system

Abstract

Provide a system. A means for loading past interview case data from a database; means for training a machine learning model based on the past interview case data; means for providing an interface for a user to input candidate information; means for transmitting the input candidate information to a server; means for analyzing the candidate information and generating initial questions; means for displaying the generated initial question to a user; a means for a user to input and transmit candidate answers to a server; means for analyzing the candidate's answer and generating a next question; A system including:

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 interviews and financial reporting situations, there is a need to generate appropriate questions to smoothly advance the dialogue, but conventional methods often rely on a fixed list of questions, making it difficult to flexibly generate questions according to the progress of the dialogue.In addition, due to the lack of a system that interactively generates and displays the next question based on data input by the user, the flow of the dialogue tends to be intermittent, making it difficult to achieve efficient and high-quality communication. [Means for solving the problem]

[0005] The present invention realizes a system for dynamically and interactively generating questions by providing a means for loading past interview case studies and financial report data from a database and training a machine learning model based on that data. The system includes a means for providing an interface for users to input candidate information and financial report content, transmitting the input data to a server, analyzing the data, generating initial questions, and displaying them to the user. The system also includes a means for generating further questions based on answers provided by candidates and managers, and providing the next questions in a timely manner, thereby smoothly advancing the flow of the dialogue. This system allows users to have more effective and natural dialogue, thereby improving the quality of interviews and financial reports.

[0006] A "database" is a system for organizing and storing information and data.

[0007] "Interview case data" is data that includes information on questions and answers related to interviews that have been conducted in the past.

[0008] A "machine learning model" is an algorithm or statistical model that uses data to learn specific patterns and rules and make predictions or judgments about new data.

[0009] "User" refers to the people or organizations that use the system, which in this case includes interviewers and managers.

[0010] "Candidate information" refers to data such as personal information, career history, and reasons for applying for the job regarding the candidate to be interviewed.

[0011] An "interface" is the means or screen through which a user interacts with a system.

[0012] An "initial question" is the first question the system provides at the start of an interview or financial report.

[0013] "Analysis" is the process of examining data or information in detail to derive meaning and relationships.

[0014] An "answer" is information in response to a question that a user provides to the system.

[0015] A "server" is a computer system that manages and processes data on a network.

[0016] "Next Questions" are the next series of questions that the system generates based on the user's answers.

[0017] "Timely" means at the right time, and done promptly and without delay.

[0018] "Interactive" refers to a form of two-way communication between the user and the system.

[0019] "Input" refers to data or information provided by a user to a system. [Brief explanation of the drawings]

[0020] [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

[0021] 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.

[0022] First, the terms used in the following description will be explained.

[0023] 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).

[0024] 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.

[0025] 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.

[0026] 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.

[0027] 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."

[0028] [First embodiment]

[0029] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0030] 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.

[0031] 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).

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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.

[0036] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0037] 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.

[0038] 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.

[0039] 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.

[0040] 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."

[0041] An embodiment of the present invention will be described below. An interview scene will be described as an example, but it should be noted that the present invention can also be applied to a financial report scene in a similar manner.

[0042] server

[0043] 1. A way to load interview case data from a database

[0044] The server loads past interview case data from a database, which includes questions asked in interviews and their answers. By retrieving data from the database, the server builds a training dataset.

[0045] 2. Means of training machine learning models

[0046] Based on the loaded interview case data, the server trains a machine learning model. Specifically, it uses natural language processing technology to learn patterns of past questions and answers and create a model that generates new questions. This model includes a question generation algorithm and can generate appropriate questions based on the candidate's information.

[0047] 3. A means of analyzing user-entered data and generating initial questions

[0048] When the user (interviewer) enters candidate information and sends it to the server, the server analyzes the data and generates initial questions, where an algorithm is applied to select the most appropriate initial questions based on the information entered by the user.

[0049] Terminal

[0050] 1. A means of providing an interface for users to input candidate information

[0051] The terminal provides an interface for the user to input candidate information, specifically, a form for inputting name, career history, reason for applying, etc.

[0052] 2. Means of sending the entered data to the server

[0053] When the user (interviewer) has completed the input, the data is sent to the server using a normal HTTP request.

[0054] 3. A means to display the generated initial questions to the user

[0055] When the terminal receives the generated initial questions from the server, it displays them to the user (interviewer). Once the questions are displayed, the user can start asking questions to the candidate.

[0056] 4. A means for users to input candidate answers and send them to the server

[0057] After the user (interviewer) receives the answer from the candidate, he / she enters the answer into the terminal and sends it to the server. This procedure is performed using an HTTP request, just like the initial question.

[0058] User

[0059] 1. Enter candidate information

[0060] The user (interviewer) enters candidate information, such as the candidate's name, highest level of education, work history, and reason for applying, according to the provided interface.

[0061] 2. Asking initial questions and getting answers

[0062] The user (interviewer) views the initial questions generated by the server, asks them to the candidate, and receives answers.

[0063] 3. Enter and submit your answers

[0064] The answer obtained is entered into the terminal and sent to the server to determine what question to ask next.

[0065] Specific examples

[0066] For example, consider an initial interview question that might be generated: "Tell me about a recent project." The interviewer asks this question to a candidate, who responds, "I worked on project X, which resulted in Y results." This response is entered into a device and sent to a server, which analyzes it and generates the next question: "What was your specific role in that project?" This process is repeated to effectively progress the interview.

[0067] In this way, users can easily generate appropriate questions based on the candidate's information, and the dialogue can proceed smoothly. This system can improve the quality of interviews.

[0068] The processing flow will be explained below.

[0069] Step 1:

[0070] The server loads the past interview case data from the database, which involves establishing a database connection and executing SQL queries to retrieve the required data.

[0071] Step 2:

[0072] The server trains a machine learning model based on the loaded interview case data. Specifically, it performs data preprocessing and trains the model using natural language processing techniques. The model learns patterns from question-answer pairs.

[0073] Step 3:

[0074] The terminal displays an interface for entering candidate information. The user (interviewer) uses this interface to enter the candidate's name, career history, reason for applying, etc.

[0075] Step 4:

[0076] Once the user has completed the input, the device sends the data as an HTTP request to the server, which receives the request and analyzes the candidate information.

[0077] Step 5:

[0078] The server generates initial questions based on the analysis results, uses a trained machine learning model to select questions that best fit the candidate information, and sends the generated questions to the device.

[0079] Step 6:

[0080] The terminal displays the initial questions received from the server to the user, who then verbally asks the candidate the questions.

[0081] Step 7:

[0082] The user receives answers from the candidates, enters them into the terminal, and sends them back to the server, again using an HTTP request.

[0083] Step 8:

[0084] The server analyzes the received answer and uses an analysis algorithm to extract important information from the answer and generate the next question.

[0085] Step 9:

[0086] The server sends the next generated question to the terminal, which then displays the received question to the user (interviewer).

[0087] Step 10:

[0088] The user asks the candidate the following questions, and this process is repeated until the interview is complete, improving the quality of the interview and facilitating a smooth dialogue.

[0089] This series of processing steps enables the present invention to provide users with dynamic and interactive question generation, improving the efficiency and quality of interviews and financial reports.

[0090] Example 1

[0091] 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."

[0092] In conventional interview systems and financial reporting systems, it is difficult for interviewers or reporters to quickly generate appropriate questions, which can lead to inefficient dialogue. Furthermore, if questions are not generated appropriately, the quality of the dialogue declines and necessary information cannot be fully elicited. The present invention aims to solve these problems by providing a system that can automatically generate appropriate and effective questions.

[0093] 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.

[0094] In this invention, the server includes means for loading past interview case data from a database, means for training a machine learning model based on the past interview case data, means for providing an interface for a user to input candidate information, means for transmitting the input candidate information to the server, means for analyzing the candidate information and generating initial questions, means for displaying the generated initial questions to the user, means for a user to input answers from the candidate and transmit them to the server, and means for analyzing the answers from the candidate and generating the next question. This enables interviewers and reporters to effectively conduct dialogue and appropriately extract necessary information.

[0095] A "database" is a system for efficiently storing, searching, and managing large amounts of data.

[0096] "Interview case data" refers to records of questions and answers from past interviews.

[0097] A "machine learning model" is a collection of algorithms that learn patterns and rules based on data and use the results of that learning to make predictions and classify new data.

[0098] "User" refers to the person who operates the system, inputs information, and views results.

[0099] An "interface" is a means, such as a screen or input form, through which a user interacts with a system.

[0100] A "server" is a computer system that processes and stores data over a network.

[0101] "Analysis" is the process of examining input data in detail and extracting meaning and patterns.

[0102] An "initial question" refers to the first question posed to a candidate or reporter when an interview or report begins.

[0103] "Answer" refers to the information or response returned in response to a question.

[0104] The "next question" is a new question that is generated based on the previous answer.

[0105] A "system" is a mechanism in which multiple components and processes work together to provide specific functions and services.

[0106] This invention relates to an interview system and a financial reporting system, and aims to effectively progress the dialogue by having a user input information about candidates and reporters, and then having a server analyze the information and automatically generate appropriate questions.

[0107] server

[0108] The server first uses a database management system to load past interview case data and financial report data from a database. This data includes past questions and their answers. The server then uses natural language processing technology to train a machine learning model based on the loaded data. Specifically, it trains a BERT-based model using frameworks such as TENSORFLOW (registered trademark) and PyTorch to build a question generation algorithm.

[0109] Terminal

[0110] The terminal provides an interface for users to input candidate information and report details, such as name, career history, reasons for applying, and financial report details, through a web form written in React. Once the user has completed the input, the data is sent to the server as an HTTP POST request.

[0111] The terminal then receives the initial questions generated by the server and displays them to the user. Similarly, when the user enters the candidate's answers and additional information, that data is also sent to the server.

[0112] User

[0113] The user uses the terminal to input candidate information and financial report details. The specific operating procedure is as follows:

[0114] 1. The user follows the provided interface to enter the candidate's name, highest level of education, work history, reason for applying, etc.

[0115] 2. Check the initial questions generated by the server and ask them to the candidate to get their answers.

[0116] 3. The answer obtained is entered into the terminal and sent to the server to generate the next question.

[0117] Specific examples

[0118] For example, an initial interview question might be generated as "Tell me about a recent project." If a user poses this question to a candidate and the candidate responds with "I worked on project X, which resulted in Y results," the response is typed into the device and sent to the server. The server then analyzes the response and generates the next question, "What was your specific role in that project?", which is returned to the device and displayed to the user. This process is repeated to effectively progress the interview.

[0119] Examples of specific prompts include the following:

[0120] Please tell us about the candidate's highest level of education.

[0121] What is your specific role in this project?

[0122] This invention allows users to quickly and efficiently generate appropriate questions for candidates and reporters, allowing for smooth dialogue, thereby improving the quality of interviews and the effectiveness of reports.

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

[0124] Specific explanation of program processing

[0125] server

[0126] Step 1:

[0127] The server loads past interview case data from a database.

[0128] Input: Database connection information

[0129] Output: Interview case data list

[0130] What it does: The server executes an SQL query to retrieve interview case data from a database, which includes past interview questions and answers.

[0131] Step 2:

[0132] The server trains a machine learning model based on the loaded interview case data.

[0133] Input: Interview case data list

[0134] Output: A trained machine learning model

[0135] How it works: The server uses TensorFlow and PyTorch to train a BERT-based natural language processing model, which involves learning question and answer patterns.

[0136] Step 3:

[0137] The server analyzes the candidate information sent by the user from the terminal and generates an initial question.

[0138] Input: Candidate information

[0139] Output: Initial question

[0140] Specific operation: The server uses a machine learning model to analyze candidate information and generate appropriate initial questions.

[0141] Terminal

[0142] Step 1:

[0143] The terminal provides an interface for the user to input candidate information.

[0144] Input: None (standby)

[0145] Output: Form for entering candidate information

[0146] Specific operation: The device uses "React" to display a web form for entering information such as name, background, and reasons for applying.

[0147] Step 2:

[0148] The terminal transmits the input candidate information to the server.

[0149] Input: Candidate information

[0150] Output: HTTP POST request

[0151] Specific operation: Generates and sends an HTTP POST request to send the candidate information entered by the user in the form to the server.

[0152] Step 3:

[0153] The terminal displays the generated initial question to the user.

[0154] Input: Initial question (response from server)

[0155] Output: Question displayed on the user interface

[0156] Specific operation: The initial question received from the server is displayed clearly to the user.

[0157] Step 4:

[0158] The user inputs the candidate's answers at the terminal and transmits them to the server.

[0159] Input: Candidate's answer

[0160] Output: HTTP POST request

[0161] Specific operation: The answers entered by the user are recorded on the terminal and sent to the server.

[0162] User

[0163] Step 1:

[0164] The user inputs the candidate information.

[0165] Input: Candidate information (name, highest level of education, work history, reason for applying, etc.)

[0166] Output: Input data (terminal)

[0167] Specific operation: The user enters each item according to the form provided on the terminal.

[0168] Step 2:

[0169] The user poses initial questions to the candidate and receives answers.

[0170] Input: Initial question (displayed on the terminal)

[0171] Output: Candidate's answers

[0172] Specific operation: The user asks the candidate initial questions displayed on the terminal and receives answers.

[0173] Step 3:

[0174] The user inputs the answers obtained and sends them to the server to generate the next question.

[0175] Input: Candidate's answer

[0176] Output: HTTP POST request

[0177] Specific operation: The user inputs the candidate's answer into the terminal and sends a request to the server to generate the next question.

[0178] (Application example 1)

[0179] 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."

[0180] Customer support and dialogue systems in modern virtual environments are based on static FAQs and scripts, making it difficult to provide appropriate responses to specific user questions and preventing efficient and effective interactive dialogue that meets user needs. A particular problem is the lack of systems that can quickly and accurately answer questions visitors have about specific products or services in virtual stores and online support.

[0181] 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.

[0182] In this invention, the server includes: means for loading past case data from a database; means for training a machine learning model based on the past case data; means for providing an interface for a user to input information; means for transmitting the input information to the server; means for analyzing the information and generating an initial question; means for displaying the generated initial question to the user; means for the user to input an answer and transmitting the answer to the server; means for analyzing the answer and generating a next question; means for answering questions about products and services from visitors through dialogue in a virtual environment; means for functioning as an application installed on a smart device; and means for dynamically generating questions and interactively displaying them through a user interface. This enables users' questions to be answered quickly and accurately, significantly improving the quality of customer support in virtual stores.

[0183] A "database" is a repository that systematically organizes information so that it can be searched and retrieved efficiently.

[0184] "Case data" is data that compiles information about specific past events or situations.

[0185] A "machine learning model" is a set of algorithms that are trained with data to perform specific tasks automatically.

[0186] An "interface" is a means or device by which a user and a system interact with each other.

[0187] A "server" is a computer that provides services and resources to other computers over a network.

[0188] "Analysis" is the process of examining data and information in detail and clarifying their structure and relationships.

[0189] An "initial question" is a question that the dialogue system generates to start the initial dialogue with the user.

[0190] "User" means a person or entity that uses a system or service.

[0191] An "answer" is a response given to a question.

[0192] The "next question" is the next question that the system generates based on the user's answer.

[0193] A "virtual environment" is a virtual space or system created by computer simulation.

[0194] "Goods" means goods or services manufactured for sale.

[0195] A "service" is an activity or undertaking that provides benefit or satisfaction to consumers.

[0196] A "smart device" is an electronic device with advanced computing power and connectivity capabilities.

[0197] An "application" is software designed for a specific purpose.

[0198] A "user interface" is the means or method by which a user interacts with a system or device.

[0199] The present invention relates to a customer support assistant system for effectively conducting dialogue with visitors in a virtual store. The system aims to respond quickly and appropriately to visitors' questions and increase their satisfaction. The following describes in detail the embodiments of the present invention.

[0200] server

[0201] The server implements the following methods:

[0202] A means of loading case data from a database

[0203] The server loads past customer support case data from the database, including past questions and their answers, which the server uses as a training data set.

[0204] A means to train machine learning models

[0205] Using the example data loaded onto the server, the server trains a machine learning model. Specifically, it uses natural language processing techniques to learn patterns of past questions and answers and create a model that generates new questions. An example of the software used is the Hugging Face transformers library.

[0206] A means of analyzing user information and generating initial questions

[0207] When the user (customer support representative) enters visitor information and sends it to the server, the server analyzes the data and applies an algorithm to generate the most appropriate initial questions, again using natural language processing technology.

[0208] Terminal

[0209] The terminal provides the means to:

[0210] A means of providing an interface for users to input information

[0211] The terminal provides an interface for the user to input visitor information, such as name, inquiry details, and past purchase history.

[0212] A means of sending the entered data to the server

[0213] Once the user has completed the input, the data is sent to the server using an HTTP request.

[0214] A means of displaying the generated initial questions to the user

[0215] Upon receiving the generated initial question from the server, the terminal displays the question to the user (customer support representative), allowing the user to begin asking questions to the visitor.

[0216] A means for the user to enter answers and send them to the server

[0217] After the user retrieves the visitor's answer, they enter it into their device and send it to the server, again using an HTTP request to send the data.

[0218] User

[0219] The role of the user (customer support representative) is to:

[0220] Entering visitor information

[0221] The user inputs visitor information using the provided interface, such as the visitor's name, inquiry details, and past purchase history.

[0222] Asking initial questions and getting answers

[0223] The user sees the initial question generated by the server, poses it to the visitor, and receives the answer.

[0224] Enter and submit your answers

[0225] The answer obtained is entered into the terminal and sent to the server to determine what question to ask next.

[0226] Examples and prompts

[0227] For example, consider a case where a visitor to a virtual store wants information about a new product. Based on the visitor's purchase history and inquiries entered through the terminal, the server generates an initial question: "What are the features of this product?" After receiving the visitor's answer, the server generates the next question: "What are the unique features of this product?" This allows for timely and accurate responses to the visitor's questions.

[0228] Example prompt sentence:

[0229] "Please tell us the features of the products you recently released."

[0230] "How have you improved after receiving customer feedback?"

[0231] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0232] Step 1:

[0233] Loading case data from a database

[0234] The server loads past customer support case data from a database, including past questions and their answers, and uses this data as a training dataset to train the machine learning model.

[0235] Input: Query request to the database

[0236] Output: Dataset of past cases

[0237] Step 2:

[0238] Train a machine learning model

[0239] The server uses the loaded example data to train a machine learning model, leveraging natural language processing techniques. Specifically, it uses Hugging Face's transformers library to learn patterns of past questions and answers and create a model that can generate new questions.

[0240] Input: Historical case dataset

[0241] Output: A trained machine learning model

[0242] Step 3:

[0243] Provides an interface for entering user information

[0244] The terminal provides an interface for the user (customer support representative) to input visitor information, which is displayed as a form for entering information such as the user's name, inquiry details, and past purchase history.

[0245] Input: None (launches user interface)

[0246] Output: Interface screen display

[0247] Step 4:

[0248] Send the entered information to the server

[0249] Once the user has entered the visitor information into the interface, the device sends the information to the server. This data transmission is done using an HTTP request.

[0250] Input: Visitor information (user input)

[0251] Output: Send data to the server (HTTP request)

[0252] Step 5:

[0253] Generate initial questions

[0254] The server analyzes the received visitor information and applies machine learning models to generate the most appropriate initial questions. Natural language processing technology selects appropriate questions based on actual visitor information.

[0255] Input: Visitor information (received by server)

[0256] Output: Initial question

[0257] Step 6:

[0258] Display the generated initial questions to the user

[0259] The terminal displays the initial question retrieved from the server to the customer support representative, and the user starts a dialogue with the visitor based on this question.

[0260] Input: Initial question (sent by server)

[0261] Output: Question displayed on the screen

[0262] Step 7:

[0263] Enter the visitor's answer and send it to the server

[0264] The user enters the answers received from the visitors into the terminal and sends the answer data to the server. This data transmission is also done using an HTTP request.

[0265] Input: Visitor's answer (user input)

[0266] Output: Send data to the server (HTTP request)

[0267] Step 8:

[0268] Generate the next question

[0269] The server analyzes the visitor's answers and generates the next question, again dynamically using a machine learning model.

[0270] Input: Visitor's response (received by the server)

[0271] Output: Next question

[0272] Step 9:

[0273] Show the next question to the user

[0274] The terminal displays the next question retrieved from the server to the customer support representative, allowing the user to pose the next question to the visitor.

[0275] Input: Next question (sent by server)

[0276] Output: Question displayed on the screen

[0277] Prompt Sentence Examples

[0278] 1. "Please tell us about the features of the products you recently released."

[0279] 2. "How have you improved after receiving customer feedback?"

[0280] 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.

[0281] The following describes an embodiment of the present invention. The following describes an interview scene as an example, but please note that the present invention can also be applied to a financial report scene using similar methods. The present invention also incorporates an emotion engine.

[0282] server

[0283] 1. A way to load past interview case data from a database

[0284] The server loads past interview case data from a database, which includes questions asked in interviews and their answers. By retrieving data from the database, the server builds a training dataset.

[0285] 2. Means of training machine learning models

[0286] Based on the loaded interview case data, the server trains a machine learning model. Specifically, it uses natural language processing technology to learn patterns of past questions and answers and create a model that generates new questions. This model includes a question generation algorithm and can generate appropriate questions based on the candidate's information.

[0287] 3. A means of analyzing user-entered data and generating initial questions

[0288] When the user (interviewer) enters candidate information and sends it to the server, the server analyzes the data and generates initial questions, where an algorithm is applied to select the most appropriate initial questions based on the information entered by the user.

[0289] Terminal

[0290] 1. A means of providing an interface for users to input candidate information

[0291] The terminal provides an interface for the user to input candidate information, specifically, a form for inputting name, career history, reason for applying, etc.

[0292] 2. Means of sending the entered data to the server

[0293] When the user (interviewer) has completed the input, the data is sent to the server using a normal HTTP request.

[0294] 3. A means to display the generated initial questions to the user

[0295] When the terminal receives the generated initial questions from the server, it displays them to the user (interviewer). Once the questions are displayed, the user can start asking questions to the candidate.

[0296] 4. A means for users to input candidate answers and send them to the server

[0297] After the user (interviewer) receives the answer from the candidate, he / she enters the answer into the terminal and sends it to the server. This procedure is performed using an HTTP request, just like the initial question.

[0298] Emotion Engine

[0299] 1. A means of recognizing emotions by analyzing the user's voice and facial expressions

[0300] The emotion engine analyzes voice and facial expression data provided by the device, allowing it to recognize the user's emotions (e.g., joy, anger, sadness, surprise, etc.).

[0301] 2. A way to generate questions based on perceived sentiment

[0302] The recognized emotion data is sent to the server, which then generates questions based on the emotion information. For example, if the user indicates a high stress level, the system generates questions to encourage relaxation.

[0303] User

[0304] 1. Enter candidate information

[0305] The user (interviewer) enters candidate information, such as the candidate's name, highest level of education, work history, and reason for applying, according to the provided interface.

[0306] 2. Asking initial questions and getting answers

[0307] The user (interviewer) views the initial questions generated by the server, asks them to the candidate, and receives answers.

[0308] 3. Enter and submit your answers

[0309] After receiving the candidate's answer, the answer is entered into the terminal and sent to the server to get the next question to ask.

[0310] 4. Emotion Recognition by Emotion Engine

[0311] During the interview, the device captures the user's voice and facial expressions, which are then analyzed by an emotion engine. The analysis results are sent to the server and used to generate the next questions.

[0312] Specific examples

[0313] For example, consider an initial interview question generated as "Tell me about a recent project." The interviewer asks this question to a candidate, who responds, "I worked on project X, which resulted in Y results." When this response is entered into a device and sent to a server, the server analyzes it and generates the next question: "What was your specific role in that project?"

[0314] During this process, if the emotion engine detects the interviewer's stress level, for example if it is determined to be high, the server can take this into account and generate relaxing questions such as, "What was the most rewarding moment in that project?"

[0315] In this way, users can easily generate appropriate questions based on the candidate's information, and the dialogue can proceed smoothly. This system improves the quality of interviews and also enables flexible responses according to the user's emotional state.

[0316] The processing flow will be explained below.

[0317] Step 1:

[0318] The server loads past interview case data from the database by establishing a database connection and executing SQL queries to retrieve the required data.

[0319] Step 2:

[0320] The server trains a machine learning model based on the loaded interview case data. Specifically, it performs data preprocessing and trains the model using natural language processing techniques.

[0321] Step 3:

[0322] The terminal provides the user (interviewer) with an interface for inputting candidate information. The user inputs the candidate's name, career history, reason for applying, etc.

[0323] Step 4:

[0324] Once the user has completed the input, the device sends the data as an HTTP request to the server, which receives the request and analyzes the candidate information.

[0325] Step 5:

[0326] The server generates initial questions based on the analysis results, uses a trained machine learning model to select questions that best fit the candidate information, and sends the generated questions to the device.

[0327] Step 6:

[0328] The terminal displays the initial questions received from the server to the user, who then verbally asks the candidate the questions.

[0329] Step 7:

[0330] The emotion engine analyzes the user's voice and facial expressions in real time to evaluate the user's emotional state, and the evaluation results are sent to the server.

[0331] Step 8:

[0332] The user receives answers from the candidates, enters these answers into the terminal, and sends them back to the server.

[0333] Step 9:

[0334] The server analyzes the received answer and also uses the emotion data from the emotion engine for analysis. The analysis algorithm takes into account the important information in the answer and the user's emotion to generate the next question.

[0335] Step 10:

[0336] The server sends the next generated question to the terminal, which then displays the received question to the user (interviewer).

[0337] Step 11:

[0338] The user asks the candidate the following questions, and this process is repeated until the interview is complete, improving the quality of the interview and facilitating a smooth dialogue.

[0339] Through this series of processing steps, the present invention provides users with dynamic and interactive question generation, improving the efficiency and quality of interviews. In addition, by combining it with an emotion engine, flexible responses according to the user's emotional state become possible.

[0340] Example 2

[0341] 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."

[0342] In traditional situations such as interviews and financial reports, efficient and effective question generation is required, but in the process, it has been difficult to generate questions that take human emotions into account and maintain a continuity of appropriate questions. In particular, the lack of a dynamic question generation method that takes into account the emotional state of the interviewer or reporter can lead to a decline in the quality of interviews and reports.

[0343] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0344] In this invention, the server includes means for loading past interview case data from a database, means for training a machine learning model based on the past interview case data, means for providing an interface for a user to input candidate information, means for transmitting the input candidate information to the server, means for analyzing the candidate information and generating initial questions, means for displaying the generated initial questions to the user, means for a user to input candidate answers and transmitting the answers to the server, means for analyzing the candidate answers and generating subsequent questions, means for analyzing the user's voice and facial expressions to recognize emotions, and means for generating questions based on the recognized emotions. This makes it possible to generate appropriate and consecutive questions while taking into account the user's emotional state during interviews and reporting situations.

[0345] A "database" is a collection of information that efficiently manages past interview case data, financial report data, etc., and can be searched and retrieved as needed.

[0346] A "machine learning model" is an algorithm that learns from past data, discovers new patterns and relationships, and automates specific tasks such as question generation.

[0347] An "interface" is a user interface component that allows the user to interact with the system and input candidate information, financial report details, and the like.

[0348] "Server" means a computer system that receives information sent by users and processes and analyzes the data.

[0349] "Analysis" is the process of examining input data in detail to understand its structure and meaning.

[0350] An "initial question" is a question that is presented to a user when they first begin interacting with the system.

[0351] The "emotion engine" is a system that analyzes the user's voice and facial expression data to recognize their emotional state at that time.

[0352] "Dynamic question generation" is the process of generating new questions in real time based on the current user situation and answers.

[0353] "Interview case data" is data on questions used in past interviews and the answers to those questions.

[0354] "Recognized emotion" refers to the emotional state of the user that has been analyzed and identified by the emotion engine.

[0355] The present invention provides a method for generating questions efficiently and effectively in interactive systems for interviews, financial reporting, etc. While the present invention will be described in detail below using an interview scene as an example, it can also be applied to financial reporting scenes using similar means.

[0356] Hardware and software used

[0357] Server: A high-performance computer system that processes and analyzes data and trains machine learning models.

[0358] Database: A management system for storing interview case data and financial report data. Example: MySQL (registered trademark).

[0359] Natural Language Processing (NLP) Library: A software library for using machine learning models. For example, "Transformers".

[0360] Emotion engine: A system that analyzes the user's voice and facial expressions. Example: "Microsoft (registered trademark) Azure (registered trademark) Emotion API."

[0361] Terminal: A device that provides an interface for users to input information. Examples include PCs and tablets.

[0362] Data Processing and Data Calculation

[0363] The server retrieves past interview case data from a database, for example using SQL queries. This data includes interview questions and their answers.

[0364] The server then uses this data to train a machine learning model, using natural language processing techniques to learn patterns of past questions and answers. Specifically, the Transformers library is used. Through this training, a model is built that can generate new questions.

[0365] When a user enters candidate information through a terminal, the information is sent to the server as an HTTP request. The server analyzes the information and generates initial questions. For example, if a user enters information about a candidate's recent projects, the server generates an initial question such as "Tell me about your recent projects."

[0366] The terminal displays the initial questions generated by the server to the user. When the user enters the candidate's answers, the answers are also sent to the server. The server analyzes the answers, generates the next appropriate questions, and sends them to the terminal.

[0367] The emotion engine analyzes the user's voice and facial expression data to recognize their emotions. This recognized emotion is sent to the server, which then generates questions based on the emotion data. For example, if a user is feeling highly stressed, the engine can generate questions such as, "What was the most rewarding moment in that project?"

[0368] Specific examples

[0369] For example, an initial interview question might be generated as "Tell me about a recent project." The interviewer asks this question to a candidate, who responds, "I worked on project X, which resulted in outcome Y." When this response is typed into a device and sent to a server, the server analyzes it and generates the next question: "What was your specific role in that project?"

[0370] As an example prompt, parse the candidate's response to the question "Tell me about a recent project": "I worked on project X, which resulted in outcome Y." Generate the following question:

[0371] The above is a specific embodiment for carrying out the invention. This invention improves the quality of interviews and reports, and enables smooth progress of dialogue.

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

[0373] Step 1:

[0374] The server loads past interview case data from a database.

[0375] Specifically, it uses an SQL query to execute a command like "SELECT FROM interview_data" to retrieve a dataset containing interview questions and their answers.

[0376] Input: Database query

[0377] Output: Dataset of past interview cases

[0378] Step 2:

[0379] The server uses the loaded interview case data to train a machine learning model.

[0380] Specifically, we use the natural language processing library "Transformers" to learn question and answer patterns, and build a question generation algorithm using models such as BERT and GPT.

[0381] Input: Past interview case dataset

[0382] Output: A trained machine learning model

[0383] Step 3:

[0384] The user inputs candidate information through the terminal.

[0385] Specifically, the process involves using a web form or application to enter information such as the candidate's name, background, and reasons for applying.

[0386] Input: Candidate information (name, career history, reason for applying, etc.)

[0387] Output: Candidate information data

[0388] Step 4:

[0389] The terminal transmits the input candidate information to the server as an HTTP request.

[0390] Specifically, a POST request is used to send data to the server. For example, an HTTP client library is used.

[0391] Input: Candidate information data

[0392] Output: Candidate information sent to the server

[0393] Step 5:

[0394] The server analyzes the received candidate information and generates initial questions.

[0395] Specifically, it analyzes candidate information using the TF-IDF algorithm and other natural language processing techniques to select the most appropriate initial questions.

[0396] Input: Candidate information data

[0397] Output: Initial question (e.g. "Tell me about your recent projects")

[0398] Step 6:

[0399] The terminal displays the initial question received from the server to the user.

[0400] Specifically, the question is displayed on the user interface using a web browser or application, and a text field for displaying the question is displayed.

[0401] Input: Initial question

[0402] Output: The initial question that is displayed to the user

[0403] Step 7:

[0404] The user inputs the answers from the candidates.

[0405] Specifically, the answer obtained is input via the terminal, for example, by using a keyboard to enter the answer in a text field.

[0406] Input: Candidate's answer

[0407] Output: Candidate's answers entered

[0408] Step 8:

[0409] The terminal transmits the input answers of the candidates to the server.

[0410] Specifically, the input data is sent to the server again via an HTTP request.

[0411] Input: Candidate's answers entered

[0412] Output: Candidate's answers sent to the server

[0413] Step 9:

[0414] The server analyzes the received answer and generates the next question.

[0415] Specifically, it uses a trained machine learning model to generate the next appropriate question, such as "What was your specific role in that project?"

[0416] Input: Candidate's answer

[0417] Output: Next question

[0418] Step 10:

[0419] The terminal displays the next question received from the server to the user.

[0420] Specifically, the new question is displayed in the user interface, such as a web browser or application.

[0421] Input: Next question

[0422] Output: The next question that is displayed to the user

[0423] Step 11:

[0424] The emotion engine analyzes the user's voice and facial expression data to recognize emotions.

[0425] Specifically, it collects audio and video data and uses analysis engines such as the Microsoft Azure Emotion API to recognize emotions.

[0426] Input: Voice and facial expression data

[0427] Output: Recognized emotion data

[0428] Step 12:

[0429] The server generates a question based on the recognized emotion data.

[0430] Specifically, the server generates relaxing questions based on emotional data. For example, in a stressful situation, it generates a question such as, "What was the most rewarding moment in that project?"

[0431] Input: Recognized emotion data

[0432] Output: Sentiment-based questions

[0433] (Application example 2)

[0434] 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."

[0435] The goal is to provide a comfortable riding experience by realizing interactive communication with passengers in autonomous vehicles and generating appropriate questions according to the passenger's emotional state. Conventional systems can only have fixed dialogues with passengers, making it difficult to respond flexibly taking into account their emotional state.

[0436] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0437] In this invention, the server includes means for loading past dialogue example data from a database, means for training a machine learning model based on the past dialogue example data, means for providing an interface for a user to input passenger information, means for transmitting the input passenger information to the server, means for analyzing the passenger information and generating initial questions, means for displaying the generated initial questions to the user, means for the user to input passenger answers and transmitting the same to the server, means for analyzing the passenger answers and generating subsequent questions, means for analyzing voice data and facial expression data to recognize emotions, and means for generating questions based on the recognized emotions. This enables interactive communication with passengers in an autonomous vehicle and realizes flexible question generation based on the passenger's emotional state.

[0438] A "database" is a system that stores various information, such as past dialogue case data and financial report data, and allows it to be searched and read.

[0439] A "machine learning model" is an algorithm that learns patterns from data and makes appropriate decisions and predictions about new data.

[0440] An "interface" is a screen or input form through which a user enters information or interacts with a system.

[0441] A "server" is a computer system that processes and stores various data, and is a device that sends and receives data in response to requests from clients.

[0442] "Analysis" refers to the process of examining input data in detail to understand its characteristics and structure.

[0443] An "initial question" is a question that is initially generated based on information entered by the user.

[0444] "Voice data" refers to audio information recorded by a microphone or other device, which can be used for voice recognition and emotion recognition.

[0445] "Facial expression data" refers to facial expression information captured by a camera or other device, and emotions can be recognized based on this information.

[0446] An "emotion engine" is software or algorithm that analyzes a user's emotional state from their voice and facial expressions.

[0447] To implement this invention, it is first necessary to properly link the main components, such as the database, machine learning model, interface, server, and emotion engine. The specific steps are described below.

[0448] server

[0449] The server loads past dialogue case data from a database, which contains a variety of dialogue case data, including interview scenes. The server uses this data to train a machine learning model. This model incorporates natural language processing technology and includes an algorithm for generating new questions. For example, it is recommended to use a Python (registered trademark)-based natural language processing library (e.g., NLTK, spaCy) or a machine learning framework (e.g., TensorFlow, PyTorch).

[0450] The server receives passenger information entered by the user and analyzes it to generate an initial question. Information from the user is usually received via an HTTP request, and the server analyzes it using an appropriate algorithm. The server then receives the passenger's answers entered by the user and analyzes them to generate the next question. During this process, an emotion engine analyzes voice and facial expression data to recognize emotions, and this data is also used to generate questions.

[0451] Terminal

[0452] The terminal provides an interface for users to enter passenger information. This interface includes text boxes and pull-down menus, allowing users to easily enter the necessary information. For example, passengers can enter their name, past ride experiences, and favorite music genres. Once the information is complete, the terminal sends the data to the server.

[0453] Next, when an initial question is generated from the server, it is displayed on the terminal. The user sees this question, asks it to the passenger, and the passenger enters their answer. The entered answer is also sent to the server, and the next question is generated in the same process. In this way, interactive communication is realized.

[0454] Emotion Engine

[0455] The emotion engine analyzes passengers' voice and facial expression data to recognize their emotions. This is done using a speech recognition library (e.g., Google® Speech-to-Text API) and a facial expression recognition library (e.g., OpenCV, Dlib). The recognized emotion information is sent to the server and taken into consideration when generating the next question.

[0456] Specific examples

[0457] For example, when a passenger boards a vehicle, the system asks, "Tell me about your recent riding experience." If the passenger answers, "It was a very comfortable and pleasant ride," the emotion engine detects the passenger's positive emotion, and the server generates subsequent questions such as, "What did you particularly like?" or "Would you like to enjoy music today?"

[0458] Prompt Sentence Examples

[0459] "If a passenger responds positively about their past ride experience, what next questions can you ask based on that positive sentiment? For example, if a passenger responds, 'It was a very comfortable and pleasant ride,' how can you best continue?"

[0460] In this way, the system can enable interactive communication with passengers inside autonomous vehicles, providing them with a comfortable and personalized riding experience.

[0461] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0462] Step 1:

[0463] The terminal provides an interface for users to input passenger information (e.g., name, past ride experiences, favorite music genres), which is then sent to the server via an HTTP request. The input here is passenger information, and the output is sent to the server.

[0464] Step 2:

[0465] The server loads past dialogue case data from a database. The loaded data is a dataset containing past interviews and other dialogue scenes, and trains a machine learning model based on this data. The input is the dialogue case data from the database, and the output is the trained machine learning model. This training uses natural language processing techniques and machine learning frameworks (e.g., TensorFlow, PyTorch).

[0466] Step 3:

[0467] The server analyzes the passenger information submitted by the user and generates an initial question. The input is the passenger information from the user, and the output is the initial question. This process uses a text analysis algorithm. For example, if the user enters positive information about their "recent ride experience," the server generates an initial question based on that information: "What did you particularly like?"

[0468] Step 4:

[0469] The terminal displays the initial question received from the server to the user. Based on the displayed initial question, the user asks the question to the passenger and receives an answer. The input is the initial question from the server, and the output is the display to the user.

[0470] Step 5:

[0471] The user inputs the passenger's answers into the terminal and sends the data to the server via an HTTP request. The input is the passenger's answers and the output is the transmission to the server.

[0472] Step 6:

[0473] The server analyzes the passenger's response and generates the next question. The input is the passenger's response from the user, and the output is the next question. The analysis uses a trained machine learning model and an emotion engine. Specifically, the next question is generated by combining emotional information extracted from voice data and facial expression data.

[0474] Step 7:

[0475] The emotion engine analyzes passengers' voice and facial expression data to recognize their emotions. The input is voice data and facial expression data, and the output is the recognized emotion information. This is done using libraries such as the Google Speech-to-Text API, OpenCV, and Dlib.

[0476] Step 8:

[0477] The server generates the next question based on the recognized emotion information. The input is the emotion information from the emotion engine and the passenger's answer, and the output is a flexibly generated next question. For example, if a passenger answers, "It was a very comfortable and pleasant ride," and a positive emotion is recognized, the next question generated will be, "What did you particularly like?"

[0478] By repeating the above steps, interactive communication with passengers inside the self-driving vehicle will proceed smoothly.

[0479] 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.

[0480] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0481] 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.

[0482] [Second embodiment]

[0483] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0484] 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.

[0485] 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).

[0486] 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.

[0487] 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.

[0488] 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).

[0489] 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.

[0490] 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.

[0491] 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.

[0492] 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.

[0493] 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.

[0494] 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."

[0495] An embodiment of the present invention will be described below. An interview scene will be described as an example, but it should be noted that the present invention can also be applied to a financial report scene in a similar manner.

[0496] server

[0497] 1. A way to load interview case data from a database

[0498] The server loads past interview case data from a database, which includes questions asked in interviews and their answers. By retrieving data from the database, the server builds a training dataset.

[0499] 2. Means of training machine learning models

[0500] Based on the loaded interview case data, the server trains a machine learning model. Specifically, it uses natural language processing technology to learn patterns of past questions and answers and create a model that generates new questions. This model includes a question generation algorithm and can generate appropriate questions based on the candidate's information.

[0501] 3. A means of analyzing user-entered data and generating initial questions

[0502] When the user (interviewer) enters candidate information and sends it to the server, the server analyzes the data and generates initial questions, where an algorithm is applied to select the most appropriate initial questions based on the information entered by the user.

[0503] Terminal

[0504] 1. A means of providing an interface for users to input candidate information

[0505] The terminal provides an interface for the user to input candidate information, specifically, a form for inputting name, career history, reason for applying, etc.

[0506] 2. Means of sending the entered data to the server

[0507] When the user (interviewer) has completed the input, the data is sent to the server using a normal HTTP request.

[0508] 3. A means to display the generated initial questions to the user

[0509] When the terminal receives the generated initial questions from the server, it displays them to the user (interviewer). Once the questions are displayed, the user can start asking questions to the candidate.

[0510] 4. A means for users to input candidate answers and send them to the server

[0511] After the user (interviewer) receives the answer from the candidate, he / she enters the answer into the terminal and sends it to the server. This procedure is performed using an HTTP request, just like the initial question.

[0512] User

[0513] 1. Enter candidate information

[0514] The user (interviewer) enters candidate information, such as the candidate's name, highest level of education, work history, and reason for applying, according to the provided interface.

[0515] 2. Asking initial questions and getting answers

[0516] The user (interviewer) views the initial questions generated by the server, asks them to the candidate, and receives answers.

[0517] 3. Enter and submit your answers

[0518] The answer obtained is entered into the terminal and sent to the server to determine what question to ask next.

[0519] Specific examples

[0520] For example, consider an initial interview question that might be generated: "Tell me about a recent project." The interviewer asks this question to a candidate, who responds, "I worked on project X, which resulted in Y results." This response is entered into a device and sent to a server, which analyzes it and generates the next question: "What was your specific role in that project?" This process is repeated to effectively progress the interview.

[0521] In this way, users can easily generate appropriate questions based on the candidate's information, and the dialogue can proceed smoothly. This system can improve the quality of interviews.

[0522] The processing flow will be explained below.

[0523] Step 1:

[0524] The server loads the past interview case data from the database, which involves establishing a database connection and executing SQL queries to retrieve the required data.

[0525] Step 2:

[0526] The server trains a machine learning model based on the loaded interview case data. Specifically, it performs data preprocessing and trains the model using natural language processing techniques. The model learns patterns from question-answer pairs.

[0527] Step 3:

[0528] The terminal displays an interface for entering candidate information. The user (interviewer) uses this interface to enter the candidate's name, career history, reason for applying, etc.

[0529] Step 4:

[0530] Once the user has completed the input, the device sends the data as an HTTP request to the server, which receives the request and analyzes the candidate information.

[0531] Step 5:

[0532] The server generates initial questions based on the analysis results, uses a trained machine learning model to select questions that best fit the candidate information, and sends the generated questions to the device.

[0533] Step 6:

[0534] The terminal displays the initial questions received from the server to the user, who then verbally asks the candidate the questions.

[0535] Step 7:

[0536] The user receives answers from the candidates, enters them into the terminal, and sends them back to the server, again using an HTTP request.

[0537] Step 8:

[0538] The server analyzes the received answer and uses an analysis algorithm to extract important information from the answer and generate the next question.

[0539] Step 9:

[0540] The server sends the next generated question to the terminal, which then displays the received question to the user (interviewer).

[0541] Step 10:

[0542] The user asks the candidate the following questions, and this process is repeated until the interview is complete, improving the quality of the interview and facilitating a smooth dialogue.

[0543] This series of processing steps enables the present invention to provide users with dynamic and interactive question generation, improving the efficiency and quality of interviews and financial reports.

[0544] Example 1

[0545] 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."

[0546] In conventional interview systems and financial reporting systems, it is difficult for interviewers or reporters to quickly generate appropriate questions, which can lead to inefficient dialogue. Furthermore, if questions are not generated appropriately, the quality of the dialogue declines and necessary information cannot be fully elicited. The present invention aims to solve these problems by providing a system that can automatically generate appropriate and effective questions.

[0547] 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.

[0548] In this invention, the server includes means for loading past interview case data from a database, means for training a machine learning model based on the past interview case data, means for providing an interface for a user to input candidate information, means for transmitting the input candidate information to the server, means for analyzing the candidate information and generating initial questions, means for displaying the generated initial questions to the user, means for a user to input answers from the candidate and transmit them to the server, and means for analyzing the answers from the candidate and generating the next question. This enables interviewers and reporters to effectively conduct dialogue and appropriately extract necessary information.

[0549] A "database" is a system for efficiently storing, searching, and managing large amounts of data.

[0550] "Interview case data" refers to records of questions and answers from past interviews.

[0551] A "machine learning model" is a collection of algorithms that learn patterns and rules based on data and use the results of that learning to make predictions and classify new data.

[0552] "User" refers to the person who operates the system, inputs information, and views results.

[0553] An "interface" is a means, such as a screen or input form, through which a user interacts with a system.

[0554] A "server" is a computer system that processes and stores data over a network.

[0555] "Analysis" is the process of examining input data in detail and extracting meaning and patterns.

[0556] An "initial question" refers to the first question posed to a candidate or reporter when an interview or report begins.

[0557] "Answer" refers to the information or response returned in response to a question.

[0558] The "next question" is a new question that is generated based on the previous answer.

[0559] A "system" is a mechanism in which multiple components and processes work together to provide specific functions and services.

[0560] This invention relates to an interview system and a financial reporting system, and aims to effectively progress the dialogue by having a user input information about candidates and reporters, and then having a server analyze the information and automatically generate appropriate questions.

[0561] server

[0562] The server first uses a database management system to load past interview case data and financial report data from a database. This data includes past questions and their answers. The server then uses natural language processing technology to train a machine learning model based on the loaded data. Specifically, it trains a BERT-based model using frameworks such as TensorFlow and PyTorch to build a question generation algorithm.

[0563] Terminal

[0564] The terminal provides an interface for users to input candidate information and report details, such as name, career history, reasons for applying, and financial report details, through a web form written in React. Once the user has completed the input, the data is sent to the server as an HTTP POST request.

[0565] The terminal then receives the initial questions generated by the server and displays them to the user. Similarly, when the user enters the candidate's answers and additional information, that data is also sent to the server.

[0566] User

[0567] The user uses the terminal to input candidate information and financial report details. The specific operating procedure is as follows:

[0568] 1. The user follows the provided interface to enter the candidate's name, highest level of education, work history, reason for applying, etc.

[0569] 2. Check the initial questions generated by the server and ask them to the candidate to get their answers.

[0570] 3. The answer obtained is entered into the terminal and sent to the server to generate the next question.

[0571] Specific examples

[0572] For example, an initial interview question might be generated as "Tell me about a recent project." If a user poses this question to a candidate and the candidate responds with "I worked on project X, which resulted in Y results," the response is typed into the device and sent to the server. The server then analyzes the response and generates the next question, "What was your specific role in that project?", which is returned to the device and displayed to the user. This process is repeated to effectively progress the interview.

[0573] Examples of specific prompts include the following:

[0574] Please tell us about the candidate's highest level of education.

[0575] What is your specific role in this project?

[0576] This invention allows users to quickly and efficiently generate appropriate questions for candidates and reporters, allowing for smooth dialogue, thereby improving the quality of interviews and the effectiveness of reports.

[0577] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0578] Specific explanation of program processing

[0579] server

[0580] Step 1:

[0581] The server loads past interview case data from a database.

[0582] Input: Database connection information

[0583] Output: Interview case data list

[0584] What it does: The server executes an SQL query to retrieve interview case data from a database, which includes past interview questions and answers.

[0585] Step 2:

[0586] The server trains a machine learning model based on the loaded interview case data.

[0587] Input: Interview case data list

[0588] Output: A trained machine learning model

[0589] How it works: The server uses TensorFlow and PyTorch to train a BERT-based natural language processing model, which involves learning question and answer patterns.

[0590] Step 3:

[0591] The server analyzes the candidate information sent by the user from the terminal and generates an initial question.

[0592] Input: Candidate information

[0593] Output: Initial question

[0594] Specific operation: The server uses a machine learning model to analyze candidate information and generate appropriate initial questions.

[0595] Terminal

[0596] Step 1:

[0597] The terminal provides an interface for the user to input candidate information.

[0598] Input: None (standby)

[0599] Output: Form for entering candidate information

[0600] Specific operation: The device uses "React" to display a web form for entering information such as name, background, and reasons for applying.

[0601] Step 2:

[0602] The terminal transmits the input candidate information to the server.

[0603] Input: Candidate information

[0604] Output: HTTP POST request

[0605] Specific operation: Generates and sends an HTTP POST request to send the candidate information entered by the user in the form to the server.

[0606] Step 3:

[0607] The terminal displays the generated initial question to the user.

[0608] Input: Initial question (response from server)

[0609] Output: Question displayed on the user interface

[0610] Specific operation: The initial question received from the server is displayed clearly to the user.

[0611] Step 4:

[0612] The user inputs the candidate's answers at the terminal and transmits them to the server.

[0613] Input: Candidate's answer

[0614] Output: HTTP POST request

[0615] Specific operation: The answers entered by the user are recorded on the terminal and sent to the server.

[0616] User

[0617] Step 1:

[0618] The user inputs the candidate information.

[0619] Input: Candidate information (name, highest level of education, work history, reason for applying, etc.)

[0620] Output: Input data (terminal)

[0621] Specific operation: The user enters each item according to the form provided on the terminal.

[0622] Step 2:

[0623] The user poses initial questions to the candidate and receives answers.

[0624] Input: Initial question (displayed on the terminal)

[0625] Output: Candidate's answers

[0626] Specific operation: The user asks the candidate initial questions displayed on the terminal and receives answers.

[0627] Step 3:

[0628] The user inputs the answers obtained and sends them to the server to generate the next question.

[0629] Input: Candidate's answer

[0630] Output: HTTP POST request

[0631] Specific operation: The user inputs the candidate's answer into the terminal and sends a request to the server to generate the next question.

[0632] (Application example 1)

[0633] 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."

[0634] Customer support and dialogue systems in modern virtual environments are based on static FAQs and scripts, making it difficult to provide appropriate responses to specific user questions and preventing efficient and effective interactive dialogue that meets user needs. A particular problem is the lack of systems that can quickly and accurately answer questions visitors have about specific products or services in virtual stores and online support.

[0635] 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.

[0636] In this invention, the server includes: means for loading past case data from a database; means for training a machine learning model based on the past case data; means for providing an interface for a user to input information; means for transmitting the input information to the server; means for analyzing the information and generating an initial question; means for displaying the generated initial question to the user; means for the user to input an answer and transmitting the answer to the server; means for analyzing the answer and generating a next question; means for answering questions about products and services from visitors through dialogue in a virtual environment; means for functioning as an application installed on a smart device; and means for dynamically generating questions and interactively displaying them through a user interface. This enables users' questions to be answered quickly and accurately, significantly improving the quality of customer support in virtual stores.

[0637] A "database" is a repository that systematically organizes information so that it can be searched and retrieved efficiently.

[0638] "Case data" is data that compiles information about specific past events or situations.

[0639] A "machine learning model" is a set of algorithms that are trained with data to perform specific tasks automatically.

[0640] An "interface" is a means or device by which a user and a system interact with each other.

[0641] A "server" is a computer that provides services and resources to other computers over a network.

[0642] "Analysis" is the process of examining data and information in detail and clarifying their structure and relationships.

[0643] An "initial question" is a question that the dialogue system generates to start the initial dialogue with the user.

[0644] "User" means a person or entity that uses a system or service.

[0645] An "answer" is a response given to a question.

[0646] The "next question" is the next question that the system generates based on the user's answer.

[0647] A "virtual environment" is a virtual space or system created by computer simulation.

[0648] "Goods" means goods or services manufactured for sale.

[0649] A "service" is an activity or undertaking that provides benefit or satisfaction to consumers.

[0650] A "smart device" is an electronic device with advanced computing power and connectivity capabilities.

[0651] An "application" is software designed for a specific purpose.

[0652] A "user interface" is the means or method by which a user interacts with a system or device.

[0653] The present invention relates to a customer support assistant system for effectively conducting dialogue with visitors in a virtual store. The system aims to respond quickly and appropriately to visitors' questions and increase their satisfaction. The following describes in detail the embodiments of the present invention.

[0654] server

[0655] The server implements the following methods:

[0656] A means of loading case data from a database

[0657] The server loads past customer support case data from the database, including past questions and their answers, which the server uses as a training data set.

[0658] A means to train machine learning models

[0659] Using the example data loaded onto the server, the server trains a machine learning model. Specifically, it uses natural language processing techniques to learn patterns of past questions and answers and create a model that generates new questions. An example of the software used is the Hugging Face transformers library.

[0660] A means of analyzing user information and generating initial questions

[0661] When the user (customer support representative) enters visitor information and sends it to the server, the server analyzes the data and applies an algorithm to generate the most appropriate initial questions, again using natural language processing technology.

[0662] Terminal

[0663] The terminal provides the means to:

[0664] A means of providing an interface for users to input information

[0665] The terminal provides an interface for the user to input visitor information, such as name, inquiry details, and past purchase history.

[0666] A means of sending the entered data to the server

[0667] Once the user has completed the input, the data is sent to the server using an HTTP request.

[0668] A means of displaying the generated initial questions to the user

[0669] Upon receiving the generated initial question from the server, the terminal displays the question to the user (customer support representative), allowing the user to begin asking questions to the visitor.

[0670] A means for the user to enter answers and send them to the server

[0671] After the user retrieves the visitor's answer, they enter it into their device and send it to the server, again using an HTTP request to send the data.

[0672] User

[0673] The role of the user (customer support representative) is to:

[0674] Entering visitor information

[0675] The user inputs visitor information using the provided interface, such as the visitor's name, inquiry details, and past purchase history.

[0676] Asking initial questions and getting answers

[0677] The user sees the initial question generated by the server, poses it to the visitor, and receives the answer.

[0678] Enter and submit your answers

[0679] The answer obtained is entered into the terminal and sent to the server to determine what question to ask next.

[0680] Examples and prompts

[0681] For example, consider a case where a visitor to a virtual store wants information about a new product. Based on the visitor's purchase history and inquiries entered through the terminal, the server generates an initial question: "What are the features of this product?" After receiving the visitor's answer, the server generates the next question: "What are the unique features of this product?" This allows for timely and accurate responses to the visitor's questions.

[0682] Example prompt sentence:

[0683] "Please tell us the features of the products you recently released."

[0684] "How have you improved after receiving customer feedback?"

[0685] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0686] Step 1:

[0687] Loading case data from a database

[0688] The server loads past customer support case data from a database, including past questions and their answers, and uses this data as a training dataset to train the machine learning model.

[0689] Input: Query request to the database

[0690] Output: Dataset of past cases

[0691] Step 2:

[0692] Train a machine learning model

[0693] The server uses the loaded example data to train a machine learning model, leveraging natural language processing techniques. Specifically, it uses Hugging Face's transformers library to learn patterns of past questions and answers and create a model that can generate new questions.

[0694] Input: Historical case dataset

[0695] Output: A trained machine learning model

[0696] Step 3:

[0697] Provides an interface for entering user information

[0698] The terminal provides an interface for the user (customer support representative) to input visitor information, which is displayed as a form for entering information such as the user's name, inquiry details, and past purchase history.

[0699] Input: None (launches user interface)

[0700] Output: Interface screen display

[0701] Step 4:

[0702] Send the entered information to the server

[0703] Once the user has entered the visitor information into the interface, the device sends the information to the server. This data transmission is done using an HTTP request.

[0704] Input: Visitor information (user input)

[0705] Output: Send data to the server (HTTP request)

[0706] Step 5:

[0707] Generate initial questions

[0708] The server analyzes the received visitor information and applies machine learning models to generate the most appropriate initial questions. Natural language processing technology selects appropriate questions based on actual visitor information.

[0709] Input: Visitor information (received by server)

[0710] Output: Initial question

[0711] Step 6:

[0712] Display the generated initial questions to the user

[0713] The terminal displays the initial question retrieved from the server to the customer support representative, and the user starts a dialogue with the visitor based on this question.

[0714] Input: Initial question (sent by server)

[0715] Output: Question displayed on the screen

[0716] Step 7:

[0717] Enter the visitor's answer and send it to the server

[0718] The user enters the answers received from the visitors into the terminal and sends the answer data to the server. This data transmission is also done using an HTTP request.

[0719] Input: Visitor's answer (user input)

[0720] Output: Send data to the server (HTTP request)

[0721] Step 8:

[0722] Generate the next question

[0723] The server analyzes the visitor's answers and generates the next question, again dynamically using a machine learning model.

[0724] Input: Visitor's response (received by the server)

[0725] Output: Next question

[0726] Step 9:

[0727] Show the next question to the user

[0728] The terminal displays the next question retrieved from the server to the customer support representative, allowing the user to pose the next question to the visitor.

[0729] Input: Next question (sent by server)

[0730] Output: Question displayed on the screen

[0731] Prompt Sentence Examples

[0732] 1. "Please tell us about the features of the products you recently released."

[0733] 2. "How have you improved after receiving customer feedback?"

[0734] 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.

[0735] The following describes an embodiment of the present invention. The following describes an interview scene as an example, but please note that the present invention can also be applied to a financial report scene using similar methods. The present invention also incorporates an emotion engine.

[0736] server

[0737] 1. A way to load past interview case data from a database

[0738] The server loads past interview case data from a database, which includes questions asked in interviews and their answers. By retrieving data from the database, the server builds a training dataset.

[0739] 2. Means of training machine learning models

[0740] Based on the loaded interview case data, the server trains a machine learning model. Specifically, it uses natural language processing technology to learn patterns of past questions and answers and create a model that generates new questions. This model includes a question generation algorithm and can generate appropriate questions based on the candidate's information.

[0741] 3. A means of analyzing user-entered data and generating initial questions

[0742] When the user (interviewer) enters candidate information and sends it to the server, the server analyzes the data and generates initial questions, where an algorithm is applied to select the most appropriate initial questions based on the information entered by the user.

[0743] Terminal

[0744] 1. A means of providing an interface for users to input candidate information

[0745] The terminal provides an interface for the user to input candidate information, specifically, a form for inputting name, career history, reason for applying, etc.

[0746] 2. Means of sending the entered data to the server

[0747] When the user (interviewer) has completed the input, the data is sent to the server using a normal HTTP request.

[0748] 3. A means to display the generated initial questions to the user

[0749] When the terminal receives the generated initial questions from the server, it displays them to the user (interviewer). Once the questions are displayed, the user can start asking questions to the candidate.

[0750] 4. A means for users to input candidate answers and send them to the server

[0751] After the user (interviewer) receives the answer from the candidate, he / she enters the answer into the terminal and sends it to the server. This procedure is performed using an HTTP request, just like the initial question.

[0752] Emotion Engine

[0753] 1. A means of recognizing emotions by analyzing the user's voice and facial expressions

[0754] The emotion engine analyzes voice and facial expression data provided by the device, allowing it to recognize the user's emotions (e.g., joy, anger, sadness, surprise, etc.).

[0755] 2. A way to generate questions based on perceived sentiment

[0756] The recognized emotion data is sent to the server, which then generates questions based on the emotion information. For example, if the user indicates a high stress level, the system generates questions to encourage relaxation.

[0757] User

[0758] 1. Enter candidate information

[0759] The user (interviewer) enters candidate information, such as the candidate's name, highest level of education, work history, and reason for applying, according to the provided interface.

[0760] 2. Asking initial questions and getting answers

[0761] The user (interviewer) views the initial questions generated by the server, asks them to the candidate, and receives answers.

[0762] 3. Enter and submit your answers

[0763] After receiving the candidate's answer, the answer is entered into the terminal and sent to the server to get the next question to ask.

[0764] 4. Emotion Recognition by Emotion Engine

[0765] During the interview, the device captures the user's voice and facial expressions, which are then analyzed by an emotion engine. The analysis results are sent to the server and used to generate the next questions.

[0766] Specific examples

[0767] For example, consider an initial interview question generated as "Tell me about a recent project." The interviewer asks this question to a candidate, who responds, "I worked on project X, which resulted in Y results." When this response is entered into a device and sent to a server, the server analyzes it and generates the next question: "What was your specific role in that project?"

[0768] During this process, if the emotion engine detects the interviewer's stress level, for example if it is determined to be high, the server can take this into account and generate relaxing questions such as, "What was the most rewarding moment in that project?"

[0769] In this way, users can easily generate appropriate questions based on the candidate's information, and the dialogue can proceed smoothly. This system improves the quality of interviews and also enables flexible responses according to the user's emotional state.

[0770] The processing flow will be explained below.

[0771] Step 1:

[0772] The server loads past interview case data from the database by establishing a database connection and executing SQL queries to retrieve the required data.

[0773] Step 2:

[0774] The server trains a machine learning model based on the loaded interview case data. Specifically, it performs data preprocessing and trains the model using natural language processing techniques.

[0775] Step 3:

[0776] The terminal provides the user (interviewer) with an interface for inputting candidate information. The user inputs the candidate's name, career history, reason for applying, etc.

[0777] Step 4:

[0778] Once the user has completed the input, the device sends the data as an HTTP request to the server, which receives the request and analyzes the candidate information.

[0779] Step 5:

[0780] The server generates initial questions based on the analysis results, uses a trained machine learning model to select questions that best fit the candidate information, and sends the generated questions to the device.

[0781] Step 6:

[0782] The terminal displays the initial questions received from the server to the user, who then verbally asks the candidate the questions.

[0783] Step 7:

[0784] The emotion engine analyzes the user's voice and facial expressions in real time to evaluate the user's emotional state, and the evaluation results are sent to the server.

[0785] Step 8:

[0786] The user receives answers from the candidates, enters these answers into the terminal, and sends them back to the server.

[0787] Step 9:

[0788] The server analyzes the received answer and also uses the emotion data from the emotion engine for analysis. The analysis algorithm takes into account the important information in the answer and the user's emotion to generate the next question.

[0789] Step 10:

[0790] The server sends the next generated question to the terminal, which then displays the received question to the user (interviewer).

[0791] Step 11:

[0792] The user asks the candidate the following questions, and this process is repeated until the interview is complete, improving the quality of the interview and facilitating a smooth dialogue.

[0793] Through this series of processing steps, the present invention provides users with dynamic and interactive question generation, improving the efficiency and quality of interviews. In addition, by combining it with an emotion engine, flexible responses according to the user's emotional state become possible.

[0794] Example 2

[0795] 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."

[0796] In traditional situations such as interviews and financial reports, efficient and effective question generation is required, but in the process, it has been difficult to generate questions that take human emotions into account and maintain a continuity of appropriate questions. In particular, the lack of a dynamic question generation method that takes into account the emotional state of the interviewer or reporter can lead to a decline in the quality of interviews and reports.

[0797] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0798] In this invention, the server includes means for loading past interview case data from a database, means for training a machine learning model based on the past interview case data, means for providing an interface for a user to input candidate information, means for transmitting the input candidate information to the server, means for analyzing the candidate information and generating initial questions, means for displaying the generated initial questions to the user, means for a user to input candidate answers and transmitting the answers to the server, means for analyzing the candidate answers and generating subsequent questions, means for analyzing the user's voice and facial expressions to recognize emotions, and means for generating questions based on the recognized emotions. This makes it possible to generate appropriate and consecutive questions while taking into account the user's emotional state during interviews and reporting situations.

[0799] A "database" is a collection of information that efficiently manages past interview case data, financial report data, etc., and can be searched and retrieved as needed.

[0800] A "machine learning model" is an algorithm that learns from past data, discovers new patterns and relationships, and automates specific tasks such as question generation.

[0801] An "interface" is a user interface component that allows the user to interact with the system and input candidate information, financial report details, and the like.

[0802] "Server" means a computer system that receives information sent by users and processes and analyzes the data.

[0803] "Analysis" is the process of examining input data in detail to understand its structure and meaning.

[0804] An "initial question" is a question that is presented to a user when they first begin interacting with the system.

[0805] The "emotion engine" is a system that analyzes the user's voice and facial expression data to recognize their emotional state at that time.

[0806] "Dynamic question generation" is the process of generating new questions in real time based on the current user situation and answers.

[0807] "Interview case data" is data on questions used in past interviews and the answers to those questions.

[0808] "Recognized emotion" refers to the emotional state of the user that has been analyzed and identified by the emotion engine.

[0809] The present invention provides a method for generating questions efficiently and effectively in interactive systems for interviews, financial reporting, etc. While the present invention will be described in detail below using an interview scene as an example, it can also be applied to financial reporting scenes using similar means.

[0810] Hardware and software used

[0811] Server: A high-performance computer system that processes and analyzes data and trains machine learning models.

[0812] Database: A management system for storing interview case data and financial report data. Example: MySQL.

[0813] Natural Language Processing (NLP) Library: A software library for using machine learning models. For example, "Transformers".

[0814] Emotion engine: A system that analyzes the user's voice and facial expressions. Example: "Microsoft Azure Emotion API."

[0815] Terminal: A device that provides an interface for users to input information. Examples include PCs and tablets.

[0816] Data Processing and Data Calculation

[0817] The server retrieves past interview case data from a database, for example using SQL queries. This data includes interview questions and their answers.

[0818] The server then uses this data to train a machine learning model, using natural language processing techniques to learn patterns of past questions and answers. Specifically, the Transformers library is used. Through this training, a model is built that can generate new questions.

[0819] When a user enters candidate information through a terminal, the information is sent to the server as an HTTP request. The server analyzes the information and generates initial questions. For example, if a user enters information about a candidate's recent projects, the server generates an initial question such as "Tell me about your recent projects."

[0820] The terminal displays the initial questions generated by the server to the user. When the user enters the candidate's answers, the answers are also sent to the server. The server analyzes the answers, generates the next appropriate questions, and sends them to the terminal.

[0821] The emotion engine analyzes the user's voice and facial expression data to recognize their emotions. This recognized emotion is sent to the server, which then generates questions based on the emotion data. For example, if a user is feeling highly stressed, the engine can generate questions such as, "What was the most rewarding moment in that project?"

[0822] Specific examples

[0823] For example, an initial interview question might be generated as "Tell me about a recent project." The interviewer asks this question to a candidate, who responds, "I worked on project X, which resulted in outcome Y." When this response is typed into a device and sent to a server, the server analyzes it and generates the next question: "What was your specific role in that project?"

[0824] As an example prompt, parse the candidate's response to the question "Tell me about a recent project": "I worked on project X, which resulted in outcome Y." Generate the following question:

[0825] The above is a specific embodiment for carrying out the invention. This invention improves the quality of interviews and reports, and enables smooth progress of dialogue.

[0826] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0827] Step 1:

[0828] The server loads past interview case data from a database.

[0829] Specifically, it uses an SQL query to execute a command like "SELECT FROM interview_data" to retrieve a dataset containing interview questions and their answers.

[0830] Input: Database query

[0831] Output: Dataset of past interview cases

[0832] Step 2:

[0833] The server uses the loaded interview case data to train a machine learning model.

[0834] Specifically, we use the natural language processing library "Transformers" to learn question and answer patterns, and build a question generation algorithm using models such as BERT and GPT.

[0835] Input: Past interview case dataset

[0836] Output: A trained machine learning model

[0837] Step 3:

[0838] The user inputs candidate information through the terminal.

[0839] Specifically, the process involves using a web form or application to enter information such as the candidate's name, background, and reasons for applying.

[0840] Input: Candidate information (name, career history, reason for applying, etc.)

[0841] Output: Candidate information data

[0842] Step 4:

[0843] The terminal transmits the input candidate information to the server as an HTTP request.

[0844] Specifically, a POST request is used to send data to the server. For example, an HTTP client library is used.

[0845] Input: Candidate information data

[0846] Output: Candidate information sent to the server

[0847] Step 5:

[0848] The server analyzes the received candidate information and generates initial questions.

[0849] Specifically, it analyzes candidate information using the TF-IDF algorithm and other natural language processing techniques to select the most appropriate initial questions.

[0850] Input: Candidate information data

[0851] Output: Initial question (e.g. "Tell me about your recent projects")

[0852] Step 6:

[0853] The terminal displays the initial question received from the server to the user.

[0854] Specifically, the question is displayed on the user interface using a web browser or application, and a text field for displaying the question is displayed.

[0855] Input: Initial question

[0856] Output: The initial question that is displayed to the user

[0857] Step 7:

[0858] The user inputs the answers from the candidates.

[0859] Specifically, the answer obtained is input via the terminal, for example, by using a keyboard to enter the answer in a text field.

[0860] Input: Candidate's answer

[0861] Output: Candidate's answers entered

[0862] Step 8:

[0863] The terminal transmits the input answers of the candidates to the server.

[0864] Specifically, the input data is sent to the server again via an HTTP request.

[0865] Input: Candidate's answers entered

[0866] Output: Candidate's answers sent to the server

[0867] Step 9:

[0868] The server analyzes the received answer and generates the next question.

[0869] Specifically, it uses a trained machine learning model to generate the next appropriate question, such as "What was your specific role in that project?"

[0870] Input: Candidate's answer

[0871] Output: Next question

[0872] Step 10:

[0873] The terminal displays the next question received from the server to the user.

[0874] Specifically, the new question is displayed in the user interface, such as a web browser or application.

[0875] Input: Next question

[0876] Output: The next question that is displayed to the user

[0877] Step 11:

[0878] The emotion engine analyzes the user's voice and facial expression data to recognize emotions.

[0879] Specifically, it collects audio and video data and uses analysis engines such as the Microsoft Azure Emotion API to recognize emotions.

[0880] Input: Voice and facial expression data

[0881] Output: Recognized emotion data

[0882] Step 12:

[0883] The server generates a question based on the recognized emotion data.

[0884] Specifically, the server generates relaxing questions based on emotional data. For example, in a stressful situation, it generates a question such as, "What was the most rewarding moment in that project?"

[0885] Input: Recognized emotion data

[0886] Output: Sentiment-based questions

[0887] (Application example 2)

[0888] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0889] The goal is to provide a comfortable riding experience by realizing interactive communication with passengers in autonomous vehicles and generating appropriate questions according to the passenger's emotional state. Conventional systems can only have fixed dialogues with passengers, making it difficult to respond flexibly taking into account their emotional state.

[0890] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0891] In this invention, the server includes means for loading past dialogue example data from a database, means for training a machine learning model based on the past dialogue example data, means for providing an interface for a user to input passenger information, means for transmitting the input passenger information to the server, means for analyzing the passenger information and generating initial questions, means for displaying the generated initial questions to the user, means for the user to input passenger answers and transmitting the same to the server, means for analyzing the passenger answers and generating subsequent questions, means for analyzing voice data and facial expression data to recognize emotions, and means for generating questions based on the recognized emotions. This enables interactive communication with passengers in an autonomous vehicle and realizes flexible question generation based on the passenger's emotional state.

[0892] A "database" is a system that stores various information, such as past dialogue case data and financial report data, and allows it to be searched and read.

[0893] A "machine learning model" is an algorithm that learns patterns from data and makes appropriate decisions and predictions about new data.

[0894] An "interface" is a screen or input form through which a user enters information or interacts with a system.

[0895] A "server" is a computer system that processes and stores various data, and is a device that sends and receives data in response to requests from clients.

[0896] "Analysis" refers to the process of examining input data in detail to understand its characteristics and structure.

[0897] An "initial question" is a question that is initially generated based on information entered by the user.

[0898] "Voice data" refers to audio information recorded by a microphone or other device, which can be used for voice recognition and emotion recognition.

[0899] "Facial expression data" refers to facial expression information captured by a camera or other device, and emotions can be recognized based on this information.

[0900] An "emotion engine" is software or algorithm that analyzes a user's emotional state from their voice and facial expressions.

[0901] To implement this invention, it is first necessary to properly link the main components, such as the database, machine learning model, interface, server, and emotion engine. The specific steps are described below.

[0902] server

[0903] The server loads past dialogue case data from a database, which contains a variety of dialogue case data, including interview scenes. The server uses this data to train a machine learning model. This model incorporates natural language processing techniques and includes an algorithm for generating new questions. It is recommended to use a Python-based natural language processing library (e.g., NLTK, spaCy) or a machine learning framework (e.g., TensorFlow, PyTorch).

[0904] The server receives passenger information entered by the user and analyzes it to generate an initial question. Information from the user is usually received via an HTTP request, and the server analyzes it using an appropriate algorithm. The server then receives the passenger's answers entered by the user and analyzes them to generate the next question. During this process, an emotion engine analyzes voice and facial expression data to recognize emotions, and this data is also used to generate questions.

[0905] Terminal

[0906] The terminal provides an interface for users to enter passenger information. This interface includes text boxes and pull-down menus, allowing users to easily enter the necessary information. For example, passengers can enter their name, past ride experiences, and favorite music genres. Once the information is complete, the terminal sends the data to the server.

[0907] Next, when an initial question is generated from the server, it is displayed on the terminal. The user sees this question, asks it to the passenger, and the passenger enters their answer. The entered answer is also sent to the server, and the next question is generated in the same process. In this way, interactive communication is realized.

[0908] Emotion Engine

[0909] The emotion engine analyzes passengers' voice and facial expression data to recognize their emotions. This is done using a speech recognition library (e.g., Google Speech-to-Text API) and a facial expression recognition library (e.g., OpenCV, Dlib). The recognized emotion information is sent to the server and taken into consideration when generating the next question.

[0910] Specific examples

[0911] For example, when a passenger boards a vehicle, the system asks, "Tell me about your recent riding experience." If the passenger answers, "It was a very comfortable and pleasant ride," the emotion engine detects the passenger's positive emotion, and the server generates subsequent questions such as, "What did you particularly like?" or "Would you like to enjoy music today?"

[0912] Prompt Sentence Examples

[0913] "If a passenger responds positively about their past ride experience, what next questions can you ask based on that positive sentiment? For example, if a passenger responds, 'It was a very comfortable and pleasant ride,' how can you best continue?"

[0914] In this way, the system can enable interactive communication with passengers inside autonomous vehicles, providing them with a comfortable and personalized riding experience.

[0915] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0916] Step 1:

[0917] The terminal provides an interface for users to input passenger information (e.g., name, past ride experiences, favorite music genres), which is then sent to the server via an HTTP request. The input here is passenger information, and the output is sent to the server.

[0918] Step 2:

[0919] The server loads past dialogue case data from a database. The loaded data is a dataset containing past interviews and other dialogue scenes, and trains a machine learning model based on this data. The input is the dialogue case data from the database, and the output is the trained machine learning model. This training uses natural language processing techniques and machine learning frameworks (e.g., TensorFlow, PyTorch).

[0920] Step 3:

[0921] The server analyzes the passenger information submitted by the user and generates an initial question. The input is the passenger information from the user, and the output is the initial question. This process uses a text analysis algorithm. For example, if the user enters positive information about their "recent ride experience," the server generates an initial question based on that information: "What did you particularly like?"

[0922] Step 4:

[0923] The terminal displays the initial question received from the server to the user. Based on the displayed initial question, the user asks the question to the passenger and receives an answer. The input is the initial question from the server, and the output is the display to the user.

[0924] Step 5:

[0925] The user inputs the passenger's answers into the terminal and sends the data to the server via an HTTP request. The input is the passenger's answers and the output is the transmission to the server.

[0926] Step 6:

[0927] The server analyzes the passenger's response and generates the next question. The input is the passenger's response from the user, and the output is the next question. The analysis uses a trained machine learning model and an emotion engine. Specifically, the next question is generated by combining emotional information extracted from voice data and facial expression data.

[0928] Step 7:

[0929] The emotion engine analyzes passengers' voice and facial expression data to recognize their emotions. The input is voice data and facial expression data, and the output is the recognized emotion information. This is done using libraries such as the Google Speech-to-Text API, OpenCV, and Dlib.

[0930] Step 8:

[0931] The server generates the next question based on the recognized emotion information. The input is the emotion information from the emotion engine and the passenger's answer, and the output is a flexibly generated next question. For example, if a passenger answers, "It was a very comfortable and pleasant ride," and a positive emotion is recognized, the next question generated will be, "What did you particularly like?"

[0932] By repeating the above steps, interactive communication with passengers inside the self-driving vehicle will proceed smoothly.

[0933] 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.

[0934] 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.

[0935] 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.

[0936] [Third embodiment]

[0937] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0938] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0939] 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).

[0940] 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.

[0941] 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.

[0942] 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).

[0943] 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.

[0944] 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.

[0945] 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.

[0946] 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.

[0947] 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.

[0948] 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."

[0949] An embodiment of the present invention will be described below. An interview scene will be described as an example, but it should be noted that the present invention can also be applied to a financial report scene in a similar manner.

[0950] server

[0951] 1. A way to load interview case data from a database

[0952] The server loads past interview case data from a database, which includes questions asked in interviews and their answers. By retrieving data from the database, the server builds a training dataset.

[0953] 2. Means of training machine learning models

[0954] Based on the loaded interview case data, the server trains a machine learning model. Specifically, it uses natural language processing technology to learn patterns of past questions and answers and create a model that generates new questions. This model includes a question generation algorithm and can generate appropriate questions based on the candidate's information.

[0955] 3. A means of analyzing user-entered data and generating initial questions

[0956] When the user (interviewer) enters candidate information and sends it to the server, the server analyzes the data and generates initial questions, where an algorithm is applied to select the most appropriate initial questions based on the information entered by the user.

[0957] Terminal

[0958] 1. A means of providing an interface for users to input candidate information

[0959] The terminal provides an interface for the user to input candidate information, specifically, a form for inputting name, career history, reason for applying, etc.

[0960] 2. Means of sending the entered data to the server

[0961] When the user (interviewer) has completed the input, the data is sent to the server using a normal HTTP request.

[0962] 3. A means to display the generated initial questions to the user

[0963] When the terminal receives the generated initial questions from the server, it displays them to the user (interviewer). Once the questions are displayed, the user can start asking questions to the candidate.

[0964] 4. A means for users to input candidate answers and send them to the server

[0965] After the user (interviewer) receives the answer from the candidate, he / she enters the answer into the terminal and sends it to the server. This procedure is performed using an HTTP request, just like the initial question.

[0966] User

[0967] 1. Enter candidate information

[0968] The user (interviewer) enters candidate information, such as the candidate's name, highest level of education, work history, and reason for applying, according to the provided interface.

[0969] 2. Asking initial questions and getting answers

[0970] The user (interviewer) views the initial questions generated by the server, asks them to the candidate, and receives answers.

[0971] 3. Enter and submit your answers

[0972] The answer obtained is entered into the terminal and sent to the server to determine what question to ask next.

[0973] Specific examples

[0974] For example, consider an initial interview question that might be generated: "Tell me about a recent project." The interviewer asks this question to a candidate, who responds, "I worked on project X, which resulted in Y results." This response is entered into a device and sent to a server, which analyzes it and generates the next question: "What was your specific role in that project?" This process is repeated to effectively progress the interview.

[0975] In this way, users can easily generate appropriate questions based on the candidate's information, and the dialogue can proceed smoothly. This system can improve the quality of interviews.

[0976] The processing flow will be explained below.

[0977] Step 1:

[0978] The server loads the past interview case data from the database, which involves establishing a database connection and executing SQL queries to retrieve the required data.

[0979] Step 2:

[0980] The server trains a machine learning model based on the loaded interview case data. Specifically, it performs data preprocessing and trains the model using natural language processing techniques. The model learns patterns from question-answer pairs.

[0981] Step 3:

[0982] The terminal displays an interface for entering candidate information. The user (interviewer) uses this interface to enter the candidate's name, career history, reason for applying, etc.

[0983] Step 4:

[0984] Once the user has completed the input, the device sends the data as an HTTP request to the server, which receives the request and analyzes the candidate information.

[0985] Step 5:

[0986] The server generates initial questions based on the analysis results, uses a trained machine learning model to select questions that best fit the candidate information, and sends the generated questions to the device.

[0987] Step 6:

[0988] The terminal displays the initial questions received from the server to the user, who then verbally asks the candidate the questions.

[0989] Step 7:

[0990] The user receives answers from the candidates, enters them into the terminal, and sends them back to the server, again using an HTTP request.

[0991] Step 8:

[0992] The server analyzes the received answer and uses an analysis algorithm to extract important information from the answer and generate the next question.

[0993] Step 9:

[0994] The server sends the next generated question to the terminal, which then displays the received question to the user (interviewer).

[0995] Step 10:

[0996] The user asks the candidate the following questions, and this process is repeated until the interview is complete, improving the quality of the interview and facilitating a smooth dialogue.

[0997] This series of processing steps enables the present invention to provide users with dynamic and interactive question generation, improving the efficiency and quality of interviews and financial reports.

[0998] Example 1

[0999] 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."

[1000] In conventional interview systems and financial reporting systems, it is difficult for interviewers or reporters to quickly generate appropriate questions, which can lead to inefficient dialogue. Furthermore, if questions are not generated appropriately, the quality of the dialogue declines and necessary information cannot be fully elicited. The present invention aims to solve these problems by providing a system that can automatically generate appropriate and effective questions.

[1001] 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.

[1002] In this invention, the server includes means for loading past interview case data from a database, means for training a machine learning model based on the past interview case data, means for providing an interface for a user to input candidate information, means for transmitting the input candidate information to the server, means for analyzing the candidate information and generating initial questions, means for displaying the generated initial questions to the user, means for a user to input answers from the candidate and transmit them to the server, and means for analyzing the answers from the candidate and generating the next question. This enables interviewers and reporters to effectively conduct dialogue and appropriately extract necessary information.

[1003] A "database" is a system for efficiently storing, searching, and managing large amounts of data.

[1004] "Interview case data" refers to records of questions and answers from past interviews.

[1005] A "machine learning model" is a collection of algorithms that learn patterns and rules based on data and use the results of that learning to make predictions and classify new data.

[1006] "User" refers to the person who operates the system, inputs information, and views results.

[1007] An "interface" is a means, such as a screen or input form, through which a user interacts with a system.

[1008] A "server" is a computer system that processes and stores data over a network.

[1009] "Analysis" is the process of examining input data in detail and extracting meaning and patterns.

[1010] An "initial question" refers to the first question posed to a candidate or reporter when an interview or report begins.

[1011] "Answer" refers to the information or response returned in response to a question.

[1012] The "next question" is a new question that is generated based on the previous answer.

[1013] A "system" is a mechanism in which multiple components and processes work together to provide specific functions and services.

[1014] This invention relates to an interview system and a financial reporting system, and aims to effectively progress the dialogue by having a user input information about candidates and reporters, and then having a server analyze the information and automatically generate appropriate questions.

[1015] server

[1016] The server first uses a database management system to load past interview case data and financial report data from a database. This data includes past questions and their answers. The server then uses natural language processing technology to train a machine learning model based on the loaded data. Specifically, it trains a BERT-based model using frameworks such as TensorFlow and PyTorch to build a question generation algorithm.

[1017] Terminal

[1018] The terminal provides an interface for users to input candidate information and report details, such as name, career history, reasons for applying, and financial report details, through a web form written in React. Once the user has completed the input, the data is sent to the server as an HTTP POST request.

[1019] The terminal then receives the initial questions generated by the server and displays them to the user. Similarly, when the user enters the candidate's answers and additional information, that data is also sent to the server.

[1020] User

[1021] The user uses the terminal to input candidate information and financial report details. The specific operating procedure is as follows:

[1022] 1. The user follows the provided interface to enter the candidate's name, highest level of education, work history, reason for applying, etc.

[1023] 2. Check the initial questions generated by the server and ask them to the candidate to get their answers.

[1024] 3. The answer obtained is entered into the terminal and sent to the server to generate the next question.

[1025] Specific examples

[1026] For example, an initial interview question might be generated as "Tell me about a recent project." If a user poses this question to a candidate and the candidate responds with "I worked on project X, which resulted in Y results," the response is typed into the device and sent to the server. The server then analyzes the response and generates the next question, "What was your specific role in that project?", which is returned to the device and displayed to the user. This process is repeated to effectively progress the interview.

[1027] Examples of specific prompts include the following:

[1028] Please tell us about the candidate's highest level of education.

[1029] What is your specific role in this project?

[1030] This invention allows users to quickly and efficiently generate appropriate questions for candidates and reporters, allowing for smooth dialogue, thereby improving the quality of interviews and the effectiveness of reports.

[1031] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1032] Specific explanation of program processing

[1033] server

[1034] Step 1:

[1035] The server loads past interview case data from a database.

[1036] Input: Database connection information

[1037] Output: Interview case data list

[1038] What it does: The server executes an SQL query to retrieve interview case data from a database, which includes past interview questions and answers.

[1039] Step 2:

[1040] The server trains a machine learning model based on the loaded interview case data.

[1041] Input: Interview case data list

[1042] Output: A trained machine learning model

[1043] How it works: The server uses TensorFlow and PyTorch to train a BERT-based natural language processing model, which involves learning question and answer patterns.

[1044] Step 3:

[1045] The server analyzes the candidate information sent by the user from the terminal and generates an initial question.

[1046] Input: Candidate information

[1047] Output: Initial question

[1048] Specific operation: The server uses a machine learning model to analyze candidate information and generate appropriate initial questions.

[1049] Terminal

[1050] Step 1:

[1051] The terminal provides an interface for the user to input candidate information.

[1052] Input: None (standby)

[1053] Output: Form for entering candidate information

[1054] Specific operation: The device uses "React" to display a web form for entering information such as name, background, and reasons for applying.

[1055] Step 2:

[1056] The terminal transmits the input candidate information to the server.

[1057] Input: Candidate information

[1058] Output: HTTP POST request

[1059] Specific operation: Generates and sends an HTTP POST request to send the candidate information entered by the user in the form to the server.

[1060] Step 3:

[1061] The terminal displays the generated initial question to the user.

[1062] Input: Initial question (response from server)

[1063] Output: Question displayed on the user interface

[1064] Specific operation: The initial question received from the server is displayed clearly to the user.

[1065] Step 4:

[1066] The user inputs the candidate's answers at the terminal and transmits them to the server.

[1067] Input: Candidate's answer

[1068] Output: HTTP POST request

[1069] Specific operation: The answers entered by the user are recorded on the terminal and sent to the server.

[1070] User

[1071] Step 1:

[1072] The user inputs the candidate information.

[1073] Input: Candidate information (name, highest level of education, work history, reason for applying, etc.)

[1074] Output: Input data (terminal)

[1075] Specific operation: The user enters each item according to the form provided on the terminal.

[1076] Step 2:

[1077] The user poses initial questions to the candidate and receives answers.

[1078] Input: Initial question (displayed on the terminal)

[1079] Output: Candidate's answers

[1080] Specific operation: The user asks the candidate initial questions displayed on the terminal and receives answers.

[1081] Step 3:

[1082] The user inputs the answers obtained and sends them to the server to generate the next question.

[1083] Input: Candidate's answer

[1084] Output: HTTP POST request

[1085] Specific operation: The user inputs the candidate's answer into the terminal and sends a request to the server to generate the next question.

[1086] (Application example 1)

[1087] 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."

[1088] Customer support and dialogue systems in modern virtual environments are based on static FAQs and scripts, making it difficult to provide appropriate responses to specific user questions and preventing efficient and effective interactive dialogue that meets user needs. A particular problem is the lack of systems that can quickly and accurately answer questions visitors have about specific products or services in virtual stores and online support.

[1089] 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.

[1090] In this invention, the server includes: means for loading past case data from a database; means for training a machine learning model based on the past case data; means for providing an interface for a user to input information; means for transmitting the input information to the server; means for analyzing the information and generating an initial question; means for displaying the generated initial question to the user; means for the user to input an answer and transmitting the answer to the server; means for analyzing the answer and generating a next question; means for answering questions about products and services from visitors through dialogue in a virtual environment; means for functioning as an application installed on a smart device; and means for dynamically generating questions and interactively displaying them through a user interface. This enables users' questions to be answered quickly and accurately, significantly improving the quality of customer support in virtual stores.

[1091] A "database" is a repository that systematically organizes information so that it can be searched and retrieved efficiently.

[1092] "Case data" is data that compiles information about specific past events or situations.

[1093] A "machine learning model" is a set of algorithms that are trained with data to perform specific tasks automatically.

[1094] An "interface" is a means or device by which a user and a system interact with each other.

[1095] A "server" is a computer that provides services and resources to other computers over a network.

[1096] "Analysis" is the process of examining data and information in detail and clarifying their structure and relationships.

[1097] An "initial question" is a question that the dialogue system generates to start the initial dialogue with the user.

[1098] "User" means a person or entity that uses a system or service.

[1099] An "answer" is a response given to a question.

[1100] The "next question" is the next question that the system generates based on the user's answer.

[1101] A "virtual environment" is a virtual space or system created by computer simulation.

[1102] "Goods" means goods or services manufactured for sale.

[1103] A "service" is an activity or undertaking that provides benefit or satisfaction to consumers.

[1104] A "smart device" is an electronic device with advanced computing power and connectivity capabilities.

[1105] An "application" is software designed for a specific purpose.

[1106] A "user interface" is the means or method by which a user interacts with a system or device.

[1107] The present invention relates to a customer support assistant system for effectively conducting dialogue with visitors in a virtual store. The system aims to respond quickly and appropriately to visitors' questions and increase their satisfaction. The following describes in detail the embodiments of the present invention.

[1108] server

[1109] The server implements the following methods:

[1110] A means of loading case data from a database

[1111] The server loads past customer support case data from the database, including past questions and their answers, which the server uses as a training data set.

[1112] A means to train machine learning models

[1113] Using the example data loaded onto the server, the server trains a machine learning model. Specifically, it uses natural language processing techniques to learn patterns of past questions and answers and create a model that generates new questions. An example of the software used is the Hugging Face transformers library.

[1114] A means of analyzing user information and generating initial questions

[1115] When the user (customer support representative) enters visitor information and sends it to the server, the server analyzes the data and applies an algorithm to generate the most appropriate initial questions, again using natural language processing technology.

[1116] Terminal

[1117] The terminal provides the means to:

[1118] A means of providing an interface for users to input information

[1119] The terminal provides an interface for the user to input visitor information, such as name, inquiry details, and past purchase history.

[1120] A means of sending the entered data to the server

[1121] Once the user has completed the input, the data is sent to the server using an HTTP request.

[1122] A means of displaying the generated initial questions to the user

[1123] Upon receiving the generated initial question from the server, the terminal displays the question to the user (customer support representative), allowing the user to begin asking questions to the visitor.

[1124] A means for the user to enter answers and send them to the server

[1125] After the user retrieves the visitor's answer, they enter it into their device and send it to the server, again using an HTTP request to send the data.

[1126] User

[1127] The role of the user (customer support representative) is to:

[1128] Entering visitor information

[1129] The user inputs visitor information using the provided interface, such as the visitor's name, inquiry details, and past purchase history.

[1130] Asking initial questions and getting answers

[1131] The user sees the initial question generated by the server, poses it to the visitor, and receives the answer.

[1132] Enter and submit your answers

[1133] The answer obtained is entered into the terminal and sent to the server to determine what question to ask next.

[1134] Examples and prompts

[1135] For example, consider a case where a visitor to a virtual store wants information about a new product. Based on the visitor's purchase history and inquiries entered through the terminal, the server generates an initial question: "What are the features of this product?" After receiving the visitor's answer, the server generates the next question: "What are the unique features of this product?" This allows for timely and accurate responses to the visitor's questions.

[1136] Example prompt sentence:

[1137] "Please tell us the features of the products you recently released."

[1138] "How have you improved after receiving customer feedback?"

[1139] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1140] Step 1:

[1141] Loading case data from a database

[1142] The server loads past customer support case data from a database, including past questions and their answers, and uses this data as a training dataset to train the machine learning model.

[1143] Input: Query request to the database

[1144] Output: Dataset of past cases

[1145] Step 2:

[1146] Train a machine learning model

[1147] The server uses the loaded example data to train a machine learning model, leveraging natural language processing techniques. Specifically, it uses Hugging Face's transformers library to learn patterns of past questions and answers and create a model that can generate new questions.

[1148] Input: Historical case dataset

[1149] Output: A trained machine learning model

[1150] Step 3:

[1151] Provides an interface for entering user information

[1152] The terminal provides an interface for the user (customer support representative) to input visitor information, which is displayed as a form for entering information such as the user's name, inquiry details, and past purchase history.

[1153] Input: None (launches user interface)

[1154] Output: Interface screen display

[1155] Step 4:

[1156] Send the entered information to the server

[1157] Once the user has entered the visitor information into the interface, the device sends the information to the server. This data transmission is done using an HTTP request.

[1158] Input: Visitor information (user input)

[1159] Output: Send data to the server (HTTP request)

[1160] Step 5:

[1161] Generate initial questions

[1162] The server analyzes the received visitor information and applies machine learning models to generate the most appropriate initial questions. Natural language processing technology selects appropriate questions based on actual visitor information.

[1163] Input: Visitor information (received by server)

[1164] Output: Initial question

[1165] Step 6:

[1166] Display the generated initial questions to the user

[1167] The terminal displays the initial question retrieved from the server to the customer support representative, and the user starts a dialogue with the visitor based on this question.

[1168] Input: Initial question (sent by server)

[1169] Output: Question displayed on the screen

[1170] Step 7:

[1171] Enter the visitor's answer and send it to the server

[1172] The user enters the answers received from the visitors into the terminal and sends the answer data to the server. This data transmission is also done using an HTTP request.

[1173] Input: Visitor's answer (user input)

[1174] Output: Send data to the server (HTTP request)

[1175] Step 8:

[1176] Generate the next question

[1177] The server analyzes the visitor's answers and generates the next question, again dynamically using a machine learning model.

[1178] Input: Visitor's response (received by the server)

[1179] Output: Next question

[1180] Step 9:

[1181] Show the next question to the user

[1182] The terminal displays the next question retrieved from the server to the customer support representative, allowing the user to pose the next question to the visitor.

[1183] Input: Next question (sent by server)

[1184] Output: Question displayed on the screen

[1185] Prompt Sentence Examples

[1186] 1. "Please tell us about the features of the products you recently released."

[1187] 2. "How have you improved after receiving customer feedback?"

[1188] 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.

[1189] The following describes an embodiment of the present invention. The following describes an interview scene as an example, but please note that the present invention can also be applied to a financial report scene using similar methods. The present invention also incorporates an emotion engine.

[1190] server

[1191] 1. A way to load past interview case data from a database

[1192] The server loads past interview case data from a database, which includes questions asked in interviews and their answers. By retrieving data from the database, the server builds a training dataset.

[1193] 2. Means of training machine learning models

[1194] Based on the loaded interview case data, the server trains a machine learning model. Specifically, it uses natural language processing technology to learn patterns of past questions and answers and create a model that generates new questions. This model includes a question generation algorithm and can generate appropriate questions based on the candidate's information.

[1195] 3. A means of analyzing user-entered data and generating initial questions

[1196] When the user (interviewer) enters candidate information and sends it to the server, the server analyzes the data and generates initial questions, where an algorithm is applied to select the most appropriate initial questions based on the information entered by the user.

[1197] Terminal

[1198] 1. A means of providing an interface for users to input candidate information

[1199] The terminal provides an interface for the user to input candidate information, specifically, a form for inputting name, career history, reason for applying, etc.

[1200] 2. Means of sending the entered data to the server

[1201] When the user (interviewer) has completed the input, the data is sent to the server using a normal HTTP request.

[1202] 3. A means to display the generated initial questions to the user

[1203] When the terminal receives the generated initial questions from the server, it displays them to the user (interviewer). Once the questions are displayed, the user can start asking questions to the candidate.

[1204] 4. A means for users to input candidate answers and send them to the server

[1205] After the user (interviewer) receives the answer from the candidate, he / she enters the answer into the terminal and sends it to the server. This procedure is performed using an HTTP request, just like the initial question.

[1206] Emotion Engine

[1207] 1. A means of recognizing emotions by analyzing the user's voice and facial expressions

[1208] The emotion engine analyzes voice and facial expression data provided by the device, allowing it to recognize the user's emotions (e.g., joy, anger, sadness, surprise, etc.).

[1209] 2. A way to generate questions based on perceived sentiment

[1210] The recognized emotion data is sent to the server, which then generates questions based on the emotion information. For example, if the user indicates a high stress level, the system generates questions to encourage relaxation.

[1211] User

[1212] 1. Enter candidate information

[1213] The user (interviewer) enters candidate information, such as the candidate's name, highest level of education, work history, and reason for applying, according to the provided interface.

[1214] 2. Asking initial questions and getting answers

[1215] The user (interviewer) views the initial questions generated by the server, asks them to the candidate, and receives answers.

[1216] 3. Enter and submit your answers

[1217] After receiving the candidate's answer, the answer is entered into the terminal and sent to the server to get the next question to ask.

[1218] 4. Emotion Recognition by Emotion Engine

[1219] During the interview, the device captures the user's voice and facial expressions, which are then analyzed by an emotion engine. The analysis results are sent to the server and used to generate the next questions.

[1220] Specific examples

[1221] For example, consider an initial interview question generated as "Tell me about a recent project." The interviewer asks this question to a candidate, who responds, "I worked on project X, which resulted in Y results." When this response is entered into a device and sent to a server, the server analyzes it and generates the next question: "What was your specific role in that project?"

[1222] During this process, if the emotion engine detects the interviewer's stress level, for example if it is determined to be high, the server can take this into account and generate relaxing questions such as, "What was the most rewarding moment in that project?"

[1223] In this way, users can easily generate appropriate questions based on the candidate's information, and the dialogue can proceed smoothly. This system improves the quality of interviews and also enables flexible responses according to the user's emotional state.

[1224] The processing flow will be explained below.

[1225] Step 1:

[1226] The server loads past interview case data from the database by establishing a database connection and executing SQL queries to retrieve the required data.

[1227] Step 2:

[1228] The server trains a machine learning model based on the loaded interview case data. Specifically, it performs data preprocessing and trains the model using natural language processing techniques.

[1229] Step 3:

[1230] The terminal provides the user (interviewer) with an interface for inputting candidate information. The user inputs the candidate's name, career history, reason for applying, etc.

[1231] Step 4:

[1232] Once the user has completed the input, the device sends the data as an HTTP request to the server, which receives the request and analyzes the candidate information.

[1233] Step 5:

[1234] The server generates initial questions based on the analysis results, uses a trained machine learning model to select questions that best fit the candidate information, and sends the generated questions to the device.

[1235] Step 6:

[1236] The terminal displays the initial questions received from the server to the user, who then verbally asks the candidate the questions.

[1237] Step 7:

[1238] The emotion engine analyzes the user's voice and facial expressions in real time to evaluate the user's emotional state, and the evaluation results are sent to the server.

[1239] Step 8:

[1240] The user receives answers from the candidates, enters these answers into the terminal, and sends them back to the server.

[1241] Step 9:

[1242] The server analyzes the received answer and also uses the emotion data from the emotion engine for analysis. The analysis algorithm takes into account the important information in the answer and the user's emotion to generate the next question.

[1243] Step 10:

[1244] The server sends the next generated question to the terminal, which then displays the received question to the user (interviewer).

[1245] Step 11:

[1246] The user asks the candidate the following questions, and this process is repeated until the interview is complete, improving the quality of the interview and facilitating a smooth dialogue.

[1247] Through this series of processing steps, the present invention provides users with dynamic and interactive question generation, improving the efficiency and quality of interviews. In addition, by combining it with an emotion engine, flexible responses according to the user's emotional state become possible.

[1248] Example 2

[1249] 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."

[1250] In traditional situations such as interviews and financial reports, efficient and effective question generation is required, but in the process, it has been difficult to generate questions that take human emotions into account and maintain a continuity of appropriate questions. In particular, the lack of a dynamic question generation method that takes into account the emotional state of the interviewer or reporter can lead to a decline in the quality of interviews and reports.

[1251] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1252] In this invention, the server includes means for loading past interview case data from a database, means for training a machine learning model based on the past interview case data, means for providing an interface for a user to input candidate information, means for transmitting the input candidate information to the server, means for analyzing the candidate information and generating initial questions, means for displaying the generated initial questions to the user, means for a user to input candidate answers and transmitting the answers to the server, means for analyzing the candidate answers and generating subsequent questions, means for analyzing the user's voice and facial expressions to recognize emotions, and means for generating questions based on the recognized emotions. This makes it possible to generate appropriate and consecutive questions while taking into account the user's emotional state during interviews and reporting situations.

[1253] A "database" is a collection of information that efficiently manages past interview case data, financial report data, etc., and can be searched and retrieved as needed.

[1254] A "machine learning model" is an algorithm that learns from past data, discovers new patterns and relationships, and automates specific tasks such as question generation.

[1255] An "interface" is a user interface component that allows the user to interact with the system and input candidate information, financial report details, and the like.

[1256] "Server" means a computer system that receives information sent by users and processes and analyzes the data.

[1257] "Analysis" is the process of examining input data in detail to understand its structure and meaning.

[1258] An "initial question" is a question that is presented to a user when they first begin interacting with the system.

[1259] The "emotion engine" is a system that analyzes the user's voice and facial expression data to recognize their emotional state at that time.

[1260] "Dynamic question generation" is the process of generating new questions in real time based on the current user situation and answers.

[1261] "Interview case data" is data on questions used in past interviews and the answers to those questions.

[1262] "Recognized emotion" refers to the emotional state of the user that has been analyzed and identified by the emotion engine.

[1263] The present invention provides a method for generating questions efficiently and effectively in interactive systems for interviews, financial reporting, etc. While the present invention will be described in detail below using an interview scene as an example, it can also be applied to financial reporting scenes using similar means.

[1264] Hardware and software used

[1265] Server: A high-performance computer system that processes and analyzes data and trains machine learning models.

[1266] Database: A management system for storing interview case data and financial report data. Example: MySQL.

[1267] Natural Language Processing (NLP) Library: A software library for using machine learning models. For example, "Transformers".

[1268] Emotion engine: A system that analyzes the user's voice and facial expressions. Example: "Microsoft Azure Emotion API."

[1269] Terminal: A device that provides an interface for users to input information. Examples include PCs and tablets.

[1270] Data Processing and Data Calculation

[1271] The server retrieves past interview case data from a database, for example using SQL queries. This data includes interview questions and their answers.

[1272] The server then uses this data to train a machine learning model, using natural language processing techniques to learn patterns of past questions and answers. Specifically, the Transformers library is used. Through this training, a model is built that can generate new questions.

[1273] When a user enters candidate information through a terminal, the information is sent to the server as an HTTP request. The server analyzes the information and generates initial questions. For example, if a user enters information about a candidate's recent projects, the server generates an initial question such as "Tell me about your recent projects."

[1274] The terminal displays the initial questions generated by the server to the user. When the user enters the candidate's answers, the answers are also sent to the server. The server analyzes the answers, generates the next appropriate questions, and sends them to the terminal.

[1275] The emotion engine analyzes the user's voice and facial expression data to recognize their emotions. This recognized emotion is sent to the server, which then generates questions based on the emotion data. For example, if a user is feeling highly stressed, the engine can generate questions such as, "What was the most rewarding moment in that project?"

[1276] Specific examples

[1277] For example, an initial interview question might be generated as "Tell me about a recent project." The interviewer asks this question to a candidate, who responds, "I worked on project X, which resulted in outcome Y." When this response is typed into a device and sent to a server, the server analyzes it and generates the next question: "What was your specific role in that project?"

[1278] As an example prompt, parse the candidate's response to the question "Tell me about a recent project": "I worked on project X, which resulted in outcome Y." Generate the following question:

[1279] The above is a specific embodiment for carrying out the invention. This invention improves the quality of interviews and reports, and enables smooth progress of dialogue.

[1280] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1281] Step 1:

[1282] The server loads past interview case data from a database.

[1283] Specifically, it uses an SQL query to execute a command like "SELECT FROM interview_data" to retrieve a dataset containing interview questions and their answers.

[1284] Input: Database query

[1285] Output: Dataset of past interview cases

[1286] Step 2:

[1287] The server uses the loaded interview case data to train a machine learning model.

[1288] Specifically, we use the natural language processing library "Transformers" to learn question and answer patterns, and build a question generation algorithm using models such as BERT and GPT.

[1289] Input: Past interview case dataset

[1290] Output: A trained machine learning model

[1291] Step 3:

[1292] The user inputs candidate information through the terminal.

[1293] Specifically, the process involves using a web form or application to enter information such as the candidate's name, background, and reasons for applying.

[1294] Input: Candidate information (name, career history, reason for applying, etc.)

[1295] Output: Candidate information data

[1296] Step 4:

[1297] The terminal transmits the input candidate information to the server as an HTTP request.

[1298] Specifically, a POST request is used to send data to the server. For example, an HTTP client library is used.

[1299] Input: Candidate information data

[1300] Output: Candidate information sent to the server

[1301] Step 5:

[1302] The server analyzes the received candidate information and generates initial questions.

[1303] Specifically, it analyzes candidate information using the TF-IDF algorithm and other natural language processing techniques to select the most appropriate initial questions.

[1304] Input: Candidate information data

[1305] Output: Initial question (e.g. "Tell me about your recent projects")

[1306] Step 6:

[1307] The terminal displays the initial question received from the server to the user.

[1308] Specifically, the question is displayed on the user interface using a web browser or application, and a text field for displaying the question is displayed.

[1309] Input: Initial question

[1310] Output: The initial question that is displayed to the user

[1311] Step 7:

[1312] The user inputs the answers from the candidates.

[1313] Specifically, the answer obtained is input via the terminal, for example, by using a keyboard to enter the answer in a text field.

[1314] Input: Candidate's answer

[1315] Output: Candidate's answers entered

[1316] Step 8:

[1317] The terminal transmits the input answers of the candidates to the server.

[1318] Specifically, the input data is sent to the server again via an HTTP request.

[1319] Input: Candidate's answers entered

[1320] Output: Candidate's answers sent to the server

[1321] Step 9:

[1322] The server analyzes the received answer and generates the next question.

[1323] Specifically, it uses a trained machine learning model to generate the next appropriate question, such as "What was your specific role in that project?"

[1324] Input: Candidate's answer

[1325] Output: Next question

[1326] Step 10:

[1327] The terminal displays the next question received from the server to the user.

[1328] Specifically, the new question is displayed in the user interface, such as a web browser or application.

[1329] Input: Next question

[1330] Output: The next question that is displayed to the user

[1331] Step 11:

[1332] The emotion engine analyzes the user's voice and facial expression data to recognize emotions.

[1333] Specifically, it collects audio and video data and uses analysis engines such as the Microsoft Azure Emotion API to recognize emotions.

[1334] Input: Voice and facial expression data

[1335] Output: Recognized emotion data

[1336] Step 12:

[1337] The server generates a question based on the recognized emotion data.

[1338] Specifically, the server generates relaxing questions based on emotional data. For example, in a stressful situation, it generates a question such as, "What was the most rewarding moment in that project?"

[1339] Input: Recognized emotion data

[1340] Output: Sentiment-based questions

[1341] (Application example 2)

[1342] 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."

[1343] The goal is to provide a comfortable riding experience by realizing interactive communication with passengers in autonomous vehicles and generating appropriate questions according to the passenger's emotional state. Conventional systems can only have fixed dialogues with passengers, making it difficult to respond flexibly taking into account their emotional state.

[1344] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1345] In this invention, the server includes means for loading past dialogue example data from a database, means for training a machine learning model based on the past dialogue example data, means for providing an interface for a user to input passenger information, means for transmitting the input passenger information to the server, means for analyzing the passenger information and generating initial questions, means for displaying the generated initial questions to the user, means for the user to input passenger answers and transmitting the same to the server, means for analyzing the passenger answers and generating subsequent questions, means for analyzing voice data and facial expression data to recognize emotions, and means for generating questions based on the recognized emotions. This enables interactive communication with passengers in an autonomous vehicle and realizes flexible question generation based on the passenger's emotional state.

[1346] A "database" is a system that stores various information, such as past dialogue case data and financial report data, and allows it to be searched and read.

[1347] A "machine learning model" is an algorithm that learns patterns from data and makes appropriate decisions and predictions about new data.

[1348] An "interface" is a screen or input form through which a user enters information or interacts with a system.

[1349] A "server" is a computer system that processes and stores various data, and is a device that sends and receives data in response to requests from clients.

[1350] "Analysis" refers to the process of examining input data in detail to understand its characteristics and structure.

[1351] An "initial question" is a question that is initially generated based on information entered by the user.

[1352] "Voice data" refers to audio information recorded by a microphone or other device, which can be used for voice recognition and emotion recognition.

[1353] "Facial expression data" refers to facial expression information captured by a camera or other device, and emotions can be recognized based on this information.

[1354] An "emotion engine" is software or algorithm that analyzes a user's emotional state from their voice and facial expressions.

[1355] To implement this invention, it is first necessary to properly link the main components, such as the database, machine learning model, interface, server, and emotion engine. The specific steps are described below.

[1356] server

[1357] The server loads past dialogue case data from a database, which contains a variety of dialogue case data, including interview scenes. The server uses this data to train a machine learning model. This model incorporates natural language processing techniques and includes an algorithm for generating new questions. It is recommended to use a Python-based natural language processing library (e.g., NLTK, spaCy) or a machine learning framework (e.g., TensorFlow, PyTorch).

[1358] The server receives passenger information entered by the user and analyzes it to generate an initial question. Information from the user is usually received via an HTTP request, and the server analyzes it using an appropriate algorithm. The server then receives the passenger's answers entered by the user and analyzes them to generate the next question. During this process, an emotion engine analyzes voice and facial expression data to recognize emotions, and this data is also used to generate questions.

[1359] Terminal

[1360] The terminal provides an interface for users to enter passenger information. This interface includes text boxes and pull-down menus, allowing users to easily enter the necessary information. For example, passengers can enter their name, past ride experiences, and favorite music genres. Once the information is complete, the terminal sends the data to the server.

[1361] Next, when an initial question is generated from the server, it is displayed on the terminal. The user sees this question, asks it to the passenger, and the passenger enters their answer. The entered answer is also sent to the server, and the next question is generated in the same process. In this way, interactive communication is realized.

[1362] Emotion Engine

[1363] The emotion engine analyzes passengers' voice and facial expression data to recognize their emotions. This is done using a speech recognition library (e.g., Google Speech-to-Text API) and a facial expression recognition library (e.g., OpenCV, Dlib). The recognized emotion information is sent to the server and taken into consideration when generating the next question.

[1364] Specific examples

[1365] For example, when a passenger boards a vehicle, the system asks, "Tell me about your recent riding experience." If the passenger answers, "It was a very comfortable and pleasant ride," the emotion engine detects the passenger's positive emotion, and the server generates subsequent questions such as, "What did you particularly like?" or "Would you like to enjoy music today?"

[1366] Prompt Sentence Examples

[1367] "If a passenger responds positively about their past ride experience, what next questions can you ask based on that positive sentiment? For example, if a passenger responds, 'It was a very comfortable and pleasant ride,' how can you best continue?"

[1368] In this way, the system can enable interactive communication with passengers inside autonomous vehicles, providing them with a comfortable and personalized riding experience.

[1369] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1370] Step 1:

[1371] The terminal provides an interface for users to input passenger information (e.g., name, past ride experiences, favorite music genres), which is then sent to the server via an HTTP request. The input here is passenger information, and the output is sent to the server.

[1372] Step 2:

[1373] The server loads past dialogue case data from a database. The loaded data is a dataset containing past interviews and other dialogue scenes, and trains a machine learning model based on this data. The input is the dialogue case data from the database, and the output is the trained machine learning model. This training uses natural language processing techniques and machine learning frameworks (e.g., TensorFlow, PyTorch).

[1374] Step 3:

[1375] The server analyzes the passenger information submitted by the user and generates an initial question. The input is the passenger information from the user, and the output is the initial question. This process uses a text analysis algorithm. For example, if the user enters positive information about their "recent ride experience," the server generates an initial question based on that information: "What did you particularly like?"

[1376] Step 4:

[1377] The terminal displays the initial question received from the server to the user. Based on the displayed initial question, the user asks the question to the passenger and receives an answer. The input is the initial question from the server, and the output is the display to the user.

[1378] Step 5:

[1379] The user inputs the passenger's answers into the terminal and sends the data to the server via an HTTP request. The input is the passenger's answers and the output is the transmission to the server.

[1380] Step 6:

[1381] The server analyzes the passenger's response and generates the next question. The input is the passenger's response from the user, and the output is the next question. The analysis uses a trained machine learning model and an emotion engine. Specifically, the next question is generated by combining emotional information extracted from voice data and facial expression data.

[1382] Step 7:

[1383] The emotion engine analyzes passengers' voice and facial expression data to recognize their emotions. The input is voice data and facial expression data, and the output is the recognized emotion information. This is done using libraries such as the Google Speech-to-Text API, OpenCV, and Dlib.

[1384] Step 8:

[1385] The server generates the next question based on the recognized emotion information. The input is the emotion information from the emotion engine and the passenger's answer, and the output is a flexibly generated next question. For example, if a passenger answers, "It was a very comfortable and pleasant ride," and a positive emotion is recognized, the next question generated will be, "What did you particularly like?"

[1386] By repeating the above steps, interactive communication with passengers inside the self-driving vehicle will proceed smoothly.

[1387] 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.

[1388] 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.

[1389] 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.

[1390] [Fourth embodiment]

[1391] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1392] 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.

[1393] 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).

[1394] 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.

[1395] 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.

[1396] 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).

[1397] 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.

[1398] 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.

[1399] 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.

[1400] 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.

[1401] 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.

[1402] 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.

[1403] 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."

[1404] An embodiment of the present invention will be described below. An interview scene will be described as an example, but it should be noted that the present invention can also be applied to a financial report scene in a similar manner.

[1405] server

[1406] 1. A way to load interview case data from a database

[1407] The server loads past interview case data from a database, which includes questions asked in interviews and their answers. By retrieving data from the database, the server builds a training dataset.

[1408] 2. Means of training machine learning models

[1409] Based on the loaded interview case data, the server trains a machine learning model. Specifically, it uses natural language processing technology to learn patterns of past questions and answers and create a model that generates new questions. This model includes a question generation algorithm and can generate appropriate questions based on the candidate's information.

[1410] 3. A means of analyzing user-entered data and generating initial questions

[1411] When the user (interviewer) enters candidate information and sends it to the server, the server analyzes the data and generates initial questions, where an algorithm is applied to select the most appropriate initial questions based on the information entered by the user.

[1412] Terminal

[1413] 1. A means of providing an interface for users to input candidate information

[1414] The terminal provides an interface for the user to input candidate information, specifically, a form for inputting name, career history, reason for applying, etc.

[1415] 2. Means of sending the entered data to the server

[1416] When the user (interviewer) has completed the input, the data is sent to the server using a normal HTTP request.

[1417] 3. A means to display the generated initial questions to the user

[1418] When the terminal receives the generated initial questions from the server, it displays them to the user (interviewer). Once the questions are displayed, the user can start asking questions to the candidate.

[1419] 4. A means for users to input candidate answers and send them to the server

[1420] After the user (interviewer) receives the answer from the candidate, he / she enters the answer into the terminal and sends it to the server. This procedure is performed using an HTTP request, just like the initial question.

[1421] User

[1422] 1. Enter candidate information

[1423] The user (interviewer) enters candidate information, such as the candidate's name, highest level of education, work history, and reason for applying, according to the provided interface.

[1424] 2. Asking initial questions and getting answers

[1425] The user (interviewer) views the initial questions generated by the server, asks them to the candidate, and receives answers.

[1426] 3. Enter and submit your answers

[1427] The answer obtained is entered into the terminal and sent to the server to determine what question to ask next.

[1428] Specific examples

[1429] For example, consider an initial interview question that might be generated: "Tell me about a recent project." The interviewer asks this question to a candidate, who responds, "I worked on project X, which resulted in Y results." This response is entered into a device and sent to a server, which analyzes it and generates the next question: "What was your specific role in that project?" This process is repeated to effectively progress the interview.

[1430] In this way, users can easily generate appropriate questions based on the candidate's information, and the dialogue can proceed smoothly. This system can improve the quality of interviews.

[1431] The processing flow will be explained below.

[1432] Step 1:

[1433] The server loads the past interview case data from the database, which involves establishing a database connection and executing SQL queries to retrieve the required data.

[1434] Step 2:

[1435] The server trains a machine learning model based on the loaded interview case data. Specifically, it performs data preprocessing and trains the model using natural language processing techniques. The model learns patterns from question-answer pairs.

[1436] Step 3:

[1437] The terminal displays an interface for entering candidate information. The user (interviewer) uses this interface to enter the candidate's name, career history, reason for applying, etc.

[1438] Step 4:

[1439] Once the user has completed the input, the device sends the data as an HTTP request to the server, which receives the request and analyzes the candidate information.

[1440] Step 5:

[1441] The server generates initial questions based on the analysis results, uses a trained machine learning model to select questions that best fit the candidate information, and sends the generated questions to the device.

[1442] Step 6:

[1443] The terminal displays the initial questions received from the server to the user, who then verbally asks the candidate the questions.

[1444] Step 7:

[1445] The user receives answers from the candidates, enters them into the terminal, and sends them back to the server, again using an HTTP request.

[1446] Step 8:

[1447] The server analyzes the received answer and uses an analysis algorithm to extract important information from the answer and generate the next question.

[1448] Step 9:

[1449] The server sends the next generated question to the terminal, which then displays the received question to the user (interviewer).

[1450] Step 10:

[1451] The user asks the candidate the following questions, and this process is repeated until the interview is complete, improving the quality of the interview and facilitating a smooth dialogue.

[1452] This series of processing steps enables the present invention to provide users with dynamic and interactive question generation, improving the efficiency and quality of interviews and financial reports.

[1453] Example 1

[1454] 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."

[1455] In conventional interview systems and financial reporting systems, it is difficult for interviewers or reporters to quickly generate appropriate questions, which can lead to inefficient dialogue. Furthermore, if questions are not generated appropriately, the quality of the dialogue declines and necessary information cannot be fully elicited. The present invention aims to solve these problems by providing a system that can automatically generate appropriate and effective questions.

[1456] 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.

[1457] In this invention, the server includes means for loading past interview case data from a database, means for training a machine learning model based on the past interview case data, means for providing an interface for a user to input candidate information, means for transmitting the input candidate information to the server, means for analyzing the candidate information and generating initial questions, means for displaying the generated initial questions to the user, means for a user to input answers from the candidate and transmit them to the server, and means for analyzing the answers from the candidate and generating the next question. This enables interviewers and reporters to effectively conduct dialogue and appropriately extract necessary information.

[1458] A "database" is a system for efficiently storing, searching, and managing large amounts of data.

[1459] "Interview case data" refers to records of questions and answers from past interviews.

[1460] A "machine learning model" is a collection of algorithms that learn patterns and rules based on data and use the results of that learning to make predictions and classify new data.

[1461] "User" refers to the person who operates the system, inputs information, and views results.

[1462] An "interface" is a means, such as a screen or input form, through which a user interacts with a system.

[1463] A "server" is a computer system that processes and stores data over a network.

[1464] "Analysis" is the process of examining input data in detail and extracting meaning and patterns.

[1465] An "initial question" refers to the first question posed to a candidate or reporter when an interview or report begins.

[1466] "Answer" refers to the information or response returned in response to a question.

[1467] The "next question" is a new question that is generated based on the previous answer.

[1468] A "system" is a mechanism in which multiple components and processes work together to provide specific functions and services.

[1469] This invention relates to an interview system and a financial reporting system, and aims to effectively progress the dialogue by having a user input information about candidates and reporters, and then having a server analyze the information and automatically generate appropriate questions.

[1470] server

[1471] The server first uses a database management system to load past interview case data and financial report data from a database. This data includes past questions and their answers. The server then uses natural language processing technology to train a machine learning model based on the loaded data. Specifically, it trains a BERT-based model using frameworks such as TensorFlow and PyTorch to build a question generation algorithm.

[1472] Terminal

[1473] The terminal provides an interface for users to input candidate information and report details, such as name, career history, reasons for applying, and financial report details, through a web form written in React. Once the user has completed the input, the data is sent to the server as an HTTP POST request.

[1474] The terminal then receives the initial questions generated by the server and displays them to the user. Similarly, when the user enters the candidate's answers and additional information, that data is also sent to the server.

[1475] User

[1476] The user uses the terminal to input candidate information and financial report details. The specific operating procedure is as follows:

[1477] 1. The user follows the provided interface to enter the candidate's name, highest level of education, work history, reason for applying, etc.

[1478] 2. Check the initial questions generated by the server and ask them to the candidate to get their answers.

[1479] 3. The answer obtained is entered into the terminal and sent to the server to generate the next question.

[1480] Specific examples

[1481] For example, an initial interview question might be generated as "Tell me about a recent project." If a user poses this question to a candidate and the candidate responds with "I worked on project X, which resulted in Y results," the response is typed into the device and sent to the server. The server then analyzes the response and generates the next question, "What was your specific role in that project?", which is returned to the device and displayed to the user. This process is repeated to effectively progress the interview.

[1482] Examples of specific prompts include the following:

[1483] Please tell us about the candidate's highest level of education.

[1484] What is your specific role in this project?

[1485] This invention allows users to quickly and efficiently generate appropriate questions for candidates and reporters, allowing for smooth dialogue, thereby improving the quality of interviews and the effectiveness of reports.

[1486] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1487] Specific explanation of program processing

[1488] server

[1489] Step 1:

[1490] The server loads past interview case data from a database.

[1491] Input: Database connection information

[1492] Output: Interview case data list

[1493] What it does: The server executes an SQL query to retrieve interview case data from a database, which includes past interview questions and answers.

[1494] Step 2:

[1495] The server trains a machine learning model based on the loaded interview case data.

[1496] Input: Interview case data list

[1497] Output: A trained machine learning model

[1498] How it works: The server uses TensorFlow and PyTorch to train a BERT-based natural language processing model, which involves learning question and answer patterns.

[1499] Step 3:

[1500] The server analyzes the candidate information sent by the user from the terminal and generates an initial question.

[1501] Input: Candidate information

[1502] Output: Initial question

[1503] Specific operation: The server uses a machine learning model to analyze candidate information and generate appropriate initial questions.

[1504] Terminal

[1505] Step 1:

[1506] The terminal provides an interface for the user to input candidate information.

[1507] Input: None (standby)

[1508] Output: Form for entering candidate information

[1509] Specific operation: The device uses "React" to display a web form for entering information such as name, background, and reasons for applying.

[1510] Step 2:

[1511] The terminal transmits the input candidate information to the server.

[1512] Input: Candidate information

[1513] Output: HTTP POST request

[1514] Specific operation: Generates and sends an HTTP POST request to send the candidate information entered by the user in the form to the server.

[1515] Step 3:

[1516] The terminal displays the generated initial question to the user.

[1517] Input: Initial question (response from server)

[1518] Output: Question displayed on the user interface

[1519] Specific operation: The initial question received from the server is displayed clearly to the user.

[1520] Step 4:

[1521] The user inputs the candidate's answers at the terminal and transmits them to the server.

[1522] Input: Candidate's answer

[1523] Output: HTTP POST request

[1524] Specific operation: The answers entered by the user are recorded on the terminal and sent to the server.

[1525] User

[1526] Step 1:

[1527] The user inputs the candidate information.

[1528] Input: Candidate information (name, highest level of education, work history, reason for applying, etc.)

[1529] Output: Input data (terminal)

[1530] Specific operation: The user enters each item according to the form provided on the terminal.

[1531] Step 2:

[1532] The user poses initial questions to the candidate and receives answers.

[1533] Input: Initial question (displayed on the terminal)

[1534] Output: Candidate's answers

[1535] Specific operation: The user asks the candidate initial questions displayed on the terminal and receives answers.

[1536] Step 3:

[1537] The user inputs the answers obtained and sends them to the server to generate the next question.

[1538] Input: Candidate's answer

[1539] Output: HTTP POST request

[1540] Specific operation: The user inputs the candidate's answer into the terminal and sends a request to the server to generate the next question.

[1541] (Application example 1)

[1542] 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."

[1543] Customer support and dialogue systems in modern virtual environments are based on static FAQs and scripts, making it difficult to provide appropriate responses to specific user questions and preventing efficient and effective interactive dialogue that meets user needs. A particular problem is the lack of systems that can quickly and accurately answer questions visitors have about specific products or services in virtual stores and online support.

[1544] 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.

[1545] In this invention, the server includes: means for loading past case data from a database; means for training a machine learning model based on the past case data; means for providing an interface for a user to input information; means for transmitting the input information to the server; means for analyzing the information and generating an initial question; means for displaying the generated initial question to the user; means for the user to input an answer and transmitting the answer to the server; means for analyzing the answer and generating a next question; means for answering questions about products and services from visitors through dialogue in a virtual environment; means for functioning as an application installed on a smart device; and means for dynamically generating questions and interactively displaying them through a user interface. This enables users' questions to be answered quickly and accurately, significantly improving the quality of customer support in virtual stores.

[1546] A "database" is a repository that systematically organizes information so that it can be searched and retrieved efficiently.

[1547] "Case data" is data that compiles information about specific past events or situations.

[1548] A "machine learning model" is a set of algorithms that are trained with data to perform specific tasks automatically.

[1549] An "interface" is a means or device by which a user and a system interact with each other.

[1550] A "server" is a computer that provides services and resources to other computers over a network.

[1551] "Analysis" is the process of examining data and information in detail and clarifying their structure and relationships.

[1552] An "initial question" is a question that the dialogue system generates to start the initial dialogue with the user.

[1553] "User" means a person or entity that uses a system or service.

[1554] An "answer" is a response given to a question.

[1555] The "next question" is the next question that the system generates based on the user's answer.

[1556] A "virtual environment" is a virtual space or system created by computer simulation.

[1557] "Goods" means goods or services manufactured for sale.

[1558] A "service" is an activity or undertaking that provides benefit or satisfaction to consumers.

[1559] A "smart device" is an electronic device with advanced computing power and connectivity capabilities.

[1560] An "application" is software designed for a specific purpose.

[1561] A "user interface" is the means or method by which a user interacts with a system or device.

[1562] The present invention relates to a customer support assistant system for effectively conducting dialogue with visitors in a virtual store. The system aims to respond quickly and appropriately to visitors' questions and increase their satisfaction. The following describes in detail the embodiments of the present invention.

[1563] server

[1564] The server implements the following methods:

[1565] A means of loading case data from a database

[1566] The server loads past customer support case data from the database, including past questions and their answers, which the server uses as a training data set.

[1567] A means to train machine learning models

[1568] Using the example data loaded onto the server, the server trains a machine learning model. Specifically, it uses natural language processing techniques to learn patterns of past questions and answers and create a model that generates new questions. An example of the software used is the Hugging Face transformers library.

[1569] A means of analyzing user information and generating initial questions

[1570] When the user (customer support representative) enters visitor information and sends it to the server, the server analyzes the data and applies an algorithm to generate the most appropriate initial questions, again using natural language processing technology.

[1571] Terminal

[1572] The terminal provides the means to:

[1573] A means of providing an interface for users to input information

[1574] The terminal provides an interface for the user to input visitor information, such as name, inquiry details, and past purchase history.

[1575] A means of sending the entered data to the server

[1576] Once the user has completed the input, the data is sent to the server using an HTTP request.

[1577] A means of displaying the generated initial questions to the user

[1578] Upon receiving the generated initial question from the server, the terminal displays the question to the user (customer support representative), allowing the user to begin asking questions to the visitor.

[1579] A means for the user to enter answers and send them to the server

[1580] After the user retrieves the visitor's answer, they enter it into their device and send it to the server, again using an HTTP request to send the data.

[1581] User

[1582] The role of the user (customer support representative) is to:

[1583] Entering visitor information

[1584] The user inputs visitor information using the provided interface, such as the visitor's name, inquiry details, and past purchase history.

[1585] Asking initial questions and getting answers

[1586] The user sees the initial question generated by the server, poses it to the visitor, and receives the answer.

[1587] Enter and submit your answers

[1588] The answer obtained is entered into the terminal and sent to the server to determine what question to ask next.

[1589] Examples and prompts

[1590] For example, consider a case where a visitor to a virtual store wants information about a new product. Based on the visitor's purchase history and inquiries entered through the terminal, the server generates an initial question: "What are the features of this product?" After receiving the visitor's answer, the server generates the next question: "What are the unique features of this product?" This allows for timely and accurate responses to the visitor's questions.

[1591] Example prompt sentence:

[1592] "Please tell us the features of the products you recently released."

[1593] "How have you improved after receiving customer feedback?"

[1594] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1595] Step 1:

[1596] Loading case data from a database

[1597] The server loads past customer support case data from a database, including past questions and their answers, and uses this data as a training dataset to train the machine learning model.

[1598] Input: Query request to the database

[1599] Output: Dataset of past cases

[1600] Step 2:

[1601] Train a machine learning model

[1602] The server uses the loaded example data to train a machine learning model, leveraging natural language processing techniques. Specifically, it uses Hugging Face's transformers library to learn patterns of past questions and answers and create a model that can generate new questions.

[1603] Input: Historical case dataset

[1604] Output: A trained machine learning model

[1605] Step 3:

[1606] Provides an interface for entering user information

[1607] The terminal provides an interface for the user (customer support representative) to input visitor information, which is displayed as a form for entering information such as the user's name, inquiry details, and past purchase history.

[1608] Input: None (launches user interface)

[1609] Output: Interface screen display

[1610] Step 4:

[1611] Send the entered information to the server

[1612] Once the user has entered the visitor information into the interface, the device sends the information to the server. This data transmission is done using an HTTP request.

[1613] Input: Visitor information (user input)

[1614] Output: Send data to the server (HTTP request)

[1615] Step 5:

[1616] Generate initial questions

[1617] The server analyzes the received visitor information and applies machine learning models to generate the most appropriate initial questions. Natural language processing technology selects appropriate questions based on actual visitor information.

[1618] Input: Visitor information (received by server)

[1619] Output: Initial question

[1620] Step 6:

[1621] Display the generated initial questions to the user

[1622] The terminal displays the initial question retrieved from the server to the customer support representative, and the user starts a dialogue with the visitor based on this question.

[1623] Input: Initial question (sent by server)

[1624] Output: Question displayed on the screen

[1625] Step 7:

[1626] Enter the visitor's answer and send it to the server

[1627] The user enters the answers received from the visitors into the terminal and sends the answer data to the server. This data transmission is also done using an HTTP request.

[1628] Input: Visitor's answer (user input)

[1629] Output: Send data to the server (HTTP request)

[1630] Step 8:

[1631] Generate the next question

[1632] The server analyzes the visitor's answers and generates the next question, again dynamically using a machine learning model.

[1633] Input: Visitor's response (received by the server)

[1634] Output: Next question

[1635] Step 9:

[1636] Show the next question to the user

[1637] The terminal displays the next question retrieved from the server to the customer support representative, allowing the user to pose the next question to the visitor.

[1638] Input: Next question (sent by server)

[1639] Output: Question displayed on the screen

[1640] Prompt Sentence Examples

[1641] 1. "Please tell us about the features of the products you recently released."

[1642] 2. "How have you improved after receiving customer feedback?"

[1643] 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.

[1644] The following describes an embodiment of the present invention. The following describes an interview scene as an example, but please note that the present invention can also be applied to a financial report scene using similar methods. The present invention also incorporates an emotion engine.

[1645] server

[1646] 1. A way to load past interview case data from a database

[1647] The server loads past interview case data from a database, which includes questions asked in interviews and their answers. By retrieving data from the database, the server builds a training dataset.

[1648] 2. Means of training machine learning models

[1649] Based on the loaded interview case data, the server trains a machine learning model. Specifically, it uses natural language processing technology to learn patterns of past questions and answers and create a model that generates new questions. This model includes a question generation algorithm and can generate appropriate questions based on the candidate's information.

[1650] 3. A means of analyzing user-entered data and generating initial questions

[1651] When the user (interviewer) enters candidate information and sends it to the server, the server analyzes the data and generates initial questions, where an algorithm is applied to select the most appropriate initial questions based on the information entered by the user.

[1652] Terminal

[1653] 1. A means of providing an interface for users to input candidate information

[1654] The terminal provides an interface for the user to input candidate information, specifically, a form for inputting name, career history, reason for applying, etc.

[1655] 2. Means of sending the entered data to the server

[1656] When the user (interviewer) has completed the input, the data is sent to the server using a normal HTTP request.

[1657] 3. A means to display the generated initial questions to the user

[1658] When the terminal receives the generated initial questions from the server, it displays them to the user (interviewer). Once the questions are displayed, the user can start asking questions to the candidate.

[1659] 4. A means for users to input candidate answers and send them to the server

[1660] After the user (interviewer) receives the answer from the candidate, he / she enters the answer into the terminal and sends it to the server. This procedure is performed using an HTTP request, just like the initial question.

[1661] Emotion Engine

[1662] 1. A means of recognizing emotions by analyzing the user's voice and facial expressions

[1663] The emotion engine analyzes voice and facial expression data provided by the device, allowing it to recognize the user's emotions (e.g., joy, anger, sadness, surprise, etc.).

[1664] 2. A way to generate questions based on perceived sentiment

[1665] The recognized emotion data is sent to the server, which then generates questions based on the emotion information. For example, if the user indicates a high stress level, the system generates questions to encourage relaxation.

[1666] User

[1667] 1. Enter candidate information

[1668] The user (interviewer) enters candidate information, such as the candidate's name, highest level of education, work history, and reason for applying, according to the provided interface.

[1669] 2. Asking initial questions and getting answers

[1670] The user (interviewer) views the initial questions generated by the server, asks them to the candidate, and receives answers.

[1671] 3. Enter and submit your answers

[1672] After receiving the candidate's answer, the answer is entered into the terminal and sent to the server to get the next question to ask.

[1673] 4. Emotion Recognition by Emotion Engine

[1674] During the interview, the device captures the user's voice and facial expressions, which are then analyzed by an emotion engine. The analysis results are sent to the server and used to generate the next questions.

[1675] Specific examples

[1676] For example, consider an initial interview question generated as "Tell me about a recent project." The interviewer asks this question to a candidate, who responds, "I worked on project X, which resulted in Y results." When this response is entered into a device and sent to a server, the server analyzes it and generates the next question: "What was your specific role in that project?"

[1677] During this process, if the emotion engine detects the interviewer's stress level, for example if it is determined to be high, the server can take this into account and generate relaxing questions such as, "What was the most rewarding moment in that project?"

[1678] In this way, users can easily generate appropriate questions based on the candidate's information, and the dialogue can proceed smoothly. This system improves the quality of interviews and also enables flexible responses according to the user's emotional state.

[1679] The processing flow will be explained below.

[1680] Step 1:

[1681] The server loads past interview case data from the database by establishing a database connection and executing SQL queries to retrieve the required data.

[1682] Step 2:

[1683] The server trains a machine learning model based on the loaded interview case data. Specifically, it performs data preprocessing and trains the model using natural language processing techniques.

[1684] Step 3:

[1685] The terminal provides the user (interviewer) with an interface for inputting candidate information. The user inputs the candidate's name, career history, reason for applying, etc.

[1686] Step 4:

[1687] Once the user has completed the input, the device sends the data as an HTTP request to the server, which receives the request and analyzes the candidate information.

[1688] Step 5:

[1689] The server generates initial questions based on the analysis results, uses a trained machine learning model to select questions that best fit the candidate information, and sends the generated questions to the device.

[1690] Step 6:

[1691] The terminal displays the initial questions received from the server to the user, who then verbally asks the candidate the questions.

[1692] Step 7:

[1693] The emotion engine analyzes the user's voice and facial expressions in real time to evaluate the user's emotional state, and the evaluation results are sent to the server.

[1694] Step 8:

[1695] The user receives answers from the candidates, enters these answers into the terminal, and sends them back to the server.

[1696] Step 9:

[1697] The server analyzes the received answer and also uses the emotion data from the emotion engine for analysis. The analysis algorithm takes into account the important information in the answer and the user's emotion to generate the next question.

[1698] Step 10:

[1699] The server sends the next generated question to the terminal, which then displays the received question to the user (interviewer).

[1700] Step 11:

[1701] The user asks the candidate the following questions, and this process is repeated until the interview is complete, improving the quality of the interview and facilitating a smooth dialogue.

[1702] Through this series of processing steps, the present invention provides users with dynamic and interactive question generation, improving the efficiency and quality of interviews. In addition, by combining it with an emotion engine, flexible responses according to the user's emotional state become possible.

[1703] Example 2

[1704] 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."

[1705] In traditional situations such as interviews and financial reports, efficient and effective question generation is required, but in the process, it has been difficult to generate questions that take human emotions into account and maintain a continuity of appropriate questions. In particular, the lack of a dynamic question generation method that takes into account the emotional state of the interviewer or reporter can lead to a decline in the quality of interviews and reports.

[1706] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1707] In this invention, the server includes means for loading past interview case data from a database, means for training a machine learning model based on the past interview case data, means for providing an interface for a user to input candidate information, means for transmitting the input candidate information to the server, means for analyzing the candidate information and generating initial questions, means for displaying the generated initial questions to the user, means for a user to input candidate answers and transmitting the answers to the server, means for analyzing the candidate answers and generating subsequent questions, means for analyzing the user's voice and facial expressions to recognize emotions, and means for generating questions based on the recognized emotions. This makes it possible to generate appropriate and consecutive questions while taking into account the user's emotional state during interviews and reporting situations.

[1708] A "database" is a collection of information that efficiently manages past interview case data, financial report data, etc., and can be searched and retrieved as needed.

[1709] A "machine learning model" is an algorithm that learns from past data, discovers new patterns and relationships, and automates specific tasks such as question generation.

[1710] An "interface" is a user interface component that allows the user to interact with the system and input candidate information, financial report details, and the like.

[1711] "Server" means a computer system that receives information sent by users and processes and analyzes the data.

[1712] "Analysis" is the process of examining input data in detail to understand its structure and meaning.

[1713] An "initial question" is a question that is presented to a user when they first begin interacting with the system.

[1714] The "emotion engine" is a system that analyzes the user's voice and facial expression data to recognize their emotional state at that time.

[1715] "Dynamic question generation" is the process of generating new questions in real time based on the current user situation and answers.

[1716] "Interview case data" is data on questions used in past interviews and the answers to those questions.

[1717] "Recognized emotion" refers to the emotional state of the user that has been analyzed and identified by the emotion engine.

[1718] The present invention provides a method for generating questions efficiently and effectively in interactive systems for interviews, financial reporting, etc. While the present invention will be described in detail below using an interview scene as an example, it can also be applied to financial reporting scenes using similar means.

[1719] Hardware and software used

[1720] Server: A high-performance computer system that processes and analyzes data and trains machine learning models.

[1721] Database: A management system for storing interview case data and financial report data. Example: MySQL.

[1722] Natural Language Processing (NLP) Library: A software library for using machine learning models. For example, "Transformers".

[1723] Emotion engine: A system that analyzes the user's voice and facial expressions. Example: "Microsoft Azure Emotion API."

[1724] Terminal: A device that provides an interface for users to input information. Examples include PCs and tablets.

[1725] Data Processing and Data Calculation

[1726] The server retrieves past interview case data from a database, for example using SQL queries. This data includes interview questions and their answers.

[1727] The server then uses this data to train a machine learning model, using natural language processing techniques to learn patterns of past questions and answers. Specifically, the Transformers library is used. Through this training, a model is built that can generate new questions.

[1728] When a user enters candidate information through a terminal, the information is sent to the server as an HTTP request. The server analyzes the information and generates initial questions. For example, if a user enters information about a candidate's recent projects, the server generates an initial question such as "Tell me about your recent projects."

[1729] The terminal displays the initial questions generated by the server to the user. When the user enters the candidate's answers, the answers are also sent to the server. The server analyzes the answers, generates the next appropriate questions, and sends them to the terminal.

[1730] The emotion engine analyzes the user's voice and facial expression data to recognize their emotions. This recognized emotion is sent to the server, which then generates questions based on the emotion data. For example, if a user is feeling highly stressed, the engine can generate questions such as, "What was the most rewarding moment in that project?"

[1731] Specific examples

[1732] For example, an initial interview question might be generated as "Tell me about a recent project." The interviewer asks this question to a candidate, who responds, "I worked on project X, which resulted in outcome Y." When this response is typed into a device and sent to a server, the server analyzes it and generates the next question: "What was your specific role in that project?"

[1733] As an example prompt, parse the candidate's response to the question "Tell me about a recent project": "I worked on project X, which resulted in outcome Y." Generate the following question:

[1734] The above is a specific embodiment for carrying out the invention. This invention improves the quality of interviews and reports, and enables smooth progress of dialogue.

[1735] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1736] Step 1:

[1737] The server loads past interview case data from a database.

[1738] Specifically, it uses an SQL query to execute a command like "SELECT FROM interview_data" to retrieve a dataset containing interview questions and their answers.

[1739] Input: Database query

[1740] Output: Dataset of past interview cases

[1741] Step 2:

[1742] The server uses the loaded interview case data to train a machine learning model.

[1743] Specifically, we use the natural language processing library "Transformers" to learn question and answer patterns, and build a question generation algorithm using models such as BERT and GPT.

[1744] Input: Past interview case dataset

[1745] Output: A trained machine learning model

[1746] Step 3:

[1747] The user inputs candidate information through the terminal.

[1748] Specifically, the process involves using a web form or application to enter information such as the candidate's name, background, and reasons for applying.

[1749] Input: Candidate information (name, career history, reason for applying, etc.)

[1750] Output: Candidate information data

[1751] Step 4:

[1752] The terminal transmits the input candidate information to the server as an HTTP request.

[1753] Specifically, a POST request is used to send data to the server. For example, an HTTP client library is used.

[1754] Input: Candidate information data

[1755] Output: Candidate information sent to the server

[1756] Step 5:

[1757] The server analyzes the received candidate information and generates initial questions.

[1758] Specifically, it analyzes candidate information using the TF-IDF algorithm and other natural language processing techniques to select the most appropriate initial questions.

[1759] Input: Candidate information data

[1760] Output: Initial question (e.g. "Tell me about your recent projects")

[1761] Step 6:

[1762] The terminal displays the initial question received from the server to the user.

[1763] Specifically, the question is displayed on the user interface using a web browser or application, and a text field for displaying the question is displayed.

[1764] Input: Initial question

[1765] Output: The initial question that is displayed to the user

[1766] Step 7:

[1767] The user inputs the answers from the candidates.

[1768] Specifically, the answer obtained is input via the terminal, for example, by using a keyboard to enter the answer in a text field.

[1769] Input: Candidate's answer

[1770] Output: Candidate's answers entered

[1771] Step 8:

[1772] The terminal transmits the input answers of the candidates to the server.

[1773] Specifically, the input data is sent to the server again via an HTTP request.

[1774] Input: Candidate's answers entered

[1775] Output: Candidate's answers sent to the server

[1776] Step 9:

[1777] The server analyzes the received answer and generates the next question.

[1778] Specifically, it uses a trained machine learning model to generate the next appropriate question, such as "What was your specific role in that project?"

[1779] Input: Candidate's answer

[1780] Output: Next question

[1781] Step 10:

[1782] The terminal displays the next question received from the server to the user.

[1783] Specifically, the new question is displayed in the user interface, such as a web browser or application.

[1784] Input: Next question

[1785] Output: The next question that is displayed to the user

[1786] Step 11:

[1787] The emotion engine analyzes the user's voice and facial expression data to recognize emotions.

[1788] Specifically, it collects audio and video data and uses analysis engines such as the Microsoft Azure Emotion API to recognize emotions.

[1789] Input: Voice and facial expression data

[1790] Output: Recognized emotion data

[1791] Step 12:

[1792] The server generates a question based on the recognized emotion data.

[1793] Specifically, the server generates relaxing questions based on emotional data. For example, in a stressful situation, it generates a question such as, "What was the most rewarding moment in that project?"

[1794] Input: Recognized emotion data

[1795] Output: Sentiment-based questions

[1796] (Application example 2)

[1797] 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."

[1798] The goal is to provide a comfortable riding experience by realizing interactive communication with passengers in autonomous vehicles and generating appropriate questions according to the passenger's emotional state. Conventional systems can only have fixed dialogues with passengers, making it difficult to respond flexibly taking into account their emotional state.

[1799] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1800] In this invention, the server includes means for loading past dialogue example data from a database, means for training a machine learning model based on the past dialogue example data, means for providing an interface for a user to input passenger information, means for transmitting the input passenger information to the server, means for analyzing the passenger information and generating initial questions, means for displaying the generated initial questions to the user, means for the user to input passenger answers and transmitting the same to the server, means for analyzing the passenger answers and generating subsequent questions, means for analyzing voice data and facial expression data to recognize emotions, and means for generating questions based on the recognized emotions. This enables interactive communication with passengers in an autonomous vehicle and realizes flexible question generation based on the passenger's emotional state.

[1801] A "database" is a system that stores various information, such as past dialogue case data and financial report data, and allows it to be searched and read.

[1802] A "machine learning model" is an algorithm that learns patterns from data and makes appropriate decisions and predictions about new data.

[1803] An "interface" is a screen or input form through which a user enters information or interacts with a system.

[1804] A "server" is a computer system that processes and stores various data, and is a device that sends and receives data in response to requests from clients.

[1805] "Analysis" refers to the process of examining input data in detail to understand its characteristics and structure.

[1806] An "initial question" is a question that is initially generated based on information entered by the user.

[1807] "Voice data" refers to audio information recorded by a microphone or other device, which can be used for voice recognition and emotion recognition.

[1808] "Facial expression data" refers to facial expression information captured by a camera or other device, and emotions can be recognized based on this information.

[1809] An "emotion engine" is software or algorithm that analyzes a user's emotional state from their voice and facial expressions.

[1810] To implement this invention, it is first necessary to properly link the main components, such as the database, machine learning model, interface, server, and emotion engine. The specific steps are described below.

[1811] server

[1812] The server loads past dialogue case data from a database, which contains a variety of dialogue case data, including interview scenes. The server uses this data to train a machine learning model. This model incorporates natural language processing techniques and includes an algorithm for generating new questions. It is recommended to use a Python-based natural language processing library (e.g., NLTK, spaCy) or a machine learning framework (e.g., TensorFlow, PyTorch).

[1813] The server receives passenger information entered by the user and analyzes it to generate an initial question. Information from the user is usually received via an HTTP request, and the server analyzes it using an appropriate algorithm. The server then receives the passenger's answers entered by the user and analyzes them to generate the next question. During this process, an emotion engine analyzes voice and facial expression data to recognize emotions, and this data is also used to generate questions.

[1814] Terminal

[1815] The terminal provides an interface for users to enter passenger information. This interface includes text boxes and pull-down menus, allowing users to easily enter the necessary information. For example, passengers can enter their name, past ride experiences, and favorite music genres. Once the information is complete, the terminal sends the data to the server.

[1816] Next, when an initial question is generated from the server, it is displayed on the terminal. The user sees this question, asks it to the passenger, and the passenger enters their answer. The entered answer is also sent to the server, and the next question is generated in the same process. In this way, interactive communication is realized.

[1817] Emotion Engine

[1818] The emotion engine analyzes passengers' voice and facial expression data to recognize their emotions. This is done using a speech recognition library (e.g., Google Speech-to-Text API) and a facial expression recognition library (e.g., OpenCV, Dlib). The recognized emotion information is sent to the server and taken into consideration when generating the next question.

[1819] Specific examples

[1820] For example, when a passenger boards a vehicle, the system asks, "Tell me about your recent riding experience." If the passenger answers, "It was a very comfortable and pleasant ride," the emotion engine detects the passenger's positive emotion, and the server generates subsequent questions such as, "What did you particularly like?" or "Would you like to enjoy music today?"

[1821] Prompt Sentence Examples

[1822] "If a passenger responds positively about their past ride experience, what next questions can you ask based on that positive sentiment? For example, if a passenger responds, 'It was a very comfortable and pleasant ride,' how can you best continue?"

[1823] In this way, the system can enable interactive communication with passengers inside autonomous vehicles, providing them with a comfortable and personalized riding experience.

[1824] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1825] Step 1:

[1826] The terminal provides an interface for users to input passenger information (e.g., name, past ride experiences, favorite music genres), which is then sent to the server via an HTTP request. The input here is passenger information, and the output is sent to the server.

[1827] Step 2:

[1828] The server loads past dialogue case data from a database. The loaded data is a dataset containing past interviews and other dialogue scenes, and trains a machine learning model based on this data. The input is the dialogue case data from the database, and the output is the trained machine learning model. This training uses natural language processing techniques and machine learning frameworks (e.g., TensorFlow, PyTorch).

[1829] Step 3:

[1830] The server analyzes the passenger information submitted by the user and generates an initial question. The input is the passenger information from the user, and the output is the initial question. This process uses a text analysis algorithm. For example, if the user enters positive information about their "recent ride experience," the server generates an initial question based on that information: "What did you particularly like?"

[1831] Step 4:

[1832] The terminal displays the initial question received from the server to the user. Based on the displayed initial question, the user asks the question to the passenger and receives an answer. The input is the initial question from the server, and the output is the display to the user.

[1833] Step 5:

[1834] The user inputs the passenger's answers into the terminal and sends the data to the server via an HTTP request. The input is the passenger's answers and the output is the transmission to the server.

[1835] Step 6:

[1836] The server analyzes the passenger's response and generates the next question. The input is the passenger's response from the user, and the output is the next question. The analysis uses a trained machine learning model and an emotion engine. Specifically, the next question is generated by combining emotional information extracted from voice data and facial expression data.

[1837] Step 7:

[1838] The emotion engine analyzes passengers' voice and facial expression data to recognize their emotions. The input is voice data and facial expression data, and the output is the recognized emotion information. This is done using libraries such as the Google Speech-to-Text API, OpenCV, and Dlib.

[1839] Step 8:

[1840] The server generates the next question based on the recognized emotion information. The input is the emotion information from the emotion engine and the passenger's answer, and the output is a flexibly generated next question. For example, if a passenger answers, "It was a very comfortable and pleasant ride," and a positive emotion is recognized, the next question generated will be, "What did you particularly like?"

[1841] By repeating the above steps, interactive communication with passengers inside the self-driving vehicle will proceed smoothly.

[1842] 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.

[1843] 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.

[1844] 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.

[1845] 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.

[1846] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1847] 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.

[1848] 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).

[1849] 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.

[1850] 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."

[1851] 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.

[1852] 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).

[1853] 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.

[1854] 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.

[1855] 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.

[1856] 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.

[1857] The hardware resource for executing a specific process can be any of the following types of processors: A processor, for example, 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 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.

[1858] 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.

[1859] 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.

[1860] 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.

[1861] 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.

[1862] 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.

[1863] The following is further disclosed regarding the above embodiment.

[1864] (Claim 1)

[1865] A means of loading past interview case data from a database;

[1866] means for training a machine learning model based on the past interview case data;

[1867] means for providing an interface for a user to input candidate information;

[1868] means for transmitting the input candidate information to a server;

[1869] means for analyzing the candidate information and generating initial questions;

[1870] means for displaying the generated initial question to a user;

[1871] a means for a user to input and transmit candidate answers to a server;

[1872] means for analyzing the candidate's answer and generating a next question;

[1873] A system including:

[1874] (Claim 2)

[1875] a means for loading historical financial report data from a database;

[1876] means for training a machine learning model based on the historical financial report data;

[1877] means for providing an interface for a user to input financial report content;

[1878] means for transmitting the input financial statement details to a server;

[1879] means for analyzing the financial report content and generating an initial question;

[1880] means for displaying the generated initial question to a user;

[1881] A means for a user to input responses to the financial report and transmit the responses to the server;

[1882] means for analyzing the answer and generating a next question;

[1883] 10. The system of claim 1, comprising:

[1884] (Claim 3)

[1885] a means for dynamically generating the next question by analyzing the current question and answer when the candidate information and financial report contents are input;

[1886] A means for providing an appropriate question based on the analyzed data in response to a request from the user to easily ask the next question;

[1887] means for displaying the appropriate questions in a timely and interactive manner on a user interface;

[1888] 10. The system of claim 1, comprising:

[1889] "Example 1"

[1890] (Claim 1)

[1891] A means of loading past interview case data from a database;

[1892] means for training a machine learning model based on the past interview case data;

[1893] means for providing an interface for a user to input candidate information;

[1894] means for transmitting the input candidate information to a server;

[1895] means for analyzing the candidate information and generating initial questions;

[1896] means for displaying the generated initial question to a user;

[1897] a means for a user to input and transmit candidate answers to a server;

[1898] means for analyzing the candidate's answer and generating a next question;

[1899] A system including:

[1900] (Claim 2)

[1901] a means of loading historical financial report data from a database;

[1902] means for training a machine learning model based on the historical financial report data;

[1903] means for providing an interface for a user to input financial report content;

[1904] means for transmitting the input financial statement details to a server;

[1905] means for analyzing the financial report content and generating an initial question;

[1906] means for displaying the generated initial question to a user;

[1907] A means for a user to input responses to the financial report and transmit the responses to the server;

[1908] means for analyzing the answer and generating a next question;

[1909] 10. The system of claim 1, comprising:

[1910] (Claim 3)

[1911] a means for dynamically generating the next question by analyzing the current question and answer when the candidate information and financial report contents are input;

[1912] A means for providing an appropriate question based on the analyzed data in response to a request from the user to easily ask the next question;

[1913] means for displaying the appropriate questions in a timely and interactive manner on a user interface;

[1914] 10. The system of claim 1, comprising:

[1915] "Application Example 1"

[1916] (Claim 1)

[1917] a means of loading historical case data from a database;

[1918] means for training a machine learning model based on the past case data;

[1919] means for providing an interface for a user to input information;

[1920] means for transmitting the input information to a server;

[1921] means for analyzing said information and generating an initial question;

[1922] means for displaying the generated initial question to a user;

[1923] A means for the user to input and transmit answers to the server;

[1924] means for analyzing the answer and generating a next question;

[1925] A means to answer visitors' questions about products and services through virtual interaction;

[1926] A means for functioning as an application installed on a smart device;

[1927] means for dynamically generating and interactively displaying questions through a user interface;

[1928] A system including:

[1929] (Claim 2)

[1930] a means for loading historical reporting data from a database;

[1931] means for training a machine learning model based on the historical report data;

[1932] means for providing an interface for a user to input report content;

[1933] means for transmitting the input report content to a server;

[1934] means for analyzing the report content and generating an initial question;

[1935] means for displaying the generated initial question to a user;

[1936] a means for the user to input a response to the report and transmit it to the server;

[1937] means for analyzing the answer and generating a next question;

[1938] 10. The system of claim 1, comprising:

[1939] (Claim 3)

[1940] means for dynamically generating a next question by analyzing the current question and answer when the information or report content is input;

[1941] A means for providing an appropriate question based on the analyzed data in response to a request from the user to easily ask the next question;

[1942] means for displaying the appropriate questions in a timely and interactive manner on a user interface;

[1943] 10. The system of claim 1, comprising:

[1944] "Example 2: Combining Emotion Engines"

[1945] (Claim 1)

[1946] A means of loading past interview case data from a database;

[1947] means for training a machine learning model based on the past interview case data;

[1948] means for providing an interface for a user to input candidate information;

[1949] means for transmitting the input candidate information to a server;

[1950] means for analyzing the candidate information and generating initial questions;

[1951] means for displaying the generated initial question to a user;

[1952] a means for a user to input and transmit candidate answers to a server;

[1953] means for analyzing the candidate's answer and generating a next question;

[1954] A means for recognizing emotions by analyzing the user's voice and facial expressions;

[1955] means for generating a question based on the recognized emotion;

[1956] A system including:

[1957] (Claim 2)

[1958] a means for loading historical financial report data from a database;

[1959] means for training a machine learning model based on the historical financial report data;

[1960] means for providing an interface for a user to input financial report content;

[1961] means for transmitting the input financial statement details to a server;

[1962] means for analyzing the financial report content and generating an initial question;

[1963] means for displaying the generated initial question to a user;

[1964] A means for a user to input responses to the financial report and transmit the responses to the server;

[1965] means for analyzing the answer and generating a next question;

[1966] means for recognizing emotions by analyzing the user's voice and facial expressions;

[1967] means for generating a question based on the recognized emotion;

[1968] 10. The system of claim 1, comprising:

[1969] (Claim 3)

[1970] a means for dynamically generating the next question by analyzing the current question and answer when the candidate information and financial report contents are input;

[1971] A means for providing an appropriate question based on the analyzed data in response to a request from the user to easily ask the next question;

[1972] means for displaying the appropriate questions in a timely and interactive manner on a user interface;

[1973] means for adjusting the next question based on the user's emotional state;

[1974] 10. The system of claim 1, comprising:

[1975] "Application example 2 when combining emotion engines"

[1976] (Claim 1)

[1977] a means for loading past dialogue case data from a database;

[1978] means for training a machine learning model based on the past dialogue example data;

[1979] means for providing an interface for a user to input passenger information;

[1980] means for transmitting the input passenger information to a server;

[1981] means for analyzing the passenger information and generating an initial question;

[1982] means for displaying the generated initial question to a user;

[1983] means for a user to input passenger responses and transmit them to a server;

[1984] means for analyzing the passenger's response and generating a next question;

[1985] A means for recognizing emotions by analyzing voice data and facial expression data;

[1986] a means for generating questions based on the recognized sentiment;

[1987] A system including:

[1988] (Claim 2)

[1989] a means for loading historical financial report data from a database;

[1990] means for training a machine learning model based on the historical financial report data;

[1991] means for providing an interface for a user to input financial report content;

[1992] means for transmitting the input financial statement details to a server;

[1993] means for analyzing the financial report content and generating an initial question;

[1994] means for displaying the generated initial question to a user;

[1995] A means for a user to input responses to the financial report and transmit the responses to the server;

[1996] means for analyzing the answer and generating a next question;

[1997] 10. The system of claim 1, comprising:

[1998] (Claim 3)

[1999] a means for dynamically generating the next question by analyzing the current question and answer when the candidate information and financial report contents are input;

[2000] A means for providing an appropriate question based on the analyzed data in response to a request from the user to easily ask the next question;

[2001] means for displaying the appropriate questions in a timely and interactive manner on a user interface;

[2002] 10. The system of claim 1, comprising: [Explanation of symbols]

[2003] 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 loading past interview case data from a database; means for training a machine learning model based on the past interview case data; means for providing an interface for a user to input candidate information; means for transmitting the input candidate information to a server; means for analyzing the candidate information and generating initial questions; means for displaying the generated initial question to a user; a means for a user to input and transmit candidate answers to a server; means for analyzing the candidate's answer and generating a next question; A system including:

2. a means for loading historical financial report data from a database; means for training a machine learning model based on the historical financial report data; means for providing an interface for a user to input financial report content; means for transmitting the input financial statement details to a server; means for analyzing the financial report content and generating an initial question; means for displaying the generated initial question to a user; A means for a user to input responses to the financial report and transmit the responses to the server; means for analyzing the answer and generating a next question; The system of claim 1 , comprising:

3. a means for dynamically generating the next question by analyzing the current question and answer when the candidate information and financial report contents are input; A means for providing an appropriate question based on the analyzed data in response to a request from the user to easily ask the next question; means for displaying the appropriate questions in a timely and interactive manner on a user interface; The system of claim 1 , comprising:

Images