System

Abstract

A system is provided.SOLUTION: A system comprising: means for generating a question for a user using a generative model; means for analyzing an answer from the user; means for generating a next question based on the analyzed answer; means for recording the generated question and the analyzed answer; and means for analyzing the recorded data to generate a report.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Traditional user interview methods have various issues, such as inconsistent information due to dependence on the interviewer's skills, time constraints, rising costs, and inefficient analysis. In particular, it is difficult to balance the quality of the interviews with costs, making it difficult to collect accurate data that companies need to make effective decisions. [Means for solving the problem]

[0005] To solve these problems, the present invention provides a system that uses a generative model to generate questions for a user, analyzes the user's answers, and generates subsequent questions based on the analysis results. This system includes a means for recording the generated questions and analyzed answers, and a means for analyzing the recorded data to generate a report. Furthermore, the system improves the quality and efficiency of interviews and reduces costs by using a means for generating initial questions based on user profile information and a generative model that uses natural language processing technology.

[0006] A "generative model" is an algorithm that automatically creates new data and information based on past data.

[0007] A "means for generating questions" refers to a system or algorithm that automatically generates appropriate questions for users.

[0008] "Means for analyzing responses" refers to systems or algorithms that analyze responses obtained from users and understand, classify, and evaluate their content.

[0009] "Means of recording" refers to a database or storage system for saving and managing generated questions and user responses.

[0010] "Means for analysis and report generation" refers to the systems and algorithms used to evaluate and analyze the recorded data and create a report summarizing the results.

[0011] "Profile information" refers to attribute information such as a user's personal information and interests.

[0012] "Natural language processing technology" refers to the technology that allows computers to understand, analyze, and generate human language. [Brief explanation of the drawings]

[0013] [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

[0014] 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.

[0015] First, the terms used in the following description will be explained.

[0016] 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).

[0017] 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.

[0018] 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.

[0019] 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.

[0020] 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."

[0021] [First embodiment]

[0022] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0023] 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.

[0024] 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).

[0025] 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.

[0026] 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.

[0027] 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.

[0028] 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.

[0029] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0030] 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.

[0031] 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.

[0032] 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.

[0033] 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."

[0034] This invention relates to a user interview system that uses a generative model. This system generates questions through dialogue with the user, analyzes the answers, and generates subsequent questions based on the results, enabling efficient and accurate data collection. Specific program processing is explained in natural language and includes examples.

[0035] Program processing description

[0036] 1. Preparation Phase

[0037] Data collection

[0038] The server collects data sets for use in the interviews, including past interview transcripts, questionnaires, and user profile information.

[0039] Model Training

[0040] The server trains the generative model based on the collected data, using natural language processing technology to build algorithms for question generation and answer analysis.

[0041] 2. Interview begins

[0042] Interview begins

[0043] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[0044] User Registration

[0045] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[0046] The user enters the required information into the form and presses the submit button.

[0047] The server stores the submitted user information in a database.

[0048] Generating initial questions

[0049] The server generates an initial question based on the user information and sends it to the terminal.

[0050] The terminal displays the question received from the server.

[0051] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[0052] 3. Question and answer repetition

[0053] Questions and Answers

[0054] The user enters an answer and presses the send button.

[0055] The server uses the generative model to analyze the user's answers and generate the next question.

[0056] The server generates a new question and sends it to the device.

[0057] For example: "Tell me more about the design. What did you particularly like about it?"

[0058] The terminal displays the new question to the user.

[0059] The user again enters the answer and presses the send button.

[0060] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[0061] 4. Data analysis phase

[0062] Data storage and analysis

[0063] The server stores all the data from the interview.

[0064] The server analyzes the stored data using a generative model and extracts various indicators (e.g., frequently occurring keywords, sentiment analysis, etc.).

[0065] Report Generation

[0066] The server compiles the analysis results in the form of a report.

[0067] The server generates the report and provides it to the company via a download link or similar.

[0068] Specific examples

[0069] Interview example

[0070] User registration and initial inquiries

[0071] 1. The user accesses the system using a terminal and participates in the interview.

[0072] "User Registration" will appear on the screen and the user will enter the required information (name, age, interests, etc.).

[0073] 2. The server generates initial questions based on the user's profile.

[0074] You might see an initial question like, "Hello, [user's name]. Today I'd like to talk to you about a new product. First, what are your impressions of this product?"

[0075] Dialogue Progress

[0076] 3. The user enters the answer

[0077] User: "I like the design, but the functionality is a bit unclear."

[0078] 4. The server analyzes the user's answer and generates the next question

[0079] The next question will be displayed, such as "Tell us more about the design. What did you particularly like about it?"

[0080] 5. The user enters the answer again

[0081] User: "I especially like the grip and the color combination."

[0082] Data Analysis and Report Generation

[0083] 6. After the interview is over, the server analyzes the collected data.

[0084] The server analyzes the interview transcript, categorizes the user's opinions, and generates a summary.

[0085] 7. The server compiles the analysis results into a report

[0086] Companies are provided with a report containing analysis results such as, "Many users give the design high marks, but feel that further explanation is needed regarding functionality."

[0087] In this way, the entire interview process is automated using generative models, reducing costs and enabling more accurate information collection.

[0088] The processing flow will be explained below.

[0089] Step 1:

[0090] The server collects data sets for use in the interviews, including past interview transcripts, questionnaires, and user profile information.

[0091] Step 2:

[0092] The server trains the generative model based on the collected data, using natural language processing technology to build algorithms for question generation and answer analysis.

[0093] Step 3:

[0094] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[0095] Step 4:

[0096] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[0097] Step 5:

[0098] The user enters the required information into the form and presses the submit button.

[0099] Step 6:

[0100] The server stores the submitted user information in a database.

[0101] Step 7:

[0102] The server generates an initial question based on the user information and sends it to the terminal.

[0103] Step 8:

[0104] The terminal displays the question received from the server.

[0105] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[0106] Step 9:

[0107] The user enters an answer to the question and presses the send button.

[0108] Step 10:

[0109] The server uses the generative model to analyze the user's answers and generate the next question.

[0110] Step 11:

[0111] The server generates a new question and sends it to the device.

[0112] For example: "Tell me more about the design. What did you particularly like about it?"

[0113] Step 12:

[0114] The terminal displays the new question to the user.

[0115] Step 13:

[0116] The user again enters the answer and presses the send button.

[0117] Step 14:

[0118] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[0119] Step 15:

[0120] The server stores all the data from the interview.

[0121] Step 16:

[0122] The server analyzes the stored data using a generative model and extracts various indicators (e.g., frequently occurring keywords, sentiment analysis, etc.).

[0123] Step 17:

[0124] The server compiles the analysis results in the form of a report.

[0125] Step 18:

[0126] The server generates the report and provides it to the company via a download link or similar.

[0127] This automates the entire interview process using generative models, reducing costs and enabling more accurate information collection.

[0128] Example 1

[0129] 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."

[0130] In conventional user interview systems, questions are generated and answers are analyzed manually, resulting in inefficiency and limited accuracy. Data storage and analysis are also time-consuming and costly. Furthermore, there is a lack of tools and methods for effectively utilizing interview data, making comprehensive analysis difficult.

[0131] 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.

[0132] In this invention, the server includes means for generating questions for the user using a generative model, means for analyzing answers from the user, and means for generating the next question based on the analyzed answers. This automates the entire interview process, enabling efficient and accurate data collection.

[0133] A "generative model" is an artificial intelligence model used to generate questions for a user and analyze the user's answers.

[0134] The "means for generating questions" is a function that uses a generative model to automatically create questions appropriate for the user.

[0135] The "means for analyzing answers" is a function that analyzes answers obtained from users using a generative model and extracts information for generating the next question.

[0136] The "means for generating the next question" is a function for generating the next question to be asked based on the analyzed answer.

[0137] The "means for recording questions and analyzed answers" is a function for saving the generated questions, the user's answers, and the analysis results thereof.

[0138] The "means for analyzing recorded data and generating a report" is a function for analyzing saved interview data using a generative model and compiling the results in a report format.

[0139] The "means for generating initial questions" is a function for generating initial questions based on the user's profile information at the initial stage of the interview.

[0140] "Natural language processing technology" is an artificial intelligence technology for understanding and generating language data, and is used in generative model algorithms.

[0141] The present invention relates to a user interview system using a generative model. Specific embodiments of the system will be described below.

[0142] Hardware and Software Use

[0143] The system's main operations are carried out by a server, a device, and a user. The system utilizes a generative model (e.g., GPT-4) running on the server and a device (e.g., a PC or smartphone) accessed by the user. The server also uses software incorporating analysis algorithms using natural language processing techniques. A specific example of such software is the Transformers library from Hugging Face.

[0144] Explanation of program processing

[0145] The operation of the system can be divided into the following main stages: data collection, model training, interview initiation, question and answer cycle, data storage and analysis, and report generation.

[0146] Data collection

[0147] The server collects the datasets required for the interview, including past interview transcripts, questionnaires, and user profile information.

[0148] Model Training

[0149] The server trains the generative AI model based on the collected data. Natural language processing technology is used for training, and algorithms for question generation and answer analysis are built. GPU resources are used to improve processing speed.

[0150] Interview begins

[0151] Users access the system using a terminal and participate in the interview through a web interface. Users click on a URL and connect to the system from their browser.

[0152] The terminal displays a user registration form, and the user enters information such as name, age, interests, etc. The entered information is sent to the server and stored in a database.

[0153] The server generates initial questions based on the user's profile information and sends them to the terminal, which then displays the received questions to the user.

[0154] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[0155] Repeated questions and answers

[0156] The user responds to the questions on the terminal, enters the answer, and presses the send button. The answer is then sent to the server.

[0157] The server uses the generative AI model to analyze the user's answers and generate the next question, which is again sent to the device and displayed to the user.

[0158] For example: "Tell me more about the design. What did you particularly like about it?"

[0159] Data storage and analysis

[0160] The server stores all interview data in a database. After the interview is completed, the stored data is analyzed using a generative model to extract indicators such as frequently occurring keywords and sentiment analysis.

[0161] Report Generation

[0162] The server compiles the analysis results into a report, which is saved in PDF or Excel format and includes a download link for the report to be provided to the company.

[0163] As an example, the following prompt sentence is used:

[0164] User information: Name = Taro, Age = 30, Interests = Technology

[0165] Question: "Hello, Taro. Today I'd like to ask you about a new product. First of all, what are your impressions of this product?"

[0166] As described above, this invention enables efficient and accurate user interviews using generative AI models.

[0167] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0168] Step 1: Data collection

[0169] The server collects a dataset for use in the interviews, which includes past interview transcripts, questionnaires, and user profile information.

[0170] Input: Database information, interview transcripts

[0171] Output: Collected dataset

[0172] What it does: The server accesses public and internal company databases, collects and stores the required data.

[0173] Step 2: Model training

[0174] The server trains the generative AI model based on the collected data. During training, natural language processing technology is used to build algorithms for question generation and answer analysis.

[0175] Input: Collected dataset

[0176] Output: A trained generative AI model

[0177] How it works: The server pre-processes the data and trains a generative AI model using natural language processing libraries (e.g., Transformers from Hugging Face), leveraging GPU resources to speed up training.

[0178] Step 3: System Access

[0179] A user accesses the system using a terminal and clicks on a URL to reach the web interface.

[0180] Input: URL

[0181] Output: Access to the system

[0182] How it works: A user opens a browser and enters the provided URL to log into the system.

[0183] Step 4: User Registration

[0184] The device displays a user registration form, where the user enters information such as name, age, and interests. The information is sent to the server and stored in a database.

[0185] Input: User-entered information (name, age, interests)

[0186] Output: Saved user information

[0187] How it works: A registration form is displayed, the user fills in the fields, presses the submit button, and the data is sent to the server, which receives it and stores it in a database.

[0188] Step 5: Generate initial questions

[0189] The server generates an initial question based on the user information and sends it to the terminal, which then displays the received question to the user.

[0190] Input: User information

[0191] Output: Initial question

[0192] How it works: The server uses the user's profile information to generate prompts and input them into the AI ​​model. The generated initial questions are sent to the device and displayed.

[0193] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[0194] Step 6: Repeat the questions and answers

[0195] The user enters an answer to a question and presses the submit button. The answer is sent to the server and analyzed using the generative AI model. The next question is generated and sent again to the device. The device displays the next question to the user.

[0196] Input: User's answer

[0197] Output: Next question

[0198] How it works: When a user enters an answer and presses the submit button, the answer is sent to the server, which uses a generative AI model to analyze the answer and generate the next question. This question is then sent to the device and displayed to the user.

[0199] For example: "Tell me more about the design. What did you particularly like about it?"

[0200] Step 7: Data storage and analysis

[0201] The server stores all the interview data, which is then analyzed using a generative model to extract metrics such as frequently occurring keywords and sentiment analysis.

[0202] Input: Interview data

[0203] Output: Analysis results (frequent keywords, sentiment analysis, etc.)

[0204] How it works: The server stores all interview data and analyzes it using a generative model. Various indicators are extracted using natural language processing techniques.

[0205] Step 8: Generate reports

[0206] The server compiles the analysis results in a report format, which is saved in PDF or Excel format and a download link is provided.

[0207] Input: Analysis results

[0208] Output: Report (PDF or Excel format)

[0209] How it works: The server documents the analysis results and generates a report, which is then provided to the company as a download link, allowing them to easily retrieve the information they need.

[0210] In this way, by going through each processing step, efficient and accurate user interviews can be conducted using generative AI models.

[0211] (Application example 1)

[0212] 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."

[0213] In conventional user interview systems, the interview process is conducted manually, which is extremely time-consuming and labor-intensive, making it difficult to collect data efficiently. Furthermore, analyzing the collected data is also time-consuming, making it difficult to utilize feedback in real time. Furthermore, there was a lack of an effective way to integrate and analyze feedback from many different customers, making it difficult to extract customer insights.

[0214] 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.

[0215] In this invention, the server includes: means for generating questions for users using a generative model; means for analyzing responses from users; means for generating subsequent questions based on the analyzed responses; means for recording the generated questions and analyzed responses; means for analyzing the recorded data and generating a report; means for users to access the system using a terminal and participate in an interview; means for the terminal to display questions and acquire responses from users; and means for providing a report generated after the interview is completed to an administrator. This automates the entire interview process, enabling efficient and accurate data collection. Furthermore, the collected data is analyzed in real time, and feedback and improvement suggestions are provided immediately based on the results, enabling rapid response.

[0216] A "generative model" is a model that generates new data or information based on existing data, and primarily uses machine learning and artificial intelligence techniques.

[0217] "User" refers to a user who accesses the system and participates in an interview.

[0218] "Means for generating questions" refers to the function of automatically creating questions to be asked during interviews using a generative model.

[0219] "Means for analyzing responses" refers to the function of receiving responses from users, understanding their contents, and analyzing them.

[0220] "Means for generating the next question" refers to a function for generating the next question to be asked based on the previously analyzed answer.

[0221] "Means for recording generated questions and analyzed answers" refers to a function for saving questions generated during an interview, the user's answers, and the analysis results in a database or the like.

[0222] "Means for generating reports" refers to the function of compiling analysis results based on recorded data and automatically creating reports to be provided to managers and personnel in charge.

[0223] "Device" refers to the device used by the user when participating in the interview, including smartphones and tablets.

[0224] "Means of participating in an interview" refers to the process by which a user accesses the system using a terminal, inputs the information required for the interview, and interacts with the system.

[0225] "Means for displaying a question and obtaining a user's response" refers to a function for generating a question to be displayed on a terminal and receiving a response entered by a user in response to that question.

[0226] "Means to provide to the manager" refers to a function for distributing the report generated after the interview to the relevant parties and using it for feedback and consideration of improvement measures.

[0227] This invention relates to a user interview system using generative models. This system improves the efficiency of collecting and analyzing customer feedback in physical stores, and can be used to immediately improve products and services.

[0228] The system mainly includes the following components:

[0229] 1. Generative model: A generative model that uses natural language processing technology and has the ability to automatically generate questions for users.

[0230] 2. Analysis server: It has the means to analyze the answers from the user and generate the next question, as well as the means to record the generated questions and analyzed answers.

[0231] 3. Interview device: A terminal accessed by the user (smartphone, tablet, etc.) that has the function of displaying questions and obtaining responses from the user.

[0232] 4. Database: A storage device for storing analysis results and interview records.

[0233] 5. Report generation tool: Provides a means to automatically generate reports based on recorded data and provide them to managers and personnel in charge.

[0234] System Operation

[0235] 1. Preparation Phase

[0236] The server collects past interview data, questionnaires, user profile information, etc. for use in the interview.

[0237] The server trains a generative model based on the collected data and builds algorithms for question generation and answer analysis, primarily using natural language processing technology.

[0238] 2. Begin the interview

[0239] The user accesses the system using the interview device, for example, by clicking on a URL that takes the user to a web interface.

[0240] The interview device displays a user registration form, and the user enters information such as name, age, and interests. This information is sent to the server and stored in a database.

[0241] The server generates initial questions based on the collected user profile information and sends them to the interview device, such as, "Hello, [user name]. Today I'd like to ask you about a new product. First of all, what is your impression of this product?"

[0242] 3. Question and answer repetition

[0243] The user enters answers into the interview device, which are then sent to the server.

[0244] The server analyzes the answer using the generative model and generates the next question. At this time, the server performs the analysis and generation process in the form of a prompt such as "User's answer: The design is good, but the operation feels a little complicated. Prompt: Please generate the next question based on this answer."

[0245] New questions are generated and sent to the interview device, such as "Tell me more about the operation. What parts did you find particularly difficult to understand?"

[0246] This process repeats until the user responds "Done."

[0247] 4. Data analysis and report generation

[0248] After the interview is completed, the server analyzes all collected data and categorizes and summarizes the user's opinions.

[0249] The server compiles the analysis results in a report and provides it to the administrator or person in charge. For example, the report may include a statement such as, "Many users rate the design highly, but feel that further explanation is needed regarding the functionality."

[0250] Specific examples

[0251] This system can be used to efficiently collect customer feedback about new products in physical stores. For example, suppose you want to obtain feedback about a new microwave. The initial question displayed is, "Hello, [user name]. Today I'd like to ask you about our new microwave. First, what is your impression of this product?" If the user responds, "The design is good, but I found the operation a little complicated," the system generates a next question: "Tell me more about the operation. Which parts did you find particularly difficult to understand?" This allows the system to collect specific user opinions. Finally, a report is generated based on the collected data, enabling efficient product improvements.

[0252] In this way, the present invention uses a generative AI model to improve the efficiency and accuracy of customer interviews in physical stores.

[0253] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0254] Step 1: Preparation Phase

[0255] The server collects a dataset to be used in the interviews. This dataset includes past interview data, a list of questions, and user profile information. The server then trains a generative model based on this data. Specifically, it uses natural language processing technology to build algorithms for question generation and answer analysis.

[0256] Input: Past interview data, questionnaire, user profile information

[0257] Processing: Training the generative model

[0258] Output: A trained generative model

[0259] Step 2: Start the interview

[0260] The user accesses the system using a terminal and participates in the interview. The user clicks on a URL to access the web interface. The terminal displays a user registration form, and the user enters information such as name, age, and interests. The entered information is sent to the server and stored in a database.

[0261] Input: User access and registration information

[0262] Process: Enter and submit user registration information

[0263] Output: User information stored in the database

[0264] Step 3: Generate initial questions

[0265] The server generates initial questions based on the collected user profile information and sends them to the interview device. For example, a question might be, "Hello, [user name]. Today I'd like to ask you about a new product. First of all, what is your impression of this product?"

[0266] Input: User profile information

[0267] Process: Generate initial questions

[0268] Output: Initial questions sent to the interview device

[0269] Step 4: Repeat the questions and answers

[0270] The user enters their answer into the interview device and presses the send button. This is sent to the server. The server uses the generative model to analyze the answer and generate the next question. For example, if the user enters, "User answer: The design is good, but the operation felt a little complicated," the server generates the next question: "Please tell us more about the operation. Which part did you find particularly difficult to understand?"

[0271] Input: User's answer

[0272] Processing: Parsing answers and generating next questions

[0273] Output: The next question sent to the interview device

[0274] Step 5: Record the data

[0275] The server records the questions generated and the analyzed answers, so that the entire interview process is stored in a database for later analysis.

[0276] Input: Generated question, parsed answer

[0277] Process: Recording questions and answers

[0278] Output: Interview data stored in a database

[0279] Step 6: Data analysis and report generation

[0280] After the interviews are completed, the server analyzes all collected data and classifies and summarizes the user opinions. The server then compiles the analysis results into a report and provides it to the manager or person in charge. For example, the report may include statements such as, "Many users rate the design highly, but feel that further explanation is needed regarding the functionality."

[0281] Input: Collected data

[0282] Processing: Data analysis and report generation

[0283] Output: Report

[0284] In this way, a question generation and answer analysis system using generative models enables efficient and accurate interviews and data collection, enabling companies to quickly collect customer feedback and use it to effectively improve their products and services.

[0285] 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.

[0286] This invention relates to a user interview system that combines a generative model and an emotion engine. This system generates questions through dialogue with the user, analyzes the answers, and generates the next question based on the results. It also recognizes the user's emotions and adjusts the dialogue content to obtain more information.

[0287] Program processing description

[0288] 1. Preparation Phase

[0289] Data collection

[0290] The server collects datasets for use in the interviews, including past interview transcripts, questionnaires, user profile information, and sentiment data.

[0291] Model Training

[0292] The server trains the generative model and emotion engine based on the collected data, utilizing natural language processing technology to build algorithms for question generation, response analysis, and emotion analysis.

[0293] 2. Interview begins

[0294] Interview begins

[0295] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[0296] User Registration

[0297] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[0298] The user enters the required information into the form and presses the submit button.

[0299] The server stores the submitted user information in a database.

[0300] Generating initial questions

[0301] The server generates an initial question based on the user information and sends it to the terminal.

[0302] The terminal displays the question received from the server.

[0303] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[0304] 3. Question and answer repetition

[0305] Questions and Answers

[0306] The user enters an answer to the question and presses the send button.

[0307] The server uses a generative model to analyze the user's answers and generate the next question. In addition, it analyzes the user's emotions using an emotion engine and adjusts the next question based on the results.

[0308] Built-in sentiment analysis

[0309] The emotion engine analyzes the user's responses to determine their emotions, and the server adjusts the tone and content of the dialogue based on that emotion information. For example, if the user expresses anxiety, the server changes the questions to be gentler and more reassuring.

[0310] The server generates a new question and sends it to the device.

[0311] For example: "Tell me more about the design. What did you particularly like about it?"

[0312] The terminal displays the new question to the user.

[0313] The user again enters the answer and presses the send button.

[0314] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[0315] 4. Data analysis phase

[0316] Data storage and analysis

[0317] The server stores all the data from the interview.

[0318] The server analyzes the stored data using a generative model and emotion engine to extract various indicators (e.g., frequently occurring keywords, emotion trend analysis, etc.).

[0319] Report Generation

[0320] The server compiles the analysis results in the form of a report.

[0321] The server generates the report and provides it to the company via a download link or similar.

[0322] Specific examples

[0323] Interview example

[0324] User registration and initial inquiries

[0325] 1. The user accesses the system using a terminal and participates in the interview.

[0326] "User Registration" will appear on the screen and the user will enter the required information (name, age, interests, etc.).

[0327] 2. The server generates initial questions based on the user's profile.

[0328] You might see an initial question like, "Hello, [user's name]. Today I'd like to talk to you about a new product. First, what are your impressions of this product?"

[0329] Dialogue Progress

[0330] 3. The user enters the answer

[0331] User: "I like the design, but the functionality is a bit unclear."

[0332] 4. The server analyzes the user's answer and generates the next question

[0333] The next question will be displayed, such as, "Tell us more about the design. What did you like most about it?" In addition, an emotion engine will analyze the emotions from the user's answers and adjust the dialogue accordingly.

[0334] 5. The user enters the answer again

[0335] User: "I especially like the grip and the color combination."

[0336] Data Analysis and Report Generation

[0337] 6. After the interview is over, the server analyzes the collected data.

[0338] The server analyzes the interview transcript, categorizes the user's opinions, and generates a summary.

[0339] 7. The server compiles the analysis results into a report

[0340] Companies are provided with a report containing analysis results such as, "Many users give the design high marks, but feel that further explanation is needed regarding functionality."

[0341] In this way, the entire interview process is automated using generative models and emotion engines, reducing costs and enabling more accurate information collection. Emotion analysis improves the user experience by enabling appropriate responses based on the user's emotional state.

[0342] The processing flow will be explained below.

[0343] Step 1:

[0344] The server collects datasets for use in the interviews, including past interview transcripts, questionnaires, user profile information, and sentiment data.

[0345] Step 2:

[0346] The server trains the generative model and emotion engine based on the collected data, utilizing natural language processing technology to build algorithms for question generation, response analysis, and emotion analysis.

[0347] Step 3:

[0348] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[0349] Step 4:

[0350] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[0351] Step 5:

[0352] The user enters the required information into the form and presses the submit button.

[0353] Step 6:

[0354] The server stores the submitted user information in a database.

[0355] Step 7:

[0356] The server generates an initial question based on the user information and sends it to the terminal.

[0357] Step 8:

[0358] The terminal displays the question received from the server.

[0359] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[0360] Step 9:

[0361] The user enters an answer to the question and presses the send button.

[0362] Step 10:

[0363] The server uses a generative model to analyze the user's answers and generate the next question. In addition, it analyzes the user's emotions using an emotion engine and adjusts the next question based on the results.

[0364] Step 11:

[0365] The emotion engine analyzes the user's responses to determine their emotions, and the server adjusts the tone and content of the dialogue based on that emotion information. For example, if the user expresses anxiety, the server changes the questions to be gentler and more reassuring.

[0366] Step 12:

[0367] The server generates a new question and sends it to the device.

[0368] For example: "Tell me more about the design. What did you particularly like about it?"

[0369] Step 13:

[0370] The terminal displays the new question to the user.

[0371] Step 14:

[0372] The user again enters the answer and presses the send button.

[0373] Step 15:

[0374] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[0375] Step 16:

[0376] The server stores all the data from the interview.

[0377] Step 17:

[0378] The server analyzes the stored data using a generative model and emotion engine to extract various indicators (e.g., frequently occurring keywords, emotion trend analysis, etc.).

[0379] Step 18:

[0380] The server compiles the analysis results in the form of a report.

[0381] Step 19:

[0382] The server generates the report and provides it to the company via a download link or similar.

[0383] This will automate the entire interview process using generative models, reducing costs and enabling more accurate information collection. Furthermore, emotion analysis will enable appropriate responses based on the user's emotional state, improving the user experience.

[0384] Example 2

[0385] 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."

[0386] Conventional interview systems simply analyze user responses and generate the next questions, but are unable to properly consider the user's emotions or the context of the conversation. This results in poor interview quality and limited information being obtained. It is also difficult to improve the user experience, and efficient information gathering is required.

[0387] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for generating questions for the user using a generative model, means for analyzing answers from the user, means for generating the next question based on the analyzed answers, means for recording the generated questions and the analyzed answers, means for analyzing the recorded data and generating a report, means for training the generative model and the emotion analysis engine based on the collected data, and means for analyzing emotions from the user answers and adjusting the content of the dialogue. This makes it possible to conduct high-quality interviews that take user emotions into consideration and to collect detailed information efficiently.

[0388] A "generative model" is a model that uses machine learning algorithms to generate new data or information.

[0389] "Means for generating questions" refers to a function that uses a generative model to create questions for the user.

[0390] "Means for analyzing responses" refers to the technology used to analyze responses from users and understand and evaluate their content.

[0391] The "means for generating the next question" refers to a function for generating the next appropriate question based on the analyzed answer.

[0392] "Means for recording" refers to the technology used to store the generated questions and analyzed answers in a database or storage.

[0393] "Means for generating reports" refers to a function that analyzes recorded data and outputs the results in the form of a report.

[0394] An "emotion analysis engine" refers to an algorithm or software that analyzes the emotions of users' responses and determines their emotional state.

[0395] "Means for adjusting the content of the dialogue" refers to the function for appropriately changing and adjusting the tone and content of the dialogue based on the results of sentiment analysis.

[0396] MODE FOR CARRYING OUT THE INVENTION

[0397] This invention relates to a user interview system that combines a generative model and an emotion analysis engine. This system generates questions through dialogue with the user, analyzes the answers, and generates the next question based on the results. It also recognizes the user's emotions and adjusts the dialogue content to obtain more information.

[0398] The system includes the following components:

[0399] Generative Model

[0400] Sentiment Analysis Engine

[0401] Database

[0402] User Interface (Terminal)

[0403] Hardware and software used

[0404] The system uses the following hardware and software:

[0405] Servers: Used for data processing and model training. Example: Cloud servers (AWS EC2, Google Cloud)

[0406] Database: Used to store user information and interview data. Example: MySQL, MongoDB

[0407] Generative models: used for question generation and answer analysis. Example: GPT-4 (OpenAI)

[0408] Sentiment analysis engine: Used for sentiment analysis. Example: Microsoft Azure's Text Analytics API

[0409] User Interface (Terminal): The interface through which the user interacts with the system. Example: A web interface created with React.

[0410] Specific operation of the system

[0411] 1. Data collection and model training:

[0412] The server collects and stores in a database a dataset for use in the interviews, including past interview transcripts, questionnaires, user profile information, and sentiment data.

[0413] The server trains the generative model and sentiment analysis engine based on the collected data. Natural language processing technology is used for training, and algorithms for question generation, answer analysis, and sentiment analysis are built using Python libraries (TensorFlow, PyTorch).

[0414] 2. Interview begins:

[0415] A user accesses the system using a terminal and accesses the web interface.

[0416] The device displays a user registration form and asks the user to enter information such as name, age, interests, etc. The user enters the required information into the form and presses the submit button.

[0417] The server stores the user information in a database, generates an initial question, and sends it to the terminal.

[0418] The terminal displays the initial question received from the server.

[0419] Examples:

[0420] "Hi, [user name]. Today I'd like to ask you about a new product. First, what are your impressions of this product?"

[0421] 3. Repeated questions and answers:

[0422] The user enters an answer to the question and presses the send button.

[0423] The server uses a generative model to analyze the user's answers and generate the next question. The emotion analysis engine analyzes the user's emotions and adjusts the next question based on the results. For example, if the user expresses anxiety, the question will be changed to a gentler, more reassuring one.

[0424] The server generates a new question and sends it to the terminal. The terminal displays the new question. The user again enters the answer and presses the send button.

[0425] Examples:

[0426] "Tell us more about the design. What parts did you particularly like?"

[0427] 4. Data Storage and Analysis:

[0428] The server stores all the data from the interview.

[0429] The server analyzes the stored data using a generative model and sentiment analysis engine to extract various indicators (e.g., frequently occurring keywords, sentiment trend analysis, etc.).

[0430] The server compiles the analysis results in a report format that is easy to understand visually (e.g., PDF or dashboard).

[0431] 5. Report provision:

[0432] The server provides the generated report to the company via a download link or the like.

[0433] As described above, the system of the present invention uses a generative AI model and an emotion analysis engine to realize high-quality interviews that take into account the user's emotions, enabling efficient and detailed information collection.

[0434] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0435] Step 1:

[0436] Data Collection Phase

[0437] What it does: The server collects a dataset to use in the interviews. This includes past interview transcripts, questionnaires, user profile information, and sentiment data. The server uses an API to download the data from cloud storage and saves it to local storage.

[0438] Input: Interview transcripts in cloud storage, questionnaire, user profile information, and sentiment data.

[0439] Output: The initial dataset saved to local storage.

[0440] Step 2:

[0441] Data Preprocessing

[0442] Specific operation: The server preprocesses the collected data, specifically by using Python libraries (numpy, pandas, NLTK, etc.) to perform data cleaning (noise removal), text tokenization, and stop word removal.

[0443] Input: An initial dataset stored in local storage.

[0444] Output: A preprocessed and clean dataset.

[0445] Step 3:

[0446] Building and training the model

[0447] How it works: The server trains a generative model (e.g., GPT-4) and a sentiment analysis engine based on the preprocessed data. It uses libraries such as TensorFlow and PyTorch. The server exports the trained model to a file and uploads it to cloud storage.

[0448] Input: Preprocessed clean dataset.

[0449] Output: A trained generative model and a sentiment analysis engine.

[0450] Step 4:

[0451] User access and interview start

[0452] Specific operation: A user accesses the system using a terminal and accesses a web interface (e.g., developed with React). The user is redirected to the homepage.

[0453] Input: The user's access request.

[0454] Output: Web interface display.

[0455] Step 5:

[0456] Displaying the user registration form

[0457] Specific operation: The device displays a user registration form. The registration form includes fields for entering name, age, interests, etc. After the user enters the information on the device, the form data is sent to the server.

[0458] Input: User information (name, age, interests).

[0459] Output: The user information sent to the server.

[0460] Step 6:

[0461] Saving user information and generating initial questions

[0462] Specific operation: The server stores user information in a database (e.g., MySQL, MongoDB). It generates initial questions based on the user information and sends them to the device. A generative AI model is used for generation.

[0463] Input: The user information sent to the server.

[0464] Output: User information stored in the database and the initial questions generated.

[0465] Step 7:

[0466] Display initial questions

[0467] Specific behavior: The device displays the initial question received from the server. Example: "Hello, [user name]. Today I'd like to ask you about a new product. First, what are your impressions of this product?"

[0468] Input: The initial question sent by the server.

[0469] Output: The initial question displayed on the user interface.

[0470] Step 8:

[0471] User answer input

[0472] Specific operation: The user enters an answer to a question using the terminal and presses the send button. The user enters an answer in the text area and clicks the send button.

[0473] Input: The user's answer.

[0474] Output: The user's answer sent to the server.

[0475] Step 9:

[0476] Analyzing answers and generating the next question

[0477] Specific operation: The server analyzes the user's answer using a generative model and generates the next question. It then analyzes the user's emotions using a sentiment analysis engine and adjusts the next question accordingly. For example, if the user expresses anxiety, it changes the question to be more gentle. The generated question is then sent to the device.

[0478] Input: The user's answer.

[0479] Output: The parsed answer and the newly generated question.

[0480] Step 10:

[0481] View new questions and re-answer

[0482] Specific operation: The terminal displays the new question received from the server. The user enters the answer again and presses the send button. This causes steps 8 to 10 to be repeated.

[0483] Input: The new question sent by the server.

[0484] Output: New question displayed on the user interface, user's re-answer.

[0485] Step 11:

[0486] Interview data storage

[0487] What happens: The server stores all the interview questions and answers in a database.

[0488] Input: Interview data (all questions and answers).

[0489] Output: Interview data stored in a database.

[0490] Step 12:

[0491] Data analysis and reporting

[0492] Specific operation: The server analyzes the stored data using a generative model and sentiment analysis engine to extract various indicators (e.g., frequently used keywords, sentiment trend analysis, etc.), compiles the results in a report format, exports the generated report to a file, and uploads it to cloud storage.

[0493] Input: Saved interview data.

[0494] Output: The generated report.

[0495] Step 13:

[0496] Providing reports

[0497] Specific operation: The server provides the generated report to the company via a download link, etc.

[0498] Input: The generated report.

[0499] Output: A download link that will be provided to the company.

[0500] (Application example 2)

[0501] 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."

[0502] Conventional user interview systems are limited to one-way question-and-answer exchanges, which means they are unable to fully reflect user emotions and detailed feedback. This makes it difficult to accurately grasp actual user experiences and opinions about products. Furthermore, they often fail to generate appropriate questions or adjust the dialogue in response to user emotions, resulting in a poor user experience.

[0503] 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.

[0504] In this invention, the server includes means for generating questions for a user using a generative model, means for analyzing answers from the user, means for generating a next question based on the analyzed answer, means for recording the generated question and the analyzed answer, means for analyzing the recorded data and generating a report, means for analyzing the user's emotions, and means for adjusting the next question based on the result of the user's emotion analysis. This makes it possible to collect detailed and accurate feedback while adjusting the content of the dialogue according to the user's emotions.

[0505] A "generative model" is an algorithm that automatically generates questions and information based on user input.

[0506] The "means for analyzing the answer" is a technique for understanding the answer entered by the user and analyzing its content.

[0507] The "means for generating questions" is a technology that automatically creates the next question based on the analyzed answer.

[0508] The "recording means" is a technique for saving the generated questions and analyzed answers in a storage device such as a database.

[0509] The "means for generating a report" is a technique for analyzing the recorded data and outputting the results in the form of a report.

[0510] "Means for analyzing emotions" refers to technology that determines emotions from users' responses and quantifies the results.

[0511] The "means for adjusting the next question based on the emotion analysis result" is a technology that appropriately changes the content and wording of the next question based on the analyzed emotion.

[0512] "Emotion trend analysis" is a technique that analyzes the transitions in a user's emotions and identifies changes and trends in emotions over time.

[0513] "Means for collecting feedback" refers to techniques for collecting opinions and evaluations from users about products they have purchased.

[0514] The present invention relates to a user interview system that combines a generative model and an emotion engine, and is intended in particular for collecting feedback on product evaluations on online shopping sites. A specific embodiment of the system that realizes this application example is described below.

[0515] System configuration

[0516] Hardware

[0517] Device: The smartphone or smart glasses used by the user.

[0518] Server: A server for data processing, question generation, answer analysis, sentiment analysis, and data storage.

[0519] software

[0520] Generative model: Used to generate questions and analyze user responses. Specifically, it uses natural language processing technologies such as OpenAI's GPT-3.

[0521] Sentiment engine: Software for analyzing emotions from user responses. Specifically, it uses a sentiment analysis model using the Transformers library.

[0522] Database: A database for storing user profile information, interview results, sentiment analysis results, etc.

[0523] Feedback Collection Process

[0524] Initial Setup

[0525] The server first collects user profile information (such as name, purchase history, etc.) and stores it in a database, then uses a generative model to create initial questions.

[0526] Interview begins

[0527] The user accesses the system from a terminal and the interview begins. Initial questions are displayed and the user submits their answers through an input form.

[0528] Question and answer exchange

[0529] The server analyzes the user's answers and uses an emotion engine to analyze the emotions from the answers. Based on the analysis results, the server generates the next question and adjusts the question content according to the user's emotions. This allows the server to obtain detailed feedback from the user.

[0530] Data storage and analysis

[0531] After all interactions are completed, the server stores the collected data in a database and analyzes the user's emotional trends from the recorded data. Based on the analysis results, the server generates a report and provides it to the company via a download link or other means.

[0532] Specific examples

[0533] A user uses a smartphone app to provide feedback on a product they have purchased. The first question is displayed in the form, "Hello, [user name]. Please rate the product you purchased. What was your first impression?" As the user enters their answer, the answer is analyzed in real time and the next question is generated based on their emotional state. For example, the question is adjusted accordingly, such as, "Tell me more about the design. What did you particularly like about it?"

[0534] Prompt Sentence Examples

[0535] User Answer: I like the design, but the functionality is a bit unclear.

[0536] Emotion: Negative

[0537] Generate the following questions:

[0538] In this way, a system is realized that adjusts the dialogue content according to the user's emotions and collects detailed and accurate feedback.

[0539] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0540] Step 1: User registration and initial setup

[0541] A user accesses the system using a terminal and enters profile information such as the user's name and purchase history. The server stores this input data in a database and generates initial questions using a generative model. The generated initial questions are in the form of, for example, "Please rate the product you purchased. What was your first impression?"

[0542] Step 2: Displaying initial questions

[0543] The server sends the generated initial question to the terminal, which displays the question to the user, and the user inputs an answer to the question and sends it to the server via the terminal.

[0544] Step 3: Response analysis and sentiment analysis

[0545] The server receives the answer sent by the user and analyzes the answer using a generative model. At the same time, it analyzes the user's sentiment from the answer using an emotion engine. For example, if a user answers "I think the design is good, but the functionality is a little unclear," the emotion analysis engine will detect negative sentiment.

[0546] Step 4: Generate the next question

[0547] The server generates the next question based on the analyzed answer and emotion data. It uses a generative model to generate a question that corresponds to the user's emotion. For example, if a negative emotion is detected, it generates a question such as, "What specific aspects of the feature did you find unclear?" The generated question is sent from the server to the device and displayed to the user.

[0548] Step 5: Conducting the interview

[0549] The user answers the next question and sends the answer to the server via their device. This process repeats, with the server analyzing each answer and generating the next question. Based on the sentiment analysis results, the question content is adjusted accordingly.

[0550] Step 6: Data storage and final analysis

[0551] After the interview is over, the server stores all questions and answers, as well as the sentiment analysis results, in a database. The recorded data is then analyzed again for sentiment trend analysis, summarizing changes in users' sentiment and opinion trends.

[0552] Step 7: Generate reports

[0553] The server generates reports based on the accumulated data, including analysis results such as classification of user opinions and sentiment trends. The generated reports are provided to the company via a download link or other means.

[0554] In this way, detailed and accurate feedback can be collected from users, and appropriate questions can be generated based on the results of sentiment analysis, making the interviews more effective.

[0555] 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.

[0556] 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.

[0557] 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.

[0558] [Second embodiment]

[0559] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0560] 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.

[0561] 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).

[0562] 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.

[0563] 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.

[0564] 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).

[0565] 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.

[0566] 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.

[0567] 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.

[0568] 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.

[0569] 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.

[0570] 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."

[0571] This invention relates to a user interview system that uses a generative model. This system generates questions through dialogue with the user, analyzes the answers, and generates subsequent questions based on the results, enabling efficient and accurate data collection. Specific program processing is explained in natural language and includes examples.

[0572] Program processing description

[0573] 1. Preparation Phase

[0574] Data collection

[0575] The server collects data sets for use in the interviews, including past interview transcripts, questionnaires, and user profile information.

[0576] Model Training

[0577] The server trains the generative model based on the collected data, using natural language processing technology to build algorithms for question generation and answer analysis.

[0578] 2. Interview begins

[0579] Interview begins

[0580] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[0581] User Registration

[0582] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[0583] The user enters the required information into the form and presses the submit button.

[0584] The server stores the submitted user information in a database.

[0585] Generating initial questions

[0586] The server generates an initial question based on the user information and sends it to the terminal.

[0587] The terminal displays the question received from the server.

[0588] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[0589] 3. Question and answer repetition

[0590] Questions and Answers

[0591] The user enters an answer and presses the send button.

[0592] The server uses the generative model to analyze the user's answers and generate the next question.

[0593] The server generates a new question and sends it to the device.

[0594] For example: "Tell me more about the design. What did you particularly like about it?"

[0595] The terminal displays the new question to the user.

[0596] The user again enters the answer and presses the send button.

[0597] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[0598] 4. Data analysis phase

[0599] Data storage and analysis

[0600] The server stores all the data from the interview.

[0601] The server analyzes the stored data using a generative model and extracts various indicators (e.g., frequently occurring keywords, sentiment analysis, etc.).

[0602] Report Generation

[0603] The server compiles the analysis results in the form of a report.

[0604] The server generates the report and provides it to the company via a download link or similar.

[0605] Specific examples

[0606] Interview example

[0607] User registration and initial inquiries

[0608] 1. The user accesses the system using a terminal and participates in the interview.

[0609] "User Registration" will appear on the screen and the user will enter the required information (name, age, interests, etc.).

[0610] 2. The server generates initial questions based on the user's profile.

[0611] You might see an initial question like, "Hello, [user's name]. Today I'd like to talk to you about a new product. First, what are your impressions of this product?"

[0612] Dialogue Progress

[0613] 3. The user enters the answer

[0614] User: "I like the design, but the functionality is a bit unclear."

[0615] 4. The server analyzes the user's answer and generates the next question

[0616] The next question will be displayed, such as "Tell us more about the design. What did you particularly like about it?"

[0617] 5. The user enters the answer again

[0618] User: "I especially like the grip and the color combination."

[0619] Data Analysis and Report Generation

[0620] 6. After the interview is over, the server analyzes the collected data.

[0621] The server analyzes the interview transcript, categorizes the user's opinions, and generates a summary.

[0622] 7. The server compiles the analysis results into a report

[0623] Companies are provided with a report containing analysis results such as, "Many users give the design high marks, but feel that further explanation is needed regarding functionality."

[0624] In this way, the entire interview process is automated using generative models, reducing costs and enabling more accurate information collection.

[0625] The processing flow will be explained below.

[0626] Step 1:

[0627] The server collects data sets for use in the interviews, including past interview transcripts, questionnaires, and user profile information.

[0628] Step 2:

[0629] The server trains the generative model based on the collected data, using natural language processing technology to build algorithms for question generation and answer analysis.

[0630] Step 3:

[0631] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[0632] Step 4:

[0633] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[0634] Step 5:

[0635] The user enters the required information into the form and presses the submit button.

[0636] Step 6:

[0637] The server stores the submitted user information in a database.

[0638] Step 7:

[0639] The server generates an initial question based on the user information and sends it to the terminal.

[0640] Step 8:

[0641] The terminal displays the question received from the server.

[0642] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[0643] Step 9:

[0644] The user enters an answer to the question and presses the send button.

[0645] Step 10:

[0646] The server uses the generative model to analyze the user's answers and generate the next question.

[0647] Step 11:

[0648] The server generates a new question and sends it to the device.

[0649] For example: "Tell me more about the design. What did you particularly like about it?"

[0650] Step 12:

[0651] The terminal displays the new question to the user.

[0652] Step 13:

[0653] The user again enters the answer and presses the send button.

[0654] Step 14:

[0655] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[0656] Step 15:

[0657] The server stores all the data from the interview.

[0658] Step 16:

[0659] The server analyzes the stored data using a generative model and extracts various indicators (e.g., frequently occurring keywords, sentiment analysis, etc.).

[0660] Step 17:

[0661] The server compiles the analysis results in the form of a report.

[0662] Step 18:

[0663] The server generates the report and provides it to the company via a download link or similar.

[0664] This automates the entire interview process using generative models, reducing costs and enabling more accurate information collection.

[0665] Example 1

[0666] 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."

[0667] In conventional user interview systems, questions are generated and answers are analyzed manually, resulting in inefficiency and limited accuracy. Data storage and analysis are also time-consuming and costly. Furthermore, there is a lack of tools and methods for effectively utilizing interview data, making comprehensive analysis difficult.

[0668] 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.

[0669] In this invention, the server includes means for generating questions for the user using a generative model, means for analyzing answers from the user, and means for generating the next question based on the analyzed answers. This automates the entire interview process, enabling efficient and accurate data collection.

[0670] A "generative model" is an artificial intelligence model used to generate questions for a user and analyze the user's answers.

[0671] The "means for generating questions" is a function that uses a generative model to automatically create questions appropriate for the user.

[0672] The "means for analyzing answers" is a function that analyzes answers obtained from users using a generative model and extracts information for generating the next question.

[0673] The "means for generating the next question" is a function for generating the next question to be asked based on the analyzed answer.

[0674] The "means for recording questions and analyzed answers" is a function for saving the generated questions, the user's answers, and the analysis results thereof.

[0675] The "means for analyzing recorded data and generating a report" is a function for analyzing saved interview data using a generative model and compiling the results in a report format.

[0676] The "means for generating initial questions" is a function for generating initial questions based on the user's profile information at the initial stage of the interview.

[0677] "Natural language processing technology" is an artificial intelligence technology for understanding and generating language data, and is used in generative model algorithms.

[0678] The present invention relates to a user interview system using a generative model. Specific embodiments of the system will be described below.

[0679] Hardware and Software Use

[0680] The system's main operations are carried out by a server, a device, and a user. The system utilizes a generative model (e.g., GPT-4) running on the server and a device (e.g., a PC or smartphone) accessed by the user. The server also uses software incorporating analysis algorithms using natural language processing techniques. A specific example of such software is the Transformers library from Hugging Face.

[0681] Explanation of program processing

[0682] The operation of the system can be divided into the following main stages: data collection, model training, interview initiation, question and answer cycle, data storage and analysis, and report generation.

[0683] Data collection

[0684] The server collects the datasets required for the interview, including past interview transcripts, questionnaires, and user profile information.

[0685] Model Training

[0686] The server trains the generative AI model based on the collected data. Natural language processing technology is used for training, and algorithms for question generation and answer analysis are built. GPU resources are used to improve processing speed.

[0687] Interview begins

[0688] Users access the system using a terminal and participate in the interview through a web interface. Users click on a URL and connect to the system from their browser.

[0689] The terminal displays a user registration form, and the user enters information such as name, age, interests, etc. The entered information is sent to the server and stored in a database.

[0690] The server generates initial questions based on the user's profile information and sends them to the terminal, which then displays the received questions to the user.

[0691] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[0692] Repeated questions and answers

[0693] The user responds to the questions on the terminal, enters the answer, and presses the send button. The answer is then sent to the server.

[0694] The server uses the generative AI model to analyze the user's answers and generate the next question, which is again sent to the device and displayed to the user.

[0695] For example: "Tell me more about the design. What did you particularly like about it?"

[0696] Data storage and analysis

[0697] The server stores all interview data in a database. After the interview is completed, the stored data is analyzed using a generative model to extract indicators such as frequently occurring keywords and sentiment analysis.

[0698] Report Generation

[0699] The server compiles the analysis results into a report, which is saved in PDF or Excel format and includes a download link for the report to be provided to the company.

[0700] As an example, the following prompt sentence is used:

[0701] User information: Name = Taro, Age = 30, Interests = Technology

[0702] Question: "Hello, Taro. Today I'd like to ask you about a new product. First of all, what are your impressions of this product?"

[0703] As described above, this invention enables efficient and accurate user interviews using generative AI models.

[0704] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0705] Step 1: Data collection

[0706] The server collects a dataset for use in the interviews, which includes past interview transcripts, questionnaires, and user profile information.

[0707] Input: Database information, interview transcripts

[0708] Output: Collected dataset

[0709] What it does: The server accesses public and internal company databases, collects and stores the required data.

[0710] Step 2: Model training

[0711] The server trains the generative AI model based on the collected data. During training, natural language processing technology is used to build algorithms for question generation and answer analysis.

[0712] Input: Collected dataset

[0713] Output: A trained generative AI model

[0714] How it works: The server pre-processes the data and trains a generative AI model using natural language processing libraries (e.g., Transformers from Hugging Face), leveraging GPU resources to speed up training.

[0715] Step 3: System Access

[0716] A user accesses the system using a terminal and clicks on a URL to reach the web interface.

[0717] Input: URL

[0718] Output: Access to the system

[0719] How it works: A user opens a browser and enters the provided URL to log into the system.

[0720] Step 4: User Registration

[0721] The device displays a user registration form, where the user enters information such as name, age, and interests. The information is sent to the server and stored in a database.

[0722] Input: User-entered information (name, age, interests)

[0723] Output: Saved user information

[0724] How it works: A registration form is displayed, the user fills in the fields, presses the submit button, and the data is sent to the server, which receives it and stores it in a database.

[0725] Step 5: Generate initial questions

[0726] The server generates an initial question based on the user information and sends it to the terminal, which then displays the received question to the user.

[0727] Input: User information

[0728] Output: Initial question

[0729] How it works: The server uses the user's profile information to generate prompts and input them into the AI ​​model. The generated initial questions are sent to the device and displayed.

[0730] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[0731] Step 6: Repeat the questions and answers

[0732] The user enters an answer to a question and presses the submit button. The answer is sent to the server and analyzed using the generative AI model. The next question is generated and sent again to the device. The device displays the next question to the user.

[0733] Input: User's answer

[0734] Output: Next question

[0735] How it works: When a user enters an answer and presses the submit button, the answer is sent to the server, which uses a generative AI model to analyze the answer and generate the next question. This question is then sent to the device and displayed to the user.

[0736] For example: "Tell me more about the design. What did you particularly like about it?"

[0737] Step 7: Data storage and analysis

[0738] The server stores all the interview data, which is then analyzed using a generative model to extract metrics such as frequently occurring keywords and sentiment analysis.

[0739] Input: Interview data

[0740] Output: Analysis results (frequent keywords, sentiment analysis, etc.)

[0741] How it works: The server stores all interview data and analyzes it using a generative model. Various indicators are extracted using natural language processing techniques.

[0742] Step 8: Generate reports

[0743] The server compiles the analysis results in a report format, which is saved in PDF or Excel format and a download link is provided.

[0744] Input: Analysis results

[0745] Output: Report (PDF or Excel format)

[0746] How it works: The server documents the analysis results and generates a report, which is then provided to the company as a download link, allowing them to easily retrieve the information they need.

[0747] In this way, by going through each processing step, efficient and accurate user interviews can be conducted using generative AI models.

[0748] (Application example 1)

[0749] 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."

[0750] In conventional user interview systems, the interview process is conducted manually, which is extremely time-consuming and labor-intensive, making it difficult to collect data efficiently. Furthermore, analyzing the collected data is also time-consuming, making it difficult to utilize feedback in real time. Furthermore, there was a lack of an effective way to integrate and analyze feedback from many different customers, making it difficult to extract customer insights.

[0751] 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.

[0752] In this invention, the server includes: means for generating questions for users using a generative model; means for analyzing responses from users; means for generating subsequent questions based on the analyzed responses; means for recording the generated questions and analyzed responses; means for analyzing the recorded data and generating a report; means for users to access the system using a terminal and participate in an interview; means for the terminal to display questions and acquire responses from users; and means for providing a report generated after the interview is completed to an administrator. This automates the entire interview process, enabling efficient and accurate data collection. Furthermore, the collected data is analyzed in real time, and feedback and improvement suggestions are provided immediately based on the results, enabling rapid response.

[0753] A "generative model" is a model that generates new data or information based on existing data, and primarily uses machine learning and artificial intelligence techniques.

[0754] "User" refers to a user who accesses the system and participates in an interview.

[0755] "Means for generating questions" refers to the function of automatically creating questions to be asked during interviews using a generative model.

[0756] "Means for analyzing responses" refers to the function of receiving responses from users, understanding their contents, and analyzing them.

[0757] "Means for generating the next question" refers to a function for generating the next question to be asked based on the previously analyzed answer.

[0758] "Means for recording generated questions and analyzed answers" refers to a function for saving questions generated during an interview, the user's answers, and the analysis results in a database or the like.

[0759] "Means for generating reports" refers to the function of compiling analysis results based on recorded data and automatically creating reports to be provided to managers and personnel in charge.

[0760] "Device" refers to the device used by the user when participating in the interview, including smartphones and tablets.

[0761] "Means of participating in an interview" refers to the process by which a user accesses the system using a terminal, inputs the information required for the interview, and interacts with the system.

[0762] "Means for displaying a question and obtaining a user's response" refers to a function for generating a question to be displayed on a terminal and receiving a response entered by a user in response to that question.

[0763] "Means to provide to the manager" refers to a function for distributing the report generated after the interview to the relevant parties and using it for feedback and consideration of improvement measures.

[0764] This invention relates to a user interview system using generative models. This system improves the efficiency of collecting and analyzing customer feedback in physical stores, and can be used to immediately improve products and services.

[0765] The system mainly includes the following components:

[0766] 1. Generative model: A generative model that uses natural language processing technology and has the ability to automatically generate questions for users.

[0767] 2. Analysis server: It has the means to analyze the answers from the user and generate the next question, as well as the means to record the generated questions and analyzed answers.

[0768] 3. Interview device: A terminal accessed by the user (smartphone, tablet, etc.) that has the function of displaying questions and obtaining responses from the user.

[0769] 4. Database: A storage device for storing analysis results and interview records.

[0770] 5. Report generation tool: Provides a means to automatically generate reports based on recorded data and provide them to managers and personnel in charge.

[0771] System Operation

[0772] 1. Preparation Phase

[0773] The server collects past interview data, questionnaires, user profile information, etc. for use in the interview.

[0774] The server trains a generative model based on the collected data and builds algorithms for question generation and answer analysis, primarily using natural language processing technology.

[0775] 2. Begin the interview

[0776] The user accesses the system using the interview device, for example, by clicking on a URL that takes the user to a web interface.

[0777] The interview device displays a user registration form, and the user enters information such as name, age, and interests. This information is sent to the server and stored in a database.

[0778] The server generates initial questions based on the collected user profile information and sends them to the interview device, such as, "Hello, [user name]. Today I'd like to ask you about a new product. First of all, what is your impression of this product?"

[0779] 3. Question and answer repetition

[0780] The user enters answers into the interview device, which are then sent to the server.

[0781] The server analyzes the answer using the generative model and generates the next question. At this time, the server performs the analysis and generation process in the form of a prompt such as "User's answer: The design is good, but the operation feels a little complicated. Prompt: Please generate the next question based on this answer."

[0782] New questions are generated and sent to the interview device, such as "Tell me more about the operation. What parts did you find particularly difficult to understand?"

[0783] This process repeats until the user responds "Done."

[0784] 4. Data analysis and report generation

[0785] After the interview is completed, the server analyzes all collected data and categorizes and summarizes the user's opinions.

[0786] The server compiles the analysis results in a report and provides it to the administrator or person in charge. For example, the report may include a statement such as, "Many users rate the design highly, but feel that further explanation is needed regarding the functionality."

[0787] Specific examples

[0788] This system can be used to efficiently collect customer feedback about new products in physical stores. For example, suppose you want to obtain feedback about a new microwave. The initial question displayed is, "Hello, [user name]. Today I'd like to ask you about our new microwave. First, what is your impression of this product?" If the user responds, "The design is good, but I found the operation a little complicated," the system generates a next question: "Tell me more about the operation. Which parts did you find particularly difficult to understand?" This allows the system to collect specific user opinions. Finally, a report is generated based on the collected data, enabling efficient product improvements.

[0789] In this way, the present invention uses a generative AI model to improve the efficiency and accuracy of customer interviews in physical stores.

[0790] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0791] Step 1: Preparation Phase

[0792] The server collects a dataset to be used in the interviews. This dataset includes past interview data, a list of questions, and user profile information. The server then trains a generative model based on this data. Specifically, it uses natural language processing technology to build algorithms for question generation and answer analysis.

[0793] Input: Past interview data, questionnaire, user profile information

[0794] Processing: Training the generative model

[0795] Output: A trained generative model

[0796] Step 2: Start the interview

[0797] The user accesses the system using a terminal and participates in the interview. The user clicks on a URL to access the web interface. The terminal displays a user registration form, and the user enters information such as name, age, and interests. The entered information is sent to the server and stored in a database.

[0798] Input: User access and registration information

[0799] Process: Enter and submit user registration information

[0800] Output: User information stored in the database

[0801] Step 3: Generate initial questions

[0802] The server generates initial questions based on the collected user profile information and sends them to the interview device. For example, a question might be, "Hello, [user name]. Today I'd like to ask you about a new product. First of all, what is your impression of this product?"

[0803] Input: User profile information

[0804] Process: Generate initial questions

[0805] Output: Initial questions sent to the interview device

[0806] Step 4: Repeat the questions and answers

[0807] The user enters their answer into the interview device and presses the send button. This is sent to the server. The server uses the generative model to analyze the answer and generate the next question. For example, if the user enters, "User answer: The design is good, but the operation felt a little complicated," the server generates the next question: "Please tell us more about the operation. Which part did you find particularly difficult to understand?"

[0808] Input: User's answer

[0809] Processing: Parsing answers and generating next questions

[0810] Output: The next question sent to the interview device

[0811] Step 5: Record the data

[0812] The server records the questions generated and the analyzed answers, so that the entire interview process is stored in a database for later analysis.

[0813] Input: Generated question, parsed answer

[0814] Process: Recording questions and answers

[0815] Output: Interview data stored in a database

[0816] Step 6: Data analysis and report generation

[0817] After the interviews are completed, the server analyzes all collected data and classifies and summarizes the user opinions. The server then compiles the analysis results into a report and provides it to the manager or person in charge. For example, the report may include statements such as, "Many users rate the design highly, but feel that further explanation is needed regarding the functionality."

[0818] Input: Collected data

[0819] Processing: Data analysis and report generation

[0820] Output: Report

[0821] In this way, a question generation and answer analysis system using generative models enables efficient and accurate interviews and data collection, enabling companies to quickly collect customer feedback and use it to effectively improve their products and services.

[0822] 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.

[0823] This invention relates to a user interview system that combines a generative model and an emotion engine. This system generates questions through dialogue with the user, analyzes the answers, and generates the next question based on the results. It also recognizes the user's emotions and adjusts the dialogue content to obtain more information.

[0824] Program processing description

[0825] 1. Preparation Phase

[0826] Data collection

[0827] The server collects datasets for use in the interviews, including past interview transcripts, questionnaires, user profile information, and sentiment data.

[0828] Model Training

[0829] The server trains the generative model and emotion engine based on the collected data, utilizing natural language processing technology to build algorithms for question generation, response analysis, and emotion analysis.

[0830] 2. Interview begins

[0831] Interview begins

[0832] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[0833] User Registration

[0834] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[0835] The user enters the required information into the form and presses the submit button.

[0836] The server stores the submitted user information in a database.

[0837] Generating initial questions

[0838] The server generates an initial question based on the user information and sends it to the terminal.

[0839] The terminal displays the question received from the server.

[0840] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[0841] 3. Question and answer repetition

[0842] Questions and Answers

[0843] The user enters an answer to the question and presses the send button.

[0844] The server uses a generative model to analyze the user's answers and generate the next question. In addition, it analyzes the user's emotions using an emotion engine and adjusts the next question based on the results.

[0845] Built-in sentiment analysis

[0846] The emotion engine analyzes the user's responses to determine their emotions, and the server adjusts the tone and content of the dialogue based on that emotion information. For example, if the user expresses anxiety, the server changes the questions to be gentler and more reassuring.

[0847] The server generates a new question and sends it to the device.

[0848] For example: "Tell me more about the design. What did you particularly like about it?"

[0849] The terminal displays the new question to the user.

[0850] The user again enters the answer and presses the send button.

[0851] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[0852] 4. Data analysis phase

[0853] Data storage and analysis

[0854] The server stores all the data from the interview.

[0855] The server analyzes the stored data using a generative model and emotion engine to extract various indicators (e.g., frequently occurring keywords, emotion trend analysis, etc.).

[0856] Report Generation

[0857] The server compiles the analysis results in the form of a report.

[0858] The server generates the report and provides it to the company via a download link or similar.

[0859] Specific examples

[0860] Interview example

[0861] User registration and initial inquiries

[0862] 1. The user accesses the system using a terminal and participates in the interview.

[0863] "User Registration" will appear on the screen and the user will enter the required information (name, age, interests, etc.).

[0864] 2. The server generates initial questions based on the user's profile.

[0865] You might see an initial question like, "Hello, [user's name]. Today I'd like to talk to you about a new product. First, what are your impressions of this product?"

[0866] Dialogue Progress

[0867] 3. The user enters the answer

[0868] User: "I like the design, but the functionality is a bit unclear."

[0869] 4. The server analyzes the user's answer and generates the next question

[0870] The next question will be displayed, such as, "Tell us more about the design. What did you like most about it?" In addition, an emotion engine will analyze the emotions from the user's answers and adjust the dialogue accordingly.

[0871] 5. The user enters the answer again

[0872] User: "I especially like the grip and the color combination."

[0873] Data Analysis and Report Generation

[0874] 6. After the interview is over, the server analyzes the collected data.

[0875] The server analyzes the interview transcript, categorizes the user's opinions, and generates a summary.

[0876] 7. The server compiles the analysis results into a report

[0877] Companies are provided with a report containing analysis results such as, "Many users give the design high marks, but feel that further explanation is needed regarding functionality."

[0878] In this way, the entire interview process is automated using generative models and emotion engines, reducing costs and enabling more accurate information collection. Emotion analysis improves the user experience by enabling appropriate responses based on the user's emotional state.

[0879] The processing flow will be explained below.

[0880] Step 1:

[0881] The server collects datasets for use in the interviews, including past interview transcripts, questionnaires, user profile information, and sentiment data.

[0882] Step 2:

[0883] The server trains the generative model and emotion engine based on the collected data, utilizing natural language processing technology to build algorithms for question generation, response analysis, and emotion analysis.

[0884] Step 3:

[0885] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[0886] Step 4:

[0887] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[0888] Step 5:

[0889] The user enters the required information into the form and presses the submit button.

[0890] Step 6:

[0891] The server stores the submitted user information in a database.

[0892] Step 7:

[0893] The server generates an initial question based on the user information and sends it to the terminal.

[0894] Step 8:

[0895] The terminal displays the question received from the server.

[0896] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[0897] Step 9:

[0898] The user enters an answer to the question and presses the send button.

[0899] Step 10:

[0900] The server uses a generative model to analyze the user's answers and generate the next question. In addition, it analyzes the user's emotions using an emotion engine and adjusts the next question based on the results.

[0901] Step 11:

[0902] The emotion engine analyzes the user's responses to determine their emotions, and the server adjusts the tone and content of the dialogue based on that emotion information. For example, if the user expresses anxiety, the server changes the questions to be gentler and more reassuring.

[0903] Step 12:

[0904] The server generates a new question and sends it to the device.

[0905] For example: "Tell me more about the design. What did you particularly like about it?"

[0906] Step 13:

[0907] The terminal displays the new question to the user.

[0908] Step 14:

[0909] The user again enters the answer and presses the send button.

[0910] Step 15:

[0911] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[0912] Step 16:

[0913] The server stores all the data from the interview.

[0914] Step 17:

[0915] The server analyzes the stored data using a generative model and emotion engine to extract various indicators (e.g., frequently occurring keywords, emotion trend analysis, etc.).

[0916] Step 18:

[0917] The server compiles the analysis results in the form of a report.

[0918] Step 19:

[0919] The server generates the report and provides it to the company via a download link or similar.

[0920] This will automate the entire interview process using generative models, reducing costs and enabling more accurate information collection. Furthermore, emotion analysis will enable appropriate responses based on the user's emotional state, improving the user experience.

[0921] Example 2

[0922] 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."

[0923] Conventional interview systems simply analyze user responses and generate the next questions, but are unable to properly consider the user's emotions or the context of the conversation. This results in poor interview quality and limited information being obtained. It is also difficult to improve the user experience, and efficient information gathering is required.

[0924] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for generating questions for the user using a generative model, means for analyzing answers from the user, means for generating the next question based on the analyzed answers, means for recording the generated questions and the analyzed answers, means for analyzing the recorded data and generating a report, means for training the generative model and the emotion analysis engine based on the collected data, and means for analyzing emotions from the user answers and adjusting the content of the dialogue. This makes it possible to conduct high-quality interviews that take user emotions into consideration and to collect detailed information efficiently.

[0925] A "generative model" is a model that uses machine learning algorithms to generate new data or information.

[0926] "Means for generating questions" refers to a function that uses a generative model to create questions for the user.

[0927] "Means for analyzing responses" refers to the technology used to analyze responses from users and understand and evaluate their content.

[0928] The "means for generating the next question" refers to a function for generating the next appropriate question based on the analyzed answer.

[0929] "Means for recording" refers to the technology used to store the generated questions and analyzed answers in a database or storage.

[0930] "Means for generating reports" refers to a function that analyzes recorded data and outputs the results in the form of a report.

[0931] An "emotion analysis engine" refers to an algorithm or software that analyzes the emotions of users' responses and determines their emotional state.

[0932] "Means for adjusting the content of the dialogue" refers to the function for appropriately changing and adjusting the tone and content of the dialogue based on the results of sentiment analysis.

[0933] MODE FOR CARRYING OUT THE INVENTION

[0934] This invention relates to a user interview system that combines a generative model and an emotion analysis engine. This system generates questions through dialogue with the user, analyzes the answers, and generates the next question based on the results. It also recognizes the user's emotions and adjusts the dialogue content to obtain more information.

[0935] The system includes the following components:

[0936] Generative Model

[0937] Sentiment Analysis Engine

[0938] Database

[0939] User Interface (Terminal)

[0940] Hardware and software used

[0941] The system uses the following hardware and software:

[0942] Servers: Used for data processing and model training. Example: Cloud servers (AWS EC2, Google Cloud)

[0943] Database: Used to store user information and interview data. Example: MySQL, MongoDB

[0944] Generative models: used for question generation and answer analysis. Example: GPT-4 (OpenAI)

[0945] Sentiment analysis engine: Used for sentiment analysis. Example: Microsoft Azure's Text Analytics API

[0946] User Interface (Terminal): The interface through which the user interacts with the system. Example: A web interface created with React.

[0947] Specific operation of the system

[0948] 1. Data collection and model training:

[0949] The server collects and stores in a database a dataset for use in the interviews, including past interview transcripts, questionnaires, user profile information, and sentiment data.

[0950] The server trains the generative model and sentiment analysis engine based on the collected data. Natural language processing technology is used for training, and algorithms for question generation, answer analysis, and sentiment analysis are built using Python libraries (TensorFlow, PyTorch).

[0951] 2. Interview begins:

[0952] A user accesses the system using a terminal and accesses the web interface.

[0953] The device displays a user registration form and asks the user to enter information such as name, age, interests, etc. The user enters the required information into the form and presses the submit button.

[0954] The server stores the user information in a database, generates an initial question, and sends it to the terminal.

[0955] The terminal displays the initial question received from the server.

[0956] Examples:

[0957] "Hi, [user name]. Today I'd like to ask you about a new product. First, what are your impressions of this product?"

[0958] 3. Repeated questions and answers:

[0959] The user enters an answer to the question and presses the send button.

[0960] The server uses a generative model to analyze the user's answers and generate the next question. The emotion analysis engine analyzes the user's emotions and adjusts the next question based on the results. For example, if the user expresses anxiety, the question will be changed to a gentler, more reassuring one.

[0961] The server generates a new question and sends it to the terminal. The terminal displays the new question. The user again enters the answer and presses the send button.

[0962] Examples:

[0963] "Tell us more about the design. What parts did you particularly like?"

[0964] 4. Data Storage and Analysis:

[0965] The server stores all the data from the interview.

[0966] The server analyzes the stored data using a generative model and sentiment analysis engine to extract various indicators (e.g., frequently occurring keywords, sentiment trend analysis, etc.).

[0967] The server compiles the analysis results in a report format that is easy to understand visually (e.g., PDF or dashboard).

[0968] 5. Report provision:

[0969] The server provides the generated report to the company via a download link or the like.

[0970] As described above, the system of the present invention uses a generative AI model and an emotion analysis engine to realize high-quality interviews that take into account the user's emotions, enabling efficient and detailed information collection.

[0971] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0972] Step 1:

[0973] Data Collection Phase

[0974] What it does: The server collects a dataset to use in the interviews. This includes past interview transcripts, questionnaires, user profile information, and sentiment data. The server uses an API to download the data from cloud storage and saves it to local storage.

[0975] Input: Interview transcripts in cloud storage, questionnaire, user profile information, and sentiment data.

[0976] Output: The initial dataset saved to local storage.

[0977] Step 2:

[0978] Data Preprocessing

[0979] Specific operation: The server preprocesses the collected data, specifically by using Python libraries (numpy, pandas, NLTK, etc.) to perform data cleaning (noise removal), text tokenization, and stop word removal.

[0980] Input: An initial dataset stored in local storage.

[0981] Output: A preprocessed and clean dataset.

[0982] Step 3:

[0983] Building and training the model

[0984] How it works: The server trains a generative model (e.g., GPT-4) and a sentiment analysis engine based on the preprocessed data. It uses libraries such as TensorFlow and PyTorch. The server exports the trained model to a file and uploads it to cloud storage.

[0985] Input: Preprocessed clean dataset.

[0986] Output: A trained generative model and a sentiment analysis engine.

[0987] Step 4:

[0988] User access and interview start

[0989] Specific operation: A user accesses the system using a terminal and accesses a web interface (e.g., developed with React). The user is redirected to the homepage.

[0990] Input: The user's access request.

[0991] Output: Web interface display.

[0992] Step 5:

[0993] Displaying the user registration form

[0994] Specific operation: The device displays a user registration form. The registration form includes fields for entering name, age, interests, etc. After the user enters the information on the device, the form data is sent to the server.

[0995] Input: User information (name, age, interests).

[0996] Output: The user information sent to the server.

[0997] Step 6:

[0998] Saving user information and generating initial questions

[0999] Specific operation: The server stores user information in a database (e.g., MySQL, MongoDB). It generates initial questions based on the user information and sends them to the device. A generative AI model is used for generation.

[1000] Input: The user information sent to the server.

[1001] Output: User information stored in the database and the initial questions generated.

[1002] Step 7:

[1003] Display initial questions

[1004] Specific behavior: The device displays the initial question received from the server. Example: "Hello, [user name]. Today I'd like to ask you about a new product. First, what are your impressions of this product?"

[1005] Input: The initial question sent by the server.

[1006] Output: The initial question displayed on the user interface.

[1007] Step 8:

[1008] User answer input

[1009] Specific operation: The user enters an answer to a question using the terminal and presses the send button. The user enters an answer in the text area and clicks the send button.

[1010] Input: The user's answer.

[1011] Output: The user's answer sent to the server.

[1012] Step 9:

[1013] Analyzing answers and generating the next question

[1014] Specific operation: The server analyzes the user's answer using a generative model and generates the next question. It then analyzes the user's emotions using a sentiment analysis engine and adjusts the next question accordingly. For example, if the user expresses anxiety, it changes the question to be more gentle. The generated question is then sent to the device.

[1015] Input: The user's answer.

[1016] Output: The parsed answer and the newly generated question.

[1017] Step 10:

[1018] View new questions and re-answer

[1019] Specific operation: The terminal displays the new question received from the server. The user enters the answer again and presses the send button. This causes steps 8 to 10 to be repeated.

[1020] Input: The new question sent by the server.

[1021] Output: New question displayed on the user interface, user's re-answer.

[1022] Step 11:

[1023] Interview data storage

[1024] What happens: The server stores all the interview questions and answers in a database.

[1025] Input: Interview data (all questions and answers).

[1026] Output: Interview data stored in a database.

[1027] Step 12:

[1028] Data analysis and reporting

[1029] Specific operation: The server analyzes the stored data using a generative model and sentiment analysis engine to extract various indicators (e.g., frequently used keywords, sentiment trend analysis, etc.), compiles the results in a report format, exports the generated report to a file, and uploads it to cloud storage.

[1030] Input: Saved interview data.

[1031] Output: The generated report.

[1032] Step 13:

[1033] Providing reports

[1034] Specific operation: The server provides the generated report to the company via a download link, etc.

[1035] Input: The generated report.

[1036] Output: A download link that will be provided to the company.

[1037] (Application example 2)

[1038] 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."

[1039] Conventional user interview systems are limited to one-way question-and-answer exchanges, which means they are unable to fully reflect user emotions and detailed feedback. This makes it difficult to accurately grasp actual user experiences and opinions about products. Furthermore, they often fail to generate appropriate questions or adjust the dialogue in response to user emotions, resulting in a poor user experience.

[1040] 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.

[1041] In this invention, the server includes means for generating questions for a user using a generative model, means for analyzing answers from the user, means for generating a next question based on the analyzed answer, means for recording the generated question and the analyzed answer, means for analyzing the recorded data and generating a report, means for analyzing the user's emotions, and means for adjusting the next question based on the result of the user's emotion analysis. This makes it possible to collect detailed and accurate feedback while adjusting the content of the dialogue according to the user's emotions.

[1042] A "generative model" is an algorithm that automatically generates questions and information based on user input.

[1043] The "means for analyzing the answer" is a technique for understanding the answer entered by the user and analyzing its content.

[1044] The "means for generating questions" is a technology that automatically creates the next question based on the analyzed answer.

[1045] The "recording means" is a technique for saving the generated questions and analyzed answers in a storage device such as a database.

[1046] The "means for generating a report" is a technique for analyzing the recorded data and outputting the results in the form of a report.

[1047] "Means for analyzing emotions" refers to technology that determines emotions from users' responses and quantifies the results.

[1048] The "means for adjusting the next question based on the emotion analysis result" is a technology that appropriately changes the content and wording of the next question based on the analyzed emotion.

[1049] "Emotion trend analysis" is a technique that analyzes the transitions in a user's emotions and identifies changes and trends in emotions over time.

[1050] "Means for collecting feedback" refers to techniques for collecting opinions and evaluations from users about products they have purchased.

[1051] The present invention relates to a user interview system that combines a generative model and an emotion engine, and is intended in particular for collecting feedback on product evaluations on online shopping sites. A specific embodiment of the system that realizes this application example is described below.

[1052] System configuration

[1053] Hardware

[1054] Device: The smartphone or smart glasses used by the user.

[1055] Server: A server for data processing, question generation, answer analysis, sentiment analysis, and data storage.

[1056] software

[1057] Generative model: Used to generate questions and analyze user responses. Specifically, it uses natural language processing technologies such as OpenAI's GPT-3.

[1058] Sentiment engine: Software for analyzing emotions from user responses. Specifically, it uses a sentiment analysis model using the Transformers library.

[1059] Database: A database for storing user profile information, interview results, sentiment analysis results, etc.

[1060] Feedback Collection Process

[1061] Initial Setup

[1062] The server first collects user profile information (such as name, purchase history, etc.) and stores it in a database, then uses a generative model to create initial questions.

[1063] Interview begins

[1064] The user accesses the system from a terminal and the interview begins. Initial questions are displayed and the user submits their answers through an input form.

[1065] Question and answer exchange

[1066] The server analyzes the user's answers and uses an emotion engine to analyze the emotions from the answers. Based on the analysis results, the server generates the next question and adjusts the question content according to the user's emotions. This allows the server to obtain detailed feedback from the user.

[1067] Data storage and analysis

[1068] After all interactions are completed, the server stores the collected data in a database and analyzes the user's emotional trends from the recorded data. Based on the analysis results, the server generates a report and provides it to the company via a download link or other means.

[1069] Specific examples

[1070] A user uses a smartphone app to provide feedback on a product they have purchased. The first question is displayed in the form, "Hello, [user name]. Please rate the product you purchased. What was your first impression?" As the user enters their answer, the answer is analyzed in real time and the next question is generated based on their emotional state. For example, the question is adjusted accordingly, such as, "Tell me more about the design. What did you particularly like about it?"

[1071] Prompt Sentence Examples

[1072] User Answer: I like the design, but the functionality is a bit unclear.

[1073] Emotion: Negative

[1074] Generate the following questions:

[1075] In this way, a system is realized that adjusts the dialogue content according to the user's emotions and collects detailed and accurate feedback.

[1076] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1077] Step 1: User registration and initial setup

[1078] A user accesses the system using a terminal and enters profile information such as the user's name and purchase history. The server stores this input data in a database and generates initial questions using a generative model. The generated initial questions are in the form of, for example, "Please rate the product you purchased. What was your first impression?"

[1079] Step 2: Displaying initial questions

[1080] The server sends the generated initial question to the terminal, which displays the question to the user, and the user inputs an answer to the question and sends it to the server via the terminal.

[1081] Step 3: Response analysis and sentiment analysis

[1082] The server receives the answer sent by the user and analyzes the answer using a generative model. At the same time, it analyzes the user's sentiment from the answer using an emotion engine. For example, if a user answers "I think the design is good, but the functionality is a little unclear," the emotion analysis engine will detect negative sentiment.

[1083] Step 4: Generate the next question

[1084] The server generates the next question based on the analyzed answer and emotion data. It uses a generative model to generate a question that corresponds to the user's emotion. For example, if a negative emotion is detected, it generates a question such as, "What specific aspects of the feature did you find unclear?" The generated question is sent from the server to the device and displayed to the user.

[1085] Step 5: Conducting the interview

[1086] The user answers the next question and sends the answer to the server via their device. This process repeats, with the server analyzing each answer and generating the next question. Based on the sentiment analysis results, the question content is adjusted accordingly.

[1087] Step 6: Data storage and final analysis

[1088] After the interview is over, the server stores all questions and answers, as well as the sentiment analysis results, in a database. The recorded data is then analyzed again for sentiment trend analysis, summarizing changes in users' sentiment and opinion trends.

[1089] Step 7: Generate reports

[1090] The server generates reports based on the accumulated data, including analysis results such as classification of user opinions and sentiment trends. The generated reports are provided to the company via a download link or other means.

[1091] In this way, detailed and accurate feedback can be collected from users, and appropriate questions can be generated based on the results of sentiment analysis, making the interviews more effective.

[1092] 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.

[1093] 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.

[1094] 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.

[1095] [Third embodiment]

[1096] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1097] 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.

[1098] 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).

[1099] 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.

[1100] 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.

[1101] 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).

[1102] 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.

[1103] 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.

[1104] 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.

[1105] 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.

[1106] 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.

[1107] 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."

[1108] This invention relates to a user interview system that uses a generative model. This system generates questions through dialogue with the user, analyzes the answers, and generates subsequent questions based on the results, enabling efficient and accurate data collection. Specific program processing is explained in natural language and includes examples.

[1109] Program processing description

[1110] 1. Preparation Phase

[1111] Data collection

[1112] The server collects data sets for use in the interviews, including past interview transcripts, questionnaires, and user profile information.

[1113] Model Training

[1114] The server trains the generative model based on the collected data, using natural language processing technology to build algorithms for question generation and answer analysis.

[1115] 2. Interview begins

[1116] Interview begins

[1117] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[1118] User Registration

[1119] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[1120] The user enters the required information into the form and presses the submit button.

[1121] The server stores the submitted user information in a database.

[1122] Generating initial questions

[1123] The server generates an initial question based on the user information and sends it to the terminal.

[1124] The terminal displays the question received from the server.

[1125] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[1126] 3. Question and answer repetition

[1127] Questions and Answers

[1128] The user enters an answer and presses the send button.

[1129] The server uses the generative model to analyze the user's answers and generate the next question.

[1130] The server generates a new question and sends it to the device.

[1131] For example: "Tell me more about the design. What did you particularly like about it?"

[1132] The terminal displays the new question to the user.

[1133] The user again enters the answer and presses the send button.

[1134] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[1135] 4. Data analysis phase

[1136] Data storage and analysis

[1137] The server stores all the data from the interview.

[1138] The server analyzes the stored data using a generative model and extracts various indicators (e.g., frequently occurring keywords, sentiment analysis, etc.).

[1139] Report Generation

[1140] The server compiles the analysis results in the form of a report.

[1141] The server generates the report and provides it to the company via a download link or similar.

[1142] Specific examples

[1143] Interview example

[1144] User registration and initial inquiries

[1145] 1. The user accesses the system using a terminal and participates in the interview.

[1146] "User Registration" will appear on the screen and the user will enter the required information (name, age, interests, etc.).

[1147] 2. The server generates initial questions based on the user's profile.

[1148] You might see an initial question like, "Hello, [user's name]. Today I'd like to talk to you about a new product. First, what are your impressions of this product?"

[1149] Dialogue Progress

[1150] 3. The user enters the answer

[1151] User: "I like the design, but the functionality is a bit unclear."

[1152] 4. The server analyzes the user's answer and generates the next question

[1153] The next question will be displayed, such as "Tell us more about the design. What did you particularly like about it?"

[1154] 5. The user enters the answer again

[1155] User: "I especially like the grip and the color combination."

[1156] Data Analysis and Report Generation

[1157] 6. After the interview is over, the server analyzes the collected data.

[1158] The server analyzes the interview transcript, categorizes the user's opinions, and generates a summary.

[1159] 7. The server compiles the analysis results into a report

[1160] Companies are provided with a report containing analysis results such as, "Many users give the design high marks, but feel that further explanation is needed regarding functionality."

[1161] In this way, the entire interview process is automated using generative models, reducing costs and enabling more accurate information collection.

[1162] The processing flow will be explained below.

[1163] Step 1:

[1164] The server collects data sets for use in the interviews, including past interview transcripts, questionnaires, and user profile information.

[1165] Step 2:

[1166] The server trains the generative model based on the collected data, using natural language processing technology to build algorithms for question generation and answer analysis.

[1167] Step 3:

[1168] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[1169] Step 4:

[1170] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[1171] Step 5:

[1172] The user enters the required information into the form and presses the submit button.

[1173] Step 6:

[1174] The server stores the submitted user information in a database.

[1175] Step 7:

[1176] The server generates an initial question based on the user information and sends it to the terminal.

[1177] Step 8:

[1178] The terminal displays the question received from the server.

[1179] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[1180] Step 9:

[1181] The user enters an answer to the question and presses the send button.

[1182] Step 10:

[1183] The server uses the generative model to analyze the user's answers and generate the next question.

[1184] Step 11:

[1185] The server generates a new question and sends it to the device.

[1186] For example: "Tell me more about the design. What did you particularly like about it?"

[1187] Step 12:

[1188] The terminal displays the new question to the user.

[1189] Step 13:

[1190] The user again enters the answer and presses the send button.

[1191] Step 14:

[1192] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[1193] Step 15:

[1194] The server stores all the data from the interview.

[1195] Step 16:

[1196] The server analyzes the stored data using a generative model and extracts various indicators (e.g., frequently occurring keywords, sentiment analysis, etc.).

[1197] Step 17:

[1198] The server compiles the analysis results in the form of a report.

[1199] Step 18:

[1200] The server generates the report and provides it to the company via a download link or similar.

[1201] This automates the entire interview process using generative models, reducing costs and enabling more accurate information collection.

[1202] Example 1

[1203] 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."

[1204] In conventional user interview systems, questions are generated and answers are analyzed manually, resulting in inefficiency and limited accuracy. Data storage and analysis are also time-consuming and costly. Furthermore, there is a lack of tools and methods for effectively utilizing interview data, making comprehensive analysis difficult.

[1205] 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.

[1206] In this invention, the server includes means for generating questions for the user using a generative model, means for analyzing answers from the user, and means for generating the next question based on the analyzed answers. This automates the entire interview process, enabling efficient and accurate data collection.

[1207] A "generative model" is an artificial intelligence model used to generate questions for a user and analyze the user's answers.

[1208] The "means for generating questions" is a function that uses a generative model to automatically create questions appropriate for the user.

[1209] The "means for analyzing answers" is a function that analyzes answers obtained from users using a generative model and extracts information for generating the next question.

[1210] The "means for generating the next question" is a function for generating the next question to be asked based on the analyzed answer.

[1211] The "means for recording questions and analyzed answers" is a function for saving the generated questions, the user's answers, and the analysis results thereof.

[1212] The "means for analyzing recorded data and generating a report" is a function for analyzing saved interview data using a generative model and compiling the results in a report format.

[1213] The "means for generating initial questions" is a function for generating initial questions based on the user's profile information at the initial stage of the interview.

[1214] "Natural language processing technology" is an artificial intelligence technology for understanding and generating language data, and is used in generative model algorithms.

[1215] The present invention relates to a user interview system using a generative model. Specific embodiments of the system will be described below.

[1216] Hardware and Software Use

[1217] The system's main operations are carried out by a server, a device, and a user. The system utilizes a generative model (e.g., GPT-4) running on the server and a device (e.g., a PC or smartphone) accessed by the user. The server also uses software incorporating analysis algorithms using natural language processing techniques. A specific example of such software is the Transformers library from Hugging Face.

[1218] Explanation of program processing

[1219] The operation of the system can be divided into the following main stages: data collection, model training, interview initiation, question and answer cycle, data storage and analysis, and report generation.

[1220] Data collection

[1221] The server collects the datasets required for the interview, including past interview transcripts, questionnaires, and user profile information.

[1222] Model Training

[1223] The server trains the generative AI model based on the collected data. Natural language processing technology is used for training, and algorithms for question generation and answer analysis are built. GPU resources are used to improve processing speed.

[1224] Interview begins

[1225] Users access the system using a terminal and participate in the interview through a web interface. Users click on a URL and connect to the system from their browser.

[1226] The terminal displays a user registration form, and the user enters information such as name, age, interests, etc. The entered information is sent to the server and stored in a database.

[1227] The server generates initial questions based on the user's profile information and sends them to the terminal, which then displays the received questions to the user.

[1228] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[1229] Repeated questions and answers

[1230] The user responds to the questions on the terminal, enters the answer, and presses the send button. The answer is then sent to the server.

[1231] The server uses the generative AI model to analyze the user's answers and generate the next question, which is again sent to the device and displayed to the user.

[1232] For example: "Tell me more about the design. What did you particularly like about it?"

[1233] Data storage and analysis

[1234] The server stores all interview data in a database. After the interview is completed, the stored data is analyzed using a generative model to extract indicators such as frequently occurring keywords and sentiment analysis.

[1235] Report Generation

[1236] The server compiles the analysis results into a report, which is saved in PDF or Excel format and includes a download link for the report to be provided to the company.

[1237] As an example, the following prompt sentence is used:

[1238] User information: Name = Taro, Age = 30, Interests = Technology

[1239] Question: "Hello, Taro. Today I'd like to ask you about a new product. First of all, what are your impressions of this product?"

[1240] As described above, this invention enables efficient and accurate user interviews using generative AI models.

[1241] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1242] Step 1: Data collection

[1243] The server collects a dataset for use in the interviews, which includes past interview transcripts, questionnaires, and user profile information.

[1244] Input: Database information, interview transcripts

[1245] Output: Collected dataset

[1246] What it does: The server accesses public and internal company databases, collects and stores the required data.

[1247] Step 2: Model training

[1248] The server trains the generative AI model based on the collected data. During training, natural language processing technology is used to build algorithms for question generation and answer analysis.

[1249] Input: Collected dataset

[1250] Output: A trained generative AI model

[1251] How it works: The server pre-processes the data and trains a generative AI model using natural language processing libraries (e.g., Transformers from Hugging Face), leveraging GPU resources to speed up training.

[1252] Step 3: System Access

[1253] A user accesses the system using a terminal and clicks on a URL to reach the web interface.

[1254] Input: URL

[1255] Output: Access to the system

[1256] How it works: A user opens a browser and enters the provided URL to log into the system.

[1257] Step 4: User Registration

[1258] The device displays a user registration form, where the user enters information such as name, age, and interests. The information is sent to the server and stored in a database.

[1259] Input: User-entered information (name, age, interests)

[1260] Output: Saved user information

[1261] How it works: A registration form is displayed, the user fills in the fields, presses the submit button, and the data is sent to the server, which receives it and stores it in a database.

[1262] Step 5: Generate initial questions

[1263] The server generates an initial question based on the user information and sends it to the terminal, which then displays the received question to the user.

[1264] Input: User information

[1265] Output: Initial question

[1266] How it works: The server uses the user's profile information to generate prompts and input them into the AI ​​model. The generated initial questions are sent to the device and displayed.

[1267] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[1268] Step 6: Repeat the questions and answers

[1269] The user enters an answer to a question and presses the submit button. The answer is sent to the server and analyzed using the generative AI model. The next question is generated and sent again to the device. The device displays the next question to the user.

[1270] Input: User's answer

[1271] Output: Next question

[1272] How it works: When a user enters an answer and presses the submit button, the answer is sent to the server, which uses a generative AI model to analyze the answer and generate the next question. This question is then sent to the device and displayed to the user.

[1273] For example: "Tell me more about the design. What did you particularly like about it?"

[1274] Step 7: Data storage and analysis

[1275] The server stores all the interview data, which is then analyzed using a generative model to extract metrics such as frequently occurring keywords and sentiment analysis.

[1276] Input: Interview data

[1277] Output: Analysis results (frequent keywords, sentiment analysis, etc.)

[1278] How it works: The server stores all interview data and analyzes it using a generative model. Various indicators are extracted using natural language processing techniques.

[1279] Step 8: Generate reports

[1280] The server compiles the analysis results in a report format, which is saved in PDF or Excel format and a download link is provided.

[1281] Input: Analysis results

[1282] Output: Report (PDF or Excel format)

[1283] How it works: The server documents the analysis results and generates a report, which is then provided to the company as a download link, allowing them to easily retrieve the information they need.

[1284] In this way, by going through each processing step, efficient and accurate user interviews can be conducted using generative AI models.

[1285] (Application example 1)

[1286] 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."

[1287] In conventional user interview systems, the interview process is conducted manually, which is extremely time-consuming and labor-intensive, making it difficult to collect data efficiently. Furthermore, analyzing the collected data is also time-consuming, making it difficult to utilize feedback in real time. Furthermore, there was a lack of an effective way to integrate and analyze feedback from many different customers, making it difficult to extract customer insights.

[1288] 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.

[1289] In this invention, the server includes: means for generating questions for users using a generative model; means for analyzing responses from users; means for generating subsequent questions based on the analyzed responses; means for recording the generated questions and analyzed responses; means for analyzing the recorded data and generating a report; means for users to access the system using a terminal and participate in an interview; means for the terminal to display questions and acquire responses from users; and means for providing a report generated after the interview is completed to an administrator. This automates the entire interview process, enabling efficient and accurate data collection. Furthermore, the collected data is analyzed in real time, and feedback and improvement suggestions are provided immediately based on the results, enabling rapid response.

[1290] A "generative model" is a model that generates new data or information based on existing data, and primarily uses machine learning and artificial intelligence techniques.

[1291] "User" refers to a user who accesses the system and participates in an interview.

[1292] "Means for generating questions" refers to the function of automatically creating questions to be asked during interviews using a generative model.

[1293] "Means for analyzing responses" refers to the function of receiving responses from users, understanding their contents, and analyzing them.

[1294] "Means for generating the next question" refers to a function for generating the next question to be asked based on the previously analyzed answer.

[1295] "Means for recording generated questions and analyzed answers" refers to a function for saving questions generated during an interview, the user's answers, and the analysis results in a database or the like.

[1296] "Means for generating reports" refers to the function of compiling analysis results based on recorded data and automatically creating reports to be provided to managers and personnel in charge.

[1297] "Device" refers to the device used by the user when participating in the interview, including smartphones and tablets.

[1298] "Means of participating in an interview" refers to the process by which a user accesses the system using a terminal, inputs the information required for the interview, and interacts with the system.

[1299] "Means for displaying a question and obtaining a user's response" refers to a function for generating a question to be displayed on a terminal and receiving a response entered by a user in response to that question.

[1300] "Means to provide to the manager" refers to a function for distributing the report generated after the interview to the relevant parties and using it for feedback and consideration of improvement measures.

[1301] This invention relates to a user interview system using generative models. This system improves the efficiency of collecting and analyzing customer feedback in physical stores, and can be used to immediately improve products and services.

[1302] The system mainly includes the following components:

[1303] 1. Generative model: A generative model that uses natural language processing technology and has the ability to automatically generate questions for users.

[1304] 2. Analysis server: It has the means to analyze the answers from the user and generate the next question, as well as the means to record the generated questions and analyzed answers.

[1305] 3. Interview device: A terminal accessed by the user (smartphone, tablet, etc.) that has the function of displaying questions and obtaining responses from the user.

[1306] 4. Database: A storage device for storing analysis results and interview records.

[1307] 5. Report generation tool: Provides a means to automatically generate reports based on recorded data and provide them to managers and personnel in charge.

[1308] System Operation

[1309] 1. Preparation Phase

[1310] The server collects past interview data, questionnaires, user profile information, etc. for use in the interview.

[1311] The server trains a generative model based on the collected data and builds algorithms for question generation and answer analysis, primarily using natural language processing technology.

[1312] 2. Begin the interview

[1313] The user accesses the system using the interview device, for example, by clicking on a URL that takes the user to a web interface.

[1314] The interview device displays a user registration form, and the user enters information such as name, age, and interests. This information is sent to the server and stored in a database.

[1315] The server generates initial questions based on the collected user profile information and sends them to the interview device, such as, "Hello, [user name]. Today I'd like to ask you about a new product. First of all, what is your impression of this product?"

[1316] 3. Question and answer repetition

[1317] The user enters answers into the interview device, which are then sent to the server.

[1318] The server analyzes the answer using the generative model and generates the next question. At this time, the server performs the analysis and generation process in the form of a prompt such as "User's answer: The design is good, but the operation feels a little complicated. Prompt: Please generate the next question based on this answer."

[1319] New questions are generated and sent to the interview device, such as "Tell me more about the operation. What parts did you find particularly difficult to understand?"

[1320] This process repeats until the user responds "Done."

[1321] 4. Data analysis and report generation

[1322] After the interview is completed, the server analyzes all collected data and categorizes and summarizes the user's opinions.

[1323] The server compiles the analysis results in a report and provides it to the administrator or person in charge. For example, the report may include a statement such as, "Many users rate the design highly, but feel that further explanation is needed regarding the functionality."

[1324] Specific examples

[1325] This system can be used to efficiently collect customer feedback about new products in physical stores. For example, suppose you want to obtain feedback about a new microwave. The initial question displayed is, "Hello, [user name]. Today I'd like to ask you about our new microwave. First, what is your impression of this product?" If the user responds, "The design is good, but I found the operation a little complicated," the system generates a next question: "Tell me more about the operation. Which parts did you find particularly difficult to understand?" This allows the system to collect specific user opinions. Finally, a report is generated based on the collected data, enabling efficient product improvements.

[1326] In this way, the present invention uses a generative AI model to improve the efficiency and accuracy of customer interviews in physical stores.

[1327] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1328] Step 1: Preparation Phase

[1329] The server collects a dataset to be used in the interviews. This dataset includes past interview data, a list of questions, and user profile information. The server then trains a generative model based on this data. Specifically, it uses natural language processing technology to build algorithms for question generation and answer analysis.

[1330] Input: Past interview data, questionnaire, user profile information

[1331] Processing: Training the generative model

[1332] Output: A trained generative model

[1333] Step 2: Start the interview

[1334] The user accesses the system using a terminal and participates in the interview. The user clicks on a URL to access the web interface. The terminal displays a user registration form, and the user enters information such as name, age, and interests. The entered information is sent to the server and stored in a database.

[1335] Input: User access and registration information

[1336] Process: Enter and submit user registration information

[1337] Output: User information stored in the database

[1338] Step 3: Generate initial questions

[1339] The server generates initial questions based on the collected user profile information and sends them to the interview device. For example, a question might be, "Hello, [user name]. Today I'd like to ask you about a new product. First of all, what is your impression of this product?"

[1340] Input: User profile information

[1341] Process: Generate initial questions

[1342] Output: Initial questions sent to the interview device

[1343] Step 4: Repeat the questions and answers

[1344] The user enters their answer into the interview device and presses the send button. This is sent to the server. The server uses the generative model to analyze the answer and generate the next question. For example, if the user enters, "User answer: The design is good, but the operation felt a little complicated," the server generates the next question: "Please tell us more about the operation. Which part did you find particularly difficult to understand?"

[1345] Input: User's answer

[1346] Processing: Parsing answers and generating next questions

[1347] Output: The next question sent to the interview device

[1348] Step 5: Record the data

[1349] The server records the questions generated and the analyzed answers, so that the entire interview process is stored in a database for later analysis.

[1350] Input: Generated question, parsed answer

[1351] Process: Recording questions and answers

[1352] Output: Interview data stored in a database

[1353] Step 6: Data analysis and report generation

[1354] After the interviews are completed, the server analyzes all collected data and classifies and summarizes the user opinions. The server then compiles the analysis results into a report and provides it to the manager or person in charge. For example, the report may include statements such as, "Many users rate the design highly, but feel that further explanation is needed regarding the functionality."

[1355] Input: Collected data

[1356] Processing: Data analysis and report generation

[1357] Output: Report

[1358] In this way, a question generation and answer analysis system using generative models enables efficient and accurate interviews and data collection, enabling companies to quickly collect customer feedback and use it to effectively improve their products and services.

[1359] 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.

[1360] This invention relates to a user interview system that combines a generative model and an emotion engine. This system generates questions through dialogue with the user, analyzes the answers, and generates the next question based on the results. It also recognizes the user's emotions and adjusts the dialogue content to obtain more information.

[1361] Program processing description

[1362] 1. Preparation Phase

[1363] Data collection

[1364] The server collects datasets for use in the interviews, including past interview transcripts, questionnaires, user profile information, and sentiment data.

[1365] Model Training

[1366] The server trains the generative model and emotion engine based on the collected data, utilizing natural language processing technology to build algorithms for question generation, response analysis, and emotion analysis.

[1367] 2. Interview begins

[1368] Interview begins

[1369] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[1370] User Registration

[1371] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[1372] The user enters the required information into the form and presses the submit button.

[1373] The server stores the submitted user information in a database.

[1374] Generating initial questions

[1375] The server generates an initial question based on the user information and sends it to the terminal.

[1376] The terminal displays the question received from the server.

[1377] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[1378] 3. Question and answer repetition

[1379] Questions and Answers

[1380] The user enters an answer to the question and presses the send button.

[1381] The server uses a generative model to analyze the user's answers and generate the next question. In addition, it analyzes the user's emotions using an emotion engine and adjusts the next question based on the results.

[1382] Built-in sentiment analysis

[1383] The emotion engine analyzes the user's responses to determine their emotions, and the server adjusts the tone and content of the dialogue based on that emotion information. For example, if the user expresses anxiety, the server changes the questions to be gentler and more reassuring.

[1384] The server generates a new question and sends it to the device.

[1385] For example: "Tell me more about the design. What did you particularly like about it?"

[1386] The terminal displays the new question to the user.

[1387] The user again enters the answer and presses the send button.

[1388] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[1389] 4. Data analysis phase

[1390] Data storage and analysis

[1391] The server stores all the data from the interview.

[1392] The server analyzes the stored data using a generative model and emotion engine to extract various indicators (e.g., frequently occurring keywords, emotion trend analysis, etc.).

[1393] Report Generation

[1394] The server compiles the analysis results in the form of a report.

[1395] The server generates the report and provides it to the company via a download link or similar.

[1396] Specific examples

[1397] Interview example

[1398] User registration and initial inquiries

[1399] 1. The user accesses the system using a terminal and participates in the interview.

[1400] "User Registration" will appear on the screen and the user will enter the required information (name, age, interests, etc.).

[1401] 2. The server generates initial questions based on the user's profile.

[1402] You might see an initial question like, "Hello, [user's name]. Today I'd like to talk to you about a new product. First, what are your impressions of this product?"

[1403] Dialogue Progress

[1404] 3. The user enters the answer

[1405] User: "I like the design, but the functionality is a bit unclear."

[1406] 4. The server analyzes the user's answer and generates the next question

[1407] The next question will be displayed, such as, "Tell us more about the design. What did you like most about it?" In addition, an emotion engine will analyze the emotions from the user's answers and adjust the dialogue accordingly.

[1408] 5. The user enters the answer again

[1409] User: "I especially like the grip and the color combination."

[1410] Data Analysis and Report Generation

[1411] 6. After the interview is over, the server analyzes the collected data.

[1412] The server analyzes the interview transcript, categorizes the user's opinions, and generates a summary.

[1413] 7. The server compiles the analysis results into a report

[1414] Companies are provided with a report containing analysis results such as, "Many users give the design high marks, but feel that further explanation is needed regarding functionality."

[1415] In this way, the entire interview process is automated using generative models and emotion engines, reducing costs and enabling more accurate information collection. Emotion analysis improves the user experience by enabling appropriate responses based on the user's emotional state.

[1416] The processing flow will be explained below.

[1417] Step 1:

[1418] The server collects datasets for use in the interviews, including past interview transcripts, questionnaires, user profile information, and sentiment data.

[1419] Step 2:

[1420] The server trains the generative model and emotion engine based on the collected data, utilizing natural language processing technology to build algorithms for question generation, response analysis, and emotion analysis.

[1421] Step 3:

[1422] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[1423] Step 4:

[1424] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[1425] Step 5:

[1426] The user enters the required information into the form and presses the submit button.

[1427] Step 6:

[1428] The server stores the submitted user information in a database.

[1429] Step 7:

[1430] The server generates an initial question based on the user information and sends it to the terminal.

[1431] Step 8:

[1432] The terminal displays the question received from the server.

[1433] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[1434] Step 9:

[1435] The user enters an answer to the question and presses the send button.

[1436] Step 10:

[1437] The server uses a generative model to analyze the user's answers and generate the next question. In addition, it analyzes the user's emotions using an emotion engine and adjusts the next question based on the results.

[1438] Step 11:

[1439] The emotion engine analyzes the user's responses to determine their emotions, and the server adjusts the tone and content of the dialogue based on that emotion information. For example, if the user expresses anxiety, the server changes the questions to be gentler and more reassuring.

[1440] Step 12:

[1441] The server generates a new question and sends it to the device.

[1442] For example: "Tell me more about the design. What did you particularly like about it?"

[1443] Step 13:

[1444] The terminal displays the new question to the user.

[1445] Step 14:

[1446] The user again enters the answer and presses the send button.

[1447] Step 15:

[1448] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[1449] Step 16:

[1450] The server stores all the data from the interview.

[1451] Step 17:

[1452] The server analyzes the stored data using a generative model and emotion engine to extract various indicators (e.g., frequently occurring keywords, emotion trend analysis, etc.).

[1453] Step 18:

[1454] The server compiles the analysis results in the form of a report.

[1455] Step 19:

[1456] The server generates the report and provides it to the company via a download link or similar.

[1457] This will automate the entire interview process using generative models, reducing costs and enabling more accurate information collection. Furthermore, emotion analysis will enable appropriate responses based on the user's emotional state, improving the user experience.

[1458] Example 2

[1459] 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."

[1460] Conventional interview systems simply analyze user responses and generate the next questions, but are unable to properly consider the user's emotions or the context of the conversation. This results in poor interview quality and limited information being obtained. It is also difficult to improve the user experience, and efficient information gathering is required.

[1461] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for generating questions for the user using a generative model, means for analyzing answers from the user, means for generating the next question based on the analyzed answers, means for recording the generated questions and the analyzed answers, means for analyzing the recorded data and generating a report, means for training the generative model and the emotion analysis engine based on the collected data, and means for analyzing emotions from the user answers and adjusting the content of the dialogue. This makes it possible to conduct high-quality interviews that take user emotions into consideration and to collect detailed information efficiently.

[1462] A "generative model" is a model that uses machine learning algorithms to generate new data or information.

[1463] "Means for generating questions" refers to a function that uses a generative model to create questions for the user.

[1464] "Means for analyzing responses" refers to the technology used to analyze responses from users and understand and evaluate their content.

[1465] The "means for generating the next question" refers to a function for generating the next appropriate question based on the analyzed answer.

[1466] "Means for recording" refers to the technology used to store the generated questions and analyzed answers in a database or storage.

[1467] "Means for generating reports" refers to a function that analyzes recorded data and outputs the results in the form of a report.

[1468] An "emotion analysis engine" refers to an algorithm or software that analyzes the emotions of users' responses and determines their emotional state.

[1469] "Means for adjusting the content of the dialogue" refers to the function for appropriately changing and adjusting the tone and content of the dialogue based on the results of sentiment analysis.

[1470] MODE FOR CARRYING OUT THE INVENTION

[1471] This invention relates to a user interview system that combines a generative model and an emotion analysis engine. This system generates questions through dialogue with the user, analyzes the answers, and generates the next question based on the results. It also recognizes the user's emotions and adjusts the dialogue content to obtain more information.

[1472] The system includes the following components:

[1473] Generative Model

[1474] Sentiment Analysis Engine

[1475] Database

[1476] User Interface (Terminal)

[1477] Hardware and software used

[1478] The system uses the following hardware and software:

[1479] Servers: Used for data processing and model training. Example: Cloud servers (AWS EC2, Google Cloud)

[1480] Database: Used to store user information and interview data. Example: MySQL, MongoDB

[1481] Generative models: used for question generation and answer analysis. Example: GPT-4 (OpenAI)

[1482] Sentiment analysis engine: Used for sentiment analysis. Example: Microsoft Azure's Text Analytics API

[1483] User Interface (Terminal): The interface through which the user interacts with the system. Example: A web interface created with React.

[1484] Specific operation of the system

[1485] 1. Data collection and model training:

[1486] The server collects and stores in a database a dataset for use in the interviews, including past interview transcripts, questionnaires, user profile information, and sentiment data.

[1487] The server trains the generative model and sentiment analysis engine based on the collected data. Natural language processing technology is used for training, and algorithms for question generation, answer analysis, and sentiment analysis are built using Python libraries (TensorFlow, PyTorch).

[1488] 2. Interview begins:

[1489] A user accesses the system using a terminal and accesses the web interface.

[1490] The device displays a user registration form and asks the user to enter information such as name, age, interests, etc. The user enters the required information into the form and presses the submit button.

[1491] The server stores the user information in a database, generates an initial question, and sends it to the terminal.

[1492] The terminal displays the initial question received from the server.

[1493] Examples:

[1494] "Hi, [user name]. Today I'd like to ask you about a new product. First, what are your impressions of this product?"

[1495] 3. Repeated questions and answers:

[1496] The user enters an answer to the question and presses the send button.

[1497] The server uses a generative model to analyze the user's answers and generate the next question. The emotion analysis engine analyzes the user's emotions and adjusts the next question based on the results. For example, if the user expresses anxiety, the question will be changed to a gentler, more reassuring one.

[1498] The server generates a new question and sends it to the terminal. The terminal displays the new question. The user again enters the answer and presses the send button.

[1499] Examples:

[1500] "Tell us more about the design. What parts did you particularly like?"

[1501] 4. Data Storage and Analysis:

[1502] The server stores all the data from the interview.

[1503] The server analyzes the stored data using a generative model and sentiment analysis engine to extract various indicators (e.g., frequently occurring keywords, sentiment trend analysis, etc.).

[1504] The server compiles the analysis results in a report format that is easy to understand visually (e.g., PDF or dashboard).

[1505] 5. Report provision:

[1506] The server provides the generated report to the company via a download link or the like.

[1507] As described above, the system of the present invention uses a generative AI model and an emotion analysis engine to realize high-quality interviews that take into account the user's emotions, enabling efficient and detailed information collection.

[1508] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1509] Step 1:

[1510] Data Collection Phase

[1511] What it does: The server collects a dataset to use in the interviews. This includes past interview transcripts, questionnaires, user profile information, and sentiment data. The server uses an API to download the data from cloud storage and saves it to local storage.

[1512] Input: Interview transcripts in cloud storage, questionnaire, user profile information, and sentiment data.

[1513] Output: The initial dataset saved to local storage.

[1514] Step 2:

[1515] Data Preprocessing

[1516] Specific operation: The server preprocesses the collected data, specifically by using Python libraries (numpy, pandas, NLTK, etc.) to perform data cleaning (noise removal), text tokenization, and stop word removal.

[1517] Input: An initial dataset stored in local storage.

[1518] Output: A preprocessed and clean dataset.

[1519] Step 3:

[1520] Building and training the model

[1521] How it works: The server trains a generative model (e.g., GPT-4) and a sentiment analysis engine based on the preprocessed data. It uses libraries such as TensorFlow and PyTorch. The server exports the trained model to a file and uploads it to cloud storage.

[1522] Input: Preprocessed clean dataset.

[1523] Output: A trained generative model and a sentiment analysis engine.

[1524] Step 4:

[1525] User access and interview start

[1526] Specific operation: A user accesses the system using a terminal and accesses a web interface (e.g., developed with React). The user is redirected to the homepage.

[1527] Input: The user's access request.

[1528] Output: Web interface display.

[1529] Step 5:

[1530] Displaying the user registration form

[1531] Specific operation: The device displays a user registration form. The registration form includes fields for entering name, age, interests, etc. After the user enters the information on the device, the form data is sent to the server.

[1532] Input: User information (name, age, interests).

[1533] Output: The user information sent to the server.

[1534] Step 6:

[1535] Saving user information and generating initial questions

[1536] Specific operation: The server stores user information in a database (e.g., MySQL, MongoDB). It generates initial questions based on the user information and sends them to the device. A generative AI model is used for generation.

[1537] Input: The user information sent to the server.

[1538] Output: User information stored in the database and the initial questions generated.

[1539] Step 7:

[1540] Display initial questions

[1541] Specific behavior: The device displays the initial question received from the server. Example: "Hello, [user name]. Today I'd like to ask you about a new product. First, what are your impressions of this product?"

[1542] Input: The initial question sent by the server.

[1543] Output: The initial question displayed on the user interface.

[1544] Step 8:

[1545] User answer input

[1546] Specific operation: The user enters an answer to a question using the terminal and presses the send button. The user enters an answer in the text area and clicks the send button.

[1547] Input: The user's answer.

[1548] Output: The user's answer sent to the server.

[1549] Step 9:

[1550] Analyzing answers and generating the next question

[1551] Specific operation: The server analyzes the user's answer using a generative model and generates the next question. It then analyzes the user's emotions using a sentiment analysis engine and adjusts the next question accordingly. For example, if the user expresses anxiety, it changes the question to be more gentle. The generated question is then sent to the device.

[1552] Input: The user's answer.

[1553] Output: The parsed answer and the newly generated question.

[1554] Step 10:

[1555] View new questions and re-answer

[1556] Specific operation: The terminal displays the new question received from the server. The user enters the answer again and presses the send button. This causes steps 8 to 10 to be repeated.

[1557] Input: The new question sent by the server.

[1558] Output: New question displayed on the user interface, user's re-answer.

[1559] Step 11:

[1560] Interview data storage

[1561] What happens: The server stores all the interview questions and answers in a database.

[1562] Input: Interview data (all questions and answers).

[1563] Output: Interview data stored in a database.

[1564] Step 12:

[1565] Data analysis and reporting

[1566] Specific operation: The server analyzes the stored data using a generative model and sentiment analysis engine to extract various indicators (e.g., frequently used keywords, sentiment trend analysis, etc.), compiles the results in a report format, exports the generated report to a file, and uploads it to cloud storage.

[1567] Input: Saved interview data.

[1568] Output: The generated report.

[1569] Step 13:

[1570] Providing reports

[1571] Specific operation: The server provides the generated report to the company via a download link, etc.

[1572] Input: The generated report.

[1573] Output: A download link that will be provided to the company.

[1574] (Application example 2)

[1575] 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."

[1576] Conventional user interview systems are limited to one-way question-and-answer exchanges, which means they are unable to fully reflect user emotions and detailed feedback. This makes it difficult to accurately grasp actual user experiences and opinions about products. Furthermore, they often fail to generate appropriate questions or adjust the dialogue in response to user emotions, resulting in a poor user experience.

[1577] 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.

[1578] In this invention, the server includes means for generating questions for a user using a generative model, means for analyzing answers from the user, means for generating a next question based on the analyzed answer, means for recording the generated question and the analyzed answer, means for analyzing the recorded data and generating a report, means for analyzing the user's emotions, and means for adjusting the next question based on the result of the user's emotion analysis. This makes it possible to collect detailed and accurate feedback while adjusting the content of the dialogue according to the user's emotions.

[1579] A "generative model" is an algorithm that automatically generates questions and information based on user input.

[1580] The "means for analyzing the answer" is a technique for understanding the answer entered by the user and analyzing its content.

[1581] The "means for generating questions" is a technology that automatically creates the next question based on the analyzed answer.

[1582] The "recording means" is a technique for saving the generated questions and analyzed answers in a storage device such as a database.

[1583] The "means for generating a report" is a technique for analyzing the recorded data and outputting the results in the form of a report.

[1584] "Means for analyzing emotions" refers to technology that determines emotions from users' responses and quantifies the results.

[1585] The "means for adjusting the next question based on the emotion analysis result" is a technology that appropriately changes the content and wording of the next question based on the analyzed emotion.

[1586] "Emotion trend analysis" is a technique that analyzes the transitions in a user's emotions and identifies changes and trends in emotions over time.

[1587] "Means for collecting feedback" refers to techniques for collecting opinions and evaluations from users about products they have purchased.

[1588] The present invention relates to a user interview system that combines a generative model and an emotion engine, and is intended in particular for collecting feedback on product evaluations on online shopping sites. A specific embodiment of the system that realizes this application example is described below.

[1589] System configuration

[1590] Hardware

[1591] Device: The smartphone or smart glasses used by the user.

[1592] Server: A server for data processing, question generation, answer analysis, sentiment analysis, and data storage.

[1593] software

[1594] Generative model: Used to generate questions and analyze user responses. Specifically, it uses natural language processing technologies such as OpenAI's GPT-3.

[1595] Sentiment engine: Software for analyzing emotions from user responses. Specifically, it uses a sentiment analysis model using the Transformers library.

[1596] Database: A database for storing user profile information, interview results, sentiment analysis results, etc.

[1597] Feedback Collection Process

[1598] Initial Setup

[1599] The server first collects user profile information (such as name, purchase history, etc.) and stores it in a database, then uses a generative model to create initial questions.

[1600] Interview begins

[1601] The user accesses the system from a terminal and the interview begins. Initial questions are displayed and the user submits their answers through an input form.

[1602] Question and answer exchange

[1603] The server analyzes the user's answers and uses an emotion engine to analyze the emotions from the answers. Based on the analysis results, the server generates the next question and adjusts the question content according to the user's emotions. This allows the server to obtain detailed feedback from the user.

[1604] Data storage and analysis

[1605] After all interactions are completed, the server stores the collected data in a database and analyzes the user's emotional trends from the recorded data. Based on the analysis results, the server generates a report and provides it to the company via a download link or other means.

[1606] Specific examples

[1607] A user uses a smartphone app to provide feedback on a product they have purchased. The first question is displayed in the form, "Hello, [user name]. Please rate the product you purchased. What was your first impression?" As the user enters their answer, the answer is analyzed in real time and the next question is generated based on their emotional state. For example, the question is adjusted accordingly, such as, "Tell me more about the design. What did you particularly like about it?"

[1608] Prompt Sentence Examples

[1609] User Answer: I like the design, but the functionality is a bit unclear.

[1610] Emotion: Negative

[1611] Generate the following questions:

[1612] In this way, a system is realized that adjusts the dialogue content according to the user's emotions and collects detailed and accurate feedback.

[1613] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1614] Step 1: User registration and initial setup

[1615] A user accesses the system using a terminal and enters profile information such as the user's name and purchase history. The server stores this input data in a database and generates initial questions using a generative model. The generated initial questions are in the form of, for example, "Please rate the product you purchased. What was your first impression?"

[1616] Step 2: Displaying initial questions

[1617] The server sends the generated initial question to the terminal, which displays the question to the user, and the user inputs an answer to the question and sends it to the server via the terminal.

[1618] Step 3: Response analysis and sentiment analysis

[1619] The server receives the answer sent by the user and analyzes the answer using a generative model. At the same time, it analyzes the user's sentiment from the answer using an emotion engine. For example, if a user answers "I think the design is good, but the functionality is a little unclear," the emotion analysis engine will detect negative sentiment.

[1620] Step 4: Generate the next question

[1621] The server generates the next question based on the analyzed answer and emotion data. It uses a generative model to generate a question that corresponds to the user's emotion. For example, if a negative emotion is detected, it generates a question such as, "What specific aspects of the feature did you find unclear?" The generated question is sent from the server to the device and displayed to the user.

[1622] Step 5: Conducting the interview

[1623] The user answers the next question and sends the answer to the server via their device. This process repeats, with the server analyzing each answer and generating the next question. Based on the sentiment analysis results, the question content is adjusted accordingly.

[1624] Step 6: Data storage and final analysis

[1625] After the interview is over, the server stores all questions and answers, as well as the sentiment analysis results, in a database. The recorded data is then analyzed again for sentiment trend analysis, summarizing changes in users' sentiment and opinion trends.

[1626] Step 7: Generate reports

[1627] The server generates reports based on the accumulated data, including analysis results such as classification of user opinions and sentiment trends. The generated reports are provided to the company via a download link or other means.

[1628] In this way, detailed and accurate feedback can be collected from users, and appropriate questions can be generated based on the results of sentiment analysis, making the interviews more effective.

[1629] 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.

[1630] 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.

[1631] 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.

[1632] [Fourth embodiment]

[1633] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1634] 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.

[1635] 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).

[1636] 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.

[1637] 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.

[1638] 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).

[1639] 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.

[1640] 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.

[1641] 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.

[1642] 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.

[1643] 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.

[1644] 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.

[1645] 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."

[1646] This invention relates to a user interview system that uses a generative model. This system generates questions through dialogue with the user, analyzes the answers, and generates subsequent questions based on the results, enabling efficient and accurate data collection. Specific program processing is explained in natural language and includes examples.

[1647] Program processing description

[1648] 1. Preparation Phase

[1649] Data collection

[1650] The server collects data sets for use in the interviews, including past interview transcripts, questionnaires, and user profile information.

[1651] Model Training

[1652] The server trains the generative model based on the collected data, using natural language processing technology to build algorithms for question generation and answer analysis.

[1653] 2. Interview begins

[1654] Interview begins

[1655] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[1656] User Registration

[1657] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[1658] The user enters the required information into the form and presses the submit button.

[1659] The server stores the submitted user information in a database.

[1660] Generating initial questions

[1661] The server generates an initial question based on the user information and sends it to the terminal.

[1662] The terminal displays the question received from the server.

[1663] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[1664] 3. Question and answer repetition

[1665] Questions and Answers

[1666] The user enters an answer and presses the send button.

[1667] The server uses the generative model to analyze the user's answers and generate the next question.

[1668] The server generates a new question and sends it to the device.

[1669] For example: "Tell me more about the design. What did you particularly like about it?"

[1670] The terminal displays the new question to the user.

[1671] The user again enters the answer and presses the send button.

[1672] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[1673] 4. Data analysis phase

[1674] Data storage and analysis

[1675] The server stores all the data from the interview.

[1676] The server analyzes the stored data using a generative model and extracts various indicators (e.g., frequently occurring keywords, sentiment analysis, etc.).

[1677] Report Generation

[1678] The server compiles the analysis results in the form of a report.

[1679] The server generates the report and provides it to the company via a download link or similar.

[1680] Specific examples

[1681] Interview example

[1682] User registration and initial inquiries

[1683] 1. The user accesses the system using a terminal and participates in the interview.

[1684] "User Registration" will appear on the screen and the user will enter the required information (name, age, interests, etc.).

[1685] 2. The server generates initial questions based on the user's profile.

[1686] You might see an initial question like, "Hello, [user's name]. Today I'd like to talk to you about a new product. First, what are your impressions of this product?"

[1687] Dialogue Progress

[1688] 3. The user enters the answer

[1689] User: "I like the design, but the functionality is a bit unclear."

[1690] 4. The server analyzes the user's answer and generates the next question

[1691] The next question will be displayed, such as "Tell us more about the design. What did you particularly like about it?"

[1692] 5. The user enters the answer again

[1693] User: "I especially like the grip and the color combination."

[1694] Data Analysis and Report Generation

[1695] 6. After the interview is over, the server analyzes the collected data.

[1696] The server analyzes the interview transcript, categorizes the user's opinions, and generates a summary.

[1697] 7. The server compiles the analysis results into a report

[1698] Companies are provided with a report containing analysis results such as, "Many users give the design high marks, but feel that further explanation is needed regarding functionality."

[1699] In this way, the entire interview process is automated using generative models, reducing costs and enabling more accurate information collection.

[1700] The processing flow will be explained below.

[1701] Step 1:

[1702] The server collects data sets for use in the interviews, including past interview transcripts, questionnaires, and user profile information.

[1703] Step 2:

[1704] The server trains the generative model based on the collected data, using natural language processing technology to build algorithms for question generation and answer analysis.

[1705] Step 3:

[1706] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[1707] Step 4:

[1708] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[1709] Step 5:

[1710] The user enters the required information into the form and presses the submit button.

[1711] Step 6:

[1712] The server stores the submitted user information in a database.

[1713] Step 7:

[1714] The server generates an initial question based on the user information and sends it to the terminal.

[1715] Step 8:

[1716] The terminal displays the question received from the server.

[1717] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[1718] Step 9:

[1719] The user enters an answer to the question and presses the send button.

[1720] Step 10:

[1721] The server uses the generative model to analyze the user's answers and generate the next question.

[1722] Step 11:

[1723] The server generates a new question and sends it to the device.

[1724] For example: "Tell me more about the design. What did you particularly like about it?"

[1725] Step 12:

[1726] The terminal displays the new question to the user.

[1727] Step 13:

[1728] The user again enters the answer and presses the send button.

[1729] Step 14:

[1730] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[1731] Step 15:

[1732] The server stores all the data from the interview.

[1733] Step 16:

[1734] The server analyzes the stored data using a generative model and extracts various indicators (e.g., frequently occurring keywords, sentiment analysis, etc.).

[1735] Step 17:

[1736] The server compiles the analysis results in the form of a report.

[1737] Step 18:

[1738] The server generates the report and provides it to the company via a download link or similar.

[1739] This automates the entire interview process using generative models, reducing costs and enabling more accurate information collection.

[1740] Example 1

[1741] 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."

[1742] In conventional user interview systems, questions are generated and answers are analyzed manually, resulting in inefficiency and limited accuracy. Data storage and analysis are also time-consuming and costly. Furthermore, there is a lack of tools and methods for effectively utilizing interview data, making comprehensive analysis difficult.

[1743] 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.

[1744] In this invention, the server includes means for generating questions for the user using a generative model, means for analyzing answers from the user, and means for generating the next question based on the analyzed answers. This automates the entire interview process, enabling efficient and accurate data collection.

[1745] A "generative model" is an artificial intelligence model used to generate questions for a user and analyze the user's answers.

[1746] The "means for generating questions" is a function that uses a generative model to automatically create questions appropriate for the user.

[1747] The "means for analyzing answers" is a function that analyzes answers obtained from users using a generative model and extracts information for generating the next question.

[1748] The "means for generating the next question" is a function for generating the next question to be asked based on the analyzed answer.

[1749] The "means for recording questions and analyzed answers" is a function for saving the generated questions, the user's answers, and the analysis results thereof.

[1750] The "means for analyzing recorded data and generating a report" is a function for analyzing saved interview data using a generative model and compiling the results in a report format.

[1751] The "means for generating initial questions" is a function for generating initial questions based on the user's profile information at the initial stage of the interview.

[1752] "Natural language processing technology" is an artificial intelligence technology for understanding and generating language data, and is used in generative model algorithms.

[1753] The present invention relates to a user interview system using a generative model. Specific embodiments of the system will be described below.

[1754] Hardware and Software Use

[1755] The system's main operations are carried out by a server, a device, and a user. The system utilizes a generative model (e.g., GPT-4) running on the server and a device (e.g., a PC or smartphone) accessed by the user. The server also uses software incorporating analysis algorithms using natural language processing techniques. A specific example of such software is the Transformers library from Hugging Face.

[1756] Explanation of program processing

[1757] The operation of the system can be divided into the following main stages: data collection, model training, interview initiation, question and answer cycle, data storage and analysis, and report generation.

[1758] Data collection

[1759] The server collects the datasets required for the interview, including past interview transcripts, questionnaires, and user profile information.

[1760] Model Training

[1761] The server trains the generative AI model based on the collected data. Natural language processing technology is used for training, and algorithms for question generation and answer analysis are built. GPU resources are used to improve processing speed.

[1762] Interview begins

[1763] Users access the system using a terminal and participate in the interview through a web interface. Users click on a URL and connect to the system from their browser.

[1764] The terminal displays a user registration form, and the user enters information such as name, age, interests, etc. The entered information is sent to the server and stored in a database.

[1765] The server generates initial questions based on the user's profile information and sends them to the terminal, which then displays the received questions to the user.

[1766] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[1767] Repeated questions and answers

[1768] The user responds to the questions on the terminal, enters the answer, and presses the send button. The answer is then sent to the server.

[1769] The server uses the generative AI model to analyze the user's answers and generate the next question, which is again sent to the device and displayed to the user.

[1770] For example: "Tell me more about the design. What did you particularly like about it?"

[1771] Data storage and analysis

[1772] The server stores all interview data in a database. After the interview is completed, the stored data is analyzed using a generative model to extract indicators such as frequently occurring keywords and sentiment analysis.

[1773] Report Generation

[1774] The server compiles the analysis results into a report, which is saved in PDF or Excel format and includes a download link for the report to be provided to the company.

[1775] As an example, the following prompt sentence is used:

[1776] User information: Name = Taro, Age = 30, Interests = Technology

[1777] Question: "Hello, Taro. Today I'd like to ask you about a new product. First of all, what are your impressions of this product?"

[1778] As described above, this invention enables efficient and accurate user interviews using generative AI models.

[1779] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1780] Step 1: Data collection

[1781] The server collects a dataset for use in the interviews, which includes past interview transcripts, questionnaires, and user profile information.

[1782] Input: Database information, interview transcripts

[1783] Output: Collected dataset

[1784] What it does: The server accesses public and internal company databases, collects and stores the required data.

[1785] Step 2: Model training

[1786] The server trains the generative AI model based on the collected data. During training, natural language processing technology is used to build algorithms for question generation and answer analysis.

[1787] Input: Collected dataset

[1788] Output: A trained generative AI model

[1789] How it works: The server pre-processes the data and trains a generative AI model using natural language processing libraries (e.g., Transformers from Hugging Face), leveraging GPU resources to speed up training.

[1790] Step 3: System Access

[1791] A user accesses the system using a terminal and clicks on a URL to reach the web interface.

[1792] Input: URL

[1793] Output: Access to the system

[1794] How it works: A user opens a browser and enters the provided URL to log into the system.

[1795] Step 4: User Registration

[1796] The device displays a user registration form, where the user enters information such as name, age, and interests. The information is sent to the server and stored in a database.

[1797] Input: User-entered information (name, age, interests)

[1798] Output: Saved user information

[1799] How it works: A registration form is displayed, the user fills in the fields, presses the submit button, and the data is sent to the server, which receives it and stores it in a database.

[1800] Step 5: Generate initial questions

[1801] The server generates an initial question based on the user information and sends it to the terminal, which then displays the received question to the user.

[1802] Input: User information

[1803] Output: Initial question

[1804] How it works: The server uses the user's profile information to generate prompts and input them into the AI ​​model. The generated initial questions are sent to the device and displayed.

[1805] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[1806] Step 6: Repeat the questions and answers

[1807] The user enters an answer to a question and presses the submit button. The answer is sent to the server and analyzed using the generative AI model. The next question is generated and sent again to the device. The device displays the next question to the user.

[1808] Input: User's answer

[1809] Output: Next question

[1810] How it works: When a user enters an answer and presses the submit button, the answer is sent to the server, which uses a generative AI model to analyze the answer and generate the next question. This question is then sent to the device and displayed to the user.

[1811] For example: "Tell me more about the design. What did you particularly like about it?"

[1812] Step 7: Data storage and analysis

[1813] The server stores all the interview data, which is then analyzed using a generative model to extract metrics such as frequently occurring keywords and sentiment analysis.

[1814] Input: Interview data

[1815] Output: Analysis results (frequent keywords, sentiment analysis, etc.)

[1816] How it works: The server stores all interview data and analyzes it using a generative model. Various indicators are extracted using natural language processing techniques.

[1817] Step 8: Generate reports

[1818] The server compiles the analysis results in a report format, which is saved in PDF or Excel format and a download link is provided.

[1819] Input: Analysis results

[1820] Output: Report (PDF or Excel format)

[1821] How it works: The server documents the analysis results and generates a report, which is then provided to the company as a download link, allowing them to easily retrieve the information they need.

[1822] In this way, by going through each processing step, efficient and accurate user interviews can be conducted using generative AI models.

[1823] (Application example 1)

[1824] 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."

[1825] In conventional user interview systems, the interview process is conducted manually, which is extremely time-consuming and labor-intensive, making it difficult to collect data efficiently. Furthermore, analyzing the collected data is also time-consuming, making it difficult to utilize feedback in real time. Furthermore, there was a lack of an effective way to integrate and analyze feedback from many different customers, making it difficult to extract customer insights.

[1826] 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.

[1827] In this invention, the server includes: means for generating questions for users using a generative model; means for analyzing responses from users; means for generating subsequent questions based on the analyzed responses; means for recording the generated questions and analyzed responses; means for analyzing the recorded data and generating a report; means for users to access the system using a terminal and participate in an interview; means for the terminal to display questions and acquire responses from users; and means for providing a report generated after the interview is completed to an administrator. This automates the entire interview process, enabling efficient and accurate data collection. Furthermore, the collected data is analyzed in real time, and feedback and improvement suggestions are provided immediately based on the results, enabling rapid response.

[1828] A "generative model" is a model that generates new data or information based on existing data, and primarily uses machine learning and artificial intelligence techniques.

[1829] "User" refers to a user who accesses the system and participates in an interview.

[1830] "Means for generating questions" refers to the function of automatically creating questions to be asked during interviews using a generative model.

[1831] "Means for analyzing responses" refers to the function of receiving responses from users, understanding their contents, and analyzing them.

[1832] "Means for generating the next question" refers to a function for generating the next question to be asked based on the previously analyzed answer.

[1833] "Means for recording generated questions and analyzed answers" refers to a function for saving questions generated during an interview, the user's answers, and the analysis results in a database or the like.

[1834] "Means for generating reports" refers to the function of compiling analysis results based on recorded data and automatically creating reports to be provided to managers and personnel in charge.

[1835] "Device" refers to the device used by the user when participating in the interview, including smartphones and tablets.

[1836] "Means of participating in an interview" refers to the process by which a user accesses the system using a terminal, inputs the information required for the interview, and interacts with the system.

[1837] "Means for displaying a question and obtaining a user's response" refers to a function for generating a question to be displayed on a terminal and receiving a response entered by a user in response to that question.

[1838] "Means to provide to the manager" refers to a function for distributing the report generated after the interview to the relevant parties and using it for feedback and consideration of improvement measures.

[1839] This invention relates to a user interview system using generative models. This system improves the efficiency of collecting and analyzing customer feedback in physical stores, and can be used to immediately improve products and services.

[1840] The system mainly includes the following components:

[1841] 1. Generative model: A generative model that uses natural language processing technology and has the ability to automatically generate questions for users.

[1842] 2. Analysis server: It has the means to analyze the answers from the user and generate the next question, as well as the means to record the generated questions and analyzed answers.

[1843] 3. Interview device: A terminal accessed by the user (smartphone, tablet, etc.) that has the function of displaying questions and obtaining responses from the user.

[1844] 4. Database: A storage device for storing analysis results and interview records.

[1845] 5. Report generation tool: Provides a means to automatically generate reports based on recorded data and provide them to managers and personnel in charge.

[1846] System Operation

[1847] 1. Preparation Phase

[1848] The server collects past interview data, questionnaires, user profile information, etc. for use in the interview.

[1849] The server trains a generative model based on the collected data and builds algorithms for question generation and answer analysis, primarily using natural language processing technology.

[1850] 2. Begin the interview

[1851] The user accesses the system using the interview device, for example, by clicking on a URL that takes the user to a web interface.

[1852] The interview device displays a user registration form, and the user enters information such as name, age, and interests. This information is sent to the server and stored in a database.

[1853] The server generates initial questions based on the collected user profile information and sends them to the interview device, such as, "Hello, [user name]. Today I'd like to ask you about a new product. First of all, what is your impression of this product?"

[1854] 3. Question and answer repetition

[1855] The user enters answers into the interview device, which are then sent to the server.

[1856] The server analyzes the answer using the generative model and generates the next question. At this time, the server performs the analysis and generation process in the form of a prompt such as "User's answer: The design is good, but the operation feels a little complicated. Prompt: Please generate the next question based on this answer."

[1857] New questions are generated and sent to the interview device, such as "Tell me more about the operation. What parts did you find particularly difficult to understand?"

[1858] This process repeats until the user responds "Done."

[1859] 4. Data analysis and report generation

[1860] After the interview is completed, the server analyzes all collected data and categorizes and summarizes the user's opinions.

[1861] The server compiles the analysis results in a report and provides it to the administrator or person in charge. For example, the report may include a statement such as, "Many users rate the design highly, but feel that further explanation is needed regarding the functionality."

[1862] Specific examples

[1863] This system can be used to efficiently collect customer feedback about new products in physical stores. For example, suppose you want to obtain feedback about a new microwave. The initial question displayed is, "Hello, [user name]. Today I'd like to ask you about our new microwave. First, what is your impression of this product?" If the user responds, "The design is good, but I found the operation a little complicated," the system generates a next question: "Tell me more about the operation. Which parts did you find particularly difficult to understand?" This allows the system to collect specific user opinions. Finally, a report is generated based on the collected data, enabling efficient product improvements.

[1864] In this way, the present invention uses a generative AI model to improve the efficiency and accuracy of customer interviews in physical stores.

[1865] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1866] Step 1: Preparation Phase

[1867] The server collects a dataset to be used in the interviews. This dataset includes past interview data, a list of questions, and user profile information. The server then trains a generative model based on this data. Specifically, it uses natural language processing technology to build algorithms for question generation and answer analysis.

[1868] Input: Past interview data, questionnaire, user profile information

[1869] Processing: Training the generative model

[1870] Output: A trained generative model

[1871] Step 2: Start the interview

[1872] The user accesses the system using a terminal and participates in the interview. The user clicks on a URL to access the web interface. The terminal displays a user registration form, and the user enters information such as name, age, and interests. The entered information is sent to the server and stored in a database.

[1873] Input: User access and registration information

[1874] Process: Enter and submit user registration information

[1875] Output: User information stored in the database

[1876] Step 3: Generate initial questions

[1877] The server generates initial questions based on the collected user profile information and sends them to the interview device. For example, a question might be, "Hello, [user name]. Today I'd like to ask you about a new product. First of all, what is your impression of this product?"

[1878] Input: User profile information

[1879] Process: Generate initial questions

[1880] Output: Initial questions sent to the interview device

[1881] Step 4: Repeat the questions and answers

[1882] The user enters their answer into the interview device and presses the send button. This is sent to the server. The server uses the generative model to analyze the answer and generate the next question. For example, if the user enters, "User answer: The design is good, but the operation felt a little complicated," the server generates the next question: "Please tell us more about the operation. Which part did you find particularly difficult to understand?"

[1883] Input: User's answer

[1884] Processing: Parsing answers and generating next questions

[1885] Output: The next question sent to the interview device

[1886] Step 5: Record the data

[1887] The server records the questions generated and the analyzed answers, so that the entire interview process is stored in a database for later analysis.

[1888] Input: Generated question, parsed answer

[1889] Process: Recording questions and answers

[1890] Output: Interview data stored in a database

[1891] Step 6: Data analysis and report generation

[1892] After the interviews are completed, the server analyzes all collected data and classifies and summarizes the user opinions. The server then compiles the analysis results into a report and provides it to the manager or person in charge. For example, the report may include statements such as, "Many users rate the design highly, but feel that further explanation is needed regarding the functionality."

[1893] Input: Collected data

[1894] Processing: Data analysis and report generation

[1895] Output: Report

[1896] In this way, a question generation and answer analysis system using generative models enables efficient and accurate interviews and data collection, enabling companies to quickly collect customer feedback and use it to effectively improve their products and services.

[1897] 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.

[1898] This invention relates to a user interview system that combines a generative model and an emotion engine. This system generates questions through dialogue with the user, analyzes the answers, and generates the next question based on the results. It also recognizes the user's emotions and adjusts the dialogue content to obtain more information.

[1899] Program processing description

[1900] 1. Preparation Phase

[1901] Data collection

[1902] The server collects datasets for use in the interviews, including past interview transcripts, questionnaires, user profile information, and sentiment data.

[1903] Model Training

[1904] The server trains the generative model and emotion engine based on the collected data, utilizing natural language processing technology to build algorithms for question generation, response analysis, and emotion analysis.

[1905] 2. Interview begins

[1906] Interview begins

[1907] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[1908] User Registration

[1909] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[1910] The user enters the required information into the form and presses the submit button.

[1911] The server stores the submitted user information in a database.

[1912] Generating initial questions

[1913] The server generates an initial question based on the user information and sends it to the terminal.

[1914] The terminal displays the question received from the server.

[1915] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[1916] 3. Question and answer repetition

[1917] Questions and Answers

[1918] The user enters an answer to the question and presses the send button.

[1919] The server uses a generative model to analyze the user's answers and generate the next question. In addition, it analyzes the user's emotions using an emotion engine and adjusts the next question based on the results.

[1920] Built-in sentiment analysis

[1921] The emotion engine analyzes the user's responses to determine their emotions, and the server adjusts the tone and content of the dialogue based on that emotion information. For example, if the user expresses anxiety, the server changes the questions to be gentler and more reassuring.

[1922] The server generates a new question and sends it to the device.

[1923] For example: "Tell me more about the design. What did you particularly like about it?"

[1924] The terminal displays the new question to the user.

[1925] The user again enters the answer and presses the send button.

[1926] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[1927] 4. Data analysis phase

[1928] Data storage and analysis

[1929] The server stores all the data from the interview.

[1930] The server analyzes the stored data using a generative model and emotion engine to extract various indicators (e.g., frequently occurring keywords, emotion trend analysis, etc.).

[1931] Report Generation

[1932] The server compiles the analysis results in the form of a report.

[1933] The server generates the report and provides it to the company via a download link or similar.

[1934] Specific examples

[1935] Interview example

[1936] User registration and initial inquiries

[1937] 1. The user accesses the system using a terminal and participates in the interview.

[1938] "User Registration" will appear on the screen and the user will enter the required information (name, age, interests, etc.).

[1939] 2. The server generates initial questions based on the user's profile.

[1940] You might see an initial question like, "Hello, [user's name]. Today I'd like to talk to you about a new product. First, what are your impressions of this product?"

[1941] Dialogue Progress

[1942] 3. The user enters the answer

[1943] User: "I like the design, but the functionality is a bit unclear."

[1944] 4. The server analyzes the user's answer and generates the next question

[1945] The next question will be displayed, such as, "Tell us more about the design. What did you like most about it?" In addition, an emotion engine will analyze the emotions from the user's answers and adjust the dialogue accordingly.

[1946] 5. The user enters the answer again

[1947] User: "I especially like the grip and the color combination."

[1948] Data Analysis and Report Generation

[1949] 6. After the interview is over, the server analyzes the collected data.

[1950] The server analyzes the interview transcript, categorizes the user's opinions, and generates a summary.

[1951] 7. The server compiles the analysis results into a report

[1952] Companies are provided with a report containing analysis results such as, "Many users give the design high marks, but feel that further explanation is needed regarding functionality."

[1953] In this way, the entire interview process is automated using generative models and emotion engines, reducing costs and enabling more accurate information collection. Emotion analysis improves the user experience by enabling appropriate responses based on the user's emotional state.

[1954] The processing flow will be explained below.

[1955] Step 1:

[1956] The server collects datasets for use in the interviews, including past interview transcripts, questionnaires, user profile information, and sentiment data.

[1957] Step 2:

[1958] The server trains the generative model and emotion engine based on the collected data, utilizing natural language processing technology to build algorithms for question generation, response analysis, and emotion analysis.

[1959] Step 3:

[1960] A user accesses the system using a terminal. A user clicks on a URL to access the web interface.

[1961] Step 4:

[1962] The device will display a user registration form, asking you to enter information such as your name, age, and interests.

[1963] Step 5:

[1964] The user enters the required information into the form and presses the submit button.

[1965] Step 6:

[1966] The server stores the submitted user information in a database.

[1967] Step 7:

[1968] The server generates an initial question based on the user information and sends it to the terminal.

[1969] Step 8:

[1970] The terminal displays the question received from the server.

[1971] For example: "Hi, [user name]. Today I'd like to talk to you about a new product. First, what are your thoughts on this product?"

[1972] Step 9:

[1973] The user enters an answer to the question and presses the send button.

[1974] Step 10:

[1975] The server uses a generative model to analyze the user's answers and generate the next question. In addition, it analyzes the user's emotions using an emotion engine and adjusts the next question based on the results.

[1976] Step 11:

[1977] The emotion engine analyzes the user's responses to determine their emotions, and the server adjusts the tone and content of the dialogue based on that emotion information. For example, if the user expresses anxiety, the server changes the questions to be gentler and more reassuring.

[1978] Step 12:

[1979] The server generates a new question and sends it to the device.

[1980] For example: "Tell me more about the design. What did you particularly like about it?"

[1981] Step 13:

[1982] The terminal displays the new question to the user.

[1983] Step 14:

[1984] The user again enters the answer and presses the send button.

[1985] Step 15:

[1986] The server again analyzes the user's answers and generates the necessary questions repeatedly.

[1987] Step 16:

[1988] The server stores all the data from the interview.

[1989] Step 17:

[1990] The server analyzes the stored data using a generative model and emotion engine to extract various indicators (e.g., frequently occurring keywords, emotion trend analysis, etc.).

[1991] Step 18:

[1992] The server compiles the analysis results in the form of a report.

[1993] Step 19:

[1994] The server generates the report and provides it to the company via a download link or similar.

[1995] This will automate the entire interview process using generative models, reducing costs and enabling more accurate information collection. Furthermore, emotion analysis will enable appropriate responses based on the user's emotional state, improving the user experience.

[1996] Example 2

[1997] 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."

[1998] Conventional interview systems simply analyze user responses and generate the next questions, but are unable to properly consider the user's emotions or the context of the conversation. This results in poor interview quality and limited information being obtained. It is also difficult to improve the user experience, and efficient information gathering is required.

[1999] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for generating questions for the user using a generative model, means for analyzing answers from the user, means for generating the next question based on the analyzed answers, means for recording the generated questions and the analyzed answers, means for analyzing the recorded data and generating a report, means for training the generative model and the emotion analysis engine based on the collected data, and means for analyzing emotions from the user answers and adjusting the content of the dialogue. This makes it possible to conduct high-quality interviews that take user emotions into consideration and to collect detailed information efficiently.

[2000] A "generative model" is a model that uses machine learning algorithms to generate new data or information.

[2001] "Means for generating questions" refers to a function that uses a generative model to create questions for the user.

[2002] "Means for analyzing responses" refers to the technology used to analyze responses from users and understand and evaluate their content.

[2003] The "means for generating the next question" refers to a function for generating the next appropriate question based on the analyzed answer.

[2004] "Means for recording" refers to the technology used to store the generated questions and analyzed answers in a database or storage.

[2005] "Means for generating reports" refers to a function that analyzes recorded data and outputs the results in the form of a report.

[2006] An "emotion analysis engine" refers to an algorithm or software that analyzes the emotions of users' responses and determines their emotional state.

[2007] "Means for adjusting the content of the dialogue" refers to the function for appropriately changing and adjusting the tone and content of the dialogue based on the results of sentiment analysis.

[2008] MODE FOR CARRYING OUT THE INVENTION

[2009] This invention relates to a user interview system that combines a generative model and an emotion analysis engine. This system generates questions through dialogue with the user, analyzes the answers, and generates the next question based on the results. It also recognizes the user's emotions and adjusts the dialogue content to obtain more information.

[2010] The system includes the following components:

[2011] Generative Model

[2012] Sentiment Analysis Engine

[2013] Database

[2014] User Interface (Terminal)

[2015] Hardware and software used

[2016] The system uses the following hardware and software:

[2017] Servers: Used for data processing and model training. Example: Cloud servers (AWS EC2, Google Cloud)

[2018] Database: Used to store user information and interview data. Example: MySQL, MongoDB

[2019] Generative models: used for question generation and answer analysis. Example: GPT-4 (OpenAI)

[2020] Sentiment analysis engine: Used for sentiment analysis. Example: Microsoft Azure's Text Analytics API

[2021] User Interface (Terminal): The interface through which the user interacts with the system. Example: A web interface created with React.

[2022] Specific operation of the system

[2023] 1. Data collection and model training:

[2024] The server collects and stores in a database a dataset for use in the interviews, including past interview transcripts, questionnaires, user profile information, and sentiment data.

[2025] The server trains the generative model and sentiment analysis engine based on the collected data. Natural language processing technology is used for training, and algorithms for question generation, answer analysis, and sentiment analysis are built using Python libraries (TensorFlow, PyTorch).

[2026] 2. Interview begins:

[2027] A user accesses the system using a terminal and accesses the web interface.

[2028] The device displays a user registration form and asks the user to enter information such as name, age, interests, etc. The user enters the required information into the form and presses the submit button.

[2029] The server stores the user information in a database, generates an initial question, and sends it to the terminal.

[2030] The terminal displays the initial question received from the server.

[2031] Examples:

[2032] "Hi, [user name]. Today I'd like to ask you about a new product. First, what are your impressions of this product?"

[2033] 3. Repeated questions and answers:

[2034] The user enters an answer to the question and presses the send button.

[2035] The server uses a generative model to analyze the user's answers and generate the next question. The emotion analysis engine analyzes the user's emotions and adjusts the next question based on the results. For example, if the user expresses anxiety, the question will be changed to a gentler, more reassuring one.

[2036] The server generates a new question and sends it to the terminal. The terminal displays the new question. The user again enters the answer and presses the send button.

[2037] Examples:

[2038] "Tell us more about the design. What parts did you particularly like?"

[2039] 4. Data Storage and Analysis:

[2040] The server stores all the data from the interview.

[2041] The server analyzes the stored data using a generative model and sentiment analysis engine to extract various indicators (e.g., frequently occurring keywords, sentiment trend analysis, etc.).

[2042] The server compiles the analysis results in a report format that is easy to understand visually (e.g., PDF or dashboard).

[2043] 5. Report provision:

[2044] The server provides the generated report to the company via a download link or the like.

[2045] As described above, the system of the present invention uses a generative AI model and an emotion analysis engine to realize high-quality interviews that take into account the user's emotions, enabling efficient and detailed information collection.

[2046] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2047] Step 1:

[2048] Data Collection Phase

[2049] What it does: The server collects a dataset to use in the interviews. This includes past interview transcripts, questionnaires, user profile information, and sentiment data. The server uses an API to download the data from cloud storage and saves it to local storage.

[2050] Input: Interview transcripts in cloud storage, questionnaire, user profile information, and sentiment data.

[2051] Output: The initial dataset saved to local storage.

[2052] Step 2:

[2053] Data Preprocessing

[2054] Specific operation: The server preprocesses the collected data, specifically by using Python libraries (numpy, pandas, NLTK, etc.) to perform data cleaning (noise removal), text tokenization, and stop word removal.

[2055] Input: An initial dataset stored in local storage.

[2056] Output: A preprocessed and clean dataset.

[2057] Step 3:

[2058] Building and training the model

[2059] How it works: The server trains a generative model (e.g., GPT-4) and a sentiment analysis engine based on the preprocessed data. It uses libraries such as TensorFlow and PyTorch. The server exports the trained model to a file and uploads it to cloud storage.

[2060] Input: Preprocessed clean dataset.

[2061] Output: A trained generative model and a sentiment analysis engine.

[2062] Step 4:

[2063] User access and interview start

[2064] Specific operation: A user accesses the system using a terminal and accesses a web interface (e.g., developed with React). The user is redirected to the homepage.

[2065] Input: The user's access request.

[2066] Output: Web interface display.

[2067] Step 5:

[2068] Displaying the user registration form

[2069] Specific operation: The device displays a user registration form. The registration form includes fields for entering name, age, interests, etc. After the user enters the information on the device, the form data is sent to the server.

[2070] Input: User information (name, age, interests).

[2071] Output: The user information sent to the server.

[2072] Step 6:

[2073] Saving user information and generating initial questions

[2074] Specific operation: The server stores user information in a database (e.g., MySQL, MongoDB). It generates initial questions based on the user information and sends them to the device. A generative AI model is used for generation.

[2075] Input: The user information sent to the server.

[2076] Output: User information stored in the database and the initial questions generated.

[2077] Step 7:

[2078] Display initial questions

[2079] Specific behavior: The device displays the initial question received from the server. Example: "Hello, [user name]. Today I'd like to ask you about a new product. First, what are your impressions of this product?"

[2080] Input: The initial question sent by the server.

[2081] Output: The initial question displayed on the user interface.

[2082] Step 8:

[2083] User answer input

[2084] Specific operation: The user enters an answer to a question using the terminal and presses the send button. The user enters an answer in the text area and clicks the send button.

[2085] Input: The user's answer.

[2086] Output: The user's answer sent to the server.

[2087] Step 9:

[2088] Analyzing answers and generating the next question

[2089] Specific operation: The server analyzes the user's answer using a generative model and generates the next question. It then analyzes the user's emotions using a sentiment analysis engine and adjusts the next question accordingly. For example, if the user expresses anxiety, it changes the question to be more gentle. The generated question is then sent to the device.

[2090] Input: The user's answer.

[2091] Output: The parsed answer and the newly generated question.

[2092] Step 10:

[2093] View new questions and re-answer

[2094] Specific operation: The terminal displays the new question received from the server. The user enters the answer again and presses the send button. This causes steps 8 to 10 to be repeated.

[2095] Input: The new question sent by the server.

[2096] Output: New question displayed on the user interface, user's re-answer.

[2097] Step 11:

[2098] Interview data storage

[2099] What happens: The server stores all the interview questions and answers in a database.

[2100] Input: Interview data (all questions and answers).

[2101] Output: Interview data stored in a database.

[2102] Step 12:

[2103] Data analysis and reporting

[2104] Specific operation: The server analyzes the stored data using a generative model and sentiment analysis engine to extract various indicators (e.g., frequently used keywords, sentiment trend analysis, etc.), compiles the results in a report format, exports the generated report to a file, and uploads it to cloud storage.

[2105] Input: Saved interview data.

[2106] Output: The generated report.

[2107] Step 13:

[2108] Providing reports

[2109] Specific operation: The server provides the generated report to the company via a download link, etc.

[2110] Input: The generated report.

[2111] Output: A download link that will be provided to the company.

[2112] (Application example 2)

[2113] 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."

[2114] Conventional user interview systems are limited to one-way question-and-answer exchanges, which means they are unable to fully reflect user emotions and detailed feedback. This makes it difficult to accurately grasp actual user experiences and opinions about products. Furthermore, they often fail to generate appropriate questions or adjust the dialogue in response to user emotions, resulting in a poor user experience.

[2115] 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.

[2116] In this invention, the server includes means for generating questions for a user using a generative model, means for analyzing answers from the user, means for generating a next question based on the analyzed answer, means for recording the generated question and the analyzed answer, means for analyzing the recorded data and generating a report, means for analyzing the user's emotions, and means for adjusting the next question based on the result of the user's emotion analysis. This makes it possible to collect detailed and accurate feedback while adjusting the content of the dialogue according to the user's emotions.

[2117] A "generative model" is an algorithm that automatically generates questions and information based on user input.

[2118] The "means for analyzing the answer" is a technique for understanding the answer entered by the user and analyzing its content.

[2119] The "means for generating questions" is a technology that automatically creates the next question based on the analyzed answer.

[2120] The "recording means" is a technique for saving the generated questions and analyzed answers in a storage device such as a database.

[2121] The "means for generating a report" is a technique for analyzing the recorded data and outputting the results in the form of a report.

[2122] "Means for analyzing emotions" refers to technology that determines emotions from users' responses and quantifies the results.

[2123] The "means for adjusting the next question based on the emotion analysis result" is a technology that appropriately changes the content and wording of the next question based on the analyzed emotion.

[2124] "Emotion trend analysis" is a technique that analyzes the transitions in a user's emotions and identifies changes and trends in emotions over time.

[2125] "Means for collecting feedback" refers to techniques for collecting opinions and evaluations from users about products they have purchased.

[2126] The present invention relates to a user interview system that combines a generative model and an emotion engine, and is intended in particular for collecting feedback on product evaluations on online shopping sites. A specific embodiment of the system that realizes this application example is described below.

[2127] System configuration

[2128] Hardware

[2129] Device: The smartphone or smart glasses used by the user.

[2130] Server: A server for data processing, question generation, answer analysis, sentiment analysis, and data storage.

[2131] software

[2132] Generative model: Used to generate questions and analyze user responses. Specifically, it uses natural language processing technologies such as OpenAI's GPT-3.

[2133] Sentiment engine: Software for analyzing emotions from user responses. Specifically, it uses a sentiment analysis model using the Transformers library.

[2134] Database: A database for storing user profile information, interview results, sentiment analysis results, etc.

[2135] Feedback Collection Process

[2136] Initial Setup

[2137] The server first collects user profile information (such as name, purchase history, etc.) and stores it in a database, then uses a generative model to create initial questions.

[2138] Interview begins

[2139] The user accesses the system from a terminal and the interview begins. Initial questions are displayed and the user submits their answers through an input form.

[2140] Question and answer exchange

[2141] The server analyzes the user's answers and uses an emotion engine to analyze the emotions from the answers. Based on the analysis results, the server generates the next question and adjusts the question content according to the user's emotions. This allows the server to obtain detailed feedback from the user.

[2142] Data storage and analysis

[2143] After all interactions are completed, the server stores the collected data in a database and analyzes the user's emotional trends from the recorded data. Based on the analysis results, the server generates a report and provides it to the company via a download link or other means.

[2144] Specific examples

[2145] A user uses a smartphone app to provide feedback on a product they have purchased. The first question is displayed in the form, "Hello, [user name]. Please rate the product you purchased. What was your first impression?" As the user enters their answer, the answer is analyzed in real time and the next question is generated based on their emotional state. For example, the question is adjusted accordingly, such as, "Tell me more about the design. What did you particularly like about it?"

[2146] Prompt Sentence Examples

[2147] User Answer: I like the design, but the functionality is a bit unclear.

[2148] Emotion: Negative

[2149] Generate the following questions:

[2150] In this way, a system is realized that adjusts the dialogue content according to the user's emotions and collects detailed and accurate feedback.

[2151] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2152] Step 1: User registration and initial setup

[2153] A user accesses the system using a terminal and enters profile information such as the user's name and purchase history. The server stores this input data in a database and generates initial questions using a generative model. The generated initial questions are in the form of, for example, "Please rate the product you purchased. What was your first impression?"

[2154] Step 2: Displaying initial questions

[2155] The server sends the generated initial question to the terminal, which displays the question to the user, and the user inputs an answer to the question and sends it to the server via the terminal.

[2156] Step 3: Response analysis and sentiment analysis

[2157] The server receives the answer sent by the user and analyzes the answer using a generative model. At the same time, it analyzes the user's sentiment from the answer using an emotion engine. For example, if a user answers "I think the design is good, but the functionality is a little unclear," the emotion analysis engine will detect negative sentiment.

[2158] Step 4: Generate the next question

[2159] The server generates the next question based on the analyzed answer and emotion data. It uses a generative model to generate a question that corresponds to the user's emotion. For example, if a negative emotion is detected, it generates a question such as, "What specific aspects of the feature did you find unclear?" The generated question is sent from the server to the device and displayed to the user.

[2160] Step 5: Conducting the interview

[2161] The user answers the next question and sends the answer to the server via their device. This process repeats, with the server analyzing each answer and generating the next question. Based on the sentiment analysis results, the question content is adjusted accordingly.

[2162] Step 6: Data storage and final analysis

[2163] After the interview is over, the server stores all questions and answers, as well as the sentiment analysis results, in a database. The recorded data is then analyzed again for sentiment trend analysis, summarizing changes in users' sentiment and opinion trends.

[2164] Step 7: Generate reports

[2165] The server generates reports based on the accumulated data, including analysis results such as classification of user opinions and sentiment trends. The generated reports are provided to the company via a download link or other means.

[2166] In this way, detailed and accurate feedback can be collected from users, and appropriate questions can be generated based on the results of sentiment analysis, making the interviews more effective.

[2167] 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.

[2168] 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.

[2169] 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.

[2170] 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.

[2171] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[2172] 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.

[2173] 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).

[2174] 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.

[2175] 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."

[2176] 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.

[2177] 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).

[2178] 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.

[2179] 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.

[2180] 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.

[2181] 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.

[2182] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[2183] 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.

[2184] 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.

[2185] 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.

[2186] 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.

[2187] 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.

[2188] The following is further disclosed regarding the above embodiment.

[2189] (Claim 1)

[2190] means for generating questions for a user using the generative model;

[2191] means for analyzing responses from users;

[2192] means for generating a next question based on the parsed answer;

[2193] a means for recording the generated questions and parsed answers;

[2194] means for analyzing the recorded data and generating reports;

[2195] A system including:

[2196] (Claim 2)

[2197] 10. The system of claim 1, further comprising means for generating initial questions based on user profile information during an initial stage of an interview.

[2198] (Claim 3)

[2199] 2. The system of claim 1, wherein the generative model uses natural language processing techniques.

[2200] "Example 1"

[2201] (Claim 1)

[2202] means for generating questions for a user using the generative model;

[2203] means for analyzing responses from users;

[2204] means for generating a next question based on the parsed answer;

[2205] a means for recording the generated questions and parsed answers;

[2206] means for analyzing the recorded data and generating reports;

[2207] A means of saving all interview data, analyzing the saved data using a generative model, and extracting various indicators;

[2208] A system including:

[2209] (Claim 2)

[2210] 10. The system of claim 1, further comprising means for generating initial questions based on user profile information during an initial stage of an interview.

[2211] (Claim 3)

[2212] 2. The system of claim 1, wherein the generative model uses natural language processing techniques.

[2213] "Application Example 1"

[2214] (Claim 1)

[2215] a means for generating a question for a user using the generative model;

[2216] A means for analyzing responses from users;

[2217] means for generating a next question based on the parsed answer;

[2218] a means for recording the generated questions and parsed answers;

[2219] means for analyzing the recorded data and generating a report;

[2220] A means for users to access the system using a terminal and participate in the interview;

[2221] a means for the terminal to display questions and obtain answers from the user;

[2222] A means of providing the manager with the report generated after the interview is completed;

[2223] A system including:

[2224] (Claim 2)

[2225] 10. The system of claim 1, further comprising means for generating initial questions based on user profile information during an initial stage of an interview.

[2226] (Claim 3)

[2227] 2. The system of claim 1, wherein the generative model uses natural language processing techniques.

[2228] "Example 2: Combining Emotion Engines"

[2229] (Claim 1)

[2230] means for generating questions for a user using the generative model;

[2231] means for analyzing responses from users;

[2232] means for generating a next question based on the parsed answer;

[2233] a means for recording the generated questions and parsed answers;

[2234] means for analyzing the recorded data and generating reports;

[2235] means for training a generative model and a sentiment analysis engine based on the collected data;

[2236] A means for analyzing emotions from user responses and adjusting the content of the dialogue;

[2237] A system including:

[2238] (Claim 2)

[2239] 10. The system of claim 1, further comprising means for generating initial questions based on user profile information during an initial stage of an interview.

[2240] (Claim 3)

[2241] 2. The system of claim 1, wherein the generative model uses natural language processing techniques.

[2242] "Application example 2 when combining emotion engines"

[2243] (Claim 1)

[2244] means for generating questions for a user using the generative model;

[2245] means for analyzing responses from users;

[2246] means for generating a next question based on the parsed answer;

[2247] a means for recording the generated questions and parsed answers;

[2248] means for analyzing the recorded data and generating reports;

[2249] means for analyzing user emotions;

[2250] a means for adjusting the next question based on the result of the user's sentiment analysis;

[2251] A system including:

[2252] (Claim 2)

[2253] 10. The system of claim 1, further comprising means for generating initial questions based on user profile information during an initial stage of an interview.

[2254] (Claim 3)

[2255] 2. The system of claim 1, wherein the generative model uses natural language processing techniques.

[2256] (Claim 4)

[2257] 2. The system according to claim 1, further comprising means for analyzing user sentiment trends from the recorded data and reflecting the analysis results in a report.

[2258] (Claim 5)

[2259] 10. The system of claim 1, further comprising means for collecting feedback on products purchased by users. [Explanation of symbols]

[2260] 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. means for generating questions for a user using the generative model; means for analyzing responses from users; means for generating a next question based on the parsed answer; a means for recording the generated questions and parsed answers; means for analyzing the recorded data and generating reports; A system including:

2. 10. The system of claim 1, further comprising means for generating initial questions based on user profile information during an initial stage of the interview.

3. The system of claim 1 , wherein the generative model uses natural language processing techniques.

Images