: Collect data on past cases

[0596] The server collects past case data from call centers and customer support systems. This process includes information such as the nature of the complaint, how it was handled, and the customer's reaction. For example, data on how to handle a delayed delivery of a product can be obtained.

[0597] Step 2: Preprocessing the data

[0598] The server preprocesses the collected case data. During preprocessing, the data is cleansed, unnecessary information is removed, and necessary information is extracted and structured. Specifically, text data is tokenized and important keywords are extracted.

[0599] Step 3: Create a dataset

[0600] The server splits the preprocessed data into training and testing sets, for example, 80% of the total data as the training set and 20% as the testing set.

[0601] Step 4: Building and training an AI model

[0602] The server uses a machine learning algorithm to build a generative AI model and trains it using a training set. Specifically, it learns pairs of complaint details and response methods and has the model recognize patterns.

[0603] Step 5: Evaluate and tune the model

[0604] The server uses the test set to evaluate the trained AI model, for example, assessing the model's accuracy, suitability, and speed of response, and adjusting any inaccuracies. It also optimizes hyperparameters as necessary.

[0605] Step 6: Implementing the Emotion Engine

[0606] The server will install an emotion engine and add the ability to analyze emotions from the user's text input or voice commands. The emotion engine will recognize the user's emotions and record the data. For example, if it detects that the user is "angry," it will save this emotion data as the analysis result.

[0607] Step 7: Providing a User Interface

[0608] The server designs the user interface and provides it through the terminal. The interface includes forms and input fields, allowing users to easily enter inquiries. For example, a form is provided for entering "What to do if the product arrives damaged."

[0609] Step 8: Query entry and sentiment analysis

[0610] The user inputs a query through the interface, for example, "The product arrived damaged." At the same time, the emotion engine analyzes the user's emotions and detects "anger."

[0611] Step 9: Generate an answer with the AI ​​model

[0612] The server receives the query and uses a generative AI model to generate the optimal answer. Based on data from the emotion engine, the server provides an answer that takes the user's emotions into consideration. For example, it may suggest sending an email with an apology and applying a discount on the next order.

[0613] Step 10: Provide your answers

[0614] The device displays the generated answer from the server to the user. The user checks the suggested answer and takes action. For example, the device displays specific measures such as "send an apology email and immediately ship a replacement product" as a response to the question "What to do if the product arrives damaged."

[0615] Step 11: Enter your feedback

[0616] The user provides feedback on the appropriateness of the provided answer, for example, by rating whether the proposed response was effective.

[0617] Step 12: Gather feedback

[0618] The server collects feedback from users, including the appropriateness of the response and areas for improvement. For example, specific opinions such as "the response was a little slow" are collected.

[0619] Step 13: Retraining the AI ​​model

[0620] The server retrains the generative AI model based on the collected feedback and emotion data, improving its accuracy and response capabilities. For example, it can learn how to respond quickly to anger.

[0621] Step 14: Automatic generation of training manuals

[0622] The server automatically generates training manuals using the collected data and learning results. For example, a "manual for dealing with product defects" is generated.

[0623] Step 15: Provide training manuals

[0624] The server displays the generated training manual on the terminal and makes it available to new and current employees, for example, by providing it as a downloadable PDF file.

[0625] Through these steps, the system will improve the quality and efficiency of customer service and call center operations. The introduction of an emotion engine will enable responses that take user emotions into consideration, which is expected to improve customer satisfaction.

[0626] Example 2

[0627] 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."

[0628] In recent years, there has been a demand for improved quality and efficiency in customer service and call center operations when it comes to handling complaints. However, conventional systems have struggled to respond in a way that takes user emotions into consideration, limiting the extent to which they can improve customer satisfaction. Furthermore, they lack the ability to automatically generate appropriate response manuals, making the efficiency of employee training an issue.

[0629] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting past case data, means for preprocessing the collected case data, means for training a generative AI model using the preprocessed data, means for providing an interface through which a user can input an inquiry, means for using an emotion engine to analyze the user's emotions, means for generating an optimal answer to the user's inquiry using the generative AI model, means for providing the generated answer to the user, means for collecting feedback from the user, means for retraining the generative AI model using the collected feedback, means for automatically generating a training manual, and means for providing the generated training manual. This enables improved quality of complaint handling and business efficiency. Furthermore, the introduction of an emotion engine enables responses that take user emotions into consideration, which is expected to improve customer satisfaction. Furthermore, the automatic generation of training manuals allows for efficient training of new employees and skill development of current employees.

[0630] "Past case data" refers to data collected from customer response history, including the content of complaints, response methods, and feedback from customers.

[0631] "Preprocessing" refers to the process of cleansing unnecessary noise from collected data and tokenizing important text information to convert it into structured data.

[0632] A "generative AI model" is a model that is constructed using machine learning algorithms with preprocessed data to predict the appropriate way to handle a complaint.

[0633] An "interface" is a tool or platform through which a user can enter a query and interact with a system, such as a web form or a mobile app.

[0634] An "emotion engine" is a system that analyzes emotions from a user's text input or voice commands, and has the ability to record the user's emotional state as data.

[0635] An "Answer" is an optimal response to a user's query, created using a generative AI model and emotion engine.

[0636] "Feedback" refers to opinions and ratings provided by users regarding the appropriateness of proposed answers.

[0637] "Retraining" is the process of using collected feedback to readjust the parameters of a generative AI model to improve the model's accuracy.

[0638] The "training manual" is a document that is automatically generated based on collected data and the learning results of the generating AI model, and is used to train new employees and improve the skills of current employees.

[0639] One embodiment of the present invention is described in detail below. This system is designed to improve the quality and efficiency of customer service and complaint handling in call center operations. By combining the process of collecting and preprocessing past case data, generating and training a generative AI model, and incorporating an emotion engine that recognizes user emotions, more appropriate and emotion-sensitive responses are possible.

[0640] First, the server collects past case data from call centers and customer support systems. Specifically, it obtains data including the details of complaints, how they were handled, and customer feedback. Data is collected from a database and extracted using the Python "pandas" library. Next, the server preprocesses this data. During preprocessing, unnecessary data is cleansed, and important text information is tokenized using the "NLTK" library and converted into structured data.

[0641] Using the preprocessed data, the server builds and trains a generative AI model. TensorFlow and PyTorch are used to build the model, and the model is trained using a training dataset. Specifically, the model learns to predict the appropriate response method using paired data of the complaint content and the response method. For example, the following prompt sentence is used:

[0642] "How should you respond if a customer complains about the following? Complaint: ____"

[0643] The server then deploys an emotion engine, which uses IBM Watson and Microsoft Azure Emotion API to analyze emotions from the user's text input and voice commands. This engine records the user's emotional state as data and reflects it in the generative AI model.

[0644] The server designs an interface that allows users to enter inquiries and provides it via the terminal. The interface is provided as a web-based form or mobile app, and is designed so that users can easily access the system and enter inquiries. For example, a customer service representative can inquire about "what to do if a customer's order is delayed" through the system.

[0645] When a user inputs an inquiry into the system, the server uses a generative AI model to generate the optimal response to that inquiry. At the same time, an emotion engine analyzes the user's emotions and optimizes the response based on that emotional data. For example, if the user expresses anger, the system will provide a response that includes a more polite apology. As a specific example, when asked what to do if an order is delayed, the system will suggest responses such as "send an apology email and offer a coupon for free shipping on the next order."

[0646] The device displays the answer generated by the server to the user, who can then act on the suggested answer. The user can also provide feedback on whether the suggested answer is appropriate. The server collects the feedback and uses it to retrain the generative AI model to improve the quality of the answer. This improves the model's accuracy and makes it better able to respond to future inquiries.

[0647] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals can be used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" can be provided in PDF format.

[0648] The above is an embodiment of the system of the present invention, which improves the quality and efficiency of customer service and complaint handling in call center operations. The introduction of an emotion engine enables responses that take user emotions into consideration, which is expected to improve customer satisfaction. Furthermore, the automatic generation of training manuals allows for efficient training of new employees and skill development for current employees.

[0649] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0650] Step 1: Data collection

[0651] The server collects past case data from call centers and customer support systems. This includes data on complaints, response methods, and customer feedback. Specifically, it extracts data from a database using SQL queries. The input is past case data, and the output is the extracted raw data.

[0652] Step 2: Data Preprocessing

[0653] The server preprocesses the collected data. Specifically, it cleanses unnecessary data using the Python "pandas" library and tokenizes text information using the "NLTK" library to convert it into structured data. The input is raw data, and the output is cleansed and tokenized structured data.

[0654] Step 3: Generate an AI model

[0655] The server uses the preprocessed data to build a generative AI model. It uses the machine learning library "TensorFlow" or "PyTorch" to define a neural network model. For example, it uses pairs of complaints and response methods as input data. The input is structured training data, and the output is an initialized generative AI model.

[0656] Step 4: Training the AI ​​model

[0657] The server trains the generative AI model. Specifically, it uses the preprocessed data to apply a learning algorithm (e.g., batch learning, setting the number of epochs, etc.) to improve the accuracy of the model. The input is the training data and the initialized AI model, and the output is the trained generative AI model.

[0658] Step 5: Implementing the Emotion Engine

[0659] The server installs an emotion engine and analyzes emotions from the user's text input or voice commands. IBM Watson and Microsoft Azure Emotion API are used as emotion engines. The input is text or voice data from the user, and the output is analyzed emotional data.

[0660] Step 6: Providing a User Interface

[0661] The server designs a query input interface and provides it through the terminal, such as a web form or a mobile app. The input is the query from the user, and the output is the query data sent to the system.

[0662] Step 7: Receiving and analyzing inquiries

[0663] A user inputs a query into the system. The server receives the query and uses a generative AI model to generate the optimal answer to that query. Specifically, a prompt such as "What should we do if a customer's order is delayed?" is used. The input is the user's query, and the output is the generated answer.

[0664] Step 8: Analyze sentiment data and optimize responses

[0665] The server uses an emotion engine to analyze the user's emotions and optimizes the answer of the generative AI model based on the emotion data. The input is the query and the generated answer, and the output is an optimized answer that takes emotions into consideration.

[0666] Step 9: Provide your answers

[0667] The device displays the generated answer from the server to the user, specifically on a web page or in the chat screen of the app. The input is the optimized answer, and the output is what is displayed to the user.

[0668] Step 10: Gather feedback

[0669] The user provides feedback on the appropriateness of the response. The server collects this feedback and stores it in a database. The input is the user feedback and the output is the collected feedback data.

[0670] Step 11: Retraining the AI ​​model

[0671] The server uses the collected feedback to retrain the generative AI model. Specifically, it uses the feedback data as new training data and readjusts the model's parameters. The inputs are the feedback data and the existing generative AI model, and the output is the retrained generative AI model.

[0672] Step 12: Automatic generation of training manuals

[0673] The server automatically generates training manuals based on the training data and model results. Specifically, it uses a Python script to analyze past cases and their effectiveness, and generates training manuals. The input is the training data and the model's learning results, and the output is the generated training manual (e.g., in PDF format).

[0674] Step 13: Provide training manuals

[0675] The terminal provides the generated training manual to the user. Specifically, it is provided as a download link or email attachment. The input is the generated training manual, and the output is the manual provided to the user.

[0676] (Application example 2)

[0677] 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."

[0678] When it comes to customer service for autonomous vehicles, there are cases where responses to user complaints or problems are delayed or inappropriate. In such cases, measures that take user feelings into consideration are required, and improving customer satisfaction is a challenge.

[0679] 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 collecting past response case data, means for preprocessing the collected response case data, means for training a generative AI model using the preprocessed data, means for providing an interface through which a user can input a query, means including an emotion engine for analyzing the user's emotions, means for generating an optimal answer to the user's query using the generative AI model, means for optimizing the generated answer based on user emotion analysis, means for providing the optimized answer to the user, means for collecting feedback from the user, means for retraining the generative AI model using the collected feedback, means for automatically generating a training manual, and means for providing the generated training manual. This enables user complaints and problems regarding autonomous vehicles to be handled quickly and appropriately, thereby improving customer satisfaction.

[0680] "Past case data" refers to data that includes details of past complaints and troubles, the methods used to deal with them, and feedback from customers.

[0681] "Preprocessing" is the process of removing unnecessary data from collected data and structuring important text information.

[0682] A "generative AI model" is a model trained using machine learning algorithms and used to predict optimal responses based on past response patterns.

[0683] A "user interface" is an interface through which a user inputs queries into a system, and supports communication between the user and the system.

[0684] The "emotion engine" is a system that analyzes emotions from the user's text and voice input and records that emotional state as data.

[0685] "Feedback" means opinions or evaluations provided by users regarding the appropriateness or quality of a system's response.

[0686] "Retraining" is the process of retraining a generative AI model using newly collected feedback to improve the model's accuracy and responsiveness.

[0687] An "education manual" is a document automatically generated based on collected data and learning results, and is used to train new employees and improve the skills of current employees.

[0688] The present invention is a system for quickly and appropriately responding to user complaints and problems with autonomous vehicles. This system collects and analyzes past case data and provides optimal countermeasures that take user emotions into consideration. An embodiment of the present invention is described in detail below.

[0689] Data collection and preprocessing

[0690] First, the server collects past case data from the autonomous vehicle's operation management system and customer response system. This includes complaint details, response methods, and customer feedback. Next, the server preprocesses this data, removing unnecessary data and tokenizing and structuring useful text information.

[0691] AI model generation and training

[0692] The server uses the preprocessed data to build a generative AI model and trains it using machine learning algorithms, allowing it to learn past response patterns and predict optimal responses, such as the appropriate way to respond to complaints.

[0693] Introducing the Emotion Engine

[0694] The server then uses an emotion engine to analyze the user's text input. The emotion engine uses natural language processing tools such as TextBlob to analyze the emotion into positive, negative, and neutral, and records the emotional state as data. This allows the AI ​​model to reflect appropriate emotional responses to the user's inquiry.

[0695] User Interface

[0696] The server designs and provides an interface through which users can input inquiries via terminals, allowing users to easily report complaints and problems.

[0697] User inquiries and sentiment analysis

[0698] When a user inputs a query into the system, the server uses a generative AI model to generate the optimal response for that query. At the same time, the emotion engine analyzes the user's emotions and optimizes the response based on that emotional data. For example, if the user expresses anger, the server will provide a more polite response that includes an apology.

[0699] Examples:

[0700] User input: "My car is stuck in the middle of the road! I'm so annoyed. What can you do?"

[0701] Emotion analysis result: "negative"

[0702] Example AI response: "We apologize for the inconvenience. Our technical team is working quickly to resolve the issue. We'll offer you a 20% discount coupon for your next visit."

[0703] Providing answers

[0704] The device displays the generated answer from the server to the user, who then acts on the suggested answer.

[0705] Gathering feedback and retraining

[0706] Users can provide feedback on the appropriateness of the proposed answers, which the server collects and uses to retrain the generative AI model, a process that improves the model's accuracy and makes it better able to respond to future inquiries.

[0707] Automatic generation of training manuals

[0708] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are used to train new employees and improve the skills of current employees. For example, a "vehicle malfunction response manual" is generated and provided to the terminal in PDF format.

[0709] In this way, the system of the present invention improves the quality of customer service in self-driving vehicles and is expected to improve customer satisfaction by enabling emotionally sensitive service.

[0710] Example prompt sentence:

[0711] Prompt: "User's complaint: My car got stuck in the middle of the road! I'm so upset. What can you do about it? Emotion: negative Suggest the best solution:"

[0712] This embodiment makes it possible to specifically implement the present invention.

[0713] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0714] Step 1:

[0715] The server collects past case data from the autonomous vehicle's operation management system and customer response system. The collected data includes the details of the complaint, how it was handled, and customer feedback. The input is raw data obtained from the operation management system database, and the output is sent to pre-processing through the collection step.

[0716] Step 2:

[0717] The server preprocesses the collected data. During this process, unnecessary data is cleansed and useful text information is tokenized and structured. Processes such as text normalization, stop word removal, and tokenization are performed. The input is the collected raw data, and the output is cleansed and structured data.

[0718] Step 3:

[0719] The server uses the preprocessed data to generate and train a generative AI model. The model is trained using machine learning algorithms (e.g., deep learning) to learn appropriate countermeasures based on past response patterns. The input is the preprocessed data, and the output is a trained generative AI model.

[0720] Step 4:

[0721] The server uses an emotion engine to analyze the user's emotions. In this process, the text entered by the user is analyzed to determine whether the emotion is positive, negative, or neutral, and the data is recorded. The input is the text input from the user, and the output is the analyzed emotion data.

[0722] Step 5:

[0723] The user uses an interface to input inquiries and complaints. The server designs this interface to make it easy for the user to input inquiries. The input is the complaint or inquiry content from the user, and the output is the inquiry data that is sent to the server.

[0724] Step 6:

[0725] The server uses a generative AI model to generate the optimal answer to a user's inquiry. The answer is optimized based on the user's emotional data analyzed by the emotion engine. For example, if the emotion included anger, a particularly polite response would be suggested. The input is the user's inquiry data and emotional data, and the output is an optimized answer.

[0726] Step 7:

[0727] The terminal displays the generated answer to the user, who then reviews the proposed answer and takes action. The input is the optimized answer sent by the server, and the output is the generated answer that is displayed to the user.

[0728] Step 8:

[0729] Users provide feedback on the appropriateness of the proposed answers, which the server collects and uses to retrain the generative AI model. The input is the user feedback data, and the output is the retraining data.

[0730] Step 9:

[0731] The server automatically generates a training manual based on the collected data and learning results. This manual is used to train new employees and improve the skills of current employees. The inputs are the collected data and learning results, and the output is a training manual (e.g., a PDF document).

[0732] In this way, each step works together to create a system that can quickly and appropriately handle complaints and problems with self-driving vehicles.

[0733] 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.

[0734] 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.

[0735] 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.

[0736] [Third embodiment]

[0737] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0738] 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.

[0739] 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).

[0740] 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.

[0741] 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.

[0742] 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).

[0743] 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.

[0744] 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.

[0745] 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.

[0746] 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.

[0747] 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.

[0748] 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."

[0749] MODE FOR CARRYING OUT THE INVENTION

[0750] An embodiment of the present invention will be described below.

[0751] This system is designed to improve the efficiency and quality of customer service and complaint handling in call center operations. It collects and preprocesses past case data, generates and trains AI models, provides a user interface, handles user inquiries, collects feedback and retrains, and automatically generates training manuals.

[0752] Data collection and preprocessing

[0753] First, the server collects past case data collected from call centers and customer support operations. For example, it obtains historical data on responses to complaints such as "the product arrived damaged." Next, the server preprocesses this data. In preprocessing, unnecessary data is cleansed and important data is extracted and structured. In particular, text data is tokenized, and important keywords are extracted.

[0754] AI model generation and training

[0755] Using the preprocessed data, the server builds and trains a generative AI model. This is the process of using machine learning algorithms to train the model and learn from past response patterns. For example, it can learn the appropriate way to respond to complaints and predict the optimal response.

[0756] Providing a user interface

[0757] Next, the server designs and implements a user interface, which is provided via a terminal. This interface is designed to allow users to easily access the system and input inquiries. For example, a customer service representative can ask the system, "What should we do if a customer's order is delayed?"

[0758] Responding to user inquiries

[0759] When a user inputs a query into the system, the server uses the generative AI model to generate the optimal response to that query. For example, the server suggests "how to respond to a delayed order" by suggesting "sending an apology email and offering a discount on the next order." The device then displays this response to the user.

[0760] Gathering feedback and retraining

[0761] Users can provide feedback on the appropriateness of the suggested answers, which the server collects and uses to retrain the AI ​​model, a process that improves the model's accuracy and allows it to better respond to future inquiries.

[0762] Automatic generation of training manuals

[0763] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated and provided to the terminal in PDF format.

[0764] Through the above process, this system will improve the quality of customer service and call center operations and increase operational efficiency.

[0765] The processing flow will be explained below.

[0766] Step 1: Collecting interaction data

[0767] The server collects past case data from call centers and customer support systems, specifically data including complaint details, response methods, and customer feedback.

[0768] Step 2: Preprocessing the data

[0769] The server preprocesses the collected case data, for example by cleansing unnecessary information and tokenizing and structuring important text information.

[0770] Step 3: Create a dataset

[0771] The server splits the preprocessed data into a training set and a test set, typically 80% for training and 20% for testing.

[0772] Step 4: Building and training an AI model

[0773] The server uses a machine learning algorithm to build a generative AI model and trains it using a training set. Specifically, it learns pairs of complaint details and response methods and has the model recognize patterns.

[0774] Step 5: Evaluate and tune the model

[0775] The server evaluates the trained AI model using the test set, and if the accuracy is low, adjusts the hyperparameters to optimize the model.

[0776] Step 6: Providing a User Interface

[0777] The server designs an interface that allows users to input queries and displays it on the terminal. This interface allows queries to be made by filling out a form or operating buttons.

[0778] Step 7: Enter your inquiry

[0779] A user inputs a specific complaint or inquiry into the system. For example, a user may inquire, "What should I do if my product delivery is delayed?"

[0780] Step 8: Generate an answer with the AI ​​model

[0781] The server receives the user's inquiry and uses a generative AI model to generate the optimal response, such as suggesting specific actions such as sending an apology email and offering a coupon for free shipping on the next purchase.

[0782] Step 9: Provide your answers

[0783] The device displays the generated answers from the server to the user, who then reviews the suggested answers and takes action.

[0784] Step 10: Provide feedback

[0785] The user provides feedback on the appropriateness of the provided answer, for example, rating whether the proposed action was effective or not.

[0786] Step 11: Gather feedback

[0787] The server collects feedback from users, which is used to improve the model.

[0788] Step 12: Retraining the AI ​​model

[0789] The server retrains the generative AI model based on the collected feedback, improving its accuracy and making it better able to respond to the next inquiry.

[0790] Step 13: Automatic generation of training manuals

[0791] The server automatically generates training manuals using the collected data and learning results. For example, a "manual for dealing with product defects" is created.

[0792] Step 14: Provide training manuals

[0793] The server displays the generated training manual on the terminal and makes it available to new and current employees, for example, by making it available for download in PDF format.

[0794] In this way, this system improves the quality of complaint handling and increases operational efficiency.

[0795] Example 1

[0796] 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."

[0797] In traditional customer service and call center operations, there is a demand for improved efficiency and quality in handling complaints. However, there are insufficient methods for effectively utilizing past case data, resulting in issues with the consistency and accuracy of responses. Furthermore, there is a lack of advanced functionality, such as automatic model updating based on appropriate feedback and automatic generation of training materials for new employees.

[0798] 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.

[0799] In this invention, the server includes means for collecting past case data, means for preprocessing the collected case data, means for training a machine learning model using the preprocessed data, means for providing an interface through which a user can input an inquiry, means for generating an optimal answer to the user's inquiry using the machine learning model, means for providing the generated answer to the user, means for collecting feedback from the user, means for retraining the machine learning model using the collected feedback, means for automatically generating training materials, and means for providing the generated training materials. This enables improved efficiency and quality in complaint handling, and also improves the consistency and accuracy of responses. Furthermore, automatic model updating and automatic generation of training materials are realized by utilizing appropriate feedback.

[0800] "Past case data" refers to records of past complaint and inquiry handling in call centers and customer service operations.

[0801] "Preprocessing" refers to the process of removing unnecessary information from collected data and extracting and formatting important information.

[0802] A "machine learning model" is a model that uses an algorithm to learn from collected and preprocessed data and predict the optimal response to future inquiries.

[0803] "Interface" refers to the operation screen and input form that users use to access the system and enter inquiries.

[0804] A "generative AI model" refers to an artificial intelligence model built using machine learning algorithms based on collected and pre-processed data.

[0805] "Optimal response" refers to the most appropriate response method generated based on past response cases and learning results, based on the user's inquiry.

[0806] "Feedback" refers to a user's rating or comment on the appropriateness of a proposed answer.

[0807] "Retraining" refers to the process of retraining an existing machine learning model based on collected feedback to improve its accuracy and performance.

[0808] "Educational materials" refers to guides and manuals on how to handle complaints and customer service, intended to educate new employees and improve the skills of current employees.

[0809] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the accompanying drawings. The present invention aims to improve the efficiency of customer service and complaint handling in call center operations, thereby improving the quality of responses.

[0810] First, the server collects past case data collected from call centers and customer support operations. For example, it obtains historical data on complaints such as "the product arrived damaged." This data is obtained from a database. Next, the server preprocesses this data. In preprocessing, unnecessary data such as dates and personal information is cleansed, and important data such as the type of complaint, how it was handled, and the results are extracted and structured. Furthermore, the text data is tokenized, and important keywords are extracted. Specifically, keywords such as "product," "damaged," and "arrived" are extracted.

[0811] Next, using the preprocessed data, the server uses a machine learning algorithm (e.g., using TensorFlow) to build and train a generative AI model. In this process, the server learns past complaint handling patterns and predicts the optimal response method. For example, it learns how to respond to "damaged product" cases and provides the most appropriate relevant information.

[0812] The server then designs and implements a user interface, which is provided to users through their terminals. This interface is provided through a web browser or dedicated application, and is designed to allow users to easily access the system and input inquiries. For example, a customer service representative can inquire about "what to do if a customer's order is delayed."

[0813] When a user inputs a query into the system, the server uses the generative AI model to generate the optimal response to that query. For example, the server suggests "how to respond to a delayed order" by suggesting "sending an apology email and offering a discount on the next order." The device then displays this response to the user.

[0814] Additionally, users can provide feedback on the appropriateness of the suggested answers. The server collects this feedback and retrains the AI ​​model. This process improves the model's accuracy and allows it to respond more appropriately to future inquiries.

[0815] Finally, the server automatically generates training materials based on the collected data and learning results. These materials are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated and provided to the terminal in PDF format.

[0816] As a specific example, the server collects complaint processing history data for "the product arrived damaged," tokenizes it, and extracts keywords such as "product," "damaged," and "arrived." Using the preprocessed data, a generative AI model is built using a machine learning algorithm to learn the optimal response method for "damaged product" cases. A user interface is designed and provided via a terminal, allowing the user to ask the system "how to respond if an order is late." The server generates "send an apology email and offer a discount on the next order," which is then displayed on the terminal. Based on user feedback, the server retrains the model and generates and provides a "manual for responding to product defects" in PDF format.

[0817] An example of a prompt for a generative AI model is:

[0818] Please suggest the best way to handle a complaint based on the complaint handling history below.

[0819] Complaint: The product arrived damaged.

[0820] Response history:

[0821] 1. Send new products to customers immediately

[0822] 2. Requesting a return of damaged goods at no cost to the customer

[0823] 3. Offer a discount coupon for your next order

[0824] Please suggest the best way to respond.

[0825] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0826] Step 1:

[0827] The server retrieves past case data collected from call centers and customer support operations from a database. The input is the database and its query, and the output is case data including the complaint details, response methods, results, etc. Specifically, the server uses an SQL query to extract the relevant data from the database.

[0828] Step 2:

[0829] Remove unnecessary information from the case data acquired by the server (data cleansing). The input is the case data acquired in step 1, and the output is the cleansed data. Specifically, it removes unnecessary fields such as dates and personal information.

[0830] Step 3:

[0831] The server structures the data. The input is the cleansed data from step 2, and the output is structured data. Specifically, the server reclassifies the data into fields such as complaint type, response method, and outcome.

[0832] Step 4:

[0833] The server tokenizes the text data and extracts important keywords. The input is the structured data from step 3, and the output is a list of keywords. Specifically, the server uses a natural language processing tool to tokenize the text and extract keywords such as "product," "damaged," and "arrived."

[0834] Step 5:

[0835] The server uses a machine learning algorithm (e.g., TensorFlow) to build a generative AI model. The input is the keyword list extracted in step 4 and the preprocessed data, and the output is the generated AI model. Specifically, the server inputs the data into the algorithm and trains the model.

[0836] Step 6:

[0837] The server stores the learning results and designs and implements the user interface. The input is the AI ​​model generated in step 5, and the output is the completed user interface. Specifically, the code to provide the interface through a web browser or dedicated application is written and deployed.

[0838] Step 7:

[0839] The terminal allows access to the system through a user interface. The input is a request to access the user interface, and the output is the display of the interface. Specifically, the terminal displays a login screen or a search screen.

[0840] Step 8:

[0841] A user inputs an inquiry into the system. The input is the inquiry entered by the user, and the output is the content being sent to the system. A specific operation is when the user inputs "What to do if an order is late."

[0842] Step 9:

[0843] The server uses the generative AI model to generate the optimal answer to the user's inquiry. The input is the inquiry entered in step 8, and the output is the generated answer. Specifically, the server inputs the inquiry into the model and generates an answer such as "send an apology email and offer a discount on the next order."

[0844] Step 10:

[0845] The terminal displays the generated answer to the user. The input is the answer generated in step 9, and the output is the display of the answer to the user. Specifically, the terminal displays the answer on the screen.

[0846] Step 11:

[0847] The user provides feedback on the appropriateness of the proposed answer. The input is the user's feedback, and the output is the feedback information being sent to the system. The specific behavior is that the user provides feedback as "appropriate" or "inappropriate."

[0848] Step 12:

[0849] The server collects feedback and retrains the AI ​​model. The input is user feedback and existing model data, and the output is the retrained AI model. Specifically, the server reflects the feedback in the model and retrains it.

[0850] Step 13:

[0851] The server automatically generates training materials based on the collected data and learning results. The input is the retrained AI model and learning results, and the output is the generated training manual. Specifically, the server generates content such as a "manual for dealing with product defects" in PDF format.

[0852] Step 14:

[0853] The terminal provides the generated educational materials to the user. The input is the generated educational manual, and the output is the provision of the manual. Specifically, the terminal provides a download link so that the user can download the manual.

[0854] (Application example 1)

[0855] 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."

[0856] In virtual stores, customer service requires quick and appropriate responses to complaints and inquiries, but conventional systems often make response processes cumbersome and result in inconsistent quality. Furthermore, there is a lack of feedback collection and learning processes to improve response accuracy, making continuous improvement difficult.

[0857] 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.

[0858] In this invention, the server includes means for collecting past case data, means for preprocessing the collected case data, means for training a generative AI model using the preprocessed data, means for providing an interface through which a user can input an inquiry, means for generating an optimal answer to the user's inquiry using the generative AI model, means for providing the generated answer to the user, means for collecting feedback from the user, means for retraining the generative AI model using the collected feedback, means for automatically generating a training manual, means for providing the generated training manual, means for responding to inquiries using a smart device, and means for accepting inquiries and providing answers in real time, thereby enabling improved quality and efficiency of customer service.

[0859] "Past case data" refers to data that includes the details of responses to complaints and inquiries recorded in the past in customer service operations.

[0860] "Means of collection" refers to equipment and software used to store data on past cases on a server, etc.

[0861] "Preprocessing means" refers to techniques and methods for cleansing collected data and extracting and structuring important information.

[0862] A "generative AI model" refers to an artificial intelligence model that is built using machine learning algorithms and has learned past response patterns.

[0863] "User Interface" means the screen or operating environment through which a user accesses the system and enters inquiries.

[0864] "Means for generating optimal answers" refers to technologies and functions that use generative AI models to output optimal answers to user inquiries.

[0865] "Feedback collection means" refers to any equipment or software used to collect user opinions or ratings regarding the appropriateness of responses.

[0866] "Retraining methods" are techniques or methods for retraining a generative AI model based on collected feedback to improve the model's accuracy.

[0867] "Means for automatically generating training manuals" refers to technologies and methods for automatically creating manuals for improving customer service skills based on case data and learning results.

[0868] "Smart devices" refer to portable electronic devices that can connect to the Internet, such as smartphones and head-mounted displays.

[0869] "Real-time means" refers to technologies and functions that instantly generate and provide answers to user inquiries.

[0870] MODE FOR CARRYING OUT THE INVENTION

[0871] A system for realizing the present invention in a virtual store is configured using the following hardware and software.

[0872] Hardware and software used

[0873] Hardware

[0874] Server: Used to process, store, and learn from large amounts of data.

[0875] Smart devices: Smartphones, head-mounted displays, etc., are the interface through which users input inquiries and receive responses.

[0876] software

[0877] TensorFlow: A machine learning library used to train generative AI models.

[0878] Django: Web framework, provides the user interface.

[0879] NLTK: A natural language processing library used for data preprocessing.

[0880] MySQL: Database management system used to store and manage data.

[0881] System configuration and operation

[0882] Data collection and preprocessing

[0883] The server collects past case data from call centers and customer service operations. The collected data is tokenized using NLTK to extract important keywords and phrases. The data is then cleansed of unnecessary data and stored in a structured format in a database.

[0884] Training generative AI models

[0885] Using the preprocessed data, the server uses TensorFlow to build and train a generative AI model, a process that allows the model to learn from past response patterns and predict the best response.

[0886] Providing a user interface

[0887] Using the Django framework, the server designs and implements the user interface. This interface, which runs on smart devices, is designed to allow users to easily input inquiries. For example, users can input "What should I do if my order is delayed" into the smartphone application.

[0888] Responding to user inquiries

[0889] When a user inputs a query, the server uses a generative AI model to generate the optimal answer. This answer is displayed on the smart device and provided to the user. For example, when asked how to respond to a delayed order, the server suggests sending an apology email and offering a discount on the next order.

[0890] Gathering feedback and retraining

[0891] Users can provide feedback on the appropriateness of the suggested answers, and the server collects this feedback to retrain the generative AI model. This process improves the model's accuracy, enabling it to better respond to future inquiries.

[0892] Automatic generation of training manuals

[0893] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are provided in PDF format and are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated.

[0894] Examples of specific examples and prompts

[0895] Example 1: When a customer asks, "What should we do if my order is late?" the system responds, "We will send you an apology email and offer you a discount on your next order."

[0896] Examples of prompts:

[0897] "What is the best way to handle a complaint?"

[0898] "What do you do if a customer inquires about a delayed order?"

[0899] Please tell me what to do if the product arrives damaged.

[0900] This makes it possible for this system to improve the quality and efficiency of customer service.

[0901] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0902] Step 1:

[0903] The server collects past case data from customer support history. The input is past support history data, such as complaints about "the product arrived damaged." The server saves this data in MySQL, a database management system. The output is a structured case database.

[0904] Step 2:

[0905] The server preprocesses the collected case data. The input is the stored case data, which is tokenized using NLTK to remove unnecessary information. Specific operations include cleaning the text data, extracting keywords, and structuring. The output is the preprocessed case data.

[0906] Step 3:

[0907] The server uses the preprocessed data to build and train a generative AI model. The input is the preprocessed corresponding case data, and TensorFlow is used to train the machine learning model. Specific operations include data splitting, model selection, training, and evaluation. The output is a trained generative AI model.

[0908] Step 4:

[0909] The server provides the user interface. The input is the interface code built using Django, allowing users to easily enter queries into the system. Specific operations include form design, deployment, and access management. The output is the interface available to users.

[0910] Step 5:

[0911] A user inputs a query from a terminal. The input is a query text by the user, such as "What should I do if my order is late?" The terminal sends this input to the server. The output is the query data sent to the server.

[0912] Step 6:

[0913] The server uses a generative AI model to generate an optimal answer to a user's query. The input is the user's query text, and the generative AI model predicts the optimal answer. Specific operations include performing model inference. The output is the generated answer text.

[0914] Step 7:

[0915] The terminal displays the answer received from the server to the user. The input is the answer text sent by the server, which is displayed in the user interface. Specific operations include adjusting the display layout and notifying the user. The output is the answer that the user can view.

[0916] Step 8:

[0917] The user provides feedback on the appropriateness of the proposed answer. The input is the user's feedback text, and the terminal sends this feedback to the server. The output is the feedback data to the server.

[0918] Step 9:

[0919] The server retrains the generative AI model based on the collected feedback. The input is user feedback data and collected case data, which are used to retrain the model. Specific operations include integrating the feedback data and running the retraining process. The output is an updated generative AI model.

[0920] Step 10:

[0921] The server automatically generates training manuals. The input is a trained or updated generative AI model and corresponding case data, and the training manual is created based on this. Specific operations include selecting a template, writing the content, and executing the automatic generation process. The output is a completed training manual.

[0922] Step 11:

[0923] The server provides the generated training manual. The input is the completed training manual, which is provided to the terminal in PDF format or other format. Specific operations include file generation and distribution. The output is a usable training manual.

[0924] 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.

[0925] MODE FOR CARRYING OUT THE INVENTION

[0926] An embodiment of the present invention will be described below.

[0927] This system is designed to improve the quality and efficiency of customer service and complaint handling in call center operations. By combining the process of collecting and preprocessing past case data, generating and training an AI model, and combining it with an emotion engine that recognizes user emotions, it becomes possible to provide more appropriate and emotionally sensitive responses.

[0928] Data collection and preprocessing

[0929] First, the server collects past case data from call centers and customer support systems. Specifically, it acquires data including the details of complaints, how they were handled, and customer feedback. Next, the server preprocesses this data. In preprocessing, unnecessary data is cleansed and important text information is tokenized and structured.

[0930] AI model generation and training

[0931] Using the preprocessed data, the server builds and trains a generative AI model. This is the process of using machine learning algorithms to train the model and learn from past response patterns. For example, it can learn the appropriate way to respond to complaints and predict the optimal response.

[0932] Introducing the Emotion Engine

[0933] Next, the server introduces an emotion engine, which analyzes emotions from the user's text input or voice commands and records the emotional state as data. This allows the generative AI model to reflect appropriate emotional responses to the user's inquiries.

[0934] Providing a user interface

[0935] The server designs an interface that allows users to input inquiries and provides it through a terminal. This interface is designed so that users can easily access the system and input inquiries. For example, a customer service representative can inquire about "what to do if a customer's order is delayed."

[0936] User inquiries and sentiment analysis

[0937] When a user inputs a query into the system, the server uses a generative AI model to generate the optimal response to that query. At the same time, the emotion engine analyzes the user's emotions and optimizes the response based on that emotional data. For example, if the user expresses anger, the system will provide a more polite response that includes an apology.

[0938] Providing answers

[0939] The terminal displays the generated answer from the server to the user, and the user can take action based on the suggested answer. For example, the suggested action for "what to do if an order is delayed" might be "send an apology email and offer a coupon for free shipping on the next order."

[0940] Gathering feedback and retraining

[0941] Users can provide feedback on the appropriateness of the suggested answers, which the server collects and uses to retrain the AI ​​model, a process that improves the model's accuracy and makes it better able to respond to future inquiries.

[0942] Automatic generation of training manuals

[0943] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated and provided to the terminal in PDF format.

[0944] Through the above process, this system will improve the quality and efficiency of customer service and call center operations. The introduction of an emotion engine will enable responses that take user emotions into consideration, which is expected to improve customer satisfaction.

[0945] The processing flow will be explained below.

[0946] Step 1: Collect data on past cases

[0947] The server collects past case data from call centers and customer support systems. This process includes information such as the nature of the complaint, how it was handled, and the customer's reaction. For example, data on how to handle a delayed delivery of a product can be obtained.

[0948] Step 2: Preprocessing the data

[0949] The server preprocesses the collected case data. During preprocessing, the data is cleansed, unnecessary information is removed, and necessary information is extracted and structured. Specifically, text data is tokenized and important keywords are extracted.

[0950] Step 3: Create a dataset

[0951] The server splits the preprocessed data into training and testing sets, for example, 80% of the total data as the training set and 20% as the testing set.

[0952] Step 4: Building and training an AI model

[0953] The server uses a machine learning algorithm to build a generative AI model and trains it using a training set. Specifically, it learns pairs of complaint details and response methods and has the model recognize patterns.

[0954] Step 5: Evaluate and tune the model

[0955] The server uses the test set to evaluate the trained AI model, for example, assessing the model's accuracy, suitability, and speed of response, and adjusting any inaccuracies. It also optimizes hyperparameters as necessary.

[0956] Step 6: Implementing the Emotion Engine

[0957] The server will install an emotion engine and add the ability to analyze emotions from the user's text input or voice commands. The emotion engine will recognize the user's emotions and record the data. For example, if it detects that the user is "angry," it will save this emotion data as the analysis result.

[0958] Step 7: Providing a User Interface

[0959] The server designs the user interface and provides it through the terminal. The interface includes forms and input fields, allowing users to easily enter inquiries. For example, a form is provided for entering "What to do if the product arrives damaged."

[0960] Step 8: Query entry and sentiment analysis

[0961] The user inputs a query through the interface, for example, "The product arrived damaged." At the same time, the emotion engine analyzes the user's emotions and detects "anger."

[0962] Step 9: Generate an answer with the AI ​​model

[0963] The server receives the query and uses a generative AI model to generate the optimal answer. Based on data from the emotion engine, the server provides an answer that takes the user's emotions into consideration. For example, it may suggest sending an email with an apology and applying a discount on the next order.

[0964] Step 10: Provide your answers

[0965] The device displays the generated answer from the server to the user. The user checks the suggested answer and takes action. For example, the device displays specific measures such as "send an apology email and immediately ship a replacement product" as a response to the question "What to do if the product arrives damaged."

[0966] Step 11: Enter your feedback

[0967] The user provides feedback on the appropriateness of the provided answer, for example, by rating whether the proposed response was effective.

[0968] Step 12: Gather feedback

[0969] The server collects feedback from users, including the appropriateness of the response and areas for improvement. For example, specific opinions such as "the response was a little slow" are collected.

[0970] Step 13: Retraining the AI ​​model

[0971] The server retrains the generative AI model based on the collected feedback and emotion data, improving its accuracy and response capabilities. For example, it can learn how to respond quickly to anger.

[0972] Step 14: Automatic generation of training manuals

[0973] The server automatically generates training manuals using the collected data and learning results. For example, a "manual for dealing with product defects" is generated.

[0974] Step 15: Provide training manuals

[0975] The server displays the generated training manual on the terminal and makes it available to new and current employees, for example, by providing it as a downloadable PDF file.

[0976] Through these steps, the system will improve the quality and efficiency of customer service and call center operations. The introduction of an emotion engine will enable responses that take user emotions into consideration, which is expected to improve customer satisfaction.

[0977] Example 2

[0978] 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."

[0979] In recent years, there has been a demand for improved quality and efficiency in customer service and call center operations when it comes to handling complaints. However, conventional systems have struggled to respond in a way that takes user emotions into consideration, limiting the extent to which they can improve customer satisfaction. Furthermore, they lack the ability to automatically generate appropriate response manuals, making the efficiency of employee training an issue.

[0980] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting past case data, means for preprocessing the collected case data, means for training a generative AI model using the preprocessed data, means for providing an interface through which a user can input an inquiry, means for using an emotion engine to analyze the user's emotions, means for generating an optimal answer to the user's inquiry using the generative AI model, means for providing the generated answer to the user, means for collecting feedback from the user, means for retraining the generative AI model using the collected feedback, means for automatically generating a training manual, and means for providing the generated training manual. This enables improved quality of complaint handling and business efficiency. Furthermore, the introduction of an emotion engine enables responses that take user emotions into consideration, which is expected to improve customer satisfaction. Furthermore, the automatic generation of training manuals allows for efficient training of new employees and skill development of current employees.

[0981] "Past case data" refers to data collected from customer response history, including the content of complaints, response methods, and feedback from customers.

[0982] "Preprocessing" refers to the process of cleansing unnecessary noise from collected data and tokenizing important text information to convert it into structured data.

[0983] A "generative AI model" is a model that is constructed using machine learning algorithms with preprocessed data to predict the appropriate way to handle a complaint.

[0984] An "interface" is a tool or platform through which a user can enter a query and interact with a system, such as a web form or a mobile app.

[0985] An "emotion engine" is a system that analyzes emotions from a user's text input or voice commands, and has the ability to record the user's emotional state as data.

[0986] An "Answer" is an optimal response to a user's query, created using a generative AI model and emotion engine.

[0987] "Feedback" refers to opinions and ratings provided by users regarding the appropriateness of proposed answers.

[0988] "Retraining" is the process of using collected feedback to readjust the parameters of a generative AI model to improve the model's accuracy.

[0989] The "training manual" is a document that is automatically generated based on collected data and the learning results of the generating AI model, and is used to train new employees and improve the skills of current employees.

[0990] One embodiment of the present invention is described in detail below. This system is designed to improve the quality and efficiency of customer service and complaint handling in call center operations. By combining the process of collecting and preprocessing past case data, generating and training a generative AI model, and incorporating an emotion engine that recognizes user emotions, more appropriate and emotion-sensitive responses are possible.

[0991] First, the server collects past case data from call centers and customer support systems. Specifically, it obtains data including the details of complaints, how they were handled, and customer feedback. Data is collected from a database and extracted using the Python "pandas" library. Next, the server preprocesses this data. During preprocessing, unnecessary data is cleansed, and important text information is tokenized using the "NLTK" library and converted into structured data.

[0992] Using the preprocessed data, the server builds and trains a generative AI model. TensorFlow and PyTorch are used to build the model, and the model is trained using a training dataset. Specifically, the model learns to predict the appropriate response method using paired data of the complaint content and the response method. For example, the following prompt sentence is used:

[0993] "How should you respond if a customer complains about the following? Complaint: ____"

[0994] The server then deploys an emotion engine, which uses IBM Watson and Microsoft Azure Emotion API to analyze emotions from the user's text input and voice commands. This engine records the user's emotional state as data and reflects it in the generative AI model.

[0995] The server designs an interface that allows users to enter inquiries and provides it via the terminal. The interface is provided as a web-based form or mobile app, and is designed so that users can easily access the system and enter inquiries. For example, a customer service representative can inquire about "what to do if a customer's order is delayed" through the system.

[0996] When a user inputs an inquiry into the system, the server uses a generative AI model to generate the optimal response to that inquiry. At the same time, an emotion engine analyzes the user's emotions and optimizes the response based on that emotional data. For example, if the user expresses anger, the system will provide a response that includes a more polite apology. As a specific example, when asked what to do if an order is delayed, the system will suggest responses such as "send an apology email and offer a coupon for free shipping on the next order."

[0997] The device displays the answer generated by the server to the user, who can then act on the suggested answer. The user can also provide feedback on whether the suggested answer is appropriate. The server collects the feedback and uses it to retrain the generative AI model to improve the quality of the answer. This improves the model's accuracy and makes it better able to respond to future inquiries.

[0998] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals can be used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" can be provided in PDF format.

[0999] The above is an embodiment of the system of the present invention, which improves the quality and efficiency of customer service and complaint handling in call center operations. The introduction of an emotion engine enables responses that take user emotions into consideration, which is expected to improve customer satisfaction. Furthermore, the automatic generation of training manuals allows for efficient training of new employees and skill development for current employees.

[1000] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1001] Step 1: Data collection

[1002] The server collects past case data from call centers and customer support systems. This includes data on complaints, response methods, and customer feedback. Specifically, it extracts data from a database using SQL queries. The input is past case data, and the output is the extracted raw data.

[1003] Step 2: Data Preprocessing

[1004] The server preprocesses the collected data. Specifically, it cleanses unnecessary data using the Python "pandas" library and tokenizes text information using the "NLTK" library to convert it into structured data. The input is raw data, and the output is cleansed and tokenized structured data.

[1005] Step 3: Generate an AI model

[1006] The server uses the preprocessed data to build a generative AI model. It uses the machine learning library "TensorFlow" or "PyTorch" to define a neural network model. For example, it uses pairs of complaints and response methods as input data. The input is structured training data, and the output is an initialized generative AI model.

[1007] Step 4: Training the AI ​​model

[1008] The server trains the generative AI model. Specifically, it uses the preprocessed data to apply a learning algorithm (e.g., batch learning, setting the number of epochs, etc.) to improve the accuracy of the model. The input is the training data and the initialized AI model, and the output is the trained generative AI model.

[1009] Step 5: Implementing the Emotion Engine

[1010] The server installs an emotion engine and analyzes emotions from the user's text input or voice commands. IBM Watson and Microsoft Azure Emotion API are used as emotion engines. The input is text or voice data from the user, and the output is analyzed emotional data.

[1011] Step 6: Providing a User Interface

[1012] The server designs a query input interface and provides it through the terminal, such as a web form or a mobile app. The input is the query from the user, and the output is the query data sent to the system.

[1013] Step 7: Receiving and analyzing inquiries

[1014] A user inputs a query into the system. The server receives the query and uses a generative AI model to generate the optimal answer to that query. Specifically, a prompt such as "What should we do if a customer's order is delayed?" is used. The input is the user's query, and the output is the generated answer.

[1015] Step 8: Analyze sentiment data and optimize responses

[1016] The server uses an emotion engine to analyze the user's emotions and optimizes the answer of the generative AI model based on the emotion data. The input is the query and the generated answer, and the output is an optimized answer that takes emotions into consideration.

[1017] Step 9: Provide your answers

[1018] The device displays the generated answer from the server to the user, specifically on a web page or in the chat screen of the app. The input is the optimized answer, and the output is what is displayed to the user.

[1019] Step 10: Gather feedback

[1020] The user provides feedback on the appropriateness of the response. The server collects this feedback and stores it in a database. The input is the user feedback and the output is the collected feedback data.

[1021] Step 11: Retraining the AI ​​model

[1022] The server uses the collected feedback to retrain the generative AI model. Specifically, it uses the feedback data as new training data and readjusts the model's parameters. The inputs are the feedback data and the existing generative AI model, and the output is the retrained generative AI model.

[1023] Step 12: Automatic generation of training manuals

[1024] The server automatically generates training manuals based on the training data and model results. Specifically, it uses a Python script to analyze past cases and their effectiveness, and generates training manuals. The input is the training data and the model's learning results, and the output is the generated training manual (e.g., in PDF format).

[1025] Step 13: Provide training manuals

[1026] The terminal provides the generated training manual to the user. Specifically, it is provided as a download link or email attachment. The input is the generated training manual, and the output is the manual provided to the user.

[1027] (Application example 2)

[1028] 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."

[1029] When it comes to customer service for autonomous vehicles, there are cases where responses to user complaints or problems are delayed or inappropriate. In such cases, measures that take user feelings into consideration are required, and improving customer satisfaction is a challenge.

[1030] 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 collecting past response case data, means for preprocessing the collected response case data, means for training a generative AI model using the preprocessed data, means for providing an interface through which a user can input a query, means including an emotion engine for analyzing the user's emotions, means for generating an optimal answer to the user's query using the generative AI model, means for optimizing the generated answer based on user emotion analysis, means for providing the optimized answer to the user, means for collecting feedback from the user, means for retraining the generative AI model using the collected feedback, means for automatically generating a training manual, and means for providing the generated training manual. This enables user complaints and problems regarding autonomous vehicles to be handled quickly and appropriately, thereby improving customer satisfaction.

[1031] "Past case data" refers to data that includes details of past complaints and troubles, the methods used to deal with them, and feedback from customers.

[1032] "Preprocessing" is the process of removing unnecessary data from collected data and structuring important text information.

[1033] A "generative AI model" is a model trained using machine learning algorithms and used to predict optimal responses based on past response patterns.

[1034] A "user interface" is an interface through which a user inputs queries into a system, and supports communication between the user and the system.

[1035] The "emotion engine" is a system that analyzes emotions from the user's text and voice input and records that emotional state as data.

[1036] "Feedback" means opinions or evaluations provided by users regarding the appropriateness or quality of a system's response.

[1037] "Retraining" is the process of retraining a generative AI model using newly collected feedback to improve the model's accuracy and responsiveness.

[1038] An "education manual" is a document automatically generated based on collected data and learning results, and is used to train new employees and improve the skills of current employees.

[1039] The present invention is a system for quickly and appropriately responding to user complaints and problems with autonomous vehicles. This system collects and analyzes past case data and provides optimal countermeasures that take user emotions into consideration. An embodiment of the present invention is described in detail below.

[1040] Data collection and preprocessing

[1041] First, the server collects past case data from the autonomous vehicle's operation management system and customer response system. This includes complaint details, response methods, and customer feedback. Next, the server preprocesses this data, removing unnecessary data and tokenizing and structuring useful text information.

[1042] AI model generation and training

[1043] The server uses the preprocessed data to build a generative AI model and trains it using machine learning algorithms, allowing it to learn past response patterns and predict optimal responses, such as the appropriate way to respond to complaints.

[1044] Introducing the Emotion Engine

[1045] The server then uses an emotion engine to analyze the user's text input. The emotion engine uses natural language processing tools such as TextBlob to analyze the emotion into positive, negative, and neutral, and records the emotional state as data. This allows the AI ​​model to reflect appropriate emotional responses to the user's inquiry.

[1046] User Interface

[1047] The server designs and provides an interface through which users can input inquiries via terminals, allowing users to easily report complaints and problems.

[1048] User inquiries and sentiment analysis

[1049] When a user inputs a query into the system, the server uses a generative AI model to generate the optimal response for that query. At the same time, the emotion engine analyzes the user's emotions and optimizes the response based on that emotional data. For example, if the user expresses anger, the server will provide a more polite response that includes an apology.

[1050] Examples:

[1051] User input: "My car is stuck in the middle of the road! I'm so annoyed. What can you do?"

[1052] Emotion analysis result: "negative"

[1053] Example AI response: "We apologize for the inconvenience. Our technical team is working quickly to resolve the issue. We'll offer you a 20% discount coupon for your next visit."

[1054] Providing answers

[1055] The device displays the generated answer from the server to the user, who then acts on the suggested answer.

[1056] Gathering feedback and retraining

[1057] Users can provide feedback on the appropriateness of the proposed answers, which the server collects and uses to retrain the generative AI model, a process that improves the model's accuracy and makes it better able to respond to future inquiries.

[1058] Automatic generation of training manuals

[1059] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are used to train new employees and improve the skills of current employees. For example, a "vehicle malfunction response manual" is generated and provided to the terminal in PDF format.

[1060] In this way, the system of the present invention improves the quality of customer service in self-driving vehicles and is expected to improve customer satisfaction by enabling emotionally sensitive service.

[1061] Example prompt sentence:

[1062] Prompt: "User's complaint: My car got stuck in the middle of the road! I'm so upset. What can you do about it? Emotion: negative Suggest the best solution:"

[1063] This embodiment makes it possible to specifically implement the present invention.

[1064] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1065] Step 1:

[1066] The server collects past case data from the autonomous vehicle's operation management system and customer response system. The collected data includes the details of the complaint, how it was handled, and customer feedback. The input is raw data obtained from the operation management system database, and the output is sent to pre-processing through the collection step.

[1067] Step 2:

[1068] The server preprocesses the collected data. During this process, unnecessary data is cleansed and useful text information is tokenized and structured. Processes such as text normalization, stop word removal, and tokenization are performed. The input is the collected raw data, and the output is cleansed and structured data.

[1069] Step 3:

[1070] The server uses the preprocessed data to generate and train a generative AI model. The model is trained using machine learning algorithms (e.g., deep learning) to learn appropriate countermeasures based on past response patterns. The input is the preprocessed data, and the output is a trained generative AI model.

[1071] Step 4:

[1072] The server uses an emotion engine to analyze the user's emotions. In this process, the text entered by the user is analyzed to determine whether the emotion is positive, negative, or neutral, and the data is recorded. The input is the text input from the user, and the output is the analyzed emotion data.

[1073] Step 5:

[1074] The user uses an interface to input inquiries and complaints. The server designs this interface to make it easy for the user to input inquiries. The input is the complaint or inquiry content from the user, and the output is the inquiry data that is sent to the server.

[1075] Step 6:

[1076] The server uses a generative AI model to generate the optimal answer to a user's inquiry. The answer is optimized based on the user's emotional data analyzed by the emotion engine. For example, if the emotion included anger, a particularly polite response would be suggested. The input is the user's inquiry data and emotional data, and the output is an optimized answer.

[1077] Step 7:

[1078] The terminal displays the generated answer to the user, who then reviews the proposed answer and takes action. The input is the optimized answer sent by the server, and the output is the generated answer that is displayed to the user.

[1079] Step 8:

[1080] Users provide feedback on the appropriateness of the proposed answers, which the server collects and uses to retrain the generative AI model. The input is the user feedback data, and the output is the retraining data.

[1081] Step 9:

[1082] The server automatically generates a training manual based on the collected data and learning results. This manual is used to train new employees and improve the skills of current employees. The inputs are the collected data and learning results, and the output is a training manual (e.g., a PDF document).

[1083] In this way, each step works together to create a system that can quickly and appropriately handle complaints and problems with self-driving vehicles.

[1084] 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.

[1085] 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.

[1086] 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.

[1087] [Fourth embodiment]

[1088] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1089] 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.

[1090] 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).

[1091] 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.

[1092] 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.

[1093] 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).

[1094] 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.

[1095] 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.

[1096] 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.

[1097] 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.

[1098] 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.

[1099] 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.

[1100] 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."

[1101] MODE FOR CARRYING OUT THE INVENTION

[1102] An embodiment of the present invention will be described below.

[1103] This system is designed to improve the efficiency and quality of customer service and complaint handling in call center operations. It collects and preprocesses past case data, generates and trains AI models, provides a user interface, handles user inquiries, collects feedback and retrains, and automatically generates training manuals.

[1104] Data collection and preprocessing

[1105] First, the server collects past case data collected from call centers and customer support operations. For example, it obtains historical data on responses to complaints such as "the product arrived damaged." Next, the server preprocesses this data. In preprocessing, unnecessary data is cleansed and important data is extracted and structured. In particular, text data is tokenized, and important keywords are extracted.

[1106] AI model generation and training

[1107] Using the preprocessed data, the server builds and trains a generative AI model. This is the process of using machine learning algorithms to train the model and learn from past response patterns. For example, it can learn the appropriate way to respond to complaints and predict the optimal response.

[1108] Providing a user interface

[1109] Next, the server designs and implements a user interface, which is provided via a terminal. This interface is designed to allow users to easily access the system and input inquiries. For example, a customer service representative can ask the system, "What should we do if a customer's order is delayed?"

[1110] Responding to user inquiries

[1111] When a user inputs a query into the system, the server uses the generative AI model to generate the optimal response to that query. For example, the server suggests "how to respond to a delayed order" by suggesting "sending an apology email and offering a discount on the next order." The device then displays this response to the user.

[1112] Gathering feedback and retraining

[1113] Users can provide feedback on the appropriateness of the suggested answers, which the server collects and uses to retrain the AI ​​model, a process that improves the model's accuracy and allows it to better respond to future inquiries.

[1114] Automatic generation of training manuals

[1115] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated and provided to the terminal in PDF format.

[1116] Through the above process, this system will improve the quality of customer service and call center operations and increase operational efficiency.

[1117] The processing flow will be explained below.

[1118] Step 1: Collecting interaction data

[1119] The server collects past case data from call centers and customer support systems, specifically data including complaint details, response methods, and customer feedback.

[1120] Step 2: Preprocessing the data

[1121] The server preprocesses the collected case data, for example by cleansing unnecessary information and tokenizing and structuring important text information.

[1122] Step 3: Create a dataset

[1123] The server splits the preprocessed data into a training set and a test set, typically 80% for training and 20% for testing.

[1124] Step 4: Building and training an AI model

[1125] The server uses a machine learning algorithm to build a generative AI model and trains it using a training set. Specifically, it learns pairs of complaint details and response methods and has the model recognize patterns.

[1126] Step 5: Evaluate and tune the model

[1127] The server evaluates the trained AI model using the test set, and if the accuracy is low, adjusts the hyperparameters to optimize the model.

[1128] Step 6: Providing a User Interface

[1129] The server designs an interface that allows users to input queries and displays it on the terminal. This interface allows queries to be made by filling out a form or operating buttons.

[1130] Step 7: Enter your inquiry

[1131] A user inputs a specific complaint or inquiry into the system. For example, a user may inquire, "What should I do if my product delivery is delayed?"

[1132] Step 8: Generate an answer with the AI ​​model

[1133] The server receives the user's inquiry and uses a generative AI model to generate the optimal response, such as suggesting specific actions such as sending an apology email and offering a coupon for free shipping on the next purchase.

[1134] Step 9: Provide your answers

[1135] The device displays the generated answers from the server to the user, who then reviews the suggested answers and takes action.

[1136] Step 10: Provide feedback

[1137] The user provides feedback on the appropriateness of the provided answer, for example, rating whether the proposed action was effective or not.

[1138] Step 11: Gather feedback

[1139] The server collects feedback from users, which is used to improve the model.

[1140] Step 12: Retraining the AI ​​model

[1141] The server retrains the generative AI model based on the collected feedback, improving its accuracy and making it better able to respond to the next inquiry.

[1142] Step 13: Automatic generation of training manuals

[1143] The server automatically generates training manuals using the collected data and learning results. For example, a "manual for dealing with product defects" is created.

[1144] Step 14: Provide training manuals

[1145] The server displays the generated training manual on the terminal and makes it available to new and current employees, for example, by making it available for download in PDF format.

[1146] In this way, this system improves the quality of complaint handling and increases operational efficiency.

[1147] Example 1

[1148] 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."

[1149] In traditional customer service and call center operations, there is a demand for improved efficiency and quality in handling complaints. However, there are insufficient methods for effectively utilizing past case data, resulting in issues with the consistency and accuracy of responses. Furthermore, there is a lack of advanced functionality, such as automatic model updating based on appropriate feedback and automatic generation of training materials for new employees.

[1150] 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.

[1151] In this invention, the server includes means for collecting past case data, means for preprocessing the collected case data, means for training a machine learning model using the preprocessed data, means for providing an interface through which a user can input an inquiry, means for generating an optimal answer to the user's inquiry using the machine learning model, means for providing the generated answer to the user, means for collecting feedback from the user, means for retraining the machine learning model using the collected feedback, means for automatically generating training materials, and means for providing the generated training materials. This enables improved efficiency and quality in complaint handling, and also improves the consistency and accuracy of responses. Furthermore, automatic model updating and automatic generation of training materials are realized by utilizing appropriate feedback.

[1152] "Past case data" refers to records of past complaint and inquiry handling in call centers and customer service operations.

[1153] "Preprocessing" refers to the process of removing unnecessary information from collected data and extracting and formatting important information.

[1154] A "machine learning model" is a model that uses an algorithm to learn from collected and preprocessed data and predict the optimal response to future inquiries.

[1155] "Interface" refers to the operation screen and input form that users use to access the system and enter inquiries.

[1156] A "generative AI model" refers to an artificial intelligence model built using machine learning algorithms based on collected and pre-processed data.

[1157] "Optimal response" refers to the most appropriate response method generated based on past response cases and learning results, based on the user's inquiry.

[1158] "Feedback" refers to a user's rating or comment on the appropriateness of a proposed answer.

[1159] "Retraining" refers to the process of retraining an existing machine learning model based on collected feedback to improve its accuracy and performance.

[1160] "Educational materials" refers to guides and manuals on how to handle complaints and customer service, intended to educate new employees and improve the skills of current employees.

[1161] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the accompanying drawings. The present invention aims to improve the efficiency of customer service and complaint handling in call center operations, thereby improving the quality of responses.

[1162] First, the server collects past case data collected from call centers and customer support operations. For example, it obtains historical data on complaints such as "the product arrived damaged." This data is obtained from a database. Next, the server preprocesses this data. In preprocessing, unnecessary data such as dates and personal information is cleansed, and important data such as the type of complaint, how it was handled, and the results are extracted and structured. Furthermore, the text data is tokenized, and important keywords are extracted. Specifically, keywords such as "product," "damaged," and "arrived" are extracted.

[1163] Next, using the preprocessed data, the server uses a machine learning algorithm (e.g., using TensorFlow) to build and train a generative AI model. In this process, the server learns past complaint handling patterns and predicts the optimal response method. For example, it learns how to respond to "damaged product" cases and provides the most appropriate relevant information.

[1164] The server then designs and implements a user interface, which is provided to users through their terminals. This interface is provided through a web browser or dedicated application, and is designed to allow users to easily access the system and input inquiries. For example, a customer service representative can inquire about "what to do if a customer's order is delayed."

[1165] When a user inputs a query into the system, the server uses the generative AI model to generate the optimal response to that query. For example, the server suggests "how to respond to a delayed order" by suggesting "sending an apology email and offering a discount on the next order." The device then displays this response to the user.

[1166] Additionally, users can provide feedback on the appropriateness of the suggested answers. The server collects this feedback and retrains the AI ​​model. This process improves the model's accuracy and allows it to respond more appropriately to future inquiries.

[1167] Finally, the server automatically generates training materials based on the collected data and learning results. These materials are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated and provided to the terminal in PDF format.

[1168] As a specific example, the server collects complaint processing history data for "the product arrived damaged," tokenizes it, and extracts keywords such as "product," "damaged," and "arrived." Using the preprocessed data, a generative AI model is built using a machine learning algorithm to learn the optimal response method for "damaged product" cases. A user interface is designed and provided via a terminal, allowing the user to ask the system "how to respond if an order is late." The server generates "send an apology email and offer a discount on the next order," which is then displayed on the terminal. Based on user feedback, the server retrains the model and generates and provides a "manual for responding to product defects" in PDF format.

[1169] An example of a prompt for a generative AI model is:

[1170] Please suggest the best way to handle a complaint based on the complaint handling history below.

[1171] Complaint: The product arrived damaged.

[1172] Response history:

[1173] 1. Send new products to customers immediately

[1174] 2. Requesting a return of damaged goods at no cost to the customer

[1175] 3. Offer a discount coupon for your next order

[1176] Please suggest the best way to respond.

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

[1178] Step 1:

[1179] The server retrieves past case data collected from call centers and customer support operations from a database. The input is the database and its query, and the output is case data including the complaint details, response methods, results, etc. Specifically, the server uses an SQL query to extract the relevant data from the database.

[1180] Step 2:

[1181] Remove unnecessary information from the case data acquired by the server (data cleansing). The input is the case data acquired in step 1, and the output is the cleansed data. Specifically, it removes unnecessary fields such as dates and personal information.

[1182] Step 3:

[1183] The server structures the data. The input is the cleansed data from step 2, and the output is structured data. Specifically, the server reclassifies the data into fields such as complaint type, response method, and outcome.

[1184] Step 4:

[1185] The server tokenizes the text data and extracts important keywords. The input is the structured data from step 3, and the output is a list of keywords. Specifically, the server uses a natural language processing tool to tokenize the text and extract keywords such as "product," "damaged," and "arrived."

[1186] Step 5:

[1187] The server uses a machine learning algorithm (e.g., TensorFlow) to build a generative AI model. The input is the keyword list extracted in step 4 and the preprocessed data, and the output is the generated AI model. Specifically, the server inputs the data into the algorithm and trains the model.

[1188] Step 6:

[1189] The server stores the learning results and designs and implements the user interface. The input is the AI ​​model generated in step 5, and the output is the completed user interface. Specifically, the code to provide the interface through a web browser or dedicated application is written and deployed.

[1190] Step 7:

[1191] The terminal allows access to the system through a user interface. The input is a request to access the user interface, and the output is the display of the interface. Specifically, the terminal displays a login screen or a search screen.

[1192] Step 8:

[1193] A user inputs an inquiry into the system. The input is the inquiry entered by the user, and the output is the content being sent to the system. A specific operation is when the user inputs "What to do if an order is late."

[1194] Step 9:

[1195] The server uses the generative AI model to generate the optimal answer to the user's inquiry. The input is the inquiry entered in step 8, and the output is the generated answer. Specifically, the server inputs the inquiry into the model and generates an answer such as "send an apology email and offer a discount on the next order."

[1196] Step 10:

[1197] The terminal displays the generated answer to the user. The input is the answer generated in step 9, and the output is the display of the answer to the user. Specifically, the terminal displays the answer on the screen.

[1198] Step 11:

[1199] The user provides feedback on the appropriateness of the proposed answer. The input is the user's feedback, and the output is the feedback information being sent to the system. The specific behavior is that the user provides feedback as "appropriate" or "inappropriate."

[1200] Step 12:

[1201] The server collects feedback and retrains the AI ​​model. The input is user feedback and existing model data, and the output is the retrained AI model. Specifically, the server reflects the feedback in the model and retrains it.

[1202] Step 13:

[1203] The server automatically generates training materials based on the collected data and learning results. The input is the retrained AI model and learning results, and the output is the generated training manual. Specifically, the server generates content such as a "manual for dealing with product defects" in PDF format.

[1204] Step 14:

[1205] The terminal provides the generated educational materials to the user. The input is the generated educational manual, and the output is the provision of the manual. Specifically, the terminal provides a download link so that the user can download the manual.

[1206] (Application example 1)

[1207] 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."

[1208] In virtual stores, customer service requires quick and appropriate responses to complaints and inquiries, but conventional systems often make response processes cumbersome and result in inconsistent quality. Furthermore, there is a lack of feedback collection and learning processes to improve response accuracy, making continuous improvement difficult.

[1209] 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.

[1210] In this invention, the server includes means for collecting past case data, means for preprocessing the collected case data, means for training a generative AI model using the preprocessed data, means for providing an interface through which a user can input an inquiry, means for generating an optimal answer to the user's inquiry using the generative AI model, means for providing the generated answer to the user, means for collecting feedback from the user, means for retraining the generative AI model using the collected feedback, means for automatically generating a training manual, means for providing the generated training manual, means for responding to inquiries using a smart device, and means for accepting inquiries and providing answers in real time, thereby enabling improved quality and efficiency of customer service.

[1211] "Past case data" refers to data that includes the details of responses to complaints and inquiries recorded in the past in customer service operations.

[1212] "Means of collection" refers to equipment and software used to store data on past cases on a server, etc.

[1213] "Preprocessing means" refers to techniques and methods for cleansing collected data and extracting and structuring important information.

[1214] A "generative AI model" refers to an artificial intelligence model that is built using machine learning algorithms and has learned past response patterns.

[1215] "User Interface" means the screen or operating environment through which a user accesses the system and enters inquiries.

[1216] "Means for generating optimal answers" refers to technologies and functions that use generative AI models to output optimal answers to user inquiries.

[1217] "Feedback collection means" refers to any equipment or software used to collect user opinions or ratings regarding the appropriateness of responses.

[1218] "Retraining methods" are techniques or methods for retraining a generative AI model based on collected feedback to improve the model's accuracy.

[1219] "Means for automatically generating training manuals" refers to technologies and methods for automatically creating manuals for improving customer service skills based on case data and learning results.

[1220] "Smart devices" refer to portable electronic devices that can connect to the Internet, such as smartphones and head-mounted displays.

[1221] "Real-time means" refers to technologies and functions that instantly generate and provide answers to user inquiries.

[1222] MODE FOR CARRYING OUT THE INVENTION

[1223] A system for realizing the present invention in a virtual store is configured using the following hardware and software.

[1224] Hardware and software used

[1225] Hardware

[1226] Server: Used to process, store, and learn from large amounts of data.

[1227] Smart devices: Smartphones, head-mounted displays, etc., are the interface through which users input inquiries and receive responses.

[1228] software

[1229] TensorFlow: A machine learning library used to train generative AI models.

[1230] Django: Web framework, provides the user interface.

[1231] NLTK: A natural language processing library used for data preprocessing.

[1232] MySQL: Database management system used to store and manage data.

[1233] System configuration and operation

[1234] Data collection and preprocessing

[1235] The server collects past case data from call centers and customer service operations. The collected data is tokenized using NLTK to extract important keywords and phrases. The data is then cleansed of unnecessary data and stored in a structured format in a database.

[1236] Training generative AI models

[1237] Using the preprocessed data, the server uses TensorFlow to build and train a generative AI model, a process that allows the model to learn from past response patterns and predict the best response.

[1238] Providing a user interface

[1239] Using the Django framework, the server designs and implements the user interface. This interface, which runs on smart devices, is designed to allow users to easily input inquiries. For example, users can input "What should I do if my order is delayed" into the smartphone application.

[1240] Responding to user inquiries

[1241] When a user inputs a query, the server uses a generative AI model to generate the optimal answer. This answer is displayed on the smart device and provided to the user. For example, when asked how to respond to a delayed order, the server suggests sending an apology email and offering a discount on the next order.

[1242] Gathering feedback and retraining

[1243] Users can provide feedback on the appropriateness of the suggested answers, and the server collects this feedback to retrain the generative AI model. This process improves the model's accuracy, enabling it to better respond to future inquiries.

[1244] Automatic generation of training manuals

[1245] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are provided in PDF format and are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated.

[1246] Examples of specific examples and prompts

[1247] Example 1: When a customer asks, "What should we do if my order is late?" the system responds, "We will send you an apology email and offer you a discount on your next order."

[1248] Examples of prompts:

[1249] "What is the best way to handle a complaint?"

[1250] "What do you do if a customer inquires about a delayed order?"

[1251] Please tell me what to do if the product arrives damaged.

[1252] This makes it possible for this system to improve the quality and efficiency of customer service.

[1253] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1254] Step 1:

[1255] The server collects past case data from customer support history. The input is past support history data, such as complaints about "the product arrived damaged." The server saves this data in MySQL, a database management system. The output is a structured case database.

[1256] Step 2:

[1257] The server preprocesses the collected case data. The input is the stored case data, which is tokenized using NLTK to remove unnecessary information. Specific operations include cleaning the text data, extracting keywords, and structuring. The output is the preprocessed case data.

[1258] Step 3:

[1259] The server uses the preprocessed data to build and train a generative AI model. The input is the preprocessed corresponding case data, and TensorFlow is used to train the machine learning model. Specific operations include data splitting, model selection, training, and evaluation. The output is a trained generative AI model.

[1260] Step 4:

[1261] The server provides the user interface. The input is the interface code built using Django, allowing users to easily enter queries into the system. Specific operations include form design, deployment, and access management. The output is the interface available to users.

[1262] Step 5:

[1263] A user inputs a query from a terminal. The input is a query text by the user, such as "What should I do if my order is late?" The terminal sends this input to the server. The output is the query data sent to the server.

[1264] Step 6:

[1265] The server uses a generative AI model to generate an optimal answer to a user's query. The input is the user's query text, and the generative AI model predicts the optimal answer. Specific operations include performing model inference. The output is the generated answer text.

[1266] Step 7:

[1267] The terminal displays the answer received from the server to the user. The input is the answer text sent by the server, which is displayed in the user interface. Specific operations include adjusting the display layout and notifying the user. The output is the answer that the user can view.

[1268] Step 8:

[1269] The user provides feedback on the appropriateness of the proposed answer. The input is the user's feedback text, and the terminal sends this feedback to the server. The output is the feedback data to the server.

[1270] Step 9:

[1271] The server retrains the generative AI model based on the collected feedback. The input is user feedback data and collected case data, which are used to retrain the model. Specific operations include integrating the feedback data and running the retraining process. The output is an updated generative AI model.

[1272] Step 10:

[1273] The server automatically generates training manuals. The input is a trained or updated generative AI model and corresponding case data, and the training manual is created based on this. Specific operations include selecting a template, writing the content, and executing the automatic generation process. The output is a completed training manual.

[1274] Step 11:

[1275] The server provides the generated training manual. The input is the completed training manual, which is provided to the terminal in PDF format or other format. Specific operations include file generation and distribution. The output is a usable training manual.

[1276] 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.

[1277] MODE FOR CARRYING OUT THE INVENTION

[1278] An embodiment of the present invention will be described below.

[1279] This system is designed to improve the quality and efficiency of customer service and complaint handling in call center operations. By combining the process of collecting and preprocessing past case data, generating and training an AI model, and combining it with an emotion engine that recognizes user emotions, it becomes possible to provide more appropriate and emotionally sensitive responses.

[1280] Data collection and preprocessing

[1281] First, the server collects past case data from call centers and customer support systems. Specifically, it acquires data including the details of complaints, how they were handled, and customer feedback. Next, the server preprocesses this data. In preprocessing, unnecessary data is cleansed and important text information is tokenized and structured.

[1282] AI model generation and training

[1283] Using the preprocessed data, the server builds and trains a generative AI model. This is the process of using machine learning algorithms to train the model and learn from past response patterns. For example, it can learn the appropriate way to respond to complaints and predict the optimal response.

[1284] Introducing the Emotion Engine

[1285] Next, the server introduces an emotion engine, which analyzes emotions from the user's text input or voice commands and records the emotional state as data. This allows the generative AI model to reflect appropriate emotional responses to the user's inquiries.

[1286] Providing a user interface

[1287] The server designs an interface that allows users to input inquiries and provides it through a terminal. This interface is designed so that users can easily access the system and input inquiries. For example, a customer service representative can inquire about "what to do if a customer's order is delayed."

[1288] User inquiries and sentiment analysis

[1289] When a user inputs a query into the system, the server uses a generative AI model to generate the optimal response to that query. At the same time, the emotion engine analyzes the user's emotions and optimizes the response based on that emotional data. For example, if the user expresses anger, the system will provide a more polite response that includes an apology.

[1290] Providing answers

[1291] The terminal displays the generated answer from the server to the user, and the user can take action based on the suggested answer. For example, the suggested action for "what to do if an order is delayed" might be "send an apology email and offer a coupon for free shipping on the next order."

[1292] Gathering feedback and retraining

[1293] Users can provide feedback on the appropriateness of the suggested answers, which the server collects and uses to retrain the AI ​​model, a process that improves the model's accuracy and makes it better able to respond to future inquiries.

[1294] Automatic generation of training manuals

[1295] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" is generated and provided to the terminal in PDF format.

[1296] Through the above process, this system will improve the quality and efficiency of customer service and call center operations. The introduction of an emotion engine will enable responses that take user emotions into consideration, which is expected to improve customer satisfaction.

[1297] The processing flow will be explained below.

[1298] Step 1: Collect data on past cases

[1299] The server collects past case data from call centers and customer support systems. This process includes information such as the nature of the complaint, how it was handled, and the customer's reaction. For example, data on how to handle a delayed delivery of a product can be obtained.

[1300] Step 2: Preprocessing the data

[1301] The server preprocesses the collected case data. During preprocessing, the data is cleansed, unnecessary information is removed, and necessary information is extracted and structured. Specifically, text data is tokenized and important keywords are extracted.

[1302] Step 3: Create a dataset

[1303] The server splits the preprocessed data into training and testing sets, for example, 80% of the total data as the training set and 20% as the testing set.

[1304] Step 4: Building and training an AI model

[1305] The server uses a machine learning algorithm to build a generative AI model and trains it using a training set. Specifically, it learns pairs of complaint details and response methods and has the model recognize patterns.

[1306] Step 5: Evaluate and tune the model

[1307] The server uses the test set to evaluate the trained AI model, for example, assessing the model's accuracy, suitability, and speed of response, and adjusting any inaccuracies. It also optimizes hyperparameters as necessary.

[1308] Step 6: Implementing the Emotion Engine

[1309] The server will install an emotion engine and add the ability to analyze emotions from the user's text input or voice commands. The emotion engine will recognize the user's emotions and record the data. For example, if it detects that the user is "angry," it will save this emotion data as the analysis result.

[1310] Step 7: Providing a User Interface

[1311] The server designs the user interface and provides it through the terminal. The interface includes forms and input fields, allowing users to easily enter inquiries. For example, a form is provided for entering "What to do if the product arrives damaged."

[1312] Step 8: Query entry and sentiment analysis

[1313] The user inputs a query through the interface, for example, "The product arrived damaged." At the same time, the emotion engine analyzes the user's emotions and detects "anger."

[1314] Step 9: Generate an answer with the AI ​​model

[1315] The server receives the query and uses a generative AI model to generate the optimal answer. Based on data from the emotion engine, the server provides an answer that takes the user's emotions into consideration. For example, it may suggest sending an email with an apology and applying a discount on the next order.

[1316] Step 10: Provide your answers

[1317] The device displays the generated answer from the server to the user. The user checks the suggested answer and takes action. For example, the device displays specific measures such as "send an apology email and immediately ship a replacement product" as a response to the question "What to do if the product arrives damaged."

[1318] Step 11: Enter your feedback

[1319] The user provides feedback on the appropriateness of the provided answer, for example, by rating whether the proposed response was effective.

[1320] Step 12: Gather feedback

[1321] The server collects feedback from users, including the appropriateness of the response and areas for improvement. For example, specific opinions such as "the response was a little slow" are collected.

[1322] Step 13: Retraining the AI ​​model

[1323] The server retrains the generative AI model based on the collected feedback and emotion data, improving its accuracy and response capabilities. For example, it can learn how to respond quickly to anger.

[1324] Step 14: Automatic generation of training manuals

[1325] The server automatically generates training manuals using the collected data and learning results. For example, a "manual for dealing with product defects" is generated.

[1326] Step 15: Provide training manuals

[1327] The server displays the generated training manual on the terminal and makes it available to new and current employees, for example, by providing it as a downloadable PDF file.

[1328] Through these steps, the system will improve the quality and efficiency of customer service and call center operations. The introduction of an emotion engine will enable responses that take user emotions into consideration, which is expected to improve customer satisfaction.

[1329] Example 2

[1330] 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."

[1331] In recent years, there has been a demand for improved quality and efficiency in customer service and call center operations when it comes to handling complaints. However, conventional systems have struggled to respond in a way that takes user emotions into consideration, limiting the extent to which they can improve customer satisfaction. Furthermore, they lack the ability to automatically generate appropriate response manuals, making the efficiency of employee training an issue.

[1332] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting past case data, means for preprocessing the collected case data, means for training a generative AI model using the preprocessed data, means for providing an interface through which a user can input an inquiry, means for using an emotion engine to analyze the user's emotions, means for generating an optimal answer to the user's inquiry using the generative AI model, means for providing the generated answer to the user, means for collecting feedback from the user, means for retraining the generative AI model using the collected feedback, means for automatically generating a training manual, and means for providing the generated training manual. This enables improved quality of complaint handling and business efficiency. Furthermore, the introduction of an emotion engine enables responses that take user emotions into consideration, which is expected to improve customer satisfaction. Furthermore, the automatic generation of training manuals allows for efficient training of new employees and skill development of current employees.

[1333] "Past case data" refers to data collected from customer response history, including the content of complaints, response methods, and feedback from customers.

[1334] "Preprocessing" refers to the process of cleansing unnecessary noise from collected data and tokenizing important text information to convert it into structured data.

[1335] A "generative AI model" is a model that is constructed using machine learning algorithms with preprocessed data to predict the appropriate way to handle a complaint.

[1336] An "interface" is a tool or platform through which a user can enter a query and interact with a system, such as a web form or a mobile app.

[1337] An "emotion engine" is a system that analyzes emotions from a user's text input or voice commands, and has the ability to record the user's emotional state as data.

[1338] An "Answer" is an optimal response to a user's query, created using a generative AI model and emotion engine.

[1339] "Feedback" refers to opinions and ratings provided by users regarding the appropriateness of proposed answers.

[1340] "Retraining" is the process of using collected feedback to readjust the parameters of a generative AI model to improve the model's accuracy.

[1341] The "training manual" is a document that is automatically generated based on collected data and the learning results of the generating AI model, and is used to train new employees and improve the skills of current employees.

[1342] One embodiment of the present invention is described in detail below. This system is designed to improve the quality and efficiency of customer service and complaint handling in call center operations. By combining the process of collecting and preprocessing past case data, generating and training a generative AI model, and incorporating an emotion engine that recognizes user emotions, more appropriate and emotion-sensitive responses are possible.

[1343] First, the server collects past case data from call centers and customer support systems. Specifically, it obtains data including the details of complaints, how they were handled, and customer feedback. Data is collected from a database and extracted using the Python "pandas" library. Next, the server preprocesses this data. During preprocessing, unnecessary data is cleansed, and important text information is tokenized using the "NLTK" library and converted into structured data.

[1344] Using the preprocessed data, the server builds and trains a generative AI model. TensorFlow and PyTorch are used to build the model, and the model is trained using a training dataset. Specifically, the model learns to predict the appropriate response method using paired data of the complaint content and the response method. For example, the following prompt sentence is used:

[1345] "How should you respond if a customer complains about the following? Complaint: ____"

[1346] The server then deploys an emotion engine, which uses IBM Watson and Microsoft Azure Emotion API to analyze emotions from the user's text input and voice commands. This engine records the user's emotional state as data and reflects it in the generative AI model.

[1347] The server designs an interface that allows users to enter inquiries and provides it via the terminal. The interface is provided as a web-based form or mobile app, and is designed so that users can easily access the system and enter inquiries. For example, a customer service representative can inquire about "what to do if a customer's order is delayed" through the system.

[1348] When a user inputs an inquiry into the system, the server uses a generative AI model to generate the optimal response to that inquiry. At the same time, an emotion engine analyzes the user's emotions and optimizes the response based on that emotional data. For example, if the user expresses anger, the system will provide a response that includes a more polite apology. As a specific example, when asked what to do if an order is delayed, the system will suggest responses such as "send an apology email and offer a coupon for free shipping on the next order."

[1349] The device displays the answer generated by the server to the user, who can then act on the suggested answer. The user can also provide feedback on whether the suggested answer is appropriate. The server collects the feedback and uses it to retrain the generative AI model to improve the quality of the answer. This improves the model's accuracy and makes it better able to respond to future inquiries.

[1350] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals can be used to train new employees and improve the skills of current employees. For example, a "Manual for Responding to Product Defects" can be provided in PDF format.

[1351] The above is an embodiment of the system of the present invention, which improves the quality and efficiency of customer service and complaint handling in call center operations. The introduction of an emotion engine enables responses that take user emotions into consideration, which is expected to improve customer satisfaction. Furthermore, the automatic generation of training manuals allows for efficient training of new employees and skill development for current employees.

[1352] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1353] Step 1: Data collection

[1354] The server collects past case data from call centers and customer support systems. This includes data on complaints, response methods, and customer feedback. Specifically, it extracts data from a database using SQL queries. The input is past case data, and the output is the extracted raw data.

[1355] Step 2: Data Preprocessing

[1356] The server preprocesses the collected data. Specifically, it cleanses unnecessary data using the Python "pandas" library and tokenizes text information using the "NLTK" library to convert it into structured data. The input is raw data, and the output is cleansed and tokenized structured data.

[1357] Step 3: Generate an AI model

[1358] The server uses the preprocessed data to build a generative AI model. It uses the machine learning library "TensorFlow" or "PyTorch" to define a neural network model. For example, it uses pairs of complaints and response methods as input data. The input is structured training data, and the output is an initialized generative AI model.

[1359] Step 4: Training the AI ​​model

[1360] The server trains the generative AI model. Specifically, it uses the preprocessed data to apply a learning algorithm (e.g., batch learning, setting the number of epochs, etc.) to improve the accuracy of the model. The input is the training data and the initialized AI model, and the output is the trained generative AI model.

[1361] Step 5: Implementing the Emotion Engine

[1362] The server installs an emotion engine and analyzes emotions from the user's text input or voice commands. IBM Watson and Microsoft Azure Emotion API are used as emotion engines. The input is text or voice data from the user, and the output is analyzed emotional data.

[1363] Step 6: Providing a User Interface

[1364] The server designs a query input interface and provides it through the terminal, such as a web form or a mobile app. The input is the query from the user, and the output is the query data sent to the system.

[1365] Step 7: Receiving and analyzing inquiries

[1366] A user inputs a query into the system. The server receives the query and uses a generative AI model to generate the optimal answer to that query. Specifically, a prompt such as "What should we do if a customer's order is delayed?" is used. The input is the user's query, and the output is the generated answer.

[1367] Step 8: Analyze sentiment data and optimize responses

[1368] The server uses an emotion engine to analyze the user's emotions and optimizes the answer of the generative AI model based on the emotion data. The input is the query and the generated answer, and the output is an optimized answer that takes emotions into consideration.

[1369] Step 9: Provide your answers

[1370] The device displays the generated answer from the server to the user, specifically on a web page or in the chat screen of the app. The input is the optimized answer, and the output is what is displayed to the user.

[1371] Step 10: Gather feedback

[1372] The user provides feedback on the appropriateness of the response. The server collects this feedback and stores it in a database. The input is the user feedback and the output is the collected feedback data.

[1373] Step 11: Retraining the AI ​​model

[1374] The server uses the collected feedback to retrain the generative AI model. Specifically, it uses the feedback data as new training data and readjusts the model's parameters. The inputs are the feedback data and the existing generative AI model, and the output is the retrained generative AI model.

[1375] Step 12: Automatic generation of training manuals

[1376] The server automatically generates training manuals based on the training data and model results. Specifically, it uses a Python script to analyze past cases and their effectiveness, and generates training manuals. The input is the training data and the model's learning results, and the output is the generated training manual (e.g., in PDF format).

[1377] Step 13: Provide training manuals

[1378] The terminal provides the generated training manual to the user. Specifically, it is provided as a download link or email attachment. The input is the generated training manual, and the output is the manual provided to the user.

[1379] (Application example 2)

[1380] 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."

[1381] When it comes to customer service for autonomous vehicles, there are cases where responses to user complaints or problems are delayed or inappropriate. In such cases, measures that take user feelings into consideration are required, and improving customer satisfaction is a challenge.

[1382] 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 collecting past response case data, means for preprocessing the collected response case data, means for training a generative AI model using the preprocessed data, means for providing an interface through which a user can input a query, means including an emotion engine for analyzing the user's emotions, means for generating an optimal answer to the user's query using the generative AI model, means for optimizing the generated answer based on user emotion analysis, means for providing the optimized answer to the user, means for collecting feedback from the user, means for retraining the generative AI model using the collected feedback, means for automatically generating a training manual, and means for providing the generated training manual. This enables user complaints and problems regarding autonomous vehicles to be handled quickly and appropriately, thereby improving customer satisfaction.

[1383] "Past case data" refers to data that includes details of past complaints and troubles, the methods used to deal with them, and feedback from customers.

[1384] "Preprocessing" is the process of removing unnecessary data from collected data and structuring important text information.

[1385] A "generative AI model" is a model trained using machine learning algorithms and used to predict optimal responses based on past response patterns.

[1386] A "user interface" is an interface through which a user inputs queries into a system, and supports communication between the user and the system.

[1387] The "emotion engine" is a system that analyzes emotions from the user's text and voice input and records that emotional state as data.

[1388] "Feedback" means opinions or evaluations provided by users regarding the appropriateness or quality of a system's response.

[1389] "Retraining" is the process of retraining a generative AI model using newly collected feedback to improve the model's accuracy and responsiveness.

[1390] An "education manual" is a document automatically generated based on collected data and learning results, and is used to train new employees and improve the skills of current employees.

[1391] The present invention is a system for quickly and appropriately responding to user complaints and problems with autonomous vehicles. This system collects and analyzes past case data and provides optimal countermeasures that take user emotions into consideration. An embodiment of the present invention is described in detail below.

[1392] Data collection and preprocessing

[1393] First, the server collects past case data from the autonomous vehicle's operation management system and customer response system. This includes complaint details, response methods, and customer feedback. Next, the server preprocesses this data, removing unnecessary data and tokenizing and structuring useful text information.

[1394] AI model generation and training

[1395] The server uses the preprocessed data to build a generative AI model and trains it using machine learning algorithms, allowing it to learn past response patterns and predict optimal responses, such as the appropriate way to respond to complaints.

[1396] Introducing the Emotion Engine

[1397] The server then uses an emotion engine to analyze the user's text input. The emotion engine uses natural language processing tools such as TextBlob to analyze the emotion into positive, negative, and neutral, and records the emotional state as data. This allows the AI ​​model to reflect appropriate emotional responses to the user's inquiry.

[1398] User Interface

[1399] The server designs and provides an interface through which users can input inquiries via terminals, allowing users to easily report complaints and problems.

[1400] User inquiries and sentiment analysis

[1401] When a user inputs a query into the system, the server uses a generative AI model to generate the optimal response for that query. At the same time, the emotion engine analyzes the user's emotions and optimizes the response based on that emotional data. For example, if the user expresses anger, the server will provide a more polite response that includes an apology.

[1402] Examples:

[1403] User input: "My car is stuck in the middle of the road! I'm so annoyed. What can you do?"

[1404] Emotion analysis result: "negative"

[1405] Example AI response: "We apologize for the inconvenience. Our technical team is working quickly to resolve the issue. We'll offer you a 20% discount coupon for your next visit."

[1406] Providing answers

[1407] The device displays the generated answer from the server to the user, who then acts on the suggested answer.

[1408] Gathering feedback and retraining

[1409] Users can provide feedback on the appropriateness of the proposed answers, which the server collects and uses to retrain the generative AI model, a process that improves the model's accuracy and makes it better able to respond to future inquiries.

[1410] Automatic generation of training manuals

[1411] Furthermore, the server automatically generates training manuals based on the collected data and learning results. These manuals are used to train new employees and improve the skills of current employees. For example, a "vehicle malfunction response manual" is generated and provided to the terminal in PDF format.

[1412] In this way, the system of the present invention improves the quality of customer service in self-driving vehicles and is expected to improve customer satisfaction by enabling emotionally sensitive service.

[1413] Example prompt sentence:

[1414] Prompt: "User's complaint: My car got stuck in the middle of the road! I'm so upset. What can you do about it? Emotion: negative Suggest the best solution:"

[1415] This embodiment makes it possible to specifically implement the present invention.

[1416] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1417] Step 1:

[1418] The server collects past case data from the autonomous vehicle's operation management system and customer response system. The collected data includes the details of the complaint, how it was handled, and customer feedback. The input is raw data obtained from the operation management system database, and the output is sent to pre-processing through the collection step.

[1419] Step 2:

[1420] The server preprocesses the collected data. During this process, unnecessary data is cleansed and useful text information is tokenized and structured. Processes such as text normalization, stop word removal, and tokenization are performed. The input is the collected raw data, and the output is cleansed and structured data.

[1421] Step 3:

[1422] The server uses the preprocessed data to generate and train a generative AI model. The model is trained using machine learning algorithms (e.g., deep learning) to learn appropriate countermeasures based on past response patterns. The input is the preprocessed data, and the output is a trained generative AI model.

[1423] Step 4:

[1424] The server uses an emotion engine to analyze the user's emotions. In this process, the text entered by the user is analyzed to determine whether the emotion is positive, negative, or neutral, and the data is recorded. The input is the text input from the user, and the output is the analyzed emotion data.

[1425] Step 5:

[1426] The user uses an interface to input inquiries and complaints. The server designs this interface to make it easy for the user to input inquiries. The input is the complaint or inquiry content from the user, and the output is the inquiry data that is sent to the server.

[1427] Step 6:

[1428] The server uses a generative AI model to generate the optimal answer to a user's inquiry. The answer is optimized based on the user's emotional data analyzed by the emotion engine. For example, if the emotion included anger, a particularly polite response would be suggested. The input is the user's inquiry data and emotional data, and the output is an optimized answer.

[1429] Step 7:

[1430] The terminal displays the generated answer to the user, who then reviews the proposed answer and takes action. The input is the optimized answer sent by the server, and the output is the generated answer that is displayed to the user.

[1431] Step 8:

[1432] Users provide feedback on the appropriateness of the proposed answers, which the server collects and uses to retrain the generative AI model. The input is the user feedback data, and the output is the retraining data.

[1433] Step 9:

[1434] The server automatically generates a training manual based on the collected data and learning results. This manual is used to train new employees and improve the skills of current employees. The inputs are the collected data and learning results, and the output is a training manual (e.g., a PDF document).

[1435] In this way, each step works together to create a system that can quickly and appropriately handle complaints and problems with self-driving vehicles.

[1436] 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.

[1437] 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.

[1438] 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.

[1439] 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.

[1440] 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.

[1441] 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.

[1442] 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).

[1443] 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.

[1444] 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."

[1445] 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.

[1446] 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).

[1447] 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.

[1448] 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.

[1449] 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.

[1450] 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.

[1451] 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.

[1452] 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.

[1453] 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.

[1454] 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.

[1455] 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.

[1456] 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.

[1457] The following is further disclosed regarding the above embodiment.

[1458] (Claim 1)

[1459] A means of collecting past case data;

[1460] means for preprocessing the collected corresponding case data;

[1461] means for training a generative AI model using the preprocessed data;

[1462] means for providing an interface through which a user can enter a query;

[1463] A means for generating an optimal answer to the user's inquiry using the generative AI model;

[1464] means for providing the generated answer to a user;

[1465] means for collecting feedback from said users;

[1466] means for retraining a generative AI model using the collected feedback;

[1467] A means for automatically generating an educational manual;

[1468] A system including means for providing the generated instructional manual.

[1469] (Claim 2)

[1470] 2. The system according to claim 1, wherein when generating an answer to the user's inquiry, suggestions are made based on similar past cases.

[1471] (Claim 3)

[1472] 2. The system according to claim 1, wherein the generated training manual is used to train new employees and improve the skills of current employees.

[1473] "Example 1"

[1474] (Claim 1)

[1475] A means of collecting past case data;

[1476] means for preprocessing the collected corresponding case data;

[1477] means for training a machine learning model using the preprocessed data;

[1478] means for providing an interface through which a user can enter a query;

[1479] means for generating an optimal answer to the user's query using the machine learning model;

[1480] means for providing the generated answer to a user;

[1481] means for collecting feedback from said users;

[1482] means for retraining a machine learning model using the collected feedback;

[1483] a means for automatically generating educational materials;

[1484] A system including means for providing said generated educational material.

[1485] (Claim 2)

[1486] 2. The system according to claim 1, wherein when generating an answer to the user's inquiry, suggestions are made based on similar past cases.

[1487] (Claim 3)

[1488] 2. The system according to claim 1, wherein the generated educational materials are used to train new employees and improve the skills of current employees.

[1489] "Application Example 1"

[1490] (Claim 1)

[1491] A means of collecting past case data;

[1492] means for preprocessing the collected corresponding case data;

[1493] means for training a generative AI model using the preprocessed data;

[1494] means for providing an interface through which a user can enter a query;

[1495] A means for generating an optimal answer to the user's inquiry using the generative AI model;

[1496] means for providing the generated answer to a user;

[1497] means for collecting feedback from said users;

[1498] means for retraining a generative AI model using the collected feedback;

[1499] A means for automatically generating an educational manual;

[1500] means for providing the generated training manual;

[1501] A means of responding to inquiries using a smart device;

[1502] A system that includes a means for receiving inquiries and providing responses in real time.

[1503] (Claim 2)

[1504] 2. The system according to claim 1, wherein when generating an answer to the user's inquiry, suggestions are made based on similar past cases.

[1505] (Claim 3)

[1506] 2. The system according to claim 1, wherein the generated training manual is used to train new employees and improve the skills of current employees.

[1507] "Example 2: Combining Emotion Engines"

[1508] (Claim 1)

[1509] A means of collecting past case data;

[1510] means for preprocessing the collected corresponding case data;

[1511] means for training a generative AI model using the preprocessed data;

[1512] means for providing an interface through which a user can enter a query;

[1513] A means using an emotion engine for analyzing the emotion of the user;

[1514] A means for generating an optimal answer to the user's inquiry using the generative AI model;

[1515] means for providing the generated answer to a user;

[1516] means for collecting feedback from said users;

[1517] means for retraining a generative AI model using the collected feedback;

[1518] A means for automatically generating an educational manual;

[1519] A system including means for providing the generated instructional manual.

[1520] (Claim 2)

[1521] The system according to claim 1, wherein when generating an answer to the user's inquiry, the system makes an optimal suggestion based on the user's emotions.

[1522] (Claim 3)

[1523] 2. The system according to claim 1, wherein the generated training manual is used to train new employees and improve the skills of current employees.

[1524] "Application example 2 when combining emotion engines"

[1525] (Claim 1)

[1526] A means of collecting past case data;

[1527] means for preprocessing the collected corresponding case data;

[1528] means for training a generative AI model using the preprocessed data;

[1529] means for providing an interface through which a user can enter a query;

[1530] means including an emotion engine for analyzing the emotion of the user;

[1531] A means for generating an optimal answer to the user's inquiry using the generative AI model;

[1532] means for optimizing the generated answers based on a sentiment analysis of the user;

[1533] means for providing the optimized answer to a user;

[1534] means for collecting feedback from said users;

[1535] means for retraining a generative AI model using the collected feedback;

[1536] A means for automatically generating an educational manual;

[1537] A system including means for providing the generated instructional manual.

[1538] (Claim 2)

[1539] The system according to claim 1, wherein when generating an answer to the user's inquiry, the system makes suggestions based on past similar cases and an analysis of the user's sentiment.

[1540] (Claim 3)

[1541] 2. The system according to claim 1, wherein the generated training manual is used to train new employees and improve the skills of current employees. [Explanation of symbols]

[1542] 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 collecting past case data; means for preprocessing the collected corresponding case data; means for training a generative AI model using the preprocessed data; means for providing an interface through which a user can enter a query; A means for generating an optimal answer to the user's inquiry using the generative AI model; means for providing the generated answer to a user; means for collecting feedback from said users; means for retraining a generative AI model using the collected feedback; A means for automatically generating an educational manual; A system including means for providing the generated instructional manual.

2. The system according to claim 1 , wherein when generating an answer to the user's inquiry, suggestions are made based on similar past cases.

3. The system according to claim 1 , wherein the generated training manual is used to train new employees and to improve the skills of current employees.

Images